A Downgraded ACA Baseline

In a new study published today by the Mercatus Center at George Mason University, I assess key predictions made by both government and nonprofit research organizations about the Affordable Care Act’s (ACA) impact. The misestimates include: overestimating total exchange enrollment, overestimating enrollment of higher income people who do not qualify for subsidies to reduce premiums, projecting too many healthy enrollees relative to less healthy enrollees, and underestimating premium increases. This post focuses on the Congressional Budget Office’s (CBO) estimates. As CBO accounts for the first two years of the ACA’s implementation, its 2016 revised baseline will almost certainly show a lower overall budgetary cost for the law’s subsidies even as the average subsidy amount increases to account for a more adverse risk pool than expected.

CBO Overestimated Total Exchange Enrollment

Right before the ACA passed Congress, CBO projected the law would have an average annual exchange enrollment of 8 million people in 2014, 13 million people in 2015, and 21 million people in 2016.

It turns out these estimates were significantly too high in both 2014 and 2015 as only about 5.5 million people enrolled in 2014 and only about 9.5 million people enrolled in 2015. If the Obama administration’s recent projection of 2016 enrollment turns out to be correct, then CBO’s 2010 estimate of 2016 enrollment will be too high by about 50%.

CBO Overestimated Unsubsidized Enrollment

In March 2015, CBO released downgraded expectations of exchange enrollment. In that update, CBO projected 11 million exchange enrollees with 3 million of them earning too much income to qualify for a premium subsidy.

It turns out that CBO’s projection of 8 million subsidized enrollees was fairly accurate but its projection of unsubsidized enrollees was about double actual enrollment.

Enrollees More Expensive Than CBO Projected

Under the risk corridor program, the government collects money from insurers with excess profits on ACA plans – exchange plans and ACA-compliant plans not sold through exchanges – and pays money to insurers with excess losses on these plans. In a February 2014 estimate, CBO projected that the government would receive an $8 billion windfall from the risk corridor program because of sizeable overall insurer profits. After the administration announced its intent to implement the program in a budget neutral manner, CBO projected that for the 2014 plan year profitable insurers would pay $1 billion for excess gains and unprofitable insurers would collect $1 billion for excess losses.

It turns out that CBO significantly overestimated insurer profitability and thus expected generally healthier ACA plan enrollees than enrolled. Insurers with excess profits owed about $360 million and insurers with large losses requested about $2.9 billion. Using this data, I estimate that insurers’ losses on ACA plans in 2014, even with an $8 billion subsidy to cover the majority of the costs of their high expense enrollees, equaled about 12% of premiums. Standard and Poor’sprojects that risk corridors will also run a significant deficit for the 2015 plan year.

CBO Underestimated Premium Increases

In March 2015, CBO projected private health insurance spending per enrollee would grow by an average of 4.3% per year over the 2014-2018 period and that exchange plan premium increases would “generally reflect the underlying trend in spending by private health insurers.”

It turns out that premiums will rise by double-digits next year. The weighted average increase is about 12%, with the average lowest cost silver and bronze plans increasing by 13% and 16%, respectively.

Anticipating Changes to CBO’s Revised ACA Baseline

A large amount of uncertainty surrounds ACA projections given the magnitude of the changes made by the law. Although several of CBO’s projections have turned out to be well off the mark, it is worth noting that their projections tended to be better than those made by the Centers for Medicare and Medicaid Services, The RAND Corporation, and the Urban Institute.

In its next ACA baseline, CBO will have to account for lower enrollment, particularly of people without subsidies, than it expected. Its new projections will also have to account for worse ACA plan risk pools and higher premium increases. These adjustments will cause other estimates to change, including the overall subsidy cost, the number of people with employer-sponsored insurance (ESI), and the amounts of revenue generated from the individual and employer mandates.

Continue reading