Major Matters: A Comparison of the Within‐Major Gender Pay Gap across College Majors for Early‐Career Graduates

Abstract

I use data from the 1993 National Survey of College Graduates and appended 1990 Census on about 11,000 men and women college graduates (8400 with bachelor's degrees only, and 2800 with graduate degrees) who earned degrees in a 5‐year period (1984–1988), to address questions regarding the link between college major and early‐career gender pay differentials. I look at within‐major gender pay differentials for two groups of college graduates: those whose highest degrees are bachelor's and those who hold graduate degrees. Among those whose highest degrees are bachelor's, I find that within‐major gender pay penalties are virtually zero for professional majors. Gender pay penalties are large for general studies majors: social sciences, history, and humanities, and business administration (except accounting). For these, jobs account for a large portion of the unexplained gap. Among individuals who hold graduate degrees, I find that within‐major gender pay penalties are zero for all fields. My findings suggest that pay penalties to women in the aggregate can be traced to relatively large penalties in a couple of key fields (bachelor's degree–level general studies fields): social sciences and humanities, and business administration (except accounting). These findings are important, as they contradict the prevailing view. Existing empirical research suggests that women are uniformly disadvantaged compared to men regardless of field of study and, thus, field of study explains little or none of the gender gap in pay. I conclude that supply‐side mechanisms are important factors in producing gender inequality among college graduates, and suggest that we further focus our attention on the “supply side” in understanding segregation and gender pay inequality.

Discipline and Gender: Factors Affecting Graduates’ Salaries

“…There is no consensus in the research literature on the subject of gender effect, and some claim that the wage premium for schooling is higher for women than for men (Davidovitch et al, 2013), while others claim the opposite (Black et al, 2005;Morgan, 2008;Taniguchi, 2005;Thomas & ZHANG, 2005;Wood et al, 1993), and yet others contend that there is currently no significant difference between genders (Hubbard, 2011).…”

“…In other disciplines (social sciences, history, business management, and the humanities), women's salaries were found to be lower than men's. The research claims that these findings show that women are given jobs with lower salaries than men with the same level of skills (Morgan, 2008).…”

“…She claimed that these differences may stem from the fact that women in general tend to gain less from their higher education with the passage of time (Taniguchi, 2005). Morgan (2008) explored the effect of discipline and gender on wage premium. Findings showed that women, at least in the first year of their career, earn the equivalent of men in professional disciplines such as engineering and the exact sciences.…”

“…In the current study, we sought to examine whether and to what degree baccalaureate degrees improve graduates' financial situation. In cooperation with Israel's Central Bureau of Statistics we gathered data on 6,091 graduates who completed their studies at the Ariel University Center during -2008 show that graduates earn almost three times the national average salary (and five times the median salary) and indicate employment rates of nearly 100%. The most profitable fields are computer sciences and mathematics, engineering, and architecture.…”

Preferences, constraints, and the process of sex segregation in college majors: A choice analysis

“…Because women are not disadvantaged everywhere (Morgan, 2008) and perceived discrimination is higher in male-dominated disciplines (Steele et al, 2002), anticipated unfair treatment could potentially contribute to the reproduction of horizontal sex segregation in college. Almost unanimously, however, references to this mechanism have been accompanied by cautionary warnings against the difficulty of testing it empirically (Charles and Grusky, 2004: 14;Mann andDiPrete, 2013: 1532; but see Xie and Shauman, 1997).…”