Stabilizing Individual Health Insurance Markets With Subsidized Reinsurance

Issue Brief

Stabilizing Individual Health Insurance Markets With Subsidized Reinsurance

A Comparison of Approaches

Scott E. Harrington, PhD

SUMMARY: This brief describes alternative forms of subsidized reinsurance and the mechanisms by which they spread risk and reduce premiums. It summarizes specific state initiatives and Congressional proposals that include subsidized reinsurance. It compares approaches to each other and to more direct subsidies of individual market enrollment. For a given amount of funding, a particular program’s efficacy will depend on how it affects insurers’ risk and the risk margins built into premiums, incentives for selecting or avoiding risks, incentives for coordinating and managing care, and the costs and complexity of administration. These effects warrant careful consideration by policymakers as they consider measures to achieve stability in the individual market in the long term.


The Affordable Care Act (ACA) fundamentally transformed individual health insurance markets and significantly expanded coverage. But markets in many states remain unstable, partly due to disproportionate enrollment by older and sicker enrollees. While ACA subsidies to households below 400% of poverty (Advance Premium Tax Credits, or APTCs) limit the impact of rising premiums on most current enrollees, unbalanced risk pools and resulting premiums discourage take up of coverage by many people eligible for little or no subsidy. Total individual market enrollment is much lower than projected when the ACA was passed, and it declined in 2017 as premiums jumped and insurers withdrew in many states.1

Federal and state officials are considering stabilizing the individual market through some type of subsidized reinsurance program, including so-called “invisible high risk pools.”2 The House’s American Health Care Act, the Senate’s Better Care Reconciliation Act, and several state waiver requests under Section 1332 of the ACA all include subsidized reinsurance provisions.3 Subsidized reinsurance combines significant external subsidies to the individual market with some degree of risk spreading through reinsurance. It contrasts with commercial reinsurance arrangements (and certain government authorized or mandated reinsurance programs), which spread risk among and between insurers without an external subsidy, and with possible direct subsidization of the individual market without reinsurance.

Traditional high risk pools operated in most states before 2014, when insurers could underwrite and price coverage based on health status. People unable to obtain coverage due to pre-existing conditions could do so in a separate risk pool, typically at higher than “standard” but subsidized premium rates, thus insulating the underwritten market from the cost of insuring these high risks.4 Under the ACA’s current structure, some form of supplyside subsidy to the individual market could serve a similar role, lowering premiums and increasing take up of coverage by younger/ healthier enrollees, thus putting further downward pressure on premiums.

This issue brief describes alternative forms of subsidized reinsurance to help stabilize individual health insurance, and compares them with possible direct subsidies of individual market enrollment. It summarizes specific state initiatives and Congressional proposals that include subsidized reinsurance programs. Assuming a comparable amount of external funding, it then discusses the potential effects of alternative approaches on insurers’ risk and risk margins in premiums; incentives for risk selection; incentives for care management, coordination, and cost control; and the costs and complexity of administration.

REINSURANCE AS A SUBSIDY MECHANISM
Under the ACA’s system of guaranteed issue and community rating, insurers price coverage based on average expected claim costs for the overall risk pool, with enrollee premiums tailored to age and type of plan. External funding through some form of reinsurance will reduce insurers’ average expected costs (net of the funding), with premium reductions spread broadly among enrollees. For a given amount of funding, a particular program’s effects will depend on its impact on the cost of coverage, and the ability of lower premiums to attract a healthier mix of enrollees and expand overall enrollment to help achieve economies of scale in non-claim costs.

Table 1 summarizes key features of three approaches to providing subsidized reinsurance, along with the alternative of providing direct per enrollee subsidies.

Per enrollee reinsurance (also known as “stop loss” or “specific excess” reinsurance), reimburses a percentage of an enrollee’s annual claim costs between a threshold and a maximum claim amount. It was used in the ACA’s transitional reinsurance program during 2014-16 and remains a part of Medicare Part D.5

Invisible high risk pools (a form of conditionbased reinsurance) reimburse a percentage of claims above a threshold for all enrollees with specified health conditions. Programs may also give insurers the discretion to enroll others based on enrollee health status statements.

Aggregate coinsurance (also known as “quota share” reinsurance) simply reimburses a percentage of all claims during a year, regardless of size or underlying health condition. It is roughly analogous to federalstate sharing of costs under Medicaid, and to overall federal funding of Medicare Part D.

A per enrollee subsidy pays a specified flat amount per enrollee per month or year. It is analogous to basic federal payments (before risk adjustment) to support Medicare Advantage, and to per capita allotment proposals for federal Medicaid funding.

Potential external funding sources include federal or state government funds, assessments on insured and self-insured plan enrollees, and assessments on providers. Any assessments linked to the individual market or requiring individual market insurers to pay for reinsurance would reduce the impact on premiums. To the extent that external funding reduces premiums in general and specifically for second lowest cost Silver plans, federal APTCs will decline, thus reducing or possibly eliminating the net cost of additional funding. Such “pass through funding” based on projected federal savings from lower APTCs has been integral to reinsurance proposals in Section 1332 waiver applications.6

PER ENROLLEE REINSURANCE
Table 2 provides a more detailed summary of subsidized per enrollee reinsurance programs.7 This approach has been used to partially subsidize “catastrophic” claims under Medicare Part D since its inception. As implemented, the ACA’s transitional reinsurance program reimbursed 100% of enrollee claims in a year between $45,000 and $250,000 for 2014, 55.1% of claims between $45,000 and $250,000 for 2015, and a projected 52.9% of claims between $90,000 and $250,000 for 2016. The program was funded by assessments for all individual, group, and self-insured enrollees, thus providing a net subsidy to the individual market, which is believed to have lowered premiums by 10%-15% in 2014 and declined over time.

The American Health Care Act (H.R. 1628) passed by the House in May would provide substantial federal funding for a Patient and State Stability Fund during 2018-2026, with required state matching of funding beginning in 2020. While states could design alternative programs, the default stabilization program was per enrollee reinsurance, which would reimburse 75% of an enrollee’s annual claim costs between $50,000 and $350,000. The Senate Republican substitute bill, the Better Care Reconciliation Act, included substantial funding for a shortterm reinsurance program during 2018-2021 with no state matching required and with program details to be determined by the HHS Secretary. It included additional long-term funding for state programs that could include reinsurance, with some state funding eventually required.

On the Democrat side, Senators Carper, Kaine, and others introduced the Individual Health Marketplace Improvement Act (S. 1354) to establish federally funded per enrollee reinsurance that in 2018 would reimburse 80% of an enrollee’s annual claims between $50,000 ($100,000 in 2021) and $500,000. The thresholds would be indexed to growth in average premiums after 2018. The bill does not specify the amount of funding but indicates that all necessary funding will be provided to carry out the program.

At the state level, and in the face of large premium increases and declining enrollment, Minnesota used more than $300 million in emergency funding to provide 25% premium rebates to non-APTC eligible individual market enrollees for 2017. It subsequently applied for a Section 1332 waiver (pending) to establish the Minnesota Premium Security Plan.8 This reinsurance plan would reimburse 80% of an enrollee’s annual claims between $50,000 and $250,000 in 2018, with possible changes in the sharing percentage and thresholds after 2018. (The state’s application also indicates that it will “strongly consider moving to a conditions-based model” beginning in 2019.) The projected required funding of $271 million in 2018, projected to reduce premiums by 20%, includes estimated pass through savings of $139 million to $167 million from reduced federal APTCs and reduced federal funding of Minnesota’s Basic Health Plan, with the state funding the remainder.

Facing large premium increases and the possibility of not having any Marketplace insurers in 2018, Iowa applied to HHS for emergency funding in June to stabilize its individual market.9 The state’s Proposed Stopgap Measure would reimburse 85% of annual enrollee claims between $100,000 and $3 million and 100% of claims above $3 million. Sixty percent of the reimbursement for claims above $1 million would come from the Federal High-Cost Risk Pooling Program (see note 7). The application projects that an additional cost of $80 million in 2018 can be funded primarily by reduced federal APTCs. 

INVISIBLE HIGH RISK POOLS
Table 3 provides additional details on invisible high risk pools in Maine, Alaska, and the proposed American Health Care Act.10 Facing large premium increases and declining insurer participation in its individual market with guaranteed issue and age-adjusted community rating, in 2011 Maine established the Maine Guaranteed Access Reinsurance Association (MGARA) as an invisible high risk pool. MGARA operated from July 1, 2012 until the ACA’s transitional reinsurance began on January 1, 2014.11 Insurers had to reinsure (“cede”) all enrollees with one or more of eight health conditions at the time coverage was issued. Insurers could also voluntarily cede enrollees based on an enrollee health statement. MGARA reimbursed 90% of a reinsured enrollee’s annual claims between $7,500 and $32,500 and 100% of claims above $32,500. To fund the program, insurers had to cede 90% of the premium for reinsured enrollees, and all individual, group, and third-party administered self-insured health plans were assessed $4 per member per month.

The Maine program’s impact on premiums and coverage take up is uncertain.12 While the state’s largest insurer estimated that MGARA accounted for a 20% rate reduction, its adoption coincided with expansion of the state’s age rating band and changes in required benefits and allowed cost sharing. Previous enrollees were allowed to retain coverage at the old rating rules, thus resulting in old and new risk pools, with younger enrollees more likely to migrate to the new rules and pool.

Under the Alaska Reinsurance Program, which took effect January 1, 2017, insurers had to reinsure enrollees with one or more of 33 health conditions at issue or if diagnosed during the plan year. Unlike Maine, insurers could not voluntarily reinsure other enrollees. The program reimbursed 100% of all claims for reinsured enrollees during the plan year; insurers had to cede 100% of the premiums. The program was funded in 2017 with $55 million from general state premium taxes. Alaska subsequently received a Section 1332 waiver in July to continue the program in 2018. The waiver application projects that lower premiums will allow the program to be funded with $52 million in APTC pass through savings in 2018.13

An amendment to the American Health Care Act proposed by Representative Palmer in March would have established a Federal Invisible High Risk Pool (FIHRP) patterned after the Maine program. Insurers would cede enrollees with one or more of eight health conditions (or additional conditions specified by the HHS Secretary) and could voluntary cede other enrollees based on a health statement. Annual claims above $10,000 would be fully reimbursed. Distinctively, providers would be reimbursed for claims above this amount at Medicare rates (which, according to one actuarial study, would have a major impact on the proposal’s ability to reduce premiums).14 Insurers would cede 90% of premiums for reinsured enrollees, subject to a possible adjustment based on cost control mechanisms. No specific funding was proposed. The revised amendment ultimately included in the bill would simply have established a Federal Invisible Risk Sharing Program with all program details to be determined by the HHS Secretary and $15 billion in funding during 2018-2026.

COMPARISON
With external funding, all these approaches have the potential to lower insurers’ average costs net of the subsidy, lowering required premiums, encouraging greater and more balanced take up of coverage, and thus contributing to even lower costs and premiums. However, they do so in different ways. For example, by requiring insurers to cede most of the premiums for reinsured enrollees, an invisible high risk pool is likely to permit significantly greater reimbursement than per enrollee reinsurance, while aggregate coinsurance and per enrollee subsidies will likely fund relatively modest amounts of total claim costs. Hence, per enrollee reinsurance can provide significant reimbursement of large claims regardless of cause, invisible high risk pools can provide substantial reimbursement of claims at much lower thresholds for enrollees with some conditions, and aggregate coinsurance can provide relatively small reimbursement of all claims.

Given this context, Table 4 provides a qualitative and subjective comparison of the alternatives’ potential effects on insurers’ risk and risk margins in premiums, insurers’ incentives, and program administrative costs and complexity. The focus is on their relative strengths and weaknesses as stabilization mechanisms in a general environment of guaranteed issue, age-adjusted community rating, and risk adjustment. Political considerations or interactions with other potential changes to the ACA are not considered.

Insurers’ risk and premiums. Insurers face two broad types of risk that affect premiums: 1) random (idiosyncratic) variation in the frequency and severity of claims, and 2) pricing risk that arises because insurers need to forecast trends in medical costs 6-18 months or longer in the future. Idiosyncratic risk generally declines as an insurer covers more enrollees in given market or diversifies across geographic regions and types of coverage. Commercial reinsurance can also be purchased. In contrast, pricing or trend risk cannot be reduced simply by selling more coverage in a given market, but can be reduced by diversification across markets. Actuarial and economic models of insurance pricing imply that greater risk requires insurers to include larger risk margins in premiums. In addition to the direct impact of an external subsidy, reinsurance approaches may further reduce premiums by reducing insurers’ risk, although it may crowd out commercial reinsurance.

Per enrollee reinsurance will likely have the largest impact on insurers’ idiosyncratic risk by providing substantial reimbursement of relatively large claims regardless of cause. Capping reimbursement for very large claims helps preserve incentives for purchasing commercial reinsurance. In contrast, invisible high risk pools reimburse more frequent, relatively modest, and less risky claims associated with chronic conditions, and they substantially eliminate incentives to purchase commercial reinsurance for reinsured enrollees. At the same time, they provide no reimbursement for large idiosyncratic claims for other enrollees. By reinsuring only a small percentage of total claim costs, aggregate coinsurance will have relatively little impact on idiosyncratic risk. Likewise, per enrollee subsidies will have little or no impact.

Each approach could reduce trend risk if the subsidy adjusts over time to reflect overall growth in claims costs or premiums, but the magnitude would likely be modest. Invisible high risk pools could have some advantage if trend risk is above average for the specified health conditions (for example, due to unanticipated increases in the number of enrollees with those conditions, or introduction of new and expensive treatments).

Incentives for risk selection. The ACA’s permanent risk adjustment program reduces insurers’ incentives to avoid sicker-than-average enrollees (risk selection) and the financial risk of attracting a disproportionate number of such enrollees (adverse selection). However, risk adjustment is inherently imperfect and does not eliminate these incentives and risks.

Per enrollee reinsurance can further reduce these risks by partially reimbursing relatively large claims not fully predicted by the risk adjustment model. Aggregate coinsurance with percentage reimbursement will have relatively little impact. Per enrollee subsidies will have no effect unless the subsidy amounts are risk adjusted, as is done for Medicare Advantage, and which presumably would replace the current risk adjustment program.

Mandatory reinsurance of claim costs for enrollees with specified conditions under an invisible high risk pool would reduce selection incentives and reduce the need for risk adjustment for those conditions.15 It would require significant coordination with the risk adjustment program to reduce overlap and redundancy. Permitting insurers to cede enrollees voluntarily based on information provided in a health statement could supplement existing risk adjustment, if it targets enrollees that are likely to have higher costs than predicted by the risk adjustment model. An ability to reinsure those enrollees voluntarily would reduce incentives to avoid enrolling them in the first place.

Incentives for care management, coordination, and cost control. Any stabilization program that reimburses a portion of an insurer’s claim costs could reduce its incentives for care management, coordination, and cost control. Per enrollee subsidies have a clear advantage on this dimension. The disincentive effects of aggregate coinsurance could be modest assuming that the percentage of claims reimbursed is small, but aggregate coinsurance would affect incentives for all claims. Per enrollee reinsurance would affect incentives for relatively large claims up to the cap, but many of these claims involve acute injuries or illness for which care management, coordination, and cost control could be less important.

Invisible high risk pools fare poorly on this dimension. Substantial reimbursement of claims for enrollees with specified chronic conditions (or for enrollees voluntarily reinsured based on health statements) risks significantly undermining insurers’ incentives for care management, coordination, and cost control, perhaps including incentives to negotiate favorable terms with providers. This could require extensive auditing, care protocols, or other specialized programs to mitigate the resulting disincentives.

Requiring that providers be reimbursed at Medicare rates, as proposed in the original Palmer amendment, essentially requires participating providers to help finance the subsidy to the individual market. In addition to the question of how such an arrangement would be administered, it would significantly disrupt insurer-provider contracting, possibly reducing the willingness of providers to contract with individual market plans. It could also exert upward pressure on negotiated prices for services not subject to reimbursement at Medicare rates, increasing insurers’ costs for those services, and thus reducing the program’s ability to lower premiums and expand enrollment.

Administrative costs and complexity. All approaches have administrative and compliance costs. Greater complexity will produce higher costs and increase the potential for unintended consequences (including gaming and opportunism over time). Per enrollee reinsurance has a simple structure with relatively straightforward administration and compliance, with substantial prior experience under the ACA’s transitional reinsurance program. Aggregate coinsurance has a simple structure with similar administrative and compliance issues. Invisible high risk pools are considerably more complex in structure, which would increase administrative and compliance costs. The administration and complexity of per enrollee subsidies would depend on how the amount of subsidy was determined and updated over time. 

CONCLUSION
Subsidized reinsurance represents a potentially important tool to help stabilize individual health insurance markets. External funding through some form of reinsurance would reduce insurers’ average expected costs net of the funding, lowering premiums and making coverage more affordable for people eligible for little or no premium subsidy. Enrollment by younger/healthier enrollees would increase, putting further downward pressure on premiums. Lower premiums would reduce premium subsidies under current law, reducing the need for additional funding. For a given amount of funding, a particular program’s efficacy will depend on how it affects insurers’ risk and risk margins in premiums; incentives for risk selection; incentives for care management, coordination, and cost control; and the costs and complexity of administration. These effects warrant careful consideration by policymakers and additional analysis. 


ENDNOTES
1. J. Oatman and E. Coe, “Individual Market Enrollment Dropped 10-12% Year-Over-Year as of the First Quarter 2017,” McKinsey Center for U.S. Health System Reform, July 2017.
2. See historical analysis and context in: R. Bovbjerg, “Reform of Financing for Health Coverage: What Can Reinsurance Accomplish?” Inquiry 29 (Summer 1992) and M. Hall and J. Lawlor, “State Reinsurance Pools for Small Group Health Insurance,” Journal of Insurance Regulation 19 (Summer 2001), which emphasize the importance of external subsidies, and present context in: F. Busch and P. Houchens, “Reinsurance and High-Risk Pools: Past, Present, and Future in the Individual Health Insurance Market,” Milliman, Inc., June 6, 2017, and “Stabilizing Risk Pools to Drive Individual Market Stability and Promote Affordable Coverage,” AHIP Issue Brief, June 2017.
3. H. Howard and D. Meuse, “How States are Addressing Uncertainty with 1332 Waivers,” Health Affairs Blog, June 9, 2017.
4. K. Pollitz, “High Risk Pools for Uninsurable Individuals,” Henry J. Kaiser Family Foundation Issue Brief, February 2017. Also see T. Miller and M. Pauly, “A Better (but Modest) Case for High-Risk Pools,” American Enterprise Institute, March 24, 2017.
5. S. Corlette and J. Hoadley, “Strategies to Stabilize the Affordable Care Act Marketplaces: Lessons from Medicare,” Georgetown University Center on Health Insurance Reforms and Robert Wood Johnson Foundation, August 2016.
6. J. Tolbert and K. Pollitz, “Section 1332 State Innovation Waivers: Current Status and Potential Changes,” Henry J. Kaiser Family Foundation Issue Brief, July 2017.
7. In 2018, the ACA risk adjustment program will incorporate per enrollee reinsurance (“The Federal High-Cost Risk Pooling Program”) to reimburse 60% of annual claims above $1 million per enrollee. In contrast to the programs in Table 2, it does not involve any external funding—it will be funded by a small nationwide charge as a percent of insurers’ premiums subject to risk adjustment. Therefore, it will primarily reduce insurers’ risk associated with high claims, rather than directly lowering insurers’ overall costs.
8. Minnesota 1332 Waiver Application, Minnesota Department of Commerce, May 30, 2017.
9. “The State of Iowa’s Proposed Stopgap Measure for the Individual Health Insurance Market,” June 12, 2017.
10. “Using High-Risk Pools to Cover High-Risk Enrollees,” American Academy of Actuaries, February 2017, compares the basic features of traditional high-risk pools and invisible high-risk pools.
11. “Review and Recommendations Regarding the Maine Guaranteed Access Reinsurance Association (MGARA): A Report to the Joint Standing Committee on Insurance and Financial Services of the 128th Maine Legislature,” Maine Bureau of Insurance, February 2017.
12. J. Allumbaugh, T. Bragdon, and J. Archambault, “Invisible High-Risk Pools: How Congress Can Lower Premiums and Deal with Pre-Existing Conditions,” Health Affairs Blog, March 2, 2017; M. Hall and N. Bagley, “Making Sense of ‘Invisible Risk Sharing’,” Health Affairs Blog, April 12, 2017.
13. Alaska 1332 Waiver Application, State of Alaska Department of Commerce, Community, and Economic Development and Division of Insurance, December 30, 2016.
14. K. Ely, T. Murawski, and W. Thompson, “The Federal Invisible High Risk Pool,” Milliman, Inc., April 17, 2017.
15. Risk adjustment might still be needed if enrollees with high cost chronic conditions were excluded from the risk pool. See A. Bingham and E. Morgan, “High Risk Pools and Risk Adjustment,” Wakely White Paper, March 2017. 


AUTHOR

Scott E. Harrington, PhD, is the Alan B. Miller Professor in the Health Care Management Department at the Wharton School at the University of Pennsylvania. 


 

 

 

ACKNOWLEDGEMENT


ABOUT LDI
Since 1967, the Leonard Davis Institute of Health Economics (LDI) has been the leading university institute dedicated to data-driven, policy-focused research that improves our nation’s health and health care. Originally founded to bridge the gap between scholars in business (Wharton) and medicine at the University of Pennsylvania, LDI now connects all of Penn’s schools and the Children’s Hospital of Philadelphia through its more than 250 Senior Fellows. 

LDI Issue Briefs are produced by LDI’s policy team. For more information please contact Janet Weiner at weinerja@mail.med.upenn.edu.