Marketplace Pulse: Leaky Risk Pools Sink Markets

Insurance Marketability Map

While the narrative arc of repeal and replace efforts came to a peak several weeks ago, individual market participants still lack resolution on several important issues, most notably the payment of cost sharing reductions. Although the bare county issue has improved considerably of late, there is ongoing concern about insurer participation as exits may occur up until late September.

These recent woes compound prior unfortunate events such as non-payment of risk corridors, extension of transition products, and underpricing followed in many markets by large premium increases. Overall, the individual market has ended up being smaller and less healthy than was hoped. The market shrank by an estimated 10 percent between 2016 and 2017, and recent survey data suggest that the uninsured rate crept up in Q1 2017.

However, despite the national nature of these challenges, market performance has varied greatly by state. For example, widely reported insurer exit and premium increases prompted Iowa’s insurance department to craft an emergency stopgap plan, while New Mexico, on the other hand, will most likely have four insurers in every county.

What factors might contribute to this variation? There are important state characteristics that may be risk factors for adverse market outcomes. Some risk factors are things that states cannot control, while others reflect policy decisions. The graphic above shows the association between adverse market outcomes such as low carrier participation and premium increases, and potential state risk factors. This list of outcomes and potential risk factors is intended to be illustrative, and by no means causal, but a cursory look at these data suggest that state actions may make a difference.

Two Kinds of Risk Factors


There are some things states can do little to avoid, like having a small, low-income population in relatively poor health, and a shortage of health care providers. These characteristics often describe rural states. In sparsely populated locales it is difficult to create provider networks, which is an important feature of insurer strategy. If the overall size of the market is small, insurers may find participation less desirable. Just as consumers in rural areas have fewer choices and pay higher prices in markets for many goods and services, health insurance markets in these areas may be similarly disadvantaged.


There are another set of risk factors that states directly determine through their policy decisions. These decisions affect the quality and stability of the risk pool in the individual market. One of the key policies in this category is the decision to permit transitional or “grandmothered” individual market products to persist. Another significant policy decision is to permit sales of limited coverage, such as short term or disease-specific “mini-med” plans. These policies facilitate “leakage” and erode the risk pool for ACA compliant plans. The magnitude of the effect is unknown, and depends on what enrollees in limited products would do in the absence of such opportunities, but the directionality seems clear, although of course some might argue that it is reversed, and other unmeasured factors weaken the market and create a strong demand for limited products. Finally, the decision to not expand Medicaid is also believed to adversely affect the risk pool by adding more expensive consumers to the individual market.


A few takeaways are clear:

  • Risk factors matter:
    • There were nearly no adverse outcomes in the absence of risk factors. In the 10 states with no adverse outcomes, eight had either zero or one risk factor.
    • On the other hand, of the nine states that had the greatest number of adverse market outcomes, six have at least four risk factors.
  • Policy factors seem to matter more than structural factors:
    • Of the nine states that had the maximum number of adverse market outcomes, eight exhibited two or more of the three policy risk factors.
    • There are a number of states, such as New Mexico or Nevada, that have high structural risk factors and low or no policy risk factors, and few adverse market outcomes.
  • Permitting limited coverage and transition products matter the most:
    • Of the states that have at least two adverse market outcomes, more than 70 percent have policies that permit both sales of limited coverage policies and transition products.
    • Practically no state that permits short term and transition plans escapes adverse market outcomes. Of the states that permit short term and transition plans, only one state (Arkansas) had no adverse market outcomes, and this state is atypical because its Medicaid expansion population is on the marketplace.

The Path Forward

While there are risks to the stability of their markets that states cannot well control, one important route to adverse market outcomes may be state policy decisions. There are frequent calls for more state flexibility, but these data suggest that the exercise of existing state flexibility is one way that states have visited a considerable amount of trouble upon their markets. Yet there is a hopeful note here as well, since this suggests that there are steps that states can take to improve their situation. While the appropriate degree of circumspection is warranted, and certainly there is much room to better understand the size and nature of the relationship between state policy decisions and market outcomes, the data shown here suggest that states may be able to improve their individual markets by eliminating transition products and restricting the sale of limited coverage.

John Palatucci at the Center for State Health Policy at Rutgers University provided helpful research assistance.