Income, self-selection, and return and onward interprovincial migration in Canada

Estimated returns to migration based on comparison of individual migrants may be biased owing to self-selection in the migration process. Using data derived from the 1986 Canadian census, I will study the effects of expected wage differentials in determining the return or onward migration decision of normative adults aged 20 to 64 years. Evidence was found that return migrations were in the 'right' direction, as they are observed to respond to provincial economic variables (that is, average employment growth and income levels) in a rational manner. After accounting for self-selectivity, I found that results indicate that return migrants (individuals migrating back to their province of birth) are negatively selected, and experience lower income levels, following the return migration, than onward migrants would have, had they chosen the return migration option. This drop in expected wages decreases the propensity associated with making a return migration. Despite this drop in income, the large proportion selecting the return migration option suggests the importance of the province of birth in the mental map of nonnative migrants.