Is There Really a Salmon-Bias Effect? Evidence of Selective Emigration among Primary Social Security Beneficiaries in the United States

Cassio M. Turra, Princeton University
Bert Kestenbaum, U.S. Social Security Administration (SSA)
Betty R. Ferguson, U.S. Social Security Administration (SSA)

Estimates of adult mortality have documented lower mortality for Hispanics than for NH-whites. A great deal of research has focused on the factors that may explain this puzzle. In the category of migration selection effects, one hypothesis that has been frequently offered is the salmon-bias effect. However, there is still no direct evidence that return migration is a determinant of the Hispanic mortality advantage. In this paper we use information on residential status and death records for primary Social Security beneficiaries to examine how mortality of elderly Hispanics and NH-whites residing abroad differs from those living in the U.S. We also look at how net migration rates differ among these population subgroups. Although our results provide evidence of higher mortality rates for Hispanics living abroad than for Hispanics living in the U.S., we found only small variations in the Hispanic mortality advantage when mortality rates for both groups were combined.