[Excerpt The year 2002 marked the 30th anniversary of the passage of Title IX, which prohibits discrimination by gender in any federally funded educational activity. Although the scope of Title IX includes all aspects of education, the application of Title IX to college athletics has been especially complicated because athletics programs, unlike most academic classes, usually are sex-segregated by sport. As explained in more detail below, Title IX essentially requires that all institutes of higher education provide student access to sport participation on a gender-neutral basis. As a result, athletic opportunities for female undergraduates have expanded significantly since 1972. For example, the female share of college athletes rose to 42% in 2001/02 from only 15% in 1972 (U.S. Department of Education, 1997, 2003). Despite this progress, gender equity is far from complete. Estimates from our data show that at the average institution in 2001/02, women comprised 55% of all students but only 42% of the varsity athletes.
Our research describes the level of noncompliance with Title IX, as measured by the proportionality gap, between 1995/96 and 2001/02, and then investigates why some institutions perform better than others do on this measure of gender equity. One important contribution of this article is the introduction of a new data set developed by the authors that includes information on athletic offerings and other institutional characteristics for the 1995/96 and 2001/02 academic years. Our data represent a substantial improvement over previous data because we include institutions in Divisions I, II, and III and adjust for changes in how institutions report athletic participation over the period; previous research focused solely on Division I institutions and did not adjust for reporting differences. We show that these data differences are important: Reliance on unadjusted data from Division I institutions results in large overestimates of the improvement in compliance at NCAA institutions during the late 1990s. Our data also include a rich set of explanatory variables that we use in regression analyses to explain the extent of institutional noncompliance. We examine the determinants of the proportionality gap by estimating OLS cross-section regressions (with and without conference fixed effects) at two points in time (1995/96 and 2001/02) and first-difference regressions for changes over the period.