Special Series: The Future of ACA Exchanges: Making History or Repeating It?

Insurance companies entering the exchanges in 2014 were well aware that Affordable Care Act (ACA) would face challenges in its initial years. Yet now we’re seeing an exodus of sorts from the individual exchanges. Would a closer look at the Massachusetts Health Connector have given insurers a clearer picture of what to expect?

When the ACA was implemented in 2014, health insurers were generally optimistic about the new individual insurance exchanges the act created. Despite a rocky start to the program, insurers increased involvement in the first year, raising the average number of carriers in each state from 5.0 to 6.1. 

Now, that optimism has largely dissipated. Sixteen of the 23 co-ops created as part of the ACA have left the program, and UnitedHealthcare and Aetna have withdrawn from most exchange markets. Many of the remaining issuers are signaling large rate increases, with consumers potentially facing higher prices even if they switch plans to reduce premium. 

The recent exits seem to indicate a fundamental miscalculation of the profile of the enrollees in the new individual market. The reality of that population was neither expected nor allowed for in pricing. This raises important questions for the future of the exchanges and the ACA as a whole.

Should insurers have known better? 

The ACA exchanges share many attributes with the Massachusetts Health Connector, which was signed into law in 2006. The dynamics – and challenges – of the Connector population are very similar to the ACA population; therefore, it is not surprising to see those challenges repeated in the ACA exchanges. 

Here is a closer look at the similarities between ACA and Connector populations. 

Demographics: Participants in the Connector’s individual sector were significantly older on average than participants in group sectors. The individual sector had relatively few members under age 19 (despite the availability of young adult plans) and a disproportionate number of adults aged 60 to 64. 

Something similar happened in the ACA exchanges: Although a widely cited actuarial study predicted that 18- to 34-year-olds would make up 40 percent or more of the adult individual market, this age group’s actual share of 2015 enrollment in states with federally facilitated marketplaces was closer to 31 percent. Given that ACA age bands are structured so that younger enrollees subsidize their older counterparts, a population that is older than expected results in unsustainably low rates. 

Adequacy of premiums: In 2008, newly insured participants in Massachusetts’s guaranteed-issue individual market generated 33 percent higher claims on a PMPM basis than small-group enrollees. Their median premiums, meanwhile, were only 22 percent higher than premiums in the small group market. Again, the ACA exchanges had a similar experience. In 2015, normalized risk scores for those purchasing individual ACA coverage were about 15 percent higher than the scores for small-group ACA coverage, but the average individual premium was almost 15 percent lower than the small group rate. (Some of the difference can be explained by the presence of the Federal Reinsurance Program, through which individual ACA plans are subsidized by other insurance.) 

Adverse selection: In Massachusetts, a relatively high percentage of individuals terminated coverage within the first year – 13.8 percent in 2006, growing to 24.2 percent in 2008. Individuals terminating coverage within a year had a loss ratio 8.8 percent higher than average, suggesting that adverse selection may be occurring in the ACA markets. 

 The recent exit [of insurers] indicates a fundamental miscalculation of the profile of the enrollees in the new individual market. 

Or was it impossible to predict?

Despite the similarities to the Connector population, it is not an exact replica, and there were some true surprises on the path to ACA implementation. Here are three market conditions that were unknowable to plans when they prepared their premiums for 2014. 

Risk pool: The transitional policy that allowed individuals to retain non-ACA-compliant coverage was extended for nearly four years. That may have pleased individual consumers, but it meant that previously underwritten individuals were not included in the single risk pool. In addition, short term or limited-benefit plans attracted enrollment because of low premiums. 

Risk mitigation programs: The availability of the risk corridor program to protect against insurer losses may have contributed to aggressive pricing in ACA’s initial years. But congressional limits on funding meant the program paid out benefits at just 12.6 percent of expected levels. Some plans have also experienced adverse outcomes from the risk adjustment program. 

Special enrollment periods: Under ACA, these periods were expanded (see here and here), providing potential insureds the opportunity to enter the pool only when coverage was needed. In 2015, about 10 percent of individuals acquiring coverage through healthcare.gov did so as part of a special enrollment period. 

What next? 

Lackluster financial results have created a lot of disruption in, and skepticism of, the ACA individual market. But there are some reasons for optimism:

• CMS has indicated that it will provide stronger enforcement of special enrollment periods in future years.
• The transitional policy is slated to conclude at the end of 2017, merging the two markets for individual coverage.
• Draft regulations have been proposed to limit use of alternative coverage to replace major medical coverage.
• CMS analysis of claims shows that the morbidity of the risk pool stabilized in 2015, and may be improving for exchange enrollees.
• Higher rate increases by participating plans and the departure of some less-disciplined competitors are countering some of the underpricing from initial periods, so we may be entering a period of more rational pricing. 

Definite challenges and uncertainties remain, for which issuers need to plan:

• How do plans retain those not receiving premium subsidies during a period of rapid increases in premium, or respond to changes in the single risk pool if they leave?
• Are plans prepared for disruption to the market caused by the end of transitional policies, including potential new entrants to the individual market from employees losing coverage where small employers drop coverage rather than enter the ACA market?
• Will continued investment by plans in coding completeness, coupled with updates to the risk adjustment model, improve the predictability of financial results from this line of business?
• Will issuers build capabilities to manage the care of their insured populations, including leveraging improved analytics on care consumption and supporting clinical integration? 

Time will tell if updated pricing and revisions to market rules sufficiently address the ACA markets’ challenges. If there is to be a viable individual ACA market after 2017, plans must focus on understanding, engaging, and managing this population.