Summary
Using the American Community Survey (ACS) 2005–2007 and Social Security Administration (SSA) administrative records (Disability Research File [DRF] + Continuous Work History Sample [CWHS] + Numident), this paper corrects the finding of Huang et al. (2011) that foreign-born adults have higher work disability than native-born — a measurement artifact of the flawed 2000 census work disability question. The true pattern is the opposite: the healthy immigrant effect extends to work disability, with the foreign-born having roughly half the prevalence and about two-thirds the disability insurance (DI) incidence of native-born adults. There is substantial variation by region of origin, with the former Soviet Union as the most notable exception to the general immigrant health advantage.
Key Claims
- Foreign-born adults have roughly half the work disability prevalence of native-born: 3.6% vs. 7.9% (men), 4.4% vs. 8.2% (women) per 2005–2007 ACS. Adjusted odds ratio (OR) ≈ 0.39–0.40 controlling for age, education, and marital status.
- The Huang et al. (2011) contrary finding was a measurement artifact: the 2000 census work disability question was systematically misunderstood — particularly by limited-English respondents — as asking about employment status rather than disability. Work disability prevalence dropped ~50% after the ACS redesigned the question in 2003.
- Work disability prevalence is lower for the foreign-born in every region of origin compared with native-born. The foreign-born advantage holds for both men and women and at every level of education and marital status.
- DI incidence (2001–2010 SSA administrative data): foreign-born 4.16/1,000 (men) and 4.36/1,000 (women) vs. 6.56/1,000 for U.S.-born. Age-standardized incidence by region ranges from 2.28/1,000 (East Asia) to 5.66/1,000 (Caribbean). All region-specific incidence rate ratios (IRRs) are significantly <1.0, none above 0.77.
- Work disability odds increase monotonically with duration in the U.S., consistent with the "assimilation to unhealthy American lifestyle" literature; non-citizens have the lowest odds.
- The former Soviet Union is the exception: OR 1.10 (men) / 1.32 (women) — higher-than-native work disability prevalence; modal impairment is mental health (vs. musculoskeletal elsewhere); >50% technical denial ratio; attributed to refugee composition (less positive health selection) and Soviet collapse exposure.
- Technical denial ratio (non-medical, mostly insufficient work history) averaged 25% overall, rising steeply over 2001–2010 due to SSA policy change of adjudicating all Supplemental Security Income (SSI) applications as simultaneous DI applications.
- Medical allowance rates for the foreign-born exceed the native-born overall (58.1%): Southern Europe 78.0%, former Soviet Union 71.6%, other Eastern Europe 71.3%. Case mix differences (musculoskeletal dominant everywhere except former Soviet Union) explain little of the cross-origin variation in allowance rates.
Concepts Introduced or Extended
Entities Mentioned
Quotes
"Work disability is much less likely among the foreign-born—both collectively and for each region of origin—than among those born in the United States and its territories."
"The conclusion across both analyses is the same: work disability is much less likely among the foreign-born."
My Take
The measurement correction is the paper's most durable contribution — documenting exactly how question-design artifacts can reverse a finding's sign, and why prior-period census data should be used with caution for disability questions. The DRF + CWHS + Numident linkage is methodologically valuable but rarely available; the paper essentially provides the entire national picture of DI incidence by nativity for 2001–2010. The former Soviet Union exception is the most theoretically interesting result: it shows that the healthy immigrant effect depends critically on the health-selection mechanism, which fails for refugee populations admitted on humanitarian grounds. One limitation: Social Security Disability Insurance (SSDI) only (no SSI), so the picture for groups with many working-age SSI recipients is partial.