All the king’s horses and all the king’s men: Reintroducing fractured risk pools

Congressional replacement of the Affordable Care Act threatens to undermine several of its core insurance market reforms.  One of these is a little-noticed requirement of a “single risk pool.”  Under the ACA, each insurer must base its individual or small-group premium rates on the full range of its non-grandfathered subscribers in that state.  A single risk pool is central to the notion that insurance spreads risk by pooling lower cost with higher cost people.  A single risk pool is also critical to market reforms that aim to prevent insurers from selecting and pricing people based on their health conditions.

Prior to the ACA’s requirement that insurers base their premium rates on their full range of relevant subscribers, insurers actively segmented subscribers into different risk pools based on how they designed and sold their policies.  Thus, an insurer could use a completely different rating structure for an HMO than for a PPO product, or for coverage sold through agents rather than through associations.  Those distinctions may seem logical, but insurers exploited them to separate healthier from sicker people in order to vary their prices more than regulators otherwise would have allowed.

The ACA put a stop to these risk segmentation practices, but the “repeal and replace” bill adopted last week by the House appears to reintroduce fractured risk pools.  The House bill does not overtly repeal the ACA’s single risk pool requirement, but it allows states to seek a waiver of normal community rating rules, and thus to reintroduce rating based on health status.

If a waiver state had to choose risk rating rather than community rating for all products, a single risk pool could still be maintained because each insurer’s base premium, prior to allowable adjustments, would be the same for all subscribers.  However, the House bill does not take this all-or-nothing approach; instead, it allows – in fact it requires – waiver states to maintain community rating for part of the market.  In doing so, the bill reintroduces a fractured market that will be difficult to put back together again – whether by state regulators, or by the “King’s Horses and Men.”

Under the House bill, in states that allow insurers to set premiums based on health status, insurers still must also sell some community-rated plans.  Only people who have a lapse in coverage are required to undergo risk rating.  People who maintain continuous coverage are still entitled to enroll in a community-rated plan, but they don’t have to; they can still choose risk rating if that offers a better deal.  Also, the House bill requires community rating by the multi-state plans that must be sold in each state, which provides another option.

As Matthew Fiedler has noted, selling community-rated alongside risk-rated plans is not sustainable if people can simply choose which type of plan to buy, since, naturally, the healthier people will choose risk-rated plans, and sicker people will gravitate to the community-rated option, thus driving its price out of reach.  The key reason that community-rated plans will end up with much higher prices is that insurers are not required to pool their medical costs with those of people who buy risk-rated plans.  Community rating is no help if it applies only to the community of people who cannot pass medical underwriting.

Prior Experience Under HIPAA

The inevitable collapse of the community-rated plans in this fractured market is not simply a matter of speculation.  We have direct recent experience prior to the ACA, both from a number of different states, as well as from under the Health Insurance Portability and Accountability Act (HIPAA).  HIPAA is now known mainly for its rule on medical privacy, but its original purpose was similar to the House bill in seeking to provide some coverage for people with pre-existing conditions.  In fact, the House bill essentially adopts HIPAA’s core “portability” provision.  Both laws provide for coverage of pre-existing conditions as long as people maintain continuous coverage over the past 12 months with no more than a two-month (63 day) break in “creditable” coverage.

Unlike the House bill, however, HIPAA did not allow people with pre-existing conditions to choose from any plan that was in the market.  Instead, states could adopt one of several approaches to covering pre-existing conditions.  Over half of states chose to build high risk pools for this purpose, which the House bill also contemplates for waiver states.  Alternatively, about 20 states used a different approach that required insurers to offer a specially-designated HIPAA policy to people entering the individual market who previously had group coverage.

This second approach was a total failure because states did not require insurers to rate these HIPAA polices as part of their normal risk pool.  Instead, insurers were allowed to price these policies as as part of a separate risk pool.  Thus, under HIPAA, people with previous coverage had two choices: either purchase coverage at a lower cost that excluded pre-existing conditions, or purchase the separate HIPAA-eligible continuous coverage plan.  Naturally, those with pre-existing conditions chose the latter, and those in good health chose the former.

As a result, prices skyrocketed for HIPAA’s continuous-coverage plans.  Various government reports documented insurers charging sometimes as much as five or six times more, and more often three or four times more, for their continuous-coverage HIPAA plans than for their normal risk-rated plans.  One report warned that insurers were “intentionally offering coverage at unaffordable rates in order to avoid providing coverage to HIPAA-eligible individuals and small groups while appearing to comply with guaranteed availability provisions of HIPAA.”

A government assessment of HIPAA’s fragmented markets by the Government Accountability Office also observed several other predictable, but troubling, behaviors designed to avoid enrolling sicker patients.  Insurers paid agents lower or no commissions for people enrolled in the plans that covered pre-existing conditions, and they encouraged people who were eligible for the continuous-coverage plans to enroll instead in risk-rated plans if they were fairly healthy.  Consequently, few people ever enrolled in HIPAA’s continuous-coverage plans in the states that pursued this option.  Comprehensive data are lacking, but the only government study on this point reported that “each of three national carriers estimated it had HIPAA enrollment of fewer than 200 individuals” per state.

The price of continuous-enrollment plans might be reduced, and thus enrollment increased, if states chose to heavily subsidize this separate risk pool.  But, as with high-risk pools, the cost of adequate subsidies is much greater than the “morsels” offered in the House bill.

Moreover, offering risk-rated plans alongside community-rated plans is only one of several ways that ACA repeal and replace could fragment states’ risk pools.  Waiving the requirement for essential health benefits can have, and has had, the same effect.  Other fragmenting measures under discussion, but not yet adopted, include:  allowing healthier individuals to form into “fictitious groups” (sometimes known as association health plans), and allowing insurers to sell across state lines.

Risk segmentation is to health insurers as gravity is to Humpty Dumpty – always threatening to pull down a teetering market.   Without good balance, the natural course of insurance markets is to fracture into separate pieces that undermine risk-pooling.  The House bill, and accompanying repeal and replace ideas, threaten to topple the cohesive risk pool that the ACA has managed to piece together for the first time in over half a century.