The results showed that although in a multivariate setting, no variable seemed to be responsible for the heterogeneity a consideration of univariate regressions, revealed that the coefficient of geography was positive and highly significant at the 1% level, while sample size, publication year and model type were only marginally significant (at 10% level).
← all excerpts
A systematic review and meta-analysis on the effects of ill health and health shocks on labour supply.
1
—
—