Working Paper: NBER ID: w22348
Authors: Lauren L. Schmitz; Dalton Conley
Abstract: This study examines whether the effect of job loss on body mass index (BMI) at older ages is moderated by genotype using twenty years of socio-demographic and genome-wide data from the Health and Retirement Study (HRS). To avoid any potential confounding we interact layoffs due to a plant or business closure—a plausibly exogenous environmental exposure—with a polygenic risk score for BMI in a regression-adjusted semiparametric differences-in-differences matching framework that compares the BMI of those before and after an involuntary job loss with a control group that has not been laid off. Results indicate genetically-at-risk workers who lost their job before they were eligible for Social Security benefits, or before age 62, were more likely to gain weight. Further analysis reveals heterogeneous treatment effects by demographic, health, and socioeconomic characteristics. In particular, we find high risk individuals who gained weight after a job loss were more likely to be male, in worse health, single, and at the bottom half of the wealth distribution. Across the board, effects are concentrated among high-risk individuals who were not overweight prior to job loss, indicating unemployment at older ages may trigger weight gain in otherwise healthy or normal weight populations.
Keywords: Job Loss; Body Mass Index; Genotype; Health and Retirement Study
JEL Codes: I1; J63; J69
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Involuntary job loss due to business closures (J65) | Changes in BMI (I12) |
| Genetically at-risk workers who experience job loss before age 62 (J26) | More likely to gain weight (I12) |
| Job loss (J63) | Weight gain in otherwise healthy populations (I10) |
| Job loss (J63) | Detrimental impact on long-term cardiovascular health (I12) |
| Job loss (J63) | Weight gain among specific subgroups (I14) |