Rethinking the U-Shaped hypothesis: evidence of a convex education–participation relationship in Sri Lanka

Purpose This paper aims to examine the relationship between women’s education and labour force participation in Sri Lanka, an analytically distinctive South Asian case characterised by high female educational attainment but persistently low female labour force participation. It assesses whether the relationship conforms to the U-shaped hypothesis or instead reflects alternative labour market dynamics. Design/methodology/approach The analysis uses aggregated Labour Force Survey data for 2017–2024. Beta regression models account for the bounded nature of participation rates, with orthogonal polynomial terms capturing non-linear effects of education and year fixed effects controlling for temporal variation. Wild-cluster bootstrap procedures are used to obtain robust inference given the small number of education clusters. Findings The results provide evidence of a convex (J-shaped) relationship between education and participation, rather than the conventional U-shape. Participation increases modestly through lower levels of education, declines slightly at the Ordinary Level and rises sharply at the Advanced Level and tertiary education. This pattern reflects the concentration of employment opportunities among highly educated women and the limited absorption of moderately educated women into formal labour markets. As the analysis is based on aggregated education–year data, findings should be interpreted as descriptive of population-level patterns rather than causal effects. Research limitations/implications The analysis relies on aggregated survey data and cannot capture individual-level heterogeneity or causal mechanisms. Future research using panel data could examine how educational transitions influence individual employment trajectories. Practical implications Educational expansion alone is insufficient to increase women’s labour market participation across all groups. Policy efforts should prioritise expanding accessible employment opportunities for moderately educated women, including investment in vocational pathways, regional labour markets and sector-specific demand beyond export-oriented industries. Social implications Without targeted interventions, the concentration of labour market opportunities among highly educated women may reinforce existing gender and socioeconomic inequalities. Addressing this imbalance is essential to ensure equitable returns to education and support inclusive development. Originality/value This study contributes to the literature by demonstrating that the education–participation relationship is structurally contingent on labour market conditions, challenging the universal applicability of the U-shaped hypothesis in South Asia.