Malaki Ronald Mandela

Abstract: I estimate heterogeneous returns to health capital on labour supply in South Africa us- ing five waves of the National Income Dynamics Study (2008–2017), a composite latent health stock purged of self-reporting bias, and separate dynamic models for hours worked and employment status. Unconditional returns to health are weak at both margins for most race-gender cells, a pattern consistent with structural barriers to employment bind- ing more tightly than health capacity in this labour market. The Child Support Grant interacts with health in a margin-specific way: among African female recipients, a one- standard-deviation health improvement is associated with a meaningful rise in hours, while the same improvement generates no detectable hours response for non-recipients. At the extensive margin the grant raises employment probability for African and Coloured women through an income channel that does not require health as a mediator. I caution against drawing welfare conclusions from the partial-equilibrium estimates reported here.

With Christine Eckert, Mehdi Hosseinkouchack, Natalina Zlatevskac

Abstract: Voluntary nutrition labels aim to promote healthier eating by reducing the cost of acquiring nutritional information, yet their effectiveness is often limited and varies across population groups. We argue that this effectiveness is moderated by the interaction between consumers external motivation and internal agency, and that life-stage transitions altering both simulta- neously represent a critical but underexplored source of treatment heterogeneity. We focus on parenthood as a life stage that introduces new responsibilities while reshaping motivations gov- erning dietary decisions, and examine how internal agency, operationalized as Locus of Control (LoC), influences parents use of informational cues. Using five waves of the nationally repre- sentative Household, Income and Labour Dynamics in Australia (HILDA) survey, we employ a quasi-difference-in-differences design with a triple interaction of treatment exposure, parent- hood, and a post-policy indicator, stratified by LoC tertiles, to estimate the effect of the Health Star Rating (HSR) on dietary quality. Among non-parents, the HSR improves diets for those with moderate to high internal LoC but produces no detectable effect on externally-oriented individuals. Among individuals with parental responsibilities, we document a large, significant policy response for those in the external LoC tertile, an amplified response for those in the middle LoC tertile, and no significant additional gain for those in the highest LoC tertile. Our event study estimates show that activation for externally-oriented parents emerges immediately, while amplification for moderate-LoC parents builds gradually. These findings suggest that vol- untary informational policies depend on the interplay between the perceived return on health information – a dimension of non-cognitive human capital – and life-stage-specific external obli- gations, and that accounting for this heterogeneity could substantially improve the reach and distributional equity of front-of-pack nutrition policies. From a policy perspective, our results highlight that expanding label coverage beyond the current voluntary 40% adoption ceiling, while designing complementary interventions targeting low-internal-agency parents, offers the highest marginal gains in both dietary equity and population health.

Progressive Taxation and Income Inequality: Quasi-Experimental Evidence from Two Kenyan PAYE Reforms