Social Forces · 2004 · 25 citations · 41 references
Public AffairsHealth PolicyPublic Health NutritionFood PolicyFood SecurityWelfare PolicyFood AssistanceFood Security EconomicsFood Security PolicySocial ProtectionWelfare ReformEligible PersonsFood Assistance ParticipationMortality Risk ModelingProbit RegressionPublic Food AssistanceUnobserved ConfoundingUnobserved HeterogeneityUnobserved Individual Effects
We use data from the National Health Interview Survey-Family Resources Supplement to examine the relationship between Food Stamp receipt and prospective adult mortality, among eligible households. We specify a switching probit model to adjust for observed and unobserved factors that correlate with selection into the Food Stamp Program and mortality, and to estimate mortality under counterfactual conditions that we do not observe. The average individual, based on observed characteristics, has higher mortality when participating than when not participating. But due to unobserved differences between participants and nonparticipants, those who self-select into participation experience lower mortality than if they did not participate. Our findings suggest that Food Stamps provide an important safety net that protects the health of those who are most likely to participate.
41
Sample Selection Bias as a Specification Error
James J. Heckman · Econometrica · 1979 · 28.9K citations
Stigma: Notes on the Management of Spoiled Identity
Postgraduate Medical Journal · 1969 · 14.3K citations · Full text
Models for Sample Selection Bias
Christopher Winship, Robert D. Mare · Annual Review of Sociology · 1992 · 761 citations
Poverty and the Distribution of Material Hardship
Susan Mayer, Christopher Jencks · The Journal of Human Resources · 1989 · 650 citations