While the federal government distributes nearly $150 billion in higher education student aid annually, regulations that hypothetically prevent low-quality institutions from benefiting from this aid rarely function as intended. Researchers have made progress in measuring the return on investment (ROI) of enrollment in higher education, but residual obstacles are preventing policymakers from using these measures to hold institutions accountable. This report outlines an improved methodology for estimating the ROI from college enrollment that may point the way toward better regulation of higher education institutions.

BPC’s models estimate the benefits to a student’s earnings from postsecondary education—a college earnings premium—for students at thousands of institutions in the United States. Each estimate is compared with the costs of attending that school. Adjustments are made to the models to more accurately assess the costs and benefits of enrollment; the public subsidy provided via local, state, and federal supports for institutions; and the negative effect of labor market discrimination.

This report finds that most institutions typically provide a positive ROI for their students, and it is noteworthy that the vast majority of students nationwide attend these institutions. Nonetheless, hundreds of institutions are estimated to provide a negative ROI to most of their students. Public institutions are the most likely to provide a positive estimated median ROI, followed by private nonprofit institutions and then private for-profit institutions, many of which are estimated to provide little value to the typical student. The largest estimated ROIs are found among prestigious private institutions as well as institutions with a focus on high-earning technical fields. The report concludes with policy recommendations to improve the availability of high-quality data for policymakers and prospective students, as well as to use ROI as a means of holding institutions accountable for their outcomes.

Read the full report here.