Clothing; other wants. Clothing comes last of the three to a constant or a diminishing proportion of the expenditures. In the New York families under consideration the percentage expenditure rises slightly with each increase of $100 in income until the $1100 group is reached, and thereafter remains constant at about 15 per cent. The expenditures for other purposes than these three primary necessities are kept under until these wants are met. By the time something like an equilibrium among these three has been reached, say at $800 for our New York families, the expenditure for recreation, social obligations, care of the health, and all other purposes save fuel and light, claims a larger proportion of the income. The proportion is 1 per cent higher at $700 than at $600, but at $800 it rises from 14 to 16 per cent of the total expenditure, and continues to increase without sign of stopping. That is, the culture-wants are beginning to claim their own, which, under the necessity of keeping the wolf from the door, they could not be permitted to have.

A striking example of this tendency of subsistence-wants to claim the lion's share of all increasing income is found in Engel's comparison of the Belgian returns of 1853 with those of a similar investigation made in 1891. At the latter period, although the average income had nearly doubled, the expenditure for food comprised 65.7 per cent of the total in 1891 as compared with 64.9 per cent in 1853. In fact, food, clothing, rent, and fuel and light consumed 96 per cent of the income in 1891 and only 94 per cent in 1853.

Minimum standards of consumption. The same general conclusion as to the relative insistence of the several classes of wants may be drawn from another method of handling the New York returns. A minimum standard, as exact as could be determined, was applied to the expenditures for food, clothing, and housing, and the number of families counted in each income group who came short of the standard. For food, the minimum was set at an expenditure at the rate of 22 cents per man per day, as calculated after the manner made familiar by W. O. Atwater in the Bulletins of the Department of Agriculture. This figure was reached, after an analysis of one hundred of the family reports, by Dr. Frank P. Underhill, of Yale University, a competent expert. Professor Atwater's estimate on the basis of data gathered in New York City a few years previous, when a lower scale of prices prevailed, was from 23 to 25 cents. For housing the minimum was fixed at one and one-half persons per room, that is, not more than six persons to four rooms. For clothing the minimum was set at an allowance of $100 for the assumed family of five persons, expenditures for washing being included in this sum.

For our present purpose the accuracy of these estimates of a minimum requirement for physical efficiency does not concern us, but only the variations in the departures from them that appear in the several income-groups. Measured by these standards, of the families with incomes between $400 and $500 all are underfed, 88 per cent are underclad, 63 per cent are overcrowded. That is, the want of shelter is being satisfied at the expense of food and clothing. In the next income-group ($500-$600), the underfed are 65 per cent, the under-clothed, as before, 88 per cent., the overcrowded 71 per cent. In paying more attention to the need of food, less attention is paid to shelter. A higher rental is paid, but more persons are crowded into the accommodations offered. In the next income-group ($600-$700) the underfed have fallen to 33 per cent., the underclad to 63 per cent., the overcrowded to 57 per cent. For every income-group thereafter, the overcrowded families preponderate over both the other classes. Even in the $1100 income-group 21 per cent are overcrowded, but none underfed, and only 6 per cent underclad. These figures, taken as a whole, imply that the most urgent need at the minimum income is for shelter, outclamoring not hunger perhaps, but at least the want of adequate food. "With a larger income a pause can be set to the desire for better housing, while more attention is given to the providing of food. [With an income still larger, of $900 and above, the deficiencies in diet are supplied, and at $1000 the minimum allowance for clothing has been attained by practically all the families. Not even at this point, however, does the desire for adequate housing, at the price which must be paid for it, suffice to persuade more than three-fourths of the families to go without enough of other things to secure it.

Saving. Another alternative to expansion of expenditures, for whatever purpose, as income increases, is saving. Saving becomes easier as income increases. But the point where savings begin is not necessarily the point where a standard even of physical efficiency is attained. There are families that save at the expense not only of comfort, but even of health, and there are families that no increase of income would induce to save. Of the underfed families just alluded to, one-half reported a surplus of income over expenditure of at least $25; 65 per cent of the families reckoned as under-clothed, and 44 per cent of the overcrowded likewise reported such a surplus. When this is compared with the percentage of all families that reported a surplus, namely 36.5, it seems fair to infer that the desire to save represses expenditures to meet actual physical necessities.

On the other hand, by no means all families on a larger income preferred saving to spending. Not until $1300 is reached is there a constant increase in the number of families that report a surplus of income over expenditures. This indicates that there are Micawbers on large incomes as there are misers on small incomes, but also that the social influences of New York City, at least, encourage adding to the good things included in standards of living quite as much as they encourage saving. The proportion of savers among the Russian and Italian families was found to be much higher than among families of more thoroughly Americanized stock.

Conclusions. On the whole the conclusions drawn from the New York investigation substantiate the restatement of Engel's "laws" given by Stephen Bauer in his article "Konsumtionsbudget" in Conrad's Handworterbuch, as follows :

With increase of income:

1.  The proportion spent for food, especially for vegetable food, falls.

2.  The proportion saved constantly increases.

3.  The proportion spent for housing, fuel, light, falls until a certain income is reached, then remains constant or increases.

4.  The proportion spent for animal food, drink, clothing, culture, and recreation rises until a certain income is reached, then remains constant or falls.