E N D
Abstract Method References The combined effect of two equity- and merit-based discretionary salary adjustments on gender differences in faculty salaries (n=517) was examined at a large mid-western university. The results of this study suggest that conscious attention to unexplained gender gaps in salary will be required if parity is to be achieved. • Participants • This study used publicly available salary data for all ranked faculty members employed by the university in both fiscal years 2006 and 2007 (FY06 & FY07). • Using SPSS, average salaries for female and male faculty members were computed for three time periods: FY06, mid-year after the first adjustment, and FY07 (after the second discretionary adjustment but with increases due to promotions removed). • As shown in the following figure, the final sample included 323 males (62%) and 194 females (38%) from the university’s seven colleges. This presentation is based on a thesis completed at Missouri State University by Meghan Lowery. For a full reference list, please contact mrlowery@siu.edu or jeannephelps@missouristate.edu Discussion Introduction The American Association of University Professors (AAUP, 2000) has reported that female faculty in public institutions typically earn 94.5 % of men’s salaries, with six to eight percent of this gap not explained by other variables expected to impact salary. Universities are expected to take proactive steps to correct pay inequities; however, little is known about how effectively decision-makers take gender into account when making salary decisions in the absence of specific legal pressures or clear directives to focus on gender. Consistent with the AAUP findings nationwide,unexplained gender differences in faculty salaries at this large mid-western university had been shown in two separate regression-based studies (Fischer, 1985; Fischer, 2003); however, these studies, which were not university-sponsored, cannot be considered as conclusively demonstrating discriminatory inequities and the findings were not widely disseminated. Whether discriminatory or not, the overall pay gap persists. The university’s transition in 2006 to a new compensation system provided a unique opportunity to study the role gender plays in salary decisions. Decision-makers (four female deans and three male deans) were provided with two salary pools from which to grant adjustments to faculty salaries. Deans were given the directive to distribute the first pool primarily on the basis of equity and the second pool, which was distributed six months later, on the basis of both equity and merit, as determined at the discretion of each dean. We examined whether these two unusual and substantial discretionary salary adjustments, given their general focus on equity, would narrow the long-standing salary gap between male and female faculty members. Results University-level Analysis Overall, female faculty at the university earned 89% of male faculty’s salaries. To examine the impact of gender and discretionary salary adjustments at the university level, a 2x3 mixed-model analysis of variance (ANOVA) was conducted. FY07 salary (adjusted to remove increases due to promotion) was the DV; gender was the between-subjects IV and discretionary salary adjustments was entered as a repeated measures within-subjects DV. Results are reported in the following table: College-level Analyses A series of seven identical 2x3 mixed-model ANOVAs were conducted to examine the impact of discretionary salary adjustments at the college level. There were significant main effects for salary adjustments in all colleges; however, only colleges #1, 4, and 5 revealed significant main effects for gender. As with the university-level analysis, there were no significant interactions between salary adjustments and gender in any of the seven colleges. Additional Analyses A 2 (gender) x 3 (salary adj.) x 3 (rank) mixed design ANOVA with repeated measures was conducted to test for significant gender differences in salary adjustments between ranks. There was a main effect for rank, F(2, 488) = 74.623, p = .001, and a main effect for gender, F(1, 488) = 8.036, p = .005. There was no interaction between rank and gender, F(2, 488) = .975, NS. The interaction between salary adjustments and rank was significant, F(4, 976) = 14.771, p = .001. There were no significant interactions between salary adjustments and gender, F(2, 976) = .051, NS, or salary adjustments, rank, and gender, F(4, 976) = .519, NS. Summary Table of Salary Differences by Gender Across Three Time Periods The following table shows average male and female salaries for the three time periods of interest: (1) FY06, prior to the two discretionary salary adjustments; and (2) at mid-year after the first adjustment, which was based primarily on equity considerations; and (3) FY07, which incorporated the second discretionary adjustment. The FY07 salaries used in these analyses were adjusted by removing all salary increases due to promotions, as promotions were not the focus of this study and increases due to promotions obscured the effects of the discretionary adjustments. • Results at the university level are consistent with prior research regarding salary differences between male and female faculty members. Men’s salaries were higher on average than women’s salaries across all three time periods in this study. It is important to note that some (but not all) of the salary gap can be accounted for by variables other than gender known to be related to salary. • The failure to find significant interactions between gender and salary adjustments suggests that conscious attention to gender will be required to make progress toward equitable pay. The overall results clearly show that no progress in narrowing the gender gap in salaries was achieved despite two substantial salary adjustments intended to target, in a general way, salary inequities. • There was intriguing evidence at the college level that some decision makers did pay attention to gender. The discretionary adjustments narrowed the existing gender gap in four colleges and actually widened the gap in three; however, lack of power to detect statistical significance hampers interpretation of the college-level analyses. Our findings do suggest that overall results at the university level may obscure important differences among individual colleges. • Rank influenced the amount of discretionary salary adjustments. The salaries of full professors (more of whom are men) rose more than the salaries of assistant and associate professors. • A portion of the discretionary salary pool was distributed on the basis of percentage increases, which results in the rich getting richer. Dollar-amount discretionary adjustments would be a more effective way to address inequities.