Prior to the ACA, about 12 percent of non-elderly uninsured were aged 55 to 64 years old, not yet eligible for Medicare.  At these ages, an individual without access to employment-based coverage may face underwriting and high premiums in the individual non-group market, depending on the state.  As these older individuals gain private insurance through the exchanges and enroll in Medicaid under the ACA, there may be downstream consequences on their Medicare spending.

We use the National Health Interview Survey (NHIS) data linked to Medicare claims data to identify the difference in Medicare spending at ages 66 and 67 between four groups categorized by their insurance status and the rating rules of the state in which they resided at age 64.  The relationship between Medicare spending and insurance status prior to aging into Medicare is ambiguous because the uninsured may spend more than the insured due to pent up demand for health services or may spend less due to a smaller preference for utilizing health care.  McWIlliams et al. allege that, among Medicare beneficiaries, those who were uninsured prior to aging into Medicare had higher rates of health care utilization and spending than those who were insured prior to aging into Medicare.  Decker et al performed a sensitivity analysis suggesting McWilliams’ observed higher spending for the previously uninsured is due primarily to choice of reweighting (rather than including covariates) and the inaccurate inclusion of those with SSDI as uninsured.

The Affordable Care Act changes the nature of the private market for insurance and thus the count and composition of who is uninsured.  Insurance purchased individually will be available on a guaranteed-issue basis and premiums will not vary by health status or sex.  In addition, variation by age will be limited.  These new federally mandated regulations of the nongroup market mirror existing community rating regulations in some states.  In addition to rating reforms, the ACA established tax credits for purchasing insurance in the individual market, expanded Medicaid eligibility, and required most legal U.S. residents to obtain insurance coverage or face a penalty tax.

Even if the average Medicare spending of the uninsured and insured were the same prior to the ACA, the change in the composition of who remains uninsured may affect Medicare spending.  Uninsured in states without rating regulations are more likely to be in poor health than those in states with some form of community rating.  If, through rating regulations favoring the near-elderly, tax subsidies to purchase, and an individual mandate to obtain coverage or pay a penalty tax, the composition of the uninsured shifts away from those with pent up demand for health care, the effect may be to reduce future Medicare spending.   Our paper attempts to quantify the difference in expected Medicare spending by prior insurance status and state regulation of the nongroup market.