Evidence Thin on Success of Federal College Prep Programs, Including GEAR UP

The failure of South Dakota’s GEAR UP program to produce evidence of success in getting more kids to college is not unique. Angela Kennecke finds a spring 2013 policy brief by Ron Haskins and Cecilia Elena Rouse of the Brookings Institution, who contend that evidence for success of the federal government’s efforts to help minority and low-income kids enter college is fuzzy at best:

…[T]hrough its TRIO programs (Upward Bound, Talent Search, Upward Bound Math-Science, Student Support Services, and a few smaller programs), the federal government is a dominant presence among sponsors of programs that aim to prepare disadvantaged students for college. The TRIO programs are designed to augment disadvantaged students’ academic preparation, give them direct experience with college work, or help them apply to colleges or seek financial aid. Together, all the federal collegepreparation programs (including another major program, GEAR UP) cost more than $1 billion a year, a modest sum by federal program standards but surely the largest investment in college-preparation programs from any single source.

…Mathematica reported that, on average, Upward Bound had “no detectable effect” on whether students enrolled in college, the type or selectivity of the institution they attended, or the likelihood that they would apply for or receive financial aid. For some subgroups of students, however, there were significant effects on enrollment or completion. For example, Upward Bound students who did not expect to complete a four-year degree when interviewed at baseline, usually in their middle school years, enrolled in college and finished their degrees more often than did similar students who did not participate in Upward Bound. The Institute of Education Sciences (IES)—the research arm of the U.S. Department of Education—gave the Mathematica evaluation its highest rating, “meets evidence standards without reservations.”

…The group selected from the UBMS program exceeded the matched comparison group on nearly all the important outcomes of college-preparation programs. Specifically, UBMS students were more likely to enroll in four-year schools, more likely to enroll in more selective schools, more likely to enroll in math and science courses, and more likely to graduate from four-year schools (although not with majors in math or physical science). However, the IES What Works Clearinghouse examined the evaluation and concluded that it “does not meet evidence standards.”

…Talent Search students in Florida, Indiana, and Texas were 68 percent, 33 percent, and 90 percent more likely, respectively, to become firsttime applicants for financial aid; they were 38 percent, 13 percent, and 55 percent more likely to enroll in a public college or university. In addition, Talent Search students in Florida and Texas, though not in Indiana, were more likely to complete high school. The IES What Works Clearinghouse concluded that the evaluation met standards, but “with reservations.”

…Perhaps the best study [of GEAR UP] was the one by Jennifer Bausmith of the College Board and Megan France of Santa Clara University. They selected 173 GEAR UP schools and an identical number of matched non-GEAR UP schools and compiled seven years of cohort data for all of them. They compared the improvement in scores on three measures of college readiness (the Preliminary Scholastic Aptitude Test, the Scholastic Aptitude Test, and Advanced Placement) for the two groups of schools. The results were not consistent across tests or cohorts. Further, several problems in the research design make it hard for us to interpret the results, including the fact that the data are based on all students in a given school and not just disadvantaged students [Ron Haskins and Cecilia Elena Rouse, “Time for Change: A New Federal Strategy to Prepare Disadvantaged Students for College,” Princeton–Brookings Institution: The Future of Children, Spring 2013].

Haskins and Rouse don’t mention scholarships; they advocate consolidating Uncle Sam’s various college prep programs into one grant program that requires applicants to demonstrate that they are using evidence-based interventions to improve student success. They advocate using some maximum percentage of these federal dollars (they suggest 20%) for student support services at college and another 2% for coordinated research and evaluation of the education programs using federal college-prep money.

Bringing the discussion back to GEAR UP, Kennecke notes that Iowa uses half of its GEAR UP money for scholarships. South Dakota’s District 21 legislators representing Platte, where our GEAR UP scandal erupted, proposed using 50% to 75% of GEAR UP money for scholarships for Native American students. House Democrats unanimously supported the plan, but only eight House Republicans were willing to join them in backing House Bill 1220.

Of course, scholarships alone aren’t guaranteed to increase college enrollment and success, but in general, if college costs less, more students sign up for and complete higher learning.