In this paper I investigate whether students self-select into the US for-profit colleges or whether the choice of for-profit sector is accidental or due to the reasons external to the students (geographic exposure to for-profit providers, tuition pricing, or random circumstances). The main student-level data samples come from the National Education Longitudinal Study of 1988 (NELS:88) and the associated Postsecondary Education Transcript Study (PETS:2000). I estimate a multinomial logit of college choice where student's choice set is defined across four alternatives: no college, a for-profit college, a non-profit 2-year (or less-than-2-year) college, and a non-selective non-profit 4-year college. I find that students self-select into for-profit sector. Three groups of significant factors stand out. First, choice of for-profit sector is characterized by lower parental involvement in student's schooling. Second, ceteris paribus, for-profit-bound students are more likely to display high levels of school absenteeism and to give birth as early as 10th grade. Third, the average predicted probabilities of choosing for-profit sector increase as in-state public community college tuition rises and county-specific concentration of for-profit providers grows larger.
