Boyle and Lahey 2016 — Spousal Labor Market Effects from Government Health Insurance Evidence from a Veterans Affairs Expansion

Summary

Using a difference-in-differences (DiD) design around the mid-1990s Veterans Affairs (VA) health care expansion, Boyle and Lahey (2016) find that when older male veterans gained government health insurance, their wives significantly increased labor supply — by $1$–$2$ percentage points (pp) in employment probability, $\approx 0.5$ hours/week, and $3$–$4\%$ in log earnings. The effect is concentrated entirely among low-socioeconomic-status (SES) wives (high school education or less, low non-housing wealth, no pension), who respond by $\approx 3$ pp more employment and $\approx \$23$–$27$/week more earnings. The paper identifies financial constraints — not complementarities of spousal leisure — as the dominant mechanism.

Key Claims

• VA mid-1990s expansion caused a large increase in husbands' non-employer health insurance (HI) receipt; husbands simultaneously exited employment (per Boyle-Lahey 2010: $-3.3\%$ labor force participation (LFP), $-8.4\%$ full-time work), which stripped wives of access to employer-sponsored insurance (ESI) through the husband's plan.
• Wives' employment probability rose $1$–$2$ pp ($3$–$4\%$), hours rose $\approx 0.5$/week, and log earnings rose $3$–$4\%$ on average.
• Effects driven entirely by low-SES wives: $+3$ pp employment, $+0.6$–$0.8$ hrs/week, $+\$23$–$27$/week. High-SES wives show no significant response.
• Employment transitions: non-working wives shifted toward part-time/self-employment; part-time wives shifted toward full-time; full-time wives were unchanged.
• The Added Worker Effect (AWE) is triggered by the positive insurance shock because husbands' VA enrollment caused them to exit employment, requiring wives to supply their own HI coverage.
• Financial constraints dominate complementarities of spousal leisure as the mechanism (SES heterogeneity test: only credit- and insurance-constrained wives respond).
• Data: Current Population Survey (CPS) 1992–2002, main sample; Health and Retirement Study (HRS) for supplementary validation; husband age 55–64; veteran vs. non-veteran wives as DiD comparison groups.

Concepts Introduced or Extended

• Added Worker Effect
• Employment Lock

Entities Mentioned

• Melissa A. Boyle
• Joanna N. Lahey

Quotes

"We find that wives of veterans who became eligible for VA health care significantly increase their labor supply... the increase is driven by low-SES wives, suggesting financial constraints rather than complementarities of spousal leisure drive the effect."

"The effects of husband's insurance coverage on wife's labor supply are concentrated among low-SES women... this result is consistent with the hypothesis that wives enter the labor market in order to obtain health insurance coverage."

My Take

The paper's main contribution is identifying a novel trigger for the Added Worker Effect: not a negative income shock but a positive insurance shock that changes the secondary earner's coverage incentive. The SES heterogeneity — null for high-SES wives — is strong evidence against the leisure-complementarities story and in favor of an insurance coverage-loss channel. The limitation is that the VA expansion was a specific population (older male veterans, mid-1990s), and the mechanism depends on the pre-Affordable Care Act (ACA) employer-sponsored insurance system where ESI was tightly tied to own employment. Post-ACA, the spousal coverage loss would be partially mitigated by marketplace alternatives.