The election of Donald Trump to the Presidency has accelerated and transformed health policy discussions. Another important recent development was the October 24 announcement by the Department of Health and Human Services (HHS) that the average increase in premiums for Marketplace plans would be 22 percent overall, and 25 percent for states using HealthCare.Gov. In this post, we aim to illuminate the underlying forces that led to these increases in premiums so as to clarify the historical record and provide useful information for interested parties who are dealing with the Affordable Care Act (ACA).

We have been tracking changes in employer-based insurance since 1986, first with the predecessor surveys to the annual Kaiser Family Foundation/Health Research & Educational Trust (Kaiser/HRET) Employer Health Benefits Survey (conducted at the Health Insurance Association of America and KPMG), and since 1999 with the Kaiser/HRET annual surveys themselves. With funding from the Commonwealth Fund, we also have followed changes in Marketplace plans since 2014, the first year of Marketplace operation.

It should be made clear that the 22 percent figure applies to Marketplace plans. These plans cover about 7 percent of the U.S. population (Exhibit 1). The figure does not apply to the 50 percent of the population with employer-based insurance where premiums have been growing 3-4 percent a year since 2011, a historic low. Nor does it apply to Medicare, where per capita costs also have grown at historically low levels since 2010, the year that the ACA was enacted.

Premium Trends for Marketplace Plans

In 2014, insurers set premiums at 15 percent less than the amounts the Congressional Budget Office (CBO) had estimated in 2009. In setting premiums for 2015, the second year of Marketplace operation, insurers had only four months of medical claims data on which to rely. There was extensive market entry by new insurers, with 25 percent more insurers competing per market rating area versus 2014. Nationally, there was no change in premiums from 2014 to 2015. Two-thirds of insurers incurred financial losses on their Marketplace business in both 2014 and 2015.

The number of insurers competing per rating area declined 7 percent in 2016 and premiums rose nationally by an average of 6 percent. With the withdrawal of major national carriers—United Healthcare, Aetna, and Humana—17 percent fewer carriers are competing per state in 2017. So 2017 is a year of reduced competition and catch-up pricing.

Over the past 35 years employer-based insurance has experienced what appears to be spikes in premiums similar to the current Marketplace (Exhibit 2). For example, according to the Bureau of Labor Statistics, Employer Cost Index premiums increased 23 percent in 1983. According to surveys by the America’s Health Insurance Plans and the Kaiser Family Foundation, employer-based premiums increased 18 percent in 1989 and 14 percent in 1990 and 2003. There have also been years of price stability such as a 1 percent increase in 1996, a 2 percent in 1997, and the 3-4 percent increases for the years 2012-2016 mentioned earlier.

In 2017, the ACA reinsurance program, which covers expenses for individuals incurring claims expenses above $90,000, closes. Reinsurance reduces the financial risk for insurers. The American Academy of Actuaries estimates the termination of the reinsurance program will increase premiums by 4-7 percent.

Could This Be The Underwriting Cycle?

The underwriting cycle was the pattern of profitability and pricing that the health insurance industry experienced from 1965-1991. Three years of financial gains were followed by three years of financial losses. Two years after insurers first incurred financial losses, premiums would spike and increase by double digit amounts for two more years, then cease increasing as profitability was restored.

Underlying the premium changes of 1965-1991 was this pattern: New market entry in the insurance market resulted in fierce price competition. Claims expenses rose more rapidly than premiums and many insurers suffered financial losses. Some insurers exited the market, and remaining insurers increased premiums substantially. After a few years, profitability returned and premium increases moderated.

If one views the current situation in the Marketplaces as the underwriting cycle, then when profitability is restored, many insurers should return to the Marketplaces and price stability should be reestablished. But if the current situation is due to a sicker-than-expected population leading to ongoing adverse selection, then the Marketplaces are unsustainable without legislative fixes.

Could This Be A Death Spiral?

In a death spiral, a plan or market attracts a disproportionate number of high-cost persons, which in turn forces insurers to raise premiums. Healthy persons are then disinclined to choose the plan, so medical expenses per person grow larger; this leads to higher premiums, which in turn leads to more adverse selection. The New Jersey state individual market in the 1990s provides a classic example of a death spiral. The state established pure community rating without an individual mandate; healthy households dropped coverage in the individual market; and premiums rose substantially.

Conventional wisdom holds that the population enrolled on the exchanges is sicker than expected, but supporting data is scant. The most cited evidence is that 28 percent of 2016 enrollment is under the age of 35, and analysts have estimated that at least 35 percent should be 18-35 for the risk pools to work. The Society of Actuaries (SOA) estimates from an analysis of medical claims data that the average risk score in 2015 for an enrollee with Marketplace coverage was 26 percent greater than for persons holding small group coverage.

The SOA also notes that the average risk score increased 5 percent from 2014 to 2015, but some of this increase may be attributable to better coding of diagnoses by plans. The Centers for Medicare and Medicaid Services (CMS) reports that medical claims expenses per member per month declined by 0.1 percent from 2014-2015, suggesting that the average risk score likely declined.

Unfortunately, researchers do not possess medical claims data from individual insurers within each state. Such data are necessary to unlock the mystery as to why some insurers—Kaiser Permanente, Molina, Centene, some Blue Cross/Blue Shield plans, and other Medicaid managed care plans—achieve profitability, while others do not. Such data would reveal if these plans are profitable due to favorable risk selection, better coordination of care and effective utilization management, or larger discounts from network providers. Many analysts believe that narrow networks are an important attribute for financial success because they lead to deeper provider discounts, more effective utilization management, and more favorable risk selection.

Subsidies for households earning less than 400 percent of poverty level income—roughly 87 percent of exchange enrollment—limit the possibility of a death spiral. The ratio of premiums for the benchmark second-lowest-cost silver plan to income determines the level of subsidies, so when premiums increase, subsidies increase also, thereby protecting enrollees. Unprotected are individuals earning more than 400 percent of poverty level income who enjoy no subsidy, and the majority of these individuals buy their coverage outside the exchange.

Massachusetts Experience

Comprehensive health reform in Massachusetts is 10 years old, and the ACA is largely modeled after it. The Massachusetts Exchange, called the Connector, has not experienced a spike in premiums such as is occurring nationally now, nor has it faced an exodus of large insurers. A few factors may explain differences between Massachusetts and the current situation with the ACA marketplaces.

Most plans operating in the state are regional plans based in Massachusetts. Massachusetts imposed a larger penalty for persons forgoing coverage than the ACA does. The ACA penalty is not more than 2 percent of annual income, whereas Massachusetts limits the penalty to no more than 50 percent for the average cost of single coverage — about $300 a month currently.

Corrective Measures

Both the underwriting cycle and a sicker-than-expected risk pool likely are contributing factors to the current spike in marketplace premiums. The underwriting cycle, if allowed to, should fix itself. If insurers have overestimated claims expenses for 2017, then under the ACA they are required to return a portion of their profits if their loss ratios are less than 0.85. (The loss ratio is the ratio of plan benefits and quality improvement efforts to earned premiums.) Premium growth for 2018 might be modest. Many insurers would be expected to re-enter the marketplaces once premiums are set at a level that permits modest profits.

Part of the answer to fixing the individual market lies in an understanding of the current uninsured. From a national sample of more than 4,000 Americans, Sara Collins and colleagues found that the remaining uninsured are increasingly Latinos (40 percent), and that 39 percent of the uninsured have incomes below the federal poverty level. Young adults constitute 48 percent of the uninsured, of whom 96 percent are eligible for Marketplace subsidies or Medicaid; nearly 40 percent of this group would be eligible for Medicaid but are not because they live in states that have not expanded the program. Yet, most of the uninsured say that coverage is unaffordable, and 38 percent are unaware of the Marketplaces.

Employer-based health insurance in the United States has been around for about 70 years. Since the 1970s, the share of the nation’s workers obtaining coverage from their employer has experienced a long-term decline. With contract workers constituting an increasing share of the workforce, more Americans in the future will need a vibrant individual market.