In order to accomplish President Obama’s American Graduation Initiative, which has a goal that “by 2020, this nation will once again have the highest proportion of college graduates in the world,” the United States must focus on both entry into and completion of postsecondary education.  The U.S. has made moderate progress in increasing college entry over the past four decades; the number of students enrolled in postsecondary institutions has more than doubled to 21 million students.[1]  Where the U.S. is lagging behind is in college completions.   The percentage of full-time students at four-year institutions who complete a bachelor’s degree in four years is only 37.9%, and the completion rate after six years is only 58.3%.[2]  There is also large variation in the completion rate by race and gender. 

Failing to complete a degree may negatively affect student outcomes.  Students loose years of workforce experience and may graduate with student loan debt, yet have not earned a wage-increasing credential to justify these sacrifices.  Students make their decision to enter college based on information in high school and their a priori predicted probability of completion.  Because of this uncertainty, having some students drop out of college is inevitable.  Some students may be on the margin of whether college entrance is a beneficial decision, and need to experience one year of college to obtain more information.  Other students, for whom there are large returns to college, may experience unforeseen shocks during college that cause them to drop out.  While some level of college failure is expected, the current high rates are likely harmful to students and the economy.

The goal of this paper is to educate policymakers on the types of students who drop out, and to identify correlates to college failure.    Are college dropouts observationally more similar to college completers or students who never enter college?  The answer to this question can inform administrators working to increase college completion rates.  Do we predict success for four-year college dropouts had they attended two-year schools?  Is there high school information such as prior achievement and course taking, that strongly predicts a failure to complete college?  Our analysis begins using the NLSY97, a longitudinal study of 9,000 youths who were 12-16 years old when the study began in 1997.  The richly detailed survey allows us to control for a detailed set of student characteristics and prior academic achievement in our analysis.  This paper also examines the NLSY79, a similar longitudinal data set that surveyed youth who entered college approximately 20 years before the NLSY97 students.  Using this second data set allows for comparisons of both the types of students who fail to complete college, as well as the predictors of college failure, over time.  While this is only a descriptive analysis, until more is understood about the types of students who drop out of college, there will likely be little progress in reducing the college failure rate in the U.S.