Bernard Black, José-Antonio Espín-Sánchez, Eric French & Kate Litvak, The Effect of Health Insurance on Near-Elderly Health and Mortality, Nw. L. & Econ. Research Paper Series, available at SSRN.
While advocates for the Affordable Care Act (ACA) assume it will improve the health of the uninsured, Bernard Black and co-authors observe that the link between health insurance and health is more tenuous than one may think. Partly because other factors have a bigger impact on health than does health care insurance and partly because the uninsured have always been able to rely on the health care safety net, we may see little improvement in the health of the previously uninsured from ACA.
In their study, Black et al., collected nationwide data on people who were age 50-61 in 1992. The authors looked at this “near-elderly” population because a beneficial effect of insurance would be most likely found in that group—younger people are healthier, and older people are covered by Medicare. The authors then looked at the study subjects’ access to health care and their health outcomes for the next 18 years. As expected, insured individuals used more health care resources than did uninsured people. However, there was no evidence that being insured lowered the risk of death 12-14 years into the study, and only mild evidence of a mortality benefit at 16-18 years.
As Black et al., write, even the mild benefit may have reflected unmeasured factors (e.g., diet or exercise habits) rather than health insurance status. By 16-18 years into the study, everyone would have become a Medicare recipient, and many of the study subjects would have become Medicare eligible much earlier. One would expect risk of death to have declined rather than increased once the uninsured persons became insured under Medicare, but their mortality rate rose only after they enrolled in Medicare. Other study results suggest that the lower risk of death for the insured resulted from factors other than insurance status. For example, people who had Medicaid or other public insurance had higher mortality rates than did the uninsured.
The study results are consistent with data from Oregon. After the state expanded its Medicaid program and assigned the limited new slots by lottery, it effectively created a randomized controlled study of the benefits of Medicaid coverage. When researchers analyzed data from the first two years of the expansion, they found that Medicaid coverage resulted in greater utilization of the health care system. However, there was no reduction in levels of hypertension, high cholesterol or diabetes. There was a reduction in levels of depression, but no increase in the extent to which participants reported being happy.1
To be sure, other studies have found improvements in health status that were related to improvements in insurance status. In one study, for example, researchers compared three states that had expanded their Medicaid programs between 2000 and 2005 to include childless adults with neighboring states that were similar demographically but had not undertaken similar expansions of their Medicaid programs. In the aggregate, the states with the expansions saw significant reductions in mortality rates compared to the neighboring states.2 But as Black et al., observe, only one of the states showed a significant decrease in mortality from the expansion, and the decrease was too large to be explained by the reduction in the number of uninsured. Similarly, while another study found a significant decrease in mortality rates for patients with emergency needs for health care once Medicare kicked in at age 65, the decrease in mortality was too large to be explained by changes in care for the small percentage of Americans who moved from being uninsured to being insured at age 65.3
All of this is not to say that health care does not matter. Rather, the study from Black et al., suggests that ACA will not do that much more for the health of the previously uninsured than did the pre-ACA safety net. The safety net is porous, but it may provide nearly as much benefits for health as will ACA.
- Katherine Baicker et al., The Oregon Experiment—Effects of Medicaid on Clinical Outcomes, 368 New Eng. J. Med. 1713 (2013). [↩]
- Benjamin D. Sommers, Katherine Baicker & Arnold M. Epstein, Mortality and Access to Care Among Adults After State Medicaid Expansions, 367 New Eng. J. Med. 1025, 1026, 1029-1031 (2012). [↩]
- David Card, Carlos Dobkin & Nicole Maestas, Does Medicare Save Lives?, 124 Quarterly J. Economics 597, 632-633 (2009). [↩]