Modelling Labor Market Access Among Vocational Education Graduates in Indonesia: A Heckman Selection Approach

Abstract

Understanding the labour market outcomes of vocational education graduates is essential for evaluating the effectiveness of mid-level tertiary education. This study uses the Heckman two-step selection model to estimate the returns on vocational education for graduates of Diploma 4 (D1-D4) programs using nationally representative data from Indonesia’s 2023 National Labor Force Survey (SAKERNAS). In the first stage, a probit model estimates the probability of labour market participation as a function of educational attainment, demographic characteristics, household factors, and urban residence. The second stage estimates the determinants of wages, conditional on employment, with the Inverse Mills Ratio (IMR) from the first stage included to correct for selection bias. Results indicate that higher vocational qualifications significantly increase both the likelihood of employment and monthly wages. However, female graduates and rural residents face lower employment probabilities and earnings. This study contributes methodologically by demonstrating how selection-corrected statistical modelling can improve the accuracy of returns to education estimates, particularly for vocational education. Substantively, the findings offer policy-relevant insights into the evolving role of vocational education in promoting economic inclusion in Indonesia's post-secondary education and labour market systems.