Search
Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Session Type
Personal Schedule
Sign In
Access for All
Exhibit Hall
Hotels
WiFi
Search Tips
Existing research often treats gender inequality in STEM as a single decision or privileges either cultural explanations (socialization, stereotypes, interests) or economic explanations (returns to education and work conditions), leaving the timing of mechanisms along the educational trajectory insufficiently specified. This study investigates the persistence of gender inequality in STEM by proposing a time-stratified framework that integrates theories of gendered socialization and instrumental rationality. Focusing on the German context, characterized by high STEM wage premiums yet traditional gender norms, we document a stage-specific pattern in which the determinants of STEM participation shift from cultural to structural forces over the educational trajectory.
We link longitudinal individual data from the National Educational Panel Study with regional labor-market signals (STEM wage premiums and STEM-nonSTEM working-hour gaps) derived from the German-Microcensus. We model entry into STEM and later switching-out as two stages, showing a “gendered entry, convergent exit” pattern. At the entry stage, the probit regression and Fairlie decomposition method show that the decision of STEM-choice is primarily driven by gendered interests (RIASEC). Objective labor-market signals are not significantly associated with students’ STEM entry, and subjective high-wage expectations do not encourage STEM entry. However, in the second stage, among those enrolled in STEM, the selection-corrected model shows that higher wage STEM premiums are associated with lower switching-out, whereas larger STEM-nonSTEM working-hours gaps are associated with higher switching-out. Most crucially, the responsiveness to these labor-market signals does not differ significantly by gender.
The findings challenge essentialist accounts suggesting women are less instrumentally oriented or inherently prefer less time-intensive work. Instead, our study’s time-stratified approach clarifies when and how cultural versus structural mechanisms operate, helping reconcile these two accounts by demonstrating a “structural convergence”. This implies that gender equality in STEM requires addressing not only cultural pipelines but also structural conditions of STEM and non-STEM work.