Gold Gaps: Have to vs Want To

The termination direct federal payment of Cost-Sharing Reduction (CSR) subsidies in October 2017 produced a major policy and relative price shock as insurers engaged in a practice known as “silver-loading.” CSR benefits reduce cost-sharing and increase actuarial value of silver plans purchased by individuals earning under 250% FPL (single individual ~$30,000). Silverloading is the practice by which insurers put the entire cost of providing the CSR benefit into silver premiums. This has led to subsidized non-silver plans to become relatively cheaper after subsidy than they were in 2014-2017. Many analysts, myself included, expected silver to be priced as if it was near platinum which would make gold plans from the same insurer with the same network and cost-sharing philosophy to be cheaper than the silver plan.

We have seen “Gold Gaps.” But they are not universal and they are becoming less common. Insurers are pricing silver and gold very tightly to each other.

There are two likely factors in play. The first is that the under 250% FPL market segment is the biggest ACA segment out there. For people earning under 200% FPL, silver beats gold on benefit and the pricing, even before the passage of the American Rescue Plan Act in March 2021 was not that bad. Now it is insanely good. Most active insurance buyers are buying mostly on price. This creates a winner take all market. Insurers have a very strong incentive to drive down their silver premiums even if it means non-silver premium leak up.

The second plausible theory is that the risk adjustment formula that moves money from plans that cover populations that code as healthier than state wide average to plans that cover populations that code as sicker than the state wide average may be over-paying silver plans. The CSR variants have what is known as an “induced demand/induced utilization factor” that attempts to estimate how much extra care and costs are generated because cost-sharing is low. Those factors were calibrated in 2011-2013 using group market data. Some people, myself included, suspect that the demographis of the group market who had high to very high actuarial value plans are meaningfully different than the CSR market where almost everyone earns under 200% FPL so the behavioral response to a $200 deductible will be different. If I face a $200 deductible, it is a grumble, curse and shake my fist at Loki type of annoyance before getting the procedure done. If my 23 year old self who earned 150% FPL at best faced a $200 deductible, they would be walking around with a torn meniscus in their right knee for a year or two until it healed up on its own. We think that risk adjustment overpays silver which means silver premiums are lower and not-silver premiums higher than they otherwise should be.

So why does this matter?

Several states, including Pennsylvania have mandated that insurers should price their plans so that silver is more expensive than gold unless there is one hell of a reason not to. Andrew Sprung at Xpostfactoid has been digging into one of the exceptions, Independence Blue Cross in Philadelphia:

In Philadelphia, for a 40 year-old, Independence’s lowest-cost silver plan, Keystone HMO Silver Proactive Lite, costs $84 per month less than the insurer’s lowest-cost gold plan. That gold plan, Keystone HMO Gold Proactive, is $60 per month more expensive than the more directly comparable Silver Proactive (not “Lite”). For a 60 year-old, those price spreads double (as the unsubsidized premium for a 60 year-old is twice as high as for a 40 year-old)….

In their rate filings, however, insurers report and justify a “pricing AV” that may be somewhat at variance with the official AV. In 2021, Independence accorded a pricing AV of 95% s to its gold plans (including the lowest cost gold plan, HMO Gold Proactive) and 70-73% for its silver plans (73% for HMO Silver Proactive). In other words, AV as calculated by the company was in line with the CMS formula for silver plans, but way above the norm for gold plans (and exceeding the norm for platinum plans). Ambetter and Oscar calculate “pricing AVs” much closer to those of the CMS formula, pegging gold plans in the 80-87% AV range and silver plans in the low-to-mid 70s…

Independence Blue Cross is claiming that their gold plans will spend 95% of collected premiums, net of risk adjustment, on claims. The ACA requires that across the entire book of business, an individual market insurer spends 80% of adjusted premium on claims or quality improvement expenses. Very lean insurers with significant scale can break even at 90% MLR, or even 91% MLR. An insurer with a 95% MLR is losing money.

Independence Blue Cross thinks that they will lose money on their gold plans. Insurers don’t like to lose money in general or in specific. So if they are required to offer a gold plan (and they are by black letter law), an insurer that thinks it loses money on each gold plan sold will try to find ways to not sell many if any gold plans.

Regulatory commands to price plans in strict metal ratios to each other may create a “have to” situation for insurers. Some insurers will opt out entirely and either get out of the markets entirely, or shrink their footprints. Other insurers will comply with the letter of the command while actively seeking to offer plans that no one will want to buy.
Until risk adjustment is fixed, I think commands to price with strict relativity between metals will lead to “have to” situations. If and when risk adjustment is improved, then I think insurers will be more likely to be in “want to” situations where they want to offer higher actuarial value gold plans to people earning over 200% FPL that they actually want to sell lots of them.