The Interaction Effects of Gender, Race, and Marital Status on Faculty Salaries

Abstract

Listen

Translate article

The question of whether employees receive equal pay for equal work in American society can be traced back to the women's movement at the turn of the 20th century and the civil rights movement in the 1960s. The Equal Pay legislation and Affirmative Action policies that emerged from these movements led to significant interest in measuring the extent of pay inequities within labor markets. The vast majority of studies of the general labor market have examined three major forms of pay discrimination: by gender, race/ethnicity, and marital status. Many studies have documented that in the general labor market, women tend to be paid less than men with similar characteistics, Blacks and Hispanics tend to be paid less than Whites, and unmarried workers tend to earn less than their married counterparts (Antecol & Bedard, 2004; Duncan, 1996; Korenman & Neumark, 1992; Loh, 1996; Neal & Johnson, 1996; U.S. General Accounting Office, 2003; Verdugo, 1992; Weinberger, 1998). Those employed in academic labor markets have likewise been interested in whether unexplained wage gaps exist among faculty, and numerous studies have been conducted using national as well as institution-specific data. Unlike studies of the general labor market, studies in academia have primarily focused on gender. Studies by Barbezat (1991), Bellas (1993, 1994), Ransom and Megdal (1993), Toutkoushian and Conley (2005), and others have found that female faculty members earn less than male faculty members with comparable levels of measurable characteristics, such as experience, education, and research productivity. Studies of faculty salaries have paid much less attention to possible pay discrimination by race/ethnicity (Barbezat, 2002; Hearn, 1999). This omission is often attributed to the relatively small number of faculty members of color at most institutions, which may result in unreliable estimates of pay differentials. Similarly, pay disparities by marital status have not been widely studied in academia. Bellas (1992, 1993, 1994) and Toutkoushian (1998) are among the few researchers who have analyzed the effects of marital status on faculty salaries. Their findings show that, as in the general labor market, there is a positive return on marriage in academia (i.e., married faculty earn more than unmarried faculty), at least for men. The relative scarcity of studies in this area is due in large part to the lack of available information concerning marital status in institution-specific databases. When studies of academic salaries have considered gender, race, and marital status, little attention has been paid to possible interaction effects among personal characteristics or social categories. Various theories and conjectures have been offered in the literature to suggest that these interactions may be more complex than is typically assumed in empirical studies (Collins, Maldonado, Takagi, Thorne, Weber, & Winant, 1995; Epstein, 1973; West & Fenstermaker, 1995). If the advantages and disadvantages associated with membership in any one group are not uniform for those who differ along other dimensions, faculty salaries may also vary simultaneously by gender, race/ethnicity, and marital status (and perhaps other factors). However, salary studies in academia typically restrict pay differences by race/ethnicity and marital status to be uniform for an entire sample, or for all men and women if regression equations are estimated separately for the sexes. Studies that use separate regression models for men and women allow the effects of race/ethnicity and marital status to vary by gender, but these effects are restricted to be the same for women or men of the same race/ethnicity and marital status, when in fact any effects of race/ethnicity may depend on marital status and vice versa. Large national surveys of faculty afford analysts the opportunity to examine differences in faculty salary based on combinations of all three dimensions--gender, race/ethnicity, and marital status--as well as the possible interactive effects among them. …