Association for Public Policy Analysis & Management

With the reauthorization of the Carl D. Perkins Career and Technical Education Act of 2006, the federal government changed the name of vocational education to career and technical education (CTE). Earlier, after 73 years of federal support for the complete curricular separation of vocational and academically-focused students, the Perkins Act of 1990 funded an integrated vocational, academic system for the first time. These acts followed two major policy trends in secondary vocational education: integrating vocational education with academics and the transition from the term “vocational” to “CTE.” Taking into account these major policy changes, have the outcomes of those students who participate in vocational education changed over time?

I utilize the National Education Longitudinal Studies data sets that follow students from the time they are enrolled in high school into adulthood and include nationally representative samples from the high school classes of 1972, 1980, 1992, and 2004.  The outcomes of interest are from follow-ups of the high school students six to eight years later including whether or not the student is working full time and a measure of income that is standardized by year. To track trends in the returns to vocational education over time, models are estimated using school-year fixed effects and Blinder-Oaxaca decomposition.

The school-year fixed effects models estimate differences in outcomes between students at the same school within a single survey year. I find no results for the classes of 1972 and 2004. Vocational students in the class of 1980 are predicted to have higher incomes than non-vocational students from their high school. Vocational students in the class of 1992 are slightly more likely to be working full time than non-vocational students at the same school but are predicted to have slightly lower income. 

The Blinder-Oaxaca decomposition models differ from the school fixed effects models because these estimates include between school variation and investigate changes over time. I find no significant changes in probability of working full time for pay for vocational students that are not mirrored by changes in the overall population of students. For income, vocational students in the class of 1980 are predicted to have higher income than in the class of 1972. Almost 100% of this change is not explained by changes in the covariates. However, income for vocational students in the class of 1992 is predicted to be lower than the vocational students in the class of 1980, reversing the increase between 1972 and 1980.

From both of these estimation techniques, I find vocational students from the class of 1980 have slightly better outcomes and vocational students from the class of 1992 have slightly worse outcomes than non-vocational students. This same pattern is seen when comparing vocational students both between and within schools, and these changes are not explained by changes in student or school level covariates. Thus far, I find no evidence that recent changes in federal policy have resulted in improved outcomes for vocational students.