Association for Public Policy Analysis & Management

No matter how you measure them, gaps in college enrollment (either by race or SES) are large and do not show signs of diminishing (e.g., Berkner & Chavez, 1997; Kane, 1994, 2004; Bozick & Lauff, 2007). Even within the four-year sector traditionally privileged students more likely to enroll in more selective four-year colleges than underrepresented minorities and low-SES students (Alon & Tienda 2007; Posselt et al., 2012; Reardon, Baker, & Klasik 2012). Because gaps in college enrollment are larger than academic achievement gaps (Reardon, Baker, & Klasik, 2012), part of the explanation for these disparities may lie in other dimensions of college-readiness—gaps in activities that students should undertake to be well-prepared for college. For example, racial differences in the completion of steps to college enrollment (e.g. taking the SAT or ACT, submitting a college application) grow as students move closer to enrollment (Klasik 2012). These gaps cannot be fully explained by academic achievement or sources of social, cultural, and economic capital, but are strongly related to four-year college enrollment.

We ask three main questions: (1) Are race- and SES-based gaps in college enrollment related to race- and SES-based gaps in the completion of steps to college enrollment? (2) Are certain gaps in the completion of some steps more strongly related to gaps in college enrollment than others? And (3) what college-readiness policies are in place that might affect state gaps in step completion and enrollment?

We use data from the High School Longitudinal Study, a nationally-representative sample of ninth-graders surveyed in 2009, with follow-ups in 11th grade and several months after on-time high school graduation. This data is also representative of 10 states, a feature we use to examine how gaps vary between states with different policy contexts. We examine steps similar to Klasik (2012) and Avery and Kane (2004), given what is available in our data. Although gap computation is easy when the completion of a step is binary (e.g. difference in percentage of White and Black students applying to college), it is more complicated when there are multiple, ordered ways of completing a step (e.g., level of college application, from no college to a very selective four-year college) In these cases we compute gaps using a method designed to estimate achievement gaps when test results are reported in ordinal proficiency categories (Ho & Reardon 2011.). This method can be used to measure gaps in other ordered, non-continuous phenomena. We describe gaps in step completion both nationally and within the representative states and perform simple correlation and regression analyses to look overall at the relationships between gaps in step completion and gaps in college enrollment selectivity.

Preliminary results indicate that racial and SES gaps in step completion are strongly, but quite variably, related to racial and SES gaps in college enrollment. The findings from this study highlight the activities that school leaders and policymakers ought to focus their attention on if they are interested in improving college readiness and increasing access to higher education for traditionally underrepresented groups.