Two proxies for state union attitudes

Abstract

In a recent article ["Do Right to Work Laws Have an Impact on Union Organizing Activities," The Journal o f Social and Political Studies, Vol. 4, No. 1, Spring 1979, pp. 81-93], I reported the results of a study testing the impact of state right-to-work (SRTW) laws on union activities. In the research, four union activity proxies-petitions filed for representative elections, elections held, elections won, and votes cast for unions-were respectively regressed against 14 independent variables including a SRTW dummy for the years 1950, 1960, and 1970. Only the 1950 statistical tests on SRTW's were significantly negative. It was concluded that union activities shifted downward in the RTW states as Section 14b of the 1947 TaftHartley Act was implemented and organizing resources were reallocated to non-RTW states. By 1960, a new organizing equilibrium was established and there has been no effect since then. In the study, two unique independent variables were developed to measure tastes and preferences among individual states for unions. The purpose of this note is to draw special attention to these variables so that future researchers can specify more complete models for empirical testing. A state union attitude variable (UAV) was one of these special explanatory instruments. This variable was used to measure public sentiment towards unions and was based on the average ratings of Congressmen and Senators by the Committee on Political Education (COPE) of the AFL-CIO. Of course, COPE is only one such rating organization. Other possible political indices would include the Americans for Democratic Action (ADA), the National Associations of Businessmen (NAB) and the Americans for Constitutional Actions (ACT). However, COPE was selected since it explicitly rates Congressmen and Senators according to their pro-union or anti-union stance. The other unique variable included in the model was the number of unfair labor practices (F) committed annually by employers. This variable was used as a proxy for employer sentiment toward union activities. The number of CA cases appropriately adjusted was included in the model as the measure of anti-union activities by employers. These two variables are more reflective of regional preferences than are North-South dummies. Moreover, both COPE and (F) values reflect union attitudes in individual states and thereby permit testing of whether SRTW laws have an influence on union activities independently of the political sentiment which may cause them to be enacted. Significant findings for SRTW laws in earlier studies might disappear if these variables were included in the models. In the future such specification errors need not occur.