Long-run mortality effects of Vietnam-era army service: evidence from Australia’s conscription lotteries

We estimate the effect of Vietnam era Army service on mortality, exploiting Australia’s conscription lotteries for identification. We utilise population data on deaths during 1994-2007 and military personnel records. The estimates are identified by over 51,000 compliers induced to enlist in the Army, including almost 16,000 who served in Vietnam. The implicit comparison group is the set of men who did not serve in the Army, but who would have served had their date of birth been selected in the ballot. We find no statistically significant effects on mortality overall, nor for any cause of death (by ICD-10 Chapter). Under reasonable assumptions on the death rate of compliers, the results can be expressed as relative risks (RR) of death during 1994-2007. The estimated overall RR associated with Army service is 1.03 (95% CI: 0.92, 1.19). On the assumption that Army service affected mortality only for those who served in Vietnam, the estimated RR for Vietnam Veterans is 1.06 (95% CI: 0.81, 1.51). We also find no evidence to support a hypothesis of offsetting effects due to domestic Army service (beneficial to longevity) and service in Vietnam (detrimental).