Comparing migrants’ and stayers’ happiness: Addressing incomplete information via a bounds analysis

Abstract Efforts to identify the happiness consequences of international migration for the migrants themselves are hampered by severe data limitations. In principle, the effect of migration on the happiness of migrants can be evaluated via comparison between migrants and stayers; the stayers are used as the counterfactual for migrants (i.e. had they not migrated). But analysis in this mode must consider the selectivity of migration: migrants and stayers are different in ways that are consequential for their happiness. The difficulty is that gaining leverage over selection into migration requires the use of data about core aspects of migrants’ pre-migration lives. Available data on migrants mainly gives information on their ‘current’ (post-migration) situations; data on their pre-migration situations are typically absent. In this article, I address the incompleteness of information via a ‘bounds’ approach. The aim is to identify happiness consequences for migrants via a propensity-score weighting analysis. Key variables, relevant to migration propensity, are lacking in the data I use (the European Social Survey)—in particular, pre-migration health, relationship status, and frustration with income. To overcome these absences, I estimate results using randomly assigned values for migrants, oriented to defined scenarios that draw on what we know about migrant selectivity in general. I apply this approach to an analysis of Polish migrants living in Ireland, the UK, and Germany. In comparisons of migrants and stayers adjusted via propensity-score weighting, there is no support for the idea that migration has led to any change in migrants’ happiness (in contrast to earlier published findings).