Discovering Diverse Mechanisms of Migration: The Mexico–US Stream 1970–2000
Scholars have identified diverse mechanisms that lead individuals to mi- grate. these mechanisms are analyzed in various migration theories devel- oped in multiple disciplines. in neoclassical economics, higher wages in the destination country propel the migration of individuals who expect to earn more there. in the new economics of migration, the uncertainty in the origin economy leads to migration by households or household members who face risks to domestic earnings. in cumulative causation theory, the growing web of social ties between countries of origin and destination fosters the migration of individuals who are connected to earlier migrants. in a series of seminal publications, massey et al. (1993, 1994, 1998) argued that the various causal configurations implied by different theories are not mutually exclusive. income-maximizing migrants can co-exist along- side migrants who seek to diversify risks or alongside those who join family or friends at the destination. massey and espinosa (1997) provided the first empirical application of this argument in the mexico–united States setting. associating each theory with a set of independent variables, the authors used regression analysis to compare which variables and theories best predict who migrates. this empirical approach, although commendable in combining vari- ous theories, did not fully reflect massey et al.’s (1993) conceptualization, since it treated theories as competing rather than complementary accounts of migration. the approach also did not consider the conditional nature of theories, that is, the fact that each theory applies to a specific group of indi- viduals under specific conditions. in more recent work, massey and taylor (2004: 383) critiqued their earlier approach (massey et al. 1998) for not being able to “state with any precision which theories were most important empirically in accounting for variations in the number, rate, and characteristics of immigrants over time and whether and why different theories may prove more or less efficacious in accounting PoPulation and develoPment Review 38(3): 393–433 (SePtembeR 2012) 393 394 discovering diverse MechanisMs of Migration
for immigration patterns in different times and places,” and they identified the major challenge for migration research to be “test[ing] various theoretical explanations comparatively...to determine which ones prevail under what circumstances and why.” the present study addresses the challenge of characterizing the causal heterogeneity of migration. thus it confronts a major methodological prob- lem in social science: identifying the different mechanisms at work among different groups of individuals. in an effort to emulate the natural sciences (lieberson and lynn 2002), quantitative social inquiry often focuses on (and generalizes from) an average case rather than studying the variability across cases (duncan 1982; Xie 2007). in recent years, however, new methods, such as multi-level, latent class, or growth curve models, have allowed research- ers to study the variability in outcomes across different contexts, groups, or trajectories (Raudenbush and bryk 1986; d’unger et al. 1998; muthén and muthén 2000). migration research has closely followed these developments. Studies have used split samples, interaction terms, or hierarchical models to identify the different factors influencing migration for men and women, among differ- ent ethnic groups, or in different contexts and time periods (e.g., Kanaiaupuni 2000; marcelli and Cornelius 2001; massey, Goldring, and durand 1994). but each of these studies relied on a few fixed categories, such as gender or community, to characterize the heterogeneity in migration, an approach that can be considered restrictive, even essentialist (Somers 1994). in this study, rather than dissecting or modeling data based on a few selected attributes, we seek to discover the...
Please join StudyMode to read the full document