IN the United States there has been much agitation of late for a shorter work week. In the spring of 1933 Senator Black of Alabama introduced a bill in the United States Senate to limit the work week to thirty hours. This bill passed by a vote of 53 to 20, but was held up on a motion to reconsider. Simultaneously Representative Connery introduced a similar bill in the House. Support given to these bills undoubtedly assured the passage on June 13, 1933, of the National Industrial Recovery Act, one purpose of which was to increase purchasing power by decreasing working hours (thus spreading work) and by increasing wages. Decreasing hours without increasing hourly wage rates to compensate was impossible politically. Prior to the Black and Connery Bills, the United States had been literally covered with the publicity and propaganda attendant upon the pronouncements of Technocracy. Although most of the claims of Technocracy have been proved to be false, the notion persists that great strides have been made in improving machines during the last decade, especially during the depression. In view of the debatable character of shorter work week proposals, it seems desirable at this time to examine some of the facts and to see how these are related to economic theory. In particular, it seems desirable to consider the economic experiment of the shorter work week carried out by the N.R.A., and to discover how hour shortening and attendant wage raising might lead to recovery or further depression. Figure I shows how gainful workers by the different classes of workers varied from 1880 to 1930. In 1880 about 8 per cent of all gainfully employed were professional workers compared with about 25 per cent in manufacturing and mining. About 47 per cent were engaged in agriculture and fishing. Employment percentages, however, did not remain static. The change has been very considerable in the last thirty years. For example, percentage of those engaged in agriculture and fishing in 1930 was 23 per cent, compared with 47 per cent in 1880. In manufacturing and mining a slight decrease occurred over the previous ten-year period. Table IV shows changes in workers in manufacturing and man-hours required during the period 1920-1933. During the period of industrial expansion, we were opening new manufacturing industries and building up our own industrial plants. Improved manufacturing technique and things of that sort were displacing workers and thus fewer were needed in manufacturing, but employment in trade

OF recent years there has been much speculation and argument regarding the nature of savings and investments and the rates of income received by such groups as entrepreneurs, property owners, creditors, industrial managers, the white-collar class, workingmen, and farmers. Unfortunately, arguments and opinions on these questions have not generally been supported by facts, for the reason that adequate statistical information has not been available. With the advent of the Deal in the United States and the accompanying rapid growth of new governmental agencies, often created with the avowed intention of bringing about a distribution of income and each possessing a division created to collect and interpret statistics, it was inevitable that a wealth of new data on the flow and use of income should become accessible. Furthermore, largely owing to the guidance of the new Central Statistical Board, composed of representatives of various departments and agencies, much of this information is thoroughly reliable. As a result of the activities of this Board, some of the older indexes have been revised to meet criticisms, and established departments of the government have often enlarged the scope of their work. Such new organizations as the National Recovery Administration, which required statistics on a variety of economic problems, greatly stimulated the collection of data by both governmental and private agencies. During the life of NRA,1 many of the six hundred or more code authorities which governed industries collected and analyzed statistics bearing on their particular activities. Data which had previously been gathered in inadequate samples by trade associations became vastly better because of increased coverage and many indexes were greatly improved. Under NRA, some industries preferred to hire private statistical agencies, such as Dun and Bradstreet of New York City, Real Estate Analysts, Inc., of St. Louis, Missouri, and Price, Waterhouse and Company of New York, to collect and interpret their material. In spite of the Supreme Court decision, several of these organizations are continuing to publish statistical series begun at the instance of code authorities. Indeed, the Real Estate Analyst has so vastly enlarged its

Charles F. Roos, Leonid Hurwicz, Benjamin Higgins, Tjalling Koopmans, Everett E. Hagen, K. G. Fuller, L. R. Nienstaedt, Jacob Marshak, Forecasting Postwar Demand: Discussion, Econometrica, Vol. 13, No. 1 (Jan., 1945), pp. 54-59