Turmoil in the Health Insurance Marketplaces

Issue Brief

Turmoil in the Health Insurance Marketplaces

What are the Policy Options? First in a New Issue Brief Series

SUMMARY: This brief explores the current volatility in the ACA’s Marketplaces and discusses key factors in their evolution over the past three years. As the law and Marketplaces stand now, continued proliferation of narrow network products and significant premium increases appear likely. The brief concludes with options for policymakers to address the turmoil in the Health Insurance Marketplaces. 

The first three years of the Affordable Care Act’s Health Insurance Marketplaces have been tumultuous ones, with rapid entry and exit of insurers and recent spikes in premiums. As concerns mount about the stability and viability of the Marketplaces, this brief provides some insight into the forces behind the headlines and presents six options for policymakers to consider.

Key ACA Features: Medical Underwriting Ban and Mandated Coverage

Two features of the ACA are key to understanding the dynamics of the Marketplaces:  the ban on medical underwriting and the mandate for coverage. 

The ACA prohibited plans from using medical underwriting to help determine premiums.  No longer would a 50-year-old with major health problems pay a higher premium than a healthy counterpart. As a result, people with pre-existing conditions faced premiums that were lower than what they paid before the ACA – if they could find coverage at all. The often unarticulated flip side is that the healthy 50-year-old faced higher premiums than before. Further, the law limited premium differences by age, so that premiums paid by the oldest couldn’t be more than three times that of the youngest. Thus, healthy 30-year-olds faced an average premium that reflected both the experience of the sicker 30-year-olds and the experience of the older, sicker enrollees. These high premiums gave healthier and younger people a strong incentive to forego coverage and sign up only when they became ill.

Anticipating this problem, the ACA imposed a mandate for coverage, with penalties to enforce the mandate. It also established a limited annual open-enrollment period to dissuade people from waiting until they were sick to enroll. In practice, the penalties were modest, and generous exceptions cushioned the impact of the limited enrollment window. As a consequence, the healthy and young faced blunted inducements to buy coverage.

Key Insurer Issues: Uncertainty and Selective Contracting

Insurers faced substantial uncertainty in this new market with new rules. [i]  Would people who had perhaps never had private health insurance actually buy coverage? To what extent would they make trade-offs between (often-subsidized) premiums and other features of coverage like deductibles and networks? Would the claims experience of these new enrollees track the insurer’s existing experience?

The specter of adverse selection loomed large. Would these new plans attract people with higher expected use of services, while those who believe themselves healthy would go without coverage? Concern about containing costs led insurers to offer more narrow network plans, selectively contracting with providers who would accept negotiated prices.

The use of selective contracting was one of the only cost containment tools left to insurers once medical underwriting had been banned and age-related rating was limited. Selective contracting was one of the most successful strategies used by managed care. With it, insurers contracted with certain hospitals and physician groups in a market area.[ii] These contracted providers become part of the managed care “network” of providers willing to accept negotiated prices. The essence of managed care success has been trading patient volume for price. To the extent that an insurer can direct more patients to a given provider, the provider is willing to accept a lower price.

Unfolding of Events 2014 - 2016

The first three open enrollment periods of the ACA reflected the uncertainty of a new insurance market.

Year One. Dominant insurers in each state offered coverage, often in every market in the state. [iii] Uncertain about how to price plans in this new market, but being risk averse, many insurers set premiums somewhat above the (inflation adjusted) premiums they had set in the individual market previously. Many smaller local and regional insurers, and even large national carriers with only a small presence in the local markets, set their premiums higher or decided to wait a year or two to see how the market would develop. It became apparent that new consumers were very conscious of premium differences and seemed to opt for low premiums even when somewhat higher priced policies with lower deductibles and broader networks may have served them better.[iv]  Moreover, enrollment was low; nationally roughly eight million people signed up. [v]   

Year Two.  Insurers still didn’t have very meaningful information on the utilization experience of new enrollees, but they learned two things from year one: first, that premiums matter, as lower premiums lead to higher enrollments, and second, that the number of covered lives was modest. The insurers on the periphery of the markets could experiment at bit. In Texas, for example, the number of non-BCBS plans offered in distinct counties increased by 75 percent as insurers in the marketplace expanded the number of plans they offered and/or the number of counties they served.[vi] Some carriers, like UnitedHealth Care and Assurant, entered selected markets. In other metro areas, hospital systems teamed up with national carriers to compete with the local Blue Cross plans. As this year played out, however, plans began to report losses. Assurant pulled out of all of the marketplaces in which it had newly offered coverage; UnitedHealth withdrew from a few of the markets it had entered. Other plans began to report losses.  Blue Cross Blue Shield of Texas, for example, reported a loss of $400 million on its individual market plans.  

Year Three. At the time they needed to set plans and premiums, insurers still didn’t have meaningful detailed information on the utilization experience. Some insurers responded to their losses by more aggressively turning to selective contracting. They stopped offering PPO plans with wide networks[vii]  and focused on HMOs and EPOs (exclusive provider organizations) with narrower networks and lower provider prices. Other carriers dropped their premiums relative to the dominant insurers, hoping that the lower premiums would attract healthier enrollees that would balance out the higher-cost people they feared they had attracted. 

However, by the summer of 2016 losses mounted and claims data were becoming available in sufficient magnitude to allow analysis. Actions in the Marketplaces came fast and furiously.[viii] Aetna announced it would pull out of 11 of 15 Marketplaces. Humana announced it was pulling out of eight states. Blue Cross Blue Shield effectively withdrew from its respective Marketplaces in Minnesota and Tennessee.[ix] Others announced that they would only offer HMO/EPO products, still other insurers asked for substantial rate increases. At the same time, several carriers that have focused on Medicaid populations, such as Molina, have reported success in the Marketplaces.  

What’s Likely to Happen Going Forward?

Insurers now have enough information to see that disproportionately high utilizers have driven their enrollments. Insurers who relied on a low-premium strategy discovered that they didn’t attract enough new healthy enrollees. Instead, they attracted price-sensitive unhealthy people from other insurers. If remaining insurers want to stay in the marketplace, they have two options:

  • Aggressively offer narrow network plans in the hopes of attracting the elusive healthy enrollees while cutting costs. This strategy has two components. First, more aggressive selective contracting will lower their claims costs. But, second, by excluding marquee health care providers such as big-name academic health centers and prominent specialists they can signal to unhealthy people that they should look elsewhere for care. This is a strategy that the traditional Medicaid managed care plans seem to be following in the Marketplaces. On the one hand they have many safety-net providers who appear willing to accept Medicaid payment rates, and these providers tend not to be attractive to less healthy individuals with some resources. The interesting question for them is whether they can extend their model to different populations, and whether they have the capacity to serve larger numbers of enrollees at a similar cost.

  • Raise premiums substantially to reflect the future losses they anticipate from disproportionately unhealthy enrollees. This too can work. Most enrollees are heavily subsidized and their out-of-pocket premiums don’t increase as long as their incomes don’t increase. A stable enrollment with high enough premiums will cover claims costs. However, it will not work if higher premiums lead a meaningful number of enrollees who are ineligible for subsidies to drop coverage. People who disenroll are likely to be the healthier ones. When they leave, the new premium may not cover the claims experience of those remaining, necessitating still higher premiums and more disenrollment. This is the classic insurance death spiral that ultimately leads to the insurer withdrawing from the Marketplace, and the insurance market collapsing. [x]

Policy Options to Consider

Policymakers have several options as they struggle with the viability of the ACA Marketplaces. Some of these can stand alone, while others could be implemented together.

            First, policymakers can do nothing.  If higher premiums cover the claims costs of enrollees and few people disenroll, the market may stabilize. Most Marketplace enrollees are subsidized and as such, their out-of-pocket premiums are essentially unaffected by increases in premiums. The costs fall largely on taxpayers, because the federal treasury will pay almost all of the higher premiums. But this approach may leave many relatively young and healthy Americans uninsured and subject to penalties for non-participation. It also leaves people without subsidies to face very high premiums. It may also leave many people with only a single insurer in their market area, and that insurer is likely to offer only a narrow network of providers. 

            Second, policymakers could increase the penalties for non-enrollment. The penalty for not having health insurance coverage is essentially 2.5 percent of income, only collected when someone is entitled to a federal income tax refund. This is often less than the cost of coverage, and may not be strong enough to induce healthy and younger people to sign up. Economist Uwe Reinhardt notes that European countries that mandate private coverage, such as Switzerland, impose much higher penalties and can garnish wages. [xi]  Higher penalties, rigorously imposed, would increase compliance with the mandate, but might also alienate a significant number of young and healthy voters.

            Third, policymakers could expand the age corridor of premiums and tighten up waivers for late enrollment. Under the ACA, the premium for the oldest enrollee can be no more than three times the premium of the youngest. Many actuaries have argued that the difference between the youngest and oldest is actually something between five and six times. [xii]  By allowing a broader premium corridor, younger people will see much lower premiums and many of them may opt to buy coverage. The downside is that older people will face higher premiums.

            Under the ACA’s waiver program some people who did not buy coverage during the open enrollment period are allowed to do so later in the year. The concern of insurers has been adverse selection. Those who wait to enroll disproportionately enroll because they have a health event. They are high utilizers almost by definition. Further restricting the circumstances under which one can get a waiver, beyond what has already been done, would encourage more risk averse people to buy coverage during open enrollment. Those who get sick later would be closed out.

            Fourth, policymakers could reintroduce the reinsurance program for insurers and/or modify risk adjustment to eliminate the budget neutrality provision. The ACA created mechanisms to provide short-term subsidies to insurers if their claims experience was greater than anticipated. These mechanisms expire at the end of 2016.  Although the drafters of the ACA understood that the new niche insurance markets would be uncertain, it appears they didn’t appreciate the extent of the uncertainty. One approach to stabilizing the Marketplaces would be to reintroduce these subsidies, probably at a more generous level. The ongoing risk adjustment program is budget neutral. The much higher level of morbidity in the enrolled populations across most carriers meant that there wasn’t enough revenue from carriers with low claims experience to offset the losses of carriers with high-utilization experience. Sufficiently large subsidies would reduce future premium increases and could encourage re-entry of insurers who left the market. The downside, again, is that taxpayers face higher liabilities.

            Fifth, policymakers could introduce a public option. A “public option” can take many forms; one approach would allow Marketplace customers to sign up for Medicare Parts A, B, and D. [xiii] Of the many questions that arise, the first is how to set premiums. One could require this new Medicare option to set premiums in the same fashion as current ACA Marketplace insurers. However, the Medicare Option would be price-competitive with private insurers only if it had cost advantages over private insurers. Currently, Medicare’s key advantage is that it pays hospitals and physicians less than private insurers do. Would hospitals and physicians accept these payments for the greater volume? If the payment levels are set nonetheless, care might be rationed among traditional Medicare and new ACA-sponsored enrollees. On the other hand, if the Medicare Option agrees to pay providers more, whatever cost advantage it may have over private payers would be at least partially eroded.

            Alternatively, ACA-sponsored enrollees could pay the same low prices as current Medicare beneficiaries. The current premium for Part A coverage for those who have not worked the 40 covered quarters is $411 per month. Part B premiums are $104 per month and Part D are about $40 per month. However, Parts B and D are subsidized at about 75 percent from the federal treasury, and the Part A premium is almost certainly subsidized as well. This approach would most likely dominate the private insurer offerings, but at the cost of substantially more federal spending.

            Sixth, policymakers can introduce a high-risk pool to the ACA Marketplace.  This approach harkens back to the high-risk pools used by most of the states prior to the ACA as well as to the transitional “Pre-Existing Condition Insurance Plan” used prior to the first open enrollment period. [xiv] People with serious health conditions would get coverage through the high-risk pool. They would pay no more for the coverage than they would have under existing ACA Marketplace plans. However, federal taxpayers would directly pay for the substantial losses that occur in this plan. The advantage of this approach is that it would allow private insurers to compete for enrollees who were of approximately the same level of likely claims experience. 


            There are no simple solutions to the turmoil in the ACA Health Insurance Marketplaces. The problem arises because the ACA combined people with dissimilar health status into a common risk pool and attempted to set premiums based on the experience of the average enrollee. As a result, unhealthy (older) people faced lower premiums, but healthy (younger) people faced higher ones. If policymakers want to address the problem, and not abandon the Marketplaces, there is no way of getting around some tough choices:  young and healthy people pay less, older and unhealthy people pay more, and/or taxpayers pick up the cost of greater subsidies.


AUTHOR: Michael A. Morrisey, PhD, Department Head and Professor of Health Policy & Management, Texas A&M University and Visiting Scholar, Leonard Davis Institute of Health Economics, University of Pennsylvania.



[i] Troyen A. Brennan and David M. Studdert. “How Will Health Insurers Respond To New Rules Under Health Reform?” Health Affairs 29, no.6 (2010):1147-1151 doi: 10.1377/hlthaff.2010.0421

[ii] Michael A. Morrisey. Health Insurance, 2nd ed. Health Administration Press, 2014, pp: 175-196.

[iii] Kaiser Family Foundation analysis of MLR data from The Center for Consumer Information & Oversight (CCIIO), Centers for Medicare & Medicaid Services (CMS). Public Use File for 2014, as of October 7, 2015.

[iv] Michael A. Morrisey, Alice Rivlin, Richard Nathan, Caitlin Brandt. “Early Assessment of Competition in the Health Insurance Marketplace”, Brookings Institution. 2016.

[v] Larry Levitt, Cynthia Cox , and Gary Claxton. “Data Note: How Has the Individual Insurance Market Grown Under the Affordable Care Act?" Kaiser Family Foundation. May 12, 2015

[vi] Michael A. Morrisey, Alice Rivlin, Richard Nathan, Caitlin Brandt. “Early Assessment of Competition in the Health Insurance Marketplace”, Brookings Institution. 2016.

[vii] Blue Cross Blue Shield Association. “Evolving Affordable Care Act Marketplaces: The 2015 to 2016 Transition”, The Health of America Report. 2016.

[ix] Andrew Ogles and Luke Hilgemann. “ObamaCare’s Meltdown Has Arrived”, Wall Street Journal. Oct. 6, 2016. http://www.wsj.com/articles/obamacares-meltdown-has-arrived-1475709560

[x] Wendell Potter. “Anatomy of a true health insurance death spiral” Healthinsurance.org. March 4, 2015. https://www.healthinsurance.org/blog/2015/03/04/anatomy-of-a-true-health-insurance-death-spiral/

[xi] Sarah Kliff. “This Princeton health economist thinks Obamacare’s marketplaces are doomed” Vox.  August 25, 2016. http://www.vox.com/2016/8/25/12630214/obamacare-marketplaces-death-spiral?

[xii] Lynn Nonnemaker. “Beyond Age Rating: Spreading Risk in Health Insurance Markets”, AARP Public Policy Institute, Insight on the Issues I35, 2009.

[xiii] Linda J. Blumberg and John Holahan. “Designing a Medicare Buy-In and a Public Plan Marketplace Option” Urban Institute. 2016.

[xiv] The Center for Consumer Information & Insurance Oversight, “Pre-Existing Condition Insurance Plan” Accessed October 2016. https://www.cms.gov/CCIIO/Programs-and-Initiatives/Insurance-Programs/Pre-Existing-Condition-Insurance-Plan.html