The Impact of the Graduate Education Initiative
on Attrition and Completion
The sensible thing financially and emotionally
would have been to get through the program quick-
ly, but the scholar in me was all too prone to for-
get about the issue of time.
—Student in art history who entered
in 1991 and completed PhD in 2001
THE PREVIOUS CHAPTER includes information on how attrition and completion rates changed in departments that took part in the Graduate Education Initiative (GEI) between the period prior to the start of the GEI and the period during which the GEI was in effect. Although such analyses are suggestive of the impact of the GEI, they ignore the possibility that the observed changes may have been caused by factors other than the GEI. These include changes in the characteristics of the entering students (such as test scores, race and ethnicity, gender, marital status, and citizenship status) and changes in labor market conditions. To more accurately estimate whether the GEI did in fact influence attrition and completion probabilities, one must more formally model the processes by which students leave their programs, continue in their programs, or graduate from them, and exploit the data that were collected as part of the GEI about the students enrolled in treatment and control departments.
This chapter provides estimates of the impact of the GEI on attrition rates, completion rates, and time-to-degree (TTD).1 It does this for the entire period of students' doctoral study and separately for the periods prior to and after admission to candidacy. We also compute how the GEI influenced the cost, in terms of student years, per doctoral degree granted. Our analyses are based on the systematic data on student progress collected annually from both the departments that participated in the GEI
1 Some of the material presented in this chapter is a nontechnical summary and exten-
sion of material presented in Jeffrey A. Groen, George H. Jakubson, Ronald G. Ehrenberg,
Scott Condie, and Albert Y. Liu, “Program Design and Student Outcomes in Graduate Ed-
ucation,” Economics of Education Review 27 (April 2008): 111–24.