A close study of Mr. Gabriel Kolko’s article, “The American ‘Income Revolution’ ” in the Winter 1957 issue of DISSENT shows its scientific apparatus to be faulty; but I shall confine my criticism to the tables on pages 48-49, which are central to the article. They seem to show that nearly 50 per cent of American families had an income below the so-called “maintenance level” in 1935/6 as well as in 1947 and 1950. Could anybody who has looked around seriously maintain that “the concentration of large families below maintenance level has remained constant since WPA” and that living standards during the 1935/6 depression years equalled the post-war boom years? How were these statistics arrived at?
The riddle’s solution is simple. When the WPA defined “maintenance level” in 1937 it tried to approximate the expenses of the median family ($1,261 a year). How well it succeeded in doing so is confirmed by its findings: in effect, one half of the consumer units (48.8 is an extremely good approximation) spent more and the other half spent less than the assumed median. Subsequently, while the percentage weights given to the various items in the cost of living index were kept stable, the quality definitions of the various consumer goods bought by the representative family were constantly adjusted to current consumer habits. In computing its index, the Bureau of Labor Statistics took great pains to ascertain what the median family actually bought.
No wonder, then, that in 1947 we again find approximately half of the families below and the other half above “maintenance,” then $3,300, rising to $4,166 in 1951. This increase reflects more than the price inflation. In 1950, namely, the percentage shares of various consumer goods in the total index were revised, at the insistence of the trade union lobby in Congress. As pointed out at the time, consumer habits had changed so fundamentally that new standards were needed.