Working Paper: CEPR ID: DP17138
Authors: Barbara Biasi; Song Ma
Abstract: This paper documents differences across higher education courses in the coverage of frontier knowledge. Comparing the text of 1.7M syllabi and 20M academic articles, we construct the "education-innovation gap," a syllabus’s relative proximity to old and new knowledge. We show that courses differ greatly in the extent to which they cover frontier knowledge. More selective and better funded schools, and those enrolling socio-economically advantaged students, teach more frontier knowledge. Instructors play a big role in shaping course content; research-active instructors teach more frontier knowledge. Students from schools teaching more frontier knowledge are more likely to complete aPhD, produce more patents, and earn more after graduation.
Keywords: education; innovation; syllabi; instructors; text analysis; inequality
JEL Codes: I23; I24; I26; J24; O33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| education-innovation gap (I24) | graduation rates (I23) |
| education-innovation gap (I24) | income (E25) |
| education-innovation gap (I24) | measures of innovation (O35) |
| lower education-innovation gap (I24) | completion of PhD (Y40) |
| lower education-innovation gap (I24) | production of patents (O34) |
| lower education-innovation gap (I24) | higher incomes after graduation (D29) |
| research-active instructors (A20) | coverage of frontier knowledge (Z00) |