| Abstract: | Evidence regarding the effects of health status on migration are mixed, with some concluding that there is a positive relation while others conclude a negative relation. This study examines the analytical bases for conflicting findings and proposes a model of the health-migration relation which allows for age selectivity, nonlinearities, interactions, and dynamic effects. Often in migration research we focus on migrant-nonmigrant differentials, which helps us to understand the cross-sectional differences between the 2 groups of persons, but it does little to elucidate the complex process by which nonmigrants become migrants. The sequence of migration and health events, from the US National Health Interview Survey from 1979-1980, show that for the population under age 45, there is little temporal connection between migration and health status changes. 2/3 of the younger population with at least 1 health or migration event migrated only, with health status changes neither preceding nor following the migration. Since the migration and health events for these persons are not associated in time, there is little basis for positing strong relations of either direction for the nonelderly. The age groups for which migration and health events are connected in time are those over age 45, and the temporal proximity becomes even more marked for the elderly. Among the elderly, the migrant-only group constituted only 10% of those with at least 1 event. As expected, for the elderly the dominant relation is health deterioration after migration, the pattern displayed by 65% of the elderly with events in the last 5 years. The elderly are also more likely to be in the Destabilized Migrant category, experiencing health status declines both before and after migration. This is also consistent with the theory that the elderly in poor health at the time of migration will be more likely to be stressed by their migration and experience further deteriorations in health. The conditional probabilities indicate that health status declines are unlikely to precipitate immediate migrations, except for those with pre-existing serious conditions or more severe activity limitations. |
