Author(s)  
Eric Brunner
Shaun Dougherty
Stephen L. Ross

We examine the effect of attending stand-alone technical high schools on the industry of employment choices and within industry earnings premiums of young adults using a regression discontinuity design. Our analysis is based on the universe of students that applied to the Connecticut Technical Education and Career System (CTECS) between 2006 and 2011. Admission to a CTECS school shifts male applicants towards working in higher paying industries that align with CTECS course work and programs of study. Admission to a CTECS school has a much more modest impact on the industry of employment for female applicants. Further, key industry effects observed for females shift these applicants towards lower paying industries. Surprisingly, both the overall industry earnings premiums and the treatment effects of CTECS on earnings premiums are similar and sometimes larger for female applicants in traditionally male dominated industries like manufacturing and construction. However, the number of females in these industries is too small to contribute substantially to female earnings in aggregate. Our mechanism analysis suggests that treatment effects on industry specific earnings premium vary across industries. In particular, for male applicants, treatment effects in the manufacturing and construction industry depend in part on work experience while in high school and as a young adult. Finally, we find that in the professional and office support industries CTECS treatment effects on earnings arise due to the selection into these industries of students with high 8th grade tests scores because these industries offer a higher direct return to cognitive skills for young adults.

Publication Type  
Working Paper
File Description  
First version, August 22, 2022
JEL Codes  
I25: Education and Economic Development
I26: Returns to Education
I24: Education and Inequality
J30: Wages, Compensation, and Labor Costs: General
Keywords  
technical high school
earnings premiums
young adults
regression discontinuity design