Using a High School Earnings Benchmark to Measure College Student Success
Implications for Accountability and Equity
In this report, we examine the design, performance, and implications of a high school earnings metric for accountability in higher education. This type of metric compares the earnings of former college students to the earnings of individuals with only a high school diploma. A version of this metric was recently proposed to assess student outcomes for short-term vocational programs that seek to access federal student aid. One advantage of such a metric is the potential for a clear, simple, and intuitive threshold against which to measure the gains from postsecondary education, but there has been little research about the optimal design and implications of such a measure.
Earnings measures are central to conversations about accountability in higher education—particularly for short-term and vocational programs—and it is imperative that these measures accurately reflect student outcomes. Appropriate accountability metrics must be in place to ensure that students obtain a sufficient return on investment and are not taken advantage of by predatory institutions. At the same time, these metrics must ensure that institutions are not punished simply because of the demographics of the students they serve.
This paper generates new evidence to inform debates over the use of earnings for higher education accountability and equity, with the hope of aiding the design of future measures. We first describe the theory motivating a high school earnings metrics and discuss the challenges and choices involved in its design. We then develop two simple potential high school metrics based on Census data. The lower of our two benchmarks, at $25,000, is roughly two times the poverty line for an individual and aligns with the benchmark used by the Department of Education (DoED) in the College Scorecard in 2013–14. Today, the $25,000 benchmark is close to the earnings of young adults age 25–34 who did not complete high school.
We apply this benchmark to the most recent release of program-level earnings data for undergraduate programs in the DoEd’s Gainful Employment data.2 Of 8,144 programs, just over half fail our low high school benchmark, with most failures accounted for by certificate programs (83%) and programs in for-profit colleges (76%). GE data on associate’s and bachelor’s degrees is limited to programs in for-profit colleges, but bachelor’s degree programs fare well, with just 11% failing the low benchmark.
Student demographics do not appear to be driving failure of programs on a high school metric. Conditioning on institutional characteristics and demographics, our regression analyses show that programs in HBCUs are not disproportionately likely to fail high school earnings metrics and there are only negligible associations between the percentage of students of color and failure on high school metrics. Although not causal, our analyses suggest that sector, level, and institutional size are stronger predictors of failure than demographics, with for-profits, less-than-two-year, and larger institutions more likely to have programs that fail.
We suggest that future research assess differential effects by field of study, examine the distribution of students across failing programs, consider the outcomes of more recent cohorts of students, and account for programs that have may closed since the release of the GE data we use here. However, our findings suggest that adding a national or state-specific high school benchmark to debt-to-earnings metrics under Gainful Employment could potentially generate needed accountability for institutions and ensure value to students, without penalizing institutions for the students they serve
