In recent Mercatus research, I found that the Affordable Care Act (ACA) was generally underperforming initial expectations, particularly with lower exchange enrollment and an overall risk pool with a disproportionate number of older and less healthy people. Last week, the Department of Health and Human Services (HHS) released 2016 exchange enrollment data through the first two months of the three-month open enrollment period. Although nearly one month of open enrollment remains, the new data generally supports my previous findings. Here are seven things you should know about the new data.

1) 2016 enrollment will likely be at least ten million people below expectations when the ACA was passed

In 2010, government and public policy experts projected that an average of about 24 million people would be enrolled in exchange plans in 2016. The Congressional Budget Office (CBO) estimates were on the low end, with a projection of 21 million exchange enrollees.

Last fall, HHS substantially lowered expectations for 2016, projecting between 11.0 and 14.1 million initial enrollees and end-of-year enrollment between 9.4 and 11.4 million people. Through the first two months of the 2016 open enrollment period, nearly 11.3 million people signed up for, or were reenrolled in, an exchange plan (multiple people may be on the same plan). If this year’s enrollment follows last year’s pattern, then about 14.0 million people will be enrolled by the end of the 2016 open enrollment period because about 20% of total sign-ups occurred in the final month last year.

Last year, about 12% of initial enrollees never paid their share of the premium. Therefore, of the 14.0 million initial enrollees, only about 12.3 million will likely effectuate coverage. Although people can enroll in an exchange plan at other times during the year if they qualify for a special enrollment period, net exchange enrollment declined by an average of about 2% each month in both 2014 and 2015. Assuming a 2% decline each month in enrollment after February yields about 10 million people covered in exchange plans at the end of 2016 and a little more than 11 million people, on average, enrolled in exchange plans in 2016.

It is worth noting that CBO’s most recent estimate of exchange enrollment—released in June 2015—projected 20 million exchange enrollees in 2016. Therefore, CBO’s next estimate of the ACA will undoubtedly be a dramatic departure from how it has projected the law to date.

2) People with at least middle class income still largely shunning exchanges

Nearly two thirds of people who have enrolled in an exchange plan in the 38 states using HealthCare.gov reside in households with income below 200% of the federal poverty level (FPL)—an amount equal to about $23,500 for a single person and about $32,000 for a two-person household. (Enrollment data by income was not released for states that are not using HealthCare.gov for enrollment.) Only one in eight initial sign-ups resides in a household with income above 300% of the FPL—an amount equal to about $35,300 for a single person and about $48,000 for a two-person household. As the figure below shows, this distribution, by income, is similar to what it was in 2015.

The overall exchange distribution is much more skewed toward lower-income people than experts projected. For example, in January 2015—just one year ago—the Urban Institute projected that only 36% of enrollees would be in households with income below 200% of the FPL and that 39% of enrollees would be in households with income above 300% of the FPL. These estimates are shown in the above figure as a reference point for actual enrollment. While CBO did not release detailed estimates by income grouping, the agency did estimate that a significantly greater share of enrollees would not receive subsidies, i.e. would be higher-income, than have thus far enrolled in coverage.

As I detailed in my Mercatus research paper, the ACA’s cost-sharing subsidies likely explain a large part of the skewed distributional enrollment. People with income below 200% of the FPL who purchase silver plans—plans that cover an estimated 70% of average expenses—qualify for large subsidies paid to insurers for insurers to reduce the generally high ACA plan deductibles and other out-of-pocket payments.

3) Enrollees still skewing older

In 2014, the House Committee on Oversight and Government Reform—where I worked at the time—released information from insurance companies about their expectation of the age mix for exchange plans. The insurer expectations of enrollees’ age mix, along with the 2015 and 2016 enrollment age mix, are contained in the following table. The table shows that far fewer children have enrolled in exchange plans than insurers expected and that people age 55 and older represent a significantly larger share of enrollees relative to what insurers expected.

While the 2016 age mix is actually skewed a bit older than the 2015 age mix thus far, enrollees who sign up toward the end of the open enrollment period tend to be younger than people who sign up earlier. As a result, the 2016 age mix will probably end up looking quite similar to the 2015 age mix, which is still significantly older than insurance companies initially expected during the law’s first open enrollment period. The much smaller number of enrollees who are younger, healthier, and above 200% of the FPL than predicted suggests that many experts substantially overestimated the degree to which the individual mandate would induce enrollment. 

4) Average advance premium tax credit up 12% from last year

According to HHS’ recently released information, 83% of exchange enrollees in the 38 states using HealthCare.gov will receive an advance premium tax credit (APTC). APTCs are the payments made by the Treasury Department to the insurance company chosen by the enrollee that reduce the enrollee’s share of premium. This year’s average APTC equals $294—12% more than last year’saverage APTC of $263. Since the APTCs are structured to limit how much people at certain income levels pay for their share of the premium, taxpayers bear most of the cost as premiums increase. The percentage APTC increase this year is about the same as the average premium increase for ACA plans.  

5) 90% of enrollees selected silver or bronze plans  

People who have signed up for 2016 exchange coverage have so far overwhelmingly enrolled in the two lowest cost plans—silver plans and bronze plans (plans projected to cover 60% of average expenses). Nearly 70% of the people are enrolled in silver plan coverage—likely because only people who select silver plans are eligible for the cost-sharing subsidies. Overall plan selection, by metal tier, is similar to last year although there has been a decline in the number of people with platinum plans (plans projected to cover 90% of average expenses).

6) 27% of enrollees are new sign-ups, 38% of enrollees are active reenrollees, and 33% of enrollees are automatic reenrollees 

About 3.0 of the 11.3 million enrollees are classified by HHS as new customers. The remaining 8.3 million enrollees consist of roughly 4.2 million active enrollees—people who logged onto an exchange website and actively selected a plan—and 3.7 million automatic enrollees—people who were enrolled in coverage at the end of 2015 and who were enrolled in either the same plan or a similar plan in 2016. The reenrollment type is unknown for roughly 300,000 auto-enrollees in the 13 states that are not using HealthCare.gov for enrollment purposes.

7) High auto-enrollment in states not using HealthCare.gov may lead to premium shock

The following table shows enrollment by new sign-ups and reenrollment type for both the 38 states using HealthCare.gov for enrollment and the 13 states (counting DC as a state) using their own sites for enrollment. There are significant differences in enrollment make-up between the states using HealthCare.gov for enrollment and the states using their own enrollment platform, differences which are difficult to understand and seem unlikely to be the result of chance.

The most striking data point in the table is that between 50% and 60% of enrollees in the 13 states (including DC as a state) with their own enrollment platform have been automatically reenrolled in their plan. People who are reenrolled in plans are at significant risk of paying a higher net premium in 2016 than in 2015 as plans that enrolled a greater share of people increased premiums more than the average premium increase. As a result, people in these 13 states face the potential for premium shock. This includes California, Maryland, and Massachusetts where 57%, 66%, and 79% of enrollees have been auto-enrolled, respectively.

Takeaway

With one month remaining in this year’s open enrollment period, exchange enrollment in 2016 is generally tracking previous year trends—significantly below initial expectations with a poorer and older risk pool than anticipated.