Treating Mental Health Conditions Improves Labor Market and Other Economic Outcomes in Low and Middle-Income Countries

2018, Scott et al., 2016, Singla et al., 2017) • However, an open question is whether mental health interventions also produce economic effects in low-income settings 1 Introduction Background Search Sample Empirical strategy Results Robustness Mechanisms Pooled OLS Costs Conclusions The recent increase in RCTs in LMIC has enabled meta-studies Increase in mental health studies in low and middle in. [...] Perform a Bayesian hierarchical meta-analysis (weighting by heterogeneity) In each figure that follows, each row will represent the results of a meta-analysis of an outcome group of interest (specified by the row) on an intervention of interest, within a sample affected by a given mental disorder 14 Introduction Background Search Sample Empirical strategy Results Robustness Mechanisms Pooled OLS C. [...] differences are random and effects follow a normal distribution across studies • Adjust standard errors of study-specific estimates to incorporate measure of the extent of heterogeneity, among the intervention effects in different studies (Tau-squared) Back 6 Appendix Effects are qualitatively similar to estimates in HIC Study Countries Their Effect Size Our Effect Size Aggregate Labour Supply Tim. [...] var.: work-related outcomes (Hedges’ g) Constant (τ estimate) 0.28 0.30 0.29 0.45 0.30 0.29 0.29 Standard deviation of τ̂ (0.11) (0.11) (0.11) (0.33) (0.12) (0.11) (0.11) Measurement heterogeneity Variance of error term No Yes No No Yes No No Control conditions No No Yes No Yes No No Measurement timing No No No Yes No No No Intervention and context heterogeneity Implementing party No No No No No Y. [...] var.: work-related outcomes (Hedges’ g) Constant (τ estimate) 0.27 0.30 0.28 0.27 0.30 0.30 0.30 Standard deviation of τ̂ (0.10) (0.09) (0.13) (0.12) (0.14) (0.13) (0.12) Measurement heterogeneity Variance of error term No Yes No No Yes No No Control conditions No No Yes No Yes No No Measurement timing No No No Yes No No No Intervention and context heterogeneity Implementing party No No No No No Y.