Association for Public Policy Analysis & Management

We use state-level datasets on community college students from three state systems to examine the earnings gains associated with stackable credentials. Each dataset includes detailed transcript information as well as students’ quarterly earnings pre-, during-, and post-college. Our analyses will address three broad research questions:

(1) To what extent do students earn stackable credentials, and what are the typical patterns of receipt? For example, what proportion of students earn more than one sub-baccalaureate award, and of those, what proportion earn in them in-order and within the same general field? How much time elapses between receipt of each credential? In which fields (e.g., health, manufacturing, protective services) are students most likely to earn stacked credentials? Are stacked credentials more popular at some types of community colleges (e.g., large colleges, urban colleges, those with low-SES service areas) than others?

(2) How do students who earn stacked credentials differ from those who earn only one award? In particular, how do students who earn multiple in-order awards within the same general field differ from students who earn only the highest of those awards in that field? Do they differ in age, ethnicity, or prior earnings and employment profiles?

(3) Do stacked credentials have a unique value in the labor market? For example, do students with multiple in-order awards within the same general field have higher earnings, or maintain more consistent employment across their college career, compared to students who earn only the highest of those awards in that field? Do students with multiple in-order awards receive an earnings bump with each award? Do students with multiple lower-level awards (e.g., multiple short certificates) have earnings similar to or higher than students with a single higher-level award which requires a similar number of cumulative course credits (e.g., an associate degree)?

We apply a series of statistical techniques based on matching to estimate the associations described above.