Spousal Mobility and Earnings

An important finding in the literature on migration has been that the earnings of married women typically decrease with a move, while the earnings of married men often increase with a move, suggesting that married women are more likely to act as the “trailing spouse.” This article considers a related but largely unexplored question: what is the effect of having an occupation that is associated with frequent migration on the migration decisions of a household and on the earnings of the spouse? Further, how do these effects differ between men and women? The Public Use Microdata Sample from the 2000 U.S. decennial census is used to calculate migration rates by occupation and education. The analysis estimates the effects of these occupational mobility measures on the migration of couples and the earnings of married individuals. I find that migration rates in both the husband’s and wife’s occupations affect the household migration decision, but mobility in the husband’s occupation matters considerably more. For couples in which the husband has a college degree (regardless of the wife’s educational level), a husband’s mobility has a large, significant negative effect on his wife’s earnings, whereas a wife’s mobility has no effect on her husband’s earnings. This negative effect does not exist for college-educated wives married to non-college-educated husbands.In the substantial literature on the relationship between migration and earnings, an important finding has been that the earnings of married women typically decrease with a move, while the earnings of married men often increase with a move. This is consistent with the notion that married women are more likely to act as the “trailing spouse” or to be a “tied mover.” This article considers a related but largely unexplored question: what is the effect of having an occupation that is associated with frequent migration on the migration decisions of a household as well as on the earnings of the spouse? And how do these effects differ between men and women?There are three reasons to move beyond the previous analysis of household moves to studying the effect of occupational mobility on migration and earnings. First, the analysis of changes in employment and earnings of movers is only part of a broader discovery concerning the extent to which the earnings of husbands and wives are affected by the ability to move to or stay in optimal locations. Second, the existing literature relies on the comparison of movers to nonmovers. Even longitudinal comparisons will not completely eliminate the bias in this comparison because movers likely differ in their earnings growth, not just the level of premigration earnings. Third, the methods used in the literature often do not adequately adjust for occupational differences between men and women, so it is difficult to know whether the current findings in the literature are the result of differences in jobs held by men and women, or rather are the result of differences in influence on location decisions. The question pursued in this article is, controlling for an individual’s own occupation and the earnings potential in that occupation, how does the migration rate in a spouse’s occupation affect one’s own labor market outcomes?This article uses the Public Use Microdata Sample (PUMS) from the 2000 U.S. decennial census to calculate mobility measures by occupation and education class. Mobility is measured by the fraction of workers who, in the past five years, have either (a) changed metropolitan area or (b) if in a nonmetropolitan area, changed Public Use Microdata Area (PUMA).¹ Using the sample of white, non-Hispanic married couples between the ages of 25 and 55 in the 2000 census, I perform migration and earnings analyses separately for four groups of couples: both have college degrees (“power couples”), only the husband has a college degree, only the wife has a college degree; and neither has a college degree.Results indicate that the mobility rates in both the husband’s and wife’s occupation affect the household migration decision, but mobility in the husband’s occupation matters considerably more. Comparison analysis for never-married individuals indicates that among individuals with college degrees, never-married men and women are equally responsive to occupation mobility in their migration behavior.The earnings analysis uses occupation fixed-effects and average wage in occupation-education class to control for substantial heterogeneity in earnings potential. For couples in which the husband has a college degree, the wife’s mobility has no effect on the husband’s earnings, regardless of the wife’s education. However, the husband’s mobility has a large, significant negative effect on the wife’s earnings. This negative effect does not exist for couples in which only the wife has a college degree.