States Should Take Additional Steps to Limit Adverse Selection Among Health Plans in an Exchange
June 28, 2011
Health insurance exchanges, as envisioned under the Affordable Care Act (ACA), are intended to make an array of different coverage options available to individuals and small businesses. Exchange plans will be offered at several coverage levels (Bronze, Silver, Gold and Platinum, plus a catastrophic plan open to certain people), with Platinum plans having the most comprehensive benefits and lowest cost-sharing and Bronze plans having the least comprehensive coverage. Such a structure inevitably risks adverse selection that could threaten the viability of the exchanges.
Adverse selection could occur among the exchange plans because sicker people, who cost more to cover, would tend to choose more comprehensive coverage options because they have greater health care needs. Meanwhile, healthier people would tend to enroll in less comprehensive plans with lower premium costs. As sicker people disproportionately enroll in the more comprehensive plans, premiums for such options would rise and could become increasingly unaffordable. The rise in premiums would cause healthier individuals in the more comprehensive plans to abandon them for less costly coverage, leaving behind an even more concentrated group of sicker, higher-cost people and thereby driving premium costs up further. Over time, insurers could decide not to offer the more comprehensive plans at all.
In addition, adverse selection could occur among plans within the same coverage level if certain insurers are able to use benefit design or other methods to entice healthier-than-average individuals to enroll in their plans and to deter those who are in poorer health, relative to other plans offered at that coverage level.
Several elements of the ACA, including the use of risk adjustment, will help limit the risk of adverse selection among plans within an exchange. But they are unlikely to be sufficient on their own. This paper recommends additional steps that states can take, within the flexibility available under the Affordable Care Act, to further limit the risk that adverse selection poses to the long-term success of states’ exchanges and to provide a meaningful choice of health insurance plans for individuals and small businesses.
Experience Shows the Need to Protect Against Adverse Selection Among Plans
Numerous analysts have warned that adverse selection could threaten the long-term viability of the health insurance exchanges if healthier individuals disproportionately enroll in plans offered outside the exchanges and sicker-than-average people enroll in the exchanges. States can take a number of steps to limit these risks.
What has received less attention, however, is the risk of adverse selection among plans within the exchanges. Qualified health plans offered through the exchanges have to be provided within several coverage levels (Bronze, Silver, Gold and Platinum). People with greater health care needs would be more likely to enroll in more comprehensive plans (Gold or Platinum), while healthier-than-average individuals would tend to enroll in less costly, less comprehensive plans (Bronze). That, in turn, could drive up premiums for the more comprehensive plans and make them increasingly unaffordable. Over time, insurers may be unwilling to offer any plans at the most comprehensive coverage level.
Adverse selection could also occur among plans within the same coverage level, if insurers are able to entice healthy people to enroll and deter those in poorer health, relative to other plans offered at that coverage level. For example, plans offered at the Silver level have to provide coverage with the same actuarial value (70 percent), but as long as they satisfy that overall actuarial value, plans can vary the specific cost-sharing amounts they charge for certain services. This variation could raise the risk of adverse selection if deductibles and co-payments are set in a way to attract those who are healthier-than-average and discourage people with greater medical needs.
This is not merely a theoretical concern. Severe adverse selection occurred in the past among plans offered to federal workers and retirees through the Federal Employees Health Benefits Program (FEHBP). That program offers multiple private insurance plans, and the federal government (as the employer) subsidizes the cost of coverage. In the 1980s and 1990s, premiums for certain plans offered through FEHBP rose because higher-cost people disproportionately enrolled in them. For example, Blue Cross and Blue Shield offered both a “high-option” plan and a “low-option” plan, named to reflect the comprehensiveness of the benefits. As the high-option plan increasingly attracted sicker enrollees, its annual premium cost grew to be $2,800 higher than the premium for the low-option plan by 1994, even though the difference in the actuarial value of the two plans’ benefits was only about $80. Eventually, the cost for the high-option plan climbed so much that Blue Cross stopped offering the high-option plan altogether in 2002. 
Medicare’s prescription drug plans have also experienced adverse selection, with more costly beneficiaries tending to sign up for certain plans, according to researchers from the Centers for Medicare and Medicaid Services. For example, the standard Medicare drug benefit design includes a “coverage gap,” where enrollees receive no coverage after their drug spending exceeds a threshold and before spending reaches a catastrophic level. (The Affordable Care Act will close this coverage gap, known as the “doughnut hole,” over time.) Some Medicare drug plans adopted alternative designs that included some coverage within the gap and ended up attracting enrollees who were in poorer health and had more chronic conditions and higher drug spending. The Medicare drug benefit program includes a number of features that are similar to those included for exchange plans under the ACA and are meant to minimize the impact of adverse selection on premiums — including risk adjustment, reinsurance, and risk corridors.  Nevertheless, the CMS researchers wrote that the selection patterns they found “could adversely affect future Medicare costs and should be watched carefully.”
Another example comes from Medicare Advantage, in which private plans offer coverage to seniors and people with disabilities as an alternative to traditional fee-for-service Medicare. Under Medicare Advantage, plans must cover the benefits available under Parts A and B of traditional Medicare (Part A covers inpatient services such as hospital care, while Part B covers outpatient services such as physician visits). The benefit package that a Medicare Advantage plan offers can vary from the traditional program’s benefit package as long as its overall actuarial value is not less than that of traditional Medicare. This means that Medicare Advantage plans have the flexibility to offer additional benefits, while scaling back coverage for other services or treatments (relative to the traditional Medicare benefit package), within certain constraints. 
Some private insurers have used this flexibility to design the benefits packages they offer so as to entice healthy Medicare beneficiaries, who are less costly to treat, while deterring sicker and more costly beneficiaries from enrolling. For example, some private plans scaled back certain Medicare benefits used primarily by sicker individuals by imposing substantially higher copayment charges for days in the hospital or costly treatments like chemotherapy. Analyses by the Medicare Rights Center, the Commonwealth Fund, the Kaiser Family Foundation, MedPAC, and the Government Accountability Office (GAO) have documented that Medicare Advantage beneficiaries who need hospital care, home health care, or other specialty services may end up paying more out-of-pocket for such services, or receiving less of a covered service, under the Medicare Advantage plans’ benefit packages than under fee-for-service Medicare. Eventually, CMS raised concerns that some Medicare Advantage plans were designing their scope of benefits and cost-sharing amounts to discourage higher-cost beneficiaries from enrolling. The agency has since taken steps to prevent such tactics and is now implementing an ACA provision that constrains Medicare Advantage plans from charging higher cost-sharing amounts than fee-for-service Medicare does.
None of these examples perfectly mirrors the potential risk of adverse selection among plans within an insurance exchange. However, they show the tendency of people to gravitate toward particular types of plans based on their health status or expected costs and the resulting negative consequences.
ACA Includes Tools to Protect Against Adverse Selection, But Substantial Risks Remain
The ACA includes several provisions that would help protect against adverse selection among qualified health plans offered through an exchange. Some of these provisions are intended to help protect insurers whose enrollee pools are less healthy, and hence more costly, than average. This can also help reduce the incentives for insurers to try to attract healthier people using benefit design, marketing, or other tactics.
For example, the ACA requires states to operate a risk adjustment system, in accordance with federal standards, that will distribute funds from health plans whose enrollees are of below-average risk (i.e., healthier and hence lower cost) to those plans that attract enrollees with higher-than-average risk (i.e., enrollees who are sicker and have higher medical needs and costs). In addition, the ACA requires state exchanges to hold an annual open enrollment period, much like the open enrollment periods for employer-sponsored plans. Typically, people will not be permitted to switch or join plans outside of the annual enrollment period, unless they experience certain changes in life circumstances such as a marriage or a job change that results in a loss of coverage. This protects insurers by preventing people who have less comprehensive coverage from switching to a more comprehensive plan just before they get an expensive surgery.
It is worth noting that the FEHBP, unlike Medicare Advantage, does not utilize risk adjustment, which likely would have helped mitigate the adverse selection problems that the program experienced in the 1980s and 1990s. However, risk adjustment alone is unlikely to be sufficient. The Congressional Budget Office (CBO) has noted that existing risk-adjustment systems “tend to overpredict the costs of beneficiaries who end up with low health care spending and to underpredict the costs of those who end up with high health spending.” In other words, risk adjustment is very difficult to do accurately enough that insurers are fully compensated based on the health of the populations they enroll. Even when done well, risk adjustment generally accounts for some, but not all, of the differences in health costs between healthier and less-healthy groups of beneficiaries. 
The Medicare Advantage program does utilize a risk adjustment system that is considered highly sophisticated and continues to evolve over time in an effort to improve the accuracy of payments to plans. But the system is not fully effective. If it were, it would have eliminated or sufficiently minimized the incentive for insurers to try to design their benefits or use other tactics to attract healthier enrollees and deter people with significant health needs. This has not been the case. Some Medicare Advantage insurers have continued to design their plans in ways that place limits on, or erect barriers to, certain services or treatment that people with serious health problems may need. Just as the federal government has had to take steps beyond risk adjustment to address this problem, states setting up exchanges will have to do so as well.
Other provisions of the ACA also help protect against adverse selection by restricting insurers’ ability to offer products in ways that would “cherry pick” healthier customers. For example, all plans in an exchange (as well as “non-grandfathered” plans in the outside individual and small-group markets) are required to offer a set of “essential health benefits,” to be defined at the federal level. This is intended to assure that all exchange plans at all coverage levels cover a consistent range of items and services.
Another provision of the Affordable Care Act requires all insurers participating in exchanges to offer plans in at least the Silver and Gold coverage levels. These are relatively high levels of coverage, with actuarial values of 70 percent and 80 percent respectively. This requirement will prevent insurers from only offering products in the lower coverage levels within an exchange, such as the Bronze level, with a 60 percent actuarial value, and catastrophic plans that will be available to people younger than age 30 (and certain other people). In addition, the ACA has a “single risk pool” requirement, under which each insurer will have to treat all of its enrollees in the individual market (inside and outside the exchange) as a single group when setting premiums. An insurer must also treat all its small group enrollees (inside and outside the exchange) as a single pool. In states where the individual and small group markets are merged, the pooling would occur across both markets. This should reduce the premium increases that would occur in more comprehensive plans if they were priced separately from plans with healthier enrollee pools.
Furthermore, the amount of the federally funded premium tax credits in the ACA for low- and moderate-income enrollees is linked to the cost of the second-lowest cost plan in the Silver level. And the cost-sharing subsidies that will be available to enrollees with incomes below 250 percent of the poverty line will be available only in connection with Silver plans. These features are likely to help attract a broader group of people to these plans, some of whom otherwise might have sought coverage in a lower tier (Bronze or catastrophic).
Nevertheless, significant risk of adverse selection among plans remains even with the protections in the ACA. The essential benefits and actuarial value requirements will create some consistency in benefit design from plan to plan. But insurers are likely to continue to have significant ability to vary their plans in numerous ways and may do so in ways that heighten the risk of adverse selection. For example, provider networks, utilization review practices, and prescription drug formularies are likely to differ across plans, and insurers could try to structure these features to attract healthier people and deter enrollment by those with higher costs. Also, insurers in an exchange are not required to offer products in the Platinum coverage level, the most comprehensive of ACA’s coverage levels with a 90 percent actuarial value. This means that the type of coverage most likely to attract high-cost people does not have to be offered at all; insurers could offer plans only in the lower levels, and as a result, their enrollee pools would not include any enrollees in Platinum coverage.
Finally, while exchange plans are supposed to be barred from employing marketing practices or benefit designs that have the effect of deterring enrollment by people with significant health care needs, such standards will be difficult to enforce. If insurers still are able to use the design of their products (either the benefits or the cost-sharing charges), as well as marketing tools, in much the same ways that Medicare Advantage insurers have in the past, they likely will be able to some extent to lure healthier people to their plans or to deter enrollment by those who are sicker, which will cause adverse selection. That could drive up costs for certain plans, making them less affordable or unaffordable, or drive insurers offering such plans out of the exchange entirely.
States Should Adopt Additional Protections Against Adverse Selection in Exchanges
An important purpose of the exchanges is to ensure a choice of a variety of coverage options so that individuals and families looking for health insurance can find insurance that best meets their needs. Additional protections against adverse selection should aim to make the system work as transparently as possible, so when people sort themselves into different plans based on their health status, insurers are compensated appropriately and risk is pooled effectively.
The recommendations below would increase consistency and transparency across plans. To be sure, that would increase consumers’ ability to select coverage based on health status or other factors. But it makes sense to allow people to make rational decisions, and then to use structural and other mechanisms to protect insurers, exchanges, and consumers from negative effects of adverse selection.
States establishing an exchange should consider the following steps to guard against the risk and effects of adverse selection among plans within an exchange:
- Establish consistent Essential Health Benefits. The federal government is expected to publish regulations regarding the Essential Health Benefits (EHB) that all exchange plans (as well as non-grandfathered plans in the individual and small-group markets outside the exchange) will be required to offer. The federal rules about essential health benefits also are expected to address the scope of the benefits these plans must cover, by placing some parameters around how extensive coverage of particular benefits should be. It is unclear, however, how specific or detailed the federal EHB rules will be; they could specify, for example, how many physician visits or outpatient mental health visits at a minimum must be covered each year, define what types of limitations are unacceptable for particular benefits, or they could take a different approach. To the extent that important differences in covered benefits could still be present among plans, states should consider more detailed, standardized benefit rules so that insurers cannot craft plans that scale back coverage in some areas important to people with significant health needs while expanding coverage in other areas more attractive to healthy people (in order to meet the ACA’s required actuarial value targets for a plan’s coverage level).
- For example, unless federal or state rules prevent it, a plan could limit the number of hospital days that it covers each year and still reach the required actuarial value target for its coverage level by increasing coverage in another area, such as primary care services. States setting up an exchange should consider going further, if necessary, by adding specificity and detail to benefit requirements to ensure there is one common standard in the state regarding what benefits are covered, what services and items fall within benefit categories, and what coverage limitations (e.g., limits on the number of hospital or physician visits or on the number of hospital days covered) are allowed. Insurers still could vary their products in a number of areas, such as by having differing provider networks, drug formularies, and utilization management techniques.
- Create a menu of cost-sharing options for each coverage level. This would help to narrow variation in cost-sharing design among plans within the same coverage level and would limit the ability of insurers to design plans within the same coverage level in ways that entice healthier enrollees and/or discourage sicker ones. It also would make plan differences more transparent to consumers and regulators. In addition, it would place less pressure on the risk adjustment system to address adverse selection and hence make that system more likely to be effective.
- Require insurers that want to operate in an exchange to offer products in all exchange coverage levels. This would ensure that Platinum plans are available and that insurers offering this coverage are not automatically attracting a disproportionate number of sicker enrollees compared to their competitors. A state adopting this requirement, however, should also apply it to insurers operating in the individual and small-group markets outside of the exchange, to avoid prompting adverse selection against the exchange from these outside markets.
- Conduct strong and ongoing enforcement and oversight. It will be crucial for states, as well as the federal government, to ensure that insurers do not act (in areas where insurers enjoy considerable flexibility such as the establishment of provider networks and the use of utilization review strategies) in ways that enable them to “cherry pick” healthier people or that otherwise contribute to adverse selection. In addition, regulators will need to ensure that plans abide by requirements in the ACA related to premium pricing, risk pooling, and actuarial value, while states and the federal government will need to ensure that risk adjustment works effectively and is modified as needed over time. States and the federal government will also need to undertake strong enforcement and oversight efforts, including ongoing monitoring of insurers and periodic audits of insurer data and examinations of rate filings. For example, they should carefully monitor changes in the relative health of enrollees in a plan over time to see if the plan’s marketing practices or benefit design changes have led to adverse selection. That would ensure more effective enforcement of the ACA’s prohibition against insurers employing marketing practices or benefit designs that have the effect of deterring enrollment by people with significant health care needs.
We note, however, that protecting against adverse selection within an exchange should not prevent consumers from making informed decisions based on their knowledge of their health care needs and other factors. It would be inappropriate, for example, to require people to remain in the same plan or coverage level for a number of years, to restrict their ability to move to higher coverage level during the next annual open enrollment period, or to charge an extra fee if they do so.
 Sarah Lueck, “States Should Structure Insurance Exchanges to Minimize Adverse Selection,” Center on Budget and Policy Priorities, August 17, 2010.
 The Federal Employees Health Benefits Program: What Lessons Can It Offer Policymakers?” National Health Policy Forum Issue Brief, March 12, 1998. See also, Leonard Burman, “Medical Savings Accounts and Adverse Selection,” Urban Institute, 1997 and Harry P Cain II, “Moving Medicare to the FEHB Model, Or How to Make an Elephant Fly,” Health Affairs, July/August 1999.
 Both reinsurance and risk corridors are intended to reduce the risk to insurers from higher-than-expected costs and to reduce the incentives to attract enrollment by a healthier, lower-cost population. The ACA’s temporary reinsurance provision requires insurers providing coverage in the individual market inside and outside the exchanges to contribute to a state reinsurance entity which will make payments to individual-market insurers that cover certain high-risk individuals. The ACA’s temporary risk corridor program provides payments to insurers offering exchange coverage if their costs are higher-than-expected (as measured by a target); those payments will be financed by contributions from plans whose costs are lower than expected.
 Gerald F. Riley, Jesse M. Levy, and Melissa A. Montgomery, “Adverse Selection in the Medicare Prescription Drug Program,” Health Affairs, November/December 2009.
 Edwin Park, “Informing the Debate about Curbing Medicare Advantage Overpayments,” Center on Budget and Policy Priorities, May 13, 2008.
 While Medicare generally prohibits Medicare Advantage plans from discriminating on the basis of health status, an analysis by the Commonwealth Fund found that that “compliance with this broad policy is not carefully defined and enforced by CMS....” Brian Biles, Lauren Hersch Nicholas and Stuart Guterman, “Medicare Beneficiary Out-of-Pocket Costs: Are Medicare Advantage Plans a Better Deal?” The Commonwealth Fund, May 2006.
 Medicare Rights Center, “Too Good to Be True: The Fine Print in Medicare Private Health Plan Benefits,” April 2007; Biles et al., op. cit; Medicare Payment Advisory Commission, “Report to the Congress: Benefit Design and Cost Sharing in Medicare Advantage Plans,” December 2004; Patricia Neuman, “Medicare Advantage: Key Issues and Implications for Beneficiaries,” Testimony before the House Budget Committee, Kaiser Family Foundation, June 28, 2007; and Government Accountability Office, “Medicare Advantage: Increased Spending Relative to Medicare Fee-for-Service May Not Always Reduce Beneficiary Out-of-Pocket Costs,” February 2008.
 Section 3202 of the ACA limits the cost-sharing charges that Medicare Advantage plans can impose for certain benefits (specifically, chemotherapy, renal dialysis, skilled nursing care, and other services as determined appropriate by the Health and Human Services Secretary) to no more than the cost-sharing charged for the same service under traditional Medicare. CMS has said that it would implement this provision by establishing acceptable cost-sharing levels for Part A and Part B services, above which a Medicare Advantage benefit design could be considered discriminatory. See 42 C.F.R. § 422.100.
 ACA Section 1343. See also Edwin Park, “Ensuring Effective Risk Adjustment,” Center on Budget and Policy Priorities, May 18, 2011. As noted, the Affordable Care Act also includes temporary risk corridor and reinsurance provisions to limit the risk of adverse selection.
 Congressional Budget Office, “Designing a Premium Support System for Medicare,” December 2006.
 The requirement to offer at least a Silver and Gold product applies only within an exchange, not to the individual or small-group markets outside exchanges. This could result in adverse selection against the exchange. See Lueck, op cit.
 See Lueck, op cit.
 Some insurers have supported such ideas, especially a proposal to prevent consumers from moving up more than one coverage level during the annual enrollment period. See comments from Aetna Inc., America’s Health Insurance Plans, and the Blue Cross and Blue Shield Association in response to the National Association of Insurance Commissioners’ draft white paper on adverse selection, April 4, 2011, http://www.naic.org/documents/committees_b_exchanges_comments_rcvd_110322_adverse_selection.pdf.