Journal Article The Retarded Acceptance of the Marginal Utility Theory: Comment Get access John P. Henderson John P. Henderson University of Maryland Search for other works by this author on: Oxford Academic Google Scholar The Quarterly Journal of Economics, Volume 69, Issue 3, August 1955, Pages 465–473, https://doi.org/10.2307/1885853 Published: 01 August 1955

SINCE about I950 the Bureau of Labor Statistics has been publishing a monthly series showing the industrial distribution of the employed nonfarm labor force.' In addition to supplying current estimates on nonagricultural employment, the figures have been compiled from I9I9 and yield, therefore, a continuous forty-year series on the changing pattern of the industrial distribution of employment.2 major value of this series is in evaluating current employment trends, serving as an important supplement to the Census Bureau's monthly estimate of aggregate employment. BLS figures can also be used effectively to study the changing industrial pattern of employment.3 BLS estimates have several distinct advantages over the usual statistical series used to evaluate long-run changes in employment patterns. Typically such series rely exclusively upon decennial data, to the exclusion of either monthly or yearly figures on the industrial composition of the work force.4 major weakness in such a dependence stems from the obliteration of cyclical changes in the industrial composition of employment, and what may be only short-run changes can too easily be interpreted as long-run movements.5 For example, decennial data for I930-50 show that there has been a long-run increase in employment in the tertiary industries, and this fits the pattern which Colin Clark and others hypothesize.6 However, yearly data on employment patterns, such as those of the BLS, indicate that there is considerable cyclical movement in the industrial composition of employment, with the result that service occupations rise as a percentage of total employment in recessions or periods of less than full employment. Each of the recent Census years (I930, I940, and I950) had a sizeable portion of unemployment, and it is by no means clear that the long-run trend of the tertiary industries is entirely consistent with the industrial pattern that emerges when decennial data are used. However, it is not the purpose of the present paper to discuss the degree to which decennial data are truly indicative of long-run trends, nor to evaluate the effectiveness of BLS data for estimating short-run movements. present intent is merely to contrast the industrial distribution of employment obtained from BLS data with two distributions derived from the decennial data. More specifically, the distribution derived from BLS figures is compared with two estimates prepared by researchers associated with the National Bureau of Research. First, the Carson-Barger estimates for the labor force7 and, second, John W. Kendrick's more recent employment figures for the industrial distribution 8 are juxtaposed with BLS figures. Both the Carson-Barger and the Ken'Bureau of Labor Statistics, United States Department of Labor, Handbook of Labor Statistics, 1950 Edition, Bulletin ioi6 (Washington, 1950), I-5; Measurement of Industrial Employment, Monthly Labor Review, 76 (September 1953), 968-73. 2 Bureau of Labor Statistics, Department of Labor, and Earnings, Annual Supplement (June 1958). 'Ewan Clague, The Shifting Industrial and Composition of the Work Force During the Next Ten Years, Daily Labor Report (Washington, 1958), Thursday, January i6, 1958, Special Supplement, i-II; reprinted in Monthly Labor Review, 8I (July 1958), I676-9I. ' Cf. Simon Kuznets, Quantitative Aspects of the Growth of Nations, III. Industrial Distribution of National Product and Labor Force, Development and Cultural Change, vi (July 1957), 19-32; Colin Clark, Conditions of Progress (3rd ed., New York, 1957); P. Whelpton, Occupational Groups in the United States, I820I920, Journal of the American Statistical Association (September I926), 335-43; and Daniel Carson, Changes in the Industrial Composition of Manpower Since the Civil War, Studies in Income and Wealth, Vol. xi (New York, I949), 46-I34. For an instance of such misinterpretation, see George Stigler, Trends in Employment in the Service Industries (Princeton, 1956). 6 Colin Clark, Conditions of Progress (London, 1957) ; and A. G. B. Fisher, Economic Implications of Material Progress, International Labour Review, LVII (July

A LTHOUGH the symmetry of the relation2i ship between factor and consumer markets has been questioned in studies of wage differentials,' little investigation has been undertaken to determine the directness of that relationship and the degree of symmetry which may exist. The findings of this paper offer some evidence for the conclusion that the relationship is inverse, and that where earnings are low, the cost of living tends to be high. An intercity comparison of both earnings and the cost of living indicates that the magnitude of the earnings differential among the large cities of the United States is three to four times that of the cost-of-living differential. The data presented in Tables i and 2 show an inverse relationship between earnings and cost of living which suggests that the worker in substandard earnings areas is forced to reconcile his position in terms of a lower standard of living than that prevailing in more advantageous factor markets. Table i shows the differentials in the earnings of direct labor 2 in 35 cities throughout the United States. Table 2 shows the cost-of-living differentials in 33 cities. Differentials are based on the relationships of local earnings and living costs to those in New York City during the period of full employment in I95i. The cities included are grouped into five regions: New England, Middle Atlantic, South, Middle West, and Far West. Both earnings and cost-of-living information are not available in all cases, but 23 cities common to both sets of data provide direct local comparisons. Some discrepancies in the timing of the collection of raw statistics should not be considered as hindering the reliability of comparisons, since it is proper to assume that these differentials are constant, at least in the short run.3 Differentials in the earnings of direct labor are presented in Table i as percentages of average hourly earnings in New York City for October I95 I.4 Average earnings for the month were computed by dividing the monthly earnings reported by the total number of hours worked.5 These were then converted to a simple index and ranked in column two. The method of computation included payments for premium hours nd such incentive wage systems as happen to have been in effect during the month of October. While it is not the purpose of this paper to discuss in any detail the circumstances which are responsible for the position of particular communities in the rankings, it should be noted that intercity comparisons reflect heterogeneous influences which provide a useful basis for investigation of the diverse factors involved in arnings differentials. The rankings retain the