The Increasing Pay Gap for Women in Textile and Clothing Industries: A Reexamination

Abstract

In the December 1980 issue of this JOURNAL, Paul McGouldrick and Michael Tannen analyze changes in the male-female earnings gap in the textile and clothing industries during the period 1910-1970.1 They attempt to standardize male and female wages by adjusting for differences in age, marital status, work experience, ethnicity, race, and other factors. Their results show a very small sex differential in earnings in 1910 after adjusting for these various factors. In contrast, a similar adjustment procedure for 1970 fails to explain much of the male-female earnings gap, and we are left with a rather large (33 percent) unexplained sex differential. McGouldrick and Tannen conclude that there was an increasing pay gap between the sexes in the clothing and textile industries during the period 1910-1970, and they suggest that this may have resulted from a relative decline in female human capital investment and changes in union and government benefits and taxes that discriminate against females. Based on an analysis of the sample used by McGouldrick and Tannen, it appears that these conclusions are not justified. In fact, the evidence indicates that the male-female earnings differential in the clothing and textile industries has not widened but rather has shown an amazing degree of stability since the early twentieth century. The major problem with McGouldrick and Tannen's work is that the data used for the early period are not representative and contain several sources of bias that lead to serious understatement of the actual sex differential. McGouldrick and Tannen's analysis for the early period is based on data presented in a study of employment and earnings in the manufacturing and mining industries conducted by the Immigration Commission in 1910, and this survey gives a very heavy weight to foreign-born employees.2 In the Immigration Commission report, 69 percent of the surveyed workers were foreign born and 31 percent were native born. At the time, immigrants comprised 20 percent of the total labor force. Foreign-born workers made especially high contributions to the total labor force in textiles (36.2 percent) and clothing (55.1 percent), but their share of total employment was not close to the 69 percent level found in the McGoludrick-Tannen sample. The heavy weight given to foreign-born workers introduces a substantial bias into the measurement of the sex differential in earnings. The male-female earnings gap for foreign-born workers was substantially lower than the male-female earnings gap for native-born workers. This was particularly true in the textile industry. Using data from the Immigration Commission study, we have calculated the ratio of female/male wages in the textile industry at .815 for foreign-born workers and .707 for native-born workers. In the clothing industry, both nativeand foreign-born females earned substantially less