Is Traditional College Aid Too Little, Too Late? Impacts on High School Outcomes From a Cluster Randomized Trial of an Early College Scholarship for Low-Income Students

• In Event: 60.058 - Promise Scholarship Programs in High-Poverty Communities: What Is the Impact of Promise Scholarship Programs on College Access in Kalamazoo, Michigan, and Milwaukee, Wisconsin?

Abstract

Objective
This paper will examine the impacts of The Degree Project on academic, behavioral, and psychological outcomes. Estimates are based on comparisons of the 18 treatment schools with 18 control schools.

Perspectives
Promise programs seek to increase college access by assuring students early in high school that they will have the financial resources to pay for college (Dynarski, 2000; Cornwell, Mustard, Sridhar, 2006; Fowler et al., 2009). College now seems unaffordable and this perception leads many low income students and students of color to inadequately prepare for the academic, social, and financial demands of college (Ikenberry & Hartle, 1998; McDonough, 2004; Noeth & Wimberly, 2002; Roderick, Nagaoka, & Allensworth, 2006). By increasing the real and perceived affordability of college, and clearly communicating the path to college, the theory is that these “early commitment” scholarship programs improve academic preparation and social capital during high school and thereby increase college entry and completion (Dynarski, 2000; Cornwell, Mustard, Sridhar, 2006; Fowler et al., 2009).

Methods
This is a cluster randomized trial. Estimates are covariate-adjusted using appropriate regression techniques (e.g., Hierarchical Linear Modeling) to increase power and estimate appropriate standard errors.

Data sources
We have administrative data from Milwaukee Public Schools including grades, test scores, attendance, and suspensions for nearly all of the 5,051 control and treatment students in the study. The analysis will also examine impacts on survey measures of students’ college expectations and concerns about college costs.

Results
In the first year, the program had no impact on grades, test scores, or attendance. Somewhat surprisingly, there is some evidence that the program increased the number of discipline incidents and this seems to be related to the ways in which the program was communicated to some students. On the other hand, we do find positive impacts on college expectations and reduced concern about college costs among recipients.

Significance
Only a few other studies have examined the impacts of promise scholarships with even quasi-experimental analyses and the results here are largely consistent with them. Studies of the well known Kalamazoo Promise also suggest that the program had a limited impact on most outcomes studied. However, a similar experiment in Canada found positive effects on college expectations, though did not examine academic outcomes.

Author

• Douglas N. Harris, Tulane University