TY - JOUR
T1 - Lifting up the lives of extremely disadvantaged youth
T2 - The role of staying in school longer
AU - Moschion, Julie
AU - van Ours, Jan C.
N1 - Publisher Copyright: © 2025
PY - 2025/10
Y1 - 2025/10
N2 - Using a sample of Australians who display high rates of early school-leaving, we compare the trajectories of respondents who left school at each incremental age between 14 and 17 with respondents who left at 18 years old or more, in terms of homelessness, incarceration, substance use and mental health issues. Leveraging recent methodological advances, we estimate a staggered difference-in-difference aiming to eliminate biases arising from reverse causality or unobserved time-invariant unobserved heterogeneity and account for heterogenous treatment effects across cohorts and time. Our results suggest that leaving school before age 18 increases males’ likelihood of experiencing homelessness, being incarcerated, using cannabis daily and illegal street drugs weekly several years after school-leaving. In contrast, for females the difference-in-difference strategy fully eliminates the correlations between their school-leaving age and outcomes suggesting that remaining biases would likely be gender-specific. To minimise concerns that gender-specific time-varying unobserved heterogeneity may be driving our results, we also show that our results are robust to controlling for the lags of all other outcomes. Finally, we find that while the occurrence and the timing of parental separation coincide with early school-leaving, our preferred specification also eliminates this correlation. Taken together, our findings suggest that preventing early school-leaving can help disadvantaged youth break cycles of multi-dimensional disadvantage.
AB - Using a sample of Australians who display high rates of early school-leaving, we compare the trajectories of respondents who left school at each incremental age between 14 and 17 with respondents who left at 18 years old or more, in terms of homelessness, incarceration, substance use and mental health issues. Leveraging recent methodological advances, we estimate a staggered difference-in-difference aiming to eliminate biases arising from reverse causality or unobserved time-invariant unobserved heterogeneity and account for heterogenous treatment effects across cohorts and time. Our results suggest that leaving school before age 18 increases males’ likelihood of experiencing homelessness, being incarcerated, using cannabis daily and illegal street drugs weekly several years after school-leaving. In contrast, for females the difference-in-difference strategy fully eliminates the correlations between their school-leaving age and outcomes suggesting that remaining biases would likely be gender-specific. To minimise concerns that gender-specific time-varying unobserved heterogeneity may be driving our results, we also show that our results are robust to controlling for the lags of all other outcomes. Finally, we find that while the occurrence and the timing of parental separation coincide with early school-leaving, our preferred specification also eliminates this correlation. Taken together, our findings suggest that preventing early school-leaving can help disadvantaged youth break cycles of multi-dimensional disadvantage.
UR - https://www.scopus.com/pages/publications/105013879066
U2 - 10.1016/j.jebo.2025.107205
DO - 10.1016/j.jebo.2025.107205
M3 - Article
AN - SCOPUS:105013879066
SN - 0167-2681
VL - 238
JO - Journal of Economic Behavior and Organization
JF - Journal of Economic Behavior and Organization
M1 - 107205
ER -