Working Paper: NBER ID: w14367
Authors: Bridget Terry Long; Michal Kurlaender
Abstract: Community colleges have become an important entryway for students intending to complete a baccalaureate degree. However, many question the viability of the transfer function and wonder whether students suffer a penalty for starting at a two-year institution. This paper examines how the outcomes of community college entrants compare to similar students who initially entered four-year institutions within the Ohio public higher education system. Using a detailed dataset, we track outcomes for nine years and employ multiple strategies to deal with selection issues: propensity score matching and instrumental variables. The results suggest that straightforward estimates are significantly biased, but even after accounting for selection, students who initially begin at a community college were 14.5 percent less likely to complete a bachelor's degree within nine years.
Keywords: community colleges; baccalaureate degree; transfer function; educational attainment
JEL Codes: C1; I2; J24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Selection bias (C24) | Straightforward estimates of outcomes (C51) |
| Observable differences (C90) | Baccalaureate degree completion (Y40) |
| Unobservable differences (J79) | Baccalaureate degree completion (Y40) |
| Starting at a community college (I23) | Baccalaureate degree completion (Y40) |