A Meta-Analysis of the Literature on Climate Change and Migration
Recent surveys of the literature devoted to climate change and migration emphasize the important diversity of outcomes and approaches of the empirical studies. In this paper, we carry out a meta-analysis in order to investigate the role of the various methodological choices of these empirical studies in finding some particular results concerning the role of climatic factors as drivers of human mobility. To that aim, we code 45 papers representative of the existing literature in terms of methodological approaches. The approaches are characterized by coding more than 80 variables capturing the methodology of the main dimensions of the methods.
These dimensions include among others authors’ reputation, type of mobility, measures of mobility, type of data, context of the study, econometric methods and last but not least measures of the climatic factors. We look at the influence of these characteristics on different outcomes of the regressions: probability of finding any effect, finding a displacement effect, finding an increase in immobility, finding evidence in favour of a direct versus an indirect effect. Our results highlight the role of some main methodological choices, such as the frequency of the data on mobility, the level of development of the covered area, the particular measures of human mobility and of the climatic factors as well as the econometric methodology.