A nonparametric analysis of the healthy immigrant effect

This paper uses data from the Swiss Labour Force Survey to evaluate the existence of the healthy immigrant effect (HIE) which would translate in i. a health advantage of immigrants upon their arrival in Switzerland compared to individuals with similar characteristics but Swiss-born and ii. an erosion of that advantage over the time of residence until convergence in the levels of health between these two groups. Our original contribution is to address this issue by taking a nonparametric approach in order overcome any potential danger of misspecification that would preclude valid inference. We find little empirical support for the HIE: i. no initial advantage and ii. no convergence but the health status of immigrants is shown to deteriorate more than Swiss-born individuals with similar characteristics. Significant differences appear when disaggregating among immigrants' country of origin. Interestingly, we show that a standard parametric approach, in contrast to our findings, would fully confirm the existence of the HIE with the same data set.