The Character of the Data

Until recent years the data on hourly earnings for industrial workers and weekly hours worked have been unsatisfactory both from the point of view of the number of industries covered and the frequency with which they were covered. The Bureau of Labor Statistics made various spot studies in certain industries but no attempt was made to establish a continuous series. In 1932 the Bureau began to make good the deficiencies by collecting monthly figures on earnings and hours for a considerable range of manufacturing industries and some nonmanufacturing industries. The early results of this new venture were, of course, less satisfactory than the later.

In this study both the earlier and later figures collected by the Bureau of Labor Statistics, up to 1936, were compared with estimates for 1929 in order to show the changes which had occurred over the period. The 1929 estimates were based on the employment and pay-roll data which the United States Bureau of Labor Statistics had been collecting for one week in each month for some years before 1929. By dividing the pay-roll figures by the employment figures weekly earnings were obtained. Data on weekly hours worked, however, were not available in the Bureau's records. These were obtained by multiplying the prevailing hours worked as reported by the Census for 1929 with the percent of full time worked as reported by the Bureau. Having the weekly earnings and the weekly hours, the hourly earnings were derived by simple division.

Wage and Employment Results, 1929, 1932-36

Table I presents hourly earnings, weekly hours worked, weekly earnings, workers employed, and manhours worked for 44 manufacturing industries in 1929, and annually from 1932 to 1936. The 44 industries include 15 industries, as defined by the Census, which employed, in 1935, over 100,000 persons each, 30 which employed from 25,000 to 100,000 each, and 9 under 25,000. All combined these industries employed 4,359,

1 Appendix 6 was prepared by Edward B. Mittelman, assisted by Nancy Hart and Paul A. Fischer.

000 wage earners or 60 percent of all wage earners employed in manufacturing. The large and mediumsized industries included in the table employed more than two-thirds of the wage earners employed by the large and medium-sized industries listed in appendix 7. The small industries included in the table covered less than 10 percent of the wage earners employed by industries employing less than 25,000 persons.

The hourly earnings in 1929 in each of the 44 industries were derived, as just indicated, by dividing weekly earnings as computed from the employment and payroll figures of the Bureau of Labor Statistics by the weekly hours as computed from the Census and Bureau figures. Another possible method calculating hourly earnings was to divide the Census wages by annual man-hours as derived by multiplying the weekly hours (obtained as described above) first by 52 and then by the average number employed. The results obtained by the first method are compared with the results obtained by this second method in table I-A. In 38 of the 44 industries, the hourly earnings as calculated by the second method are within 10 percent of the figures obtained by the first; in the remaining six industries they are subject to serious question. However, no conclusions from the results on all these industries have been drawn in this report which assume a greater reliability of these figures than that indicated above.

In table II, the industries are combined, first on the basis of durable goods industries, and semi- and nondurable goods industries, and then, on the basis of concentration. Industries in which the four largest enterprises employed at least 30 percent of all workers are considered concentrated; the others, not concentrated. The hourly earnings for all industries in each class are unweighted averages. Weighted averages can be derived from table I on the basis of the number employed in each industry. There is no significant difference, however, between the two types of averages.

Some of the industries for which wage data are presented combine two or more industries separately reported in the Census of Manufactures for 1935, and some exclude industries separately reported, at one time or another, between 1929 and 1935. The 44 industries given in table I are equivalent to more than 44 on the basis of the Census classification. Thus the number of wage earners in any one of the 44 industries in table I might not coincide with the number in the Census industry having the same name.

231

1 As of 1929, 1931, 1933 census classifications and carried forward for later years on a comparable basis.

2 Phonographs only.

3 Radios only for March to December.

Wages and Hours in the Motor Vehicle Industry, 1932, Monthly Labor Review, June 1933, p. 1371.

Compiled from data reported by the Silverware Manufacturers Institute to NRA, Division of Research and Planning, reproduced in Wages and Hours in American Industry, NRA Source Material in 3 volumes, by Solomon Barkin and Anne Page, Labor Studies Section, Vol. III, p. 1117, Division of Review, Washington, D. Č.,

1936.

Includes clothing (except work), mer's and clothing, work, men's on the ceusus basis for 1929 and 1931, and carried forward for later years on a comparable basis. 7 Wages and Hours in the Men's Clothing Industry, 1928, Bulletin 503, U. S. Bureau of Labor Statistics, p. 48; Ibid. 1930, Bulletin 557, p. 49.

Wage and Hours of Labor in the Men's Clothing Industry 1982, Bulletin 594, U. S. Bureau of Labor Statistics, p. 39.

10 Wages and Hours in the Boot and Shoe Industry, 1910-1932, Bulletin 579, U. S. Bureau of Labor Statistics, p. 78.

11 Cigarette, Snuff, Chewing, and Smoking Tobacco Manufacturing Industry, by Donald Yakeley, p. 53; mss. Miscellaneous NRA files 1934, p. 65. This source was used because it appeared to be more reliable than the figure derived from the procedure generally followed. The figure derived from that procedure was 33.7 cents. This was less than the hourly earnings in cigars and cigarettes, and the relationship generally is the other way around.

1 The NRA results based on data supplied by Chemical Alliance, Inc., involving 36,237 workers in 1929 and 23,597 in 1932 gave 55.4 cents for 1929 and 52.3 cents for 1932 and 51.3 hours for 1929 and 41.0 hours for 1932. Wages and Hours in American Industry, NRA Source Material in 3 volumes, by Solomon Barkin and Anne_Page, Labor Studies Section, Vol. I, pp. 213, 218, Division of Review, Washington, D. C.,

1936.