Publications, working papers, and other research using data resources from IPUMS.

DeWaard, Jack; Nawrotzki, Raphael, J 2018. Modeling migration and population displacement in response to environmental and climate change.

Claims that environn1ental and climate change (ECC) may lead to mass migration and popu­lation displacement (MPD) have helped to motivate research on the ECC-MPD relationship (Myers 2002; Stern 2007). In this chapter, we summarize an important part of these efforts by providing a brief, but targeted review of the motivations for and uses of multilevel event his­tory models. Our work builds on earlier reviews that highlighted the in1portance of multilevel approaches (Hunter et al. 2015; Kniveton et al. 2008; Mcleman 2013; Piguet 2010), including some of the tools (e.g., event history models) that demographers use to study migration (Fussell et al. 2014). One tool that received only limited attention in these reviews is that of multilevel event history models.As we later discuss, these models are characterized by several hallmarks that are particularly important in research on the ECC-MPD relationship. This chapter is organized as follows. First, in an effort to accommodate readers from different training and disciplinary backgrounds, we (re)introduce event history models. We then describe two problems that complicate efforts to use event history models to study the ECC-MPD relationship. Following a discussion of three overlapping approaches chat are commonly used to mitigate these problems, we detail their use in current empirical research. Specifically, we inven­tory and assess 20 studies published since the early 2000s in which the author(s) used multilevel event history models to study the ECC-MPD relationship. Particular attention is paid to the levels of analysis considered, the measure(s) ofECC used, and model specificacion(s). Given the targeted nature of this chapter, interested readers are encouraged to pursue more extended and advanced treatments of event history models (Allison 1982, 1984; Cox 1972; Singer and Wil­let 2003;Yamaguchi 1991), multilevel models (Gelman 2006; Kreft and De Leeuw 1998; Luke 2004; Rabe-Hesketh and Skrondol 2012; Raudenbush and Bryk 2002; Snijders and Bosker 2011), and multilevel event history models (Barber et al. 2000; Duchateau and Janssen 2010).