11 states where higher-income ACA marketplace enrollees went for cheap gold plans

Subscribe to xpostfactoid 

(Program note: the second part of this post looks at 11 states where gold plan selection was highest in 2022. If you've no use for my musings on what ACA markets should look like, skip down to the subhead "Top gold plan markets in 2022" below.)

Having spent much of Saturday looking into gold plan selection in the ACA marketplace in 2022 (it's at a record level), I found myself musing on Sunday morning that an ideal metal level distribution, in very broad outline, would be silver for low income, gold for the middle class, and bronze for the wealthy.

Silver at low incomes, because Cost Sharing Reduction (CSR), available only with silver plans, is such a huge value-add for enrollees with incomes up to 200% FPL, rendering silver plans roughly platinum-equivalent at no additional cost to the enrollee. Silver plans are now free for enrollees with incomes up to 150% FPL and available for no more than 2% of income up to 200% FPL. They should be the near-universal choice up to that threshold (alas, they're not).

Gold at middle incomes: gold plans have an actuarial value (80%) slightly below the average for employer-sponsored plans. They should be available for less than the benchmark (second cheapest) silver plan (though in most states and regions they're not), because a) silver plans are platinum-equivalent at incomes below 200% FPL, b) the vast majority of silver plan enrollees have incomes below 200% FPL, c)  silver plans therefore have a higher average actuarial value on average than gold plans, and d) Ever since Trump cut off direct reimbursement of insurers for the value of CSR, just before Open Enrollment for 2018, CSR has theoretically been priced into silver plans only -- a practice known as silver loading --  in almost all states. But the discounts so far are less than analysts who anticipated Trump's cutoff of direct reimbursement, including CBO, anticipated*: the expectation was that gold would be priced consistently below silver. 

More radically, if silver plans are priced as if they're platinum, as they now are by fiat in New Mexico, very few people with income over 200% FPL will buy them, further raising the average silver plan AV and therefore justifying even steeper discounts for gold and bronze plans. 

Bronze at high incomes: bronze plans are no fun. You pay a lot out of pocket: deductibles average more than $7,000 for a single person. If you have enough assets to cover several thousand dollars in medical costs without a major change in life circumstances, however, the odds are usually that you'll come out ahead paying the lower bronze premium (especially when silver loading widens the gap between bronze and silver premiums).  Moreover, a species of bronze plans linked to Health Savings Accounts has a lower statutory annual out-of-pocket maximum ($7,050 this year) than that of many gold plans or silver plans without CSR (up to $8,700). Those who know they'll need a lot of care can profit both from the lower bronze premium and the lower OOP. 

This choice outline is currently operative in rating areas (in many cases, whole states) where gold plans are available at a premium below that of benchmark silver. And it's only a very broad outline. For many middle class enrollees (200-400% FPL), bronze plans will prove a reasonable bet. Insurers tend to underprice silver plans,* so certain silver plans may be a good choice at higher incomes (particularly in the 200-250% FPL income bracket, where an almost negligibly weak level of CSR does reduce the highest allowable out-of-pocket maximum).  At income in the 150-200% FPL range, when gold plans are priced below silver, gold may in some cases provide more affordable access than silver to a plan with a desired provider network. [Update: before the American Rescue Plan removed the income cap on premium subsidies, actuary Greg Fann made the same metal level-by-income recommendations.]

To date, silver loading -- pricing CSR fully into silver plans -- has been partial and haphazard. Insurers tend to underprice silver plans  because  a) silver is still enrollees' dominant choice, given the marketplace's low-income skew, b) the two cheapest silver plans get outsized market share, and c) CMS's risk adjustment formula favors silver plans.

Top gold plan markets in 2022

In 2022, 9.7% of all marketplace enrollees nationally (1.4 million) chose gold plans, up from 8.2% in 2021 and 4.0% in 2017, the last year before Trump created a partially silver-loaded market by cutting off direct CSR reimbursement. At incomes over 200% FPL, 17% of enrollees nationally chose gold in 2022.

In most of the minority of states where gold plans are consistently priced below benchmark silver, gold plan selection was much higher.  The table below shows the level of gold plan selection at incomes over 200% FPL in the 11 states that had gold selection rates above 33% among enrollees with income over 200% FPL. These are mostly small states, collectively accounting for just 8% of enrollment nationally. But they account for 30% gold enrollment. 

Gold plan selection at incomes over 200% FPL
11 states with highest gold plan selection

    Source:  CMS, 2022 state-level public use files
      Not shown: enrollees who did not report income (about 9% of gold enrollment)

In these 11 states, the rate of gold plan selection at incomes over 200% FPL is almost three times the national average. In four of these states, insurance departments effectively directed insurers to price silver plans either in line with their average actuarial value (Maryland, Pennsylvania, Virginia) or as if no one with an income over 200% FPL would select silver (New Mexico, which implemented a CSR pricing factor effectively pricing silver at a platinum level). The other states are small markets mostly dominated by one insurer.

Marketplace enrollees' metal level selection is pretty sensitive to price. The chart and graph below place the degree of gold plan selection at incomes over 200% FPL next to the premium spread between the  lowest-cost gold plan and the benchmark silver plan in the largest city in each of the 11 states with the highest gold plan selection. 

The premiums shown are unsubsidized for a 47 year-old -- roughly the median age for marketplace enrollees. The spread between the gold and benchmark silver plan reflects the price difference that any 47 year-old would encounter, regardless of subsidy. The spreads are smaller at younger ages and larger at older ones -- about 2/3 as large at age 21, and nearly double at age 64.

Gold plan selection and price spreads between lowest-cost gold and benchmark silver
Unsubsidized premiums for a 47 year-old

         Sources:  CMS, 2022 state-level public use files; HealthSherpa for premiums

Many factors other than this simple premium relationship between lowest-cost gold and benchmark silver affect metal level choice. In Delaware, a monopoly state, whereas the price advantage for lowest-cost gold is minimal, the lowest-cost gold plan has a $0 deductible, compared to deductibles in the $3,000 range for the two cheapest silver plans. In New Mexico, six gold plans are priced below lowest-cost silver, boosting gold selection beyond what the price spread recorded above might predict. Pennsylvania is somewhat anomalous, and not well-represented by Philadelphia, where the dominant insurer, Independence Blue Cross, prices its lowest-cost gold plan well above benchmark. In the five-county area centered on Philadelphia, only 9% of enrollees selected gold plans (though two insurers with minimal market share offered gold plans below benchmark), compared to 37% in the state as a whole. In Pittsburgh, lowest-cost gold is priced 11.7% below benchmark silver. Still, overall, the association of gold plan pricing vis-à-vis silver and gold plan selection is pretty strong.

This is all by way of arguing once again, by demonstration, that CMS should mandate strict silver loading -- that is, require insurers to price their plans in strict proportion to their actuarial value -- which, in the case of silver plans, means their average actuarial value, which always exceeds that of gold plans. Doing so would require fixing CMS's risk adjustment formula, which compensates plans that enroll members with higher-than-average risk scores and penalizes those with lower-than-average-risk enrollees, as the current formula reportedly favors silver plans, anticipating that the high actuarial value obtained by silver plan enrollees at low incomes will induce more use of medical services than it has proved to. Getting this done would provide a major boost to affordability for about half of marketplace enrollees without  requiring legislation.

------

* By statute, the ACA requires the federal government to reimburse insurers directly for the extra value  that CSR provides to silver plans. But the law left it to Congress to appropriate the funds, and Republicans in Congress blocked the funding in advance of the marketplace's launch in 2014. The Obama administration paid the reimbursements anyway; the Republican House sued to block the payments and won a court judgment, though the presiding judge stayed the ruling pending appeal. In a December 2015 ASPE brief, HHS analyzed the likely consequences of cutting off direct funding of CSR: insurers would price CSR into silver plans only, which would raise premium subsidies (keyed to an income-adjusted silver benchmark) along with silver premiums, creating discounts in bronze and gold plans. Gold plans would become cheaper than silver, since most silver plan enrollees would have incomes below 200% FPL, where silver plans are roughly equivalent to platinum. The Urban Institute came to a similar conclusion in January 2016, as did CBO in August 2017. When Trump came into office, it was widely anticipated that he would cut off CSR reimbursement, which he did abruptly in October 2017, creating a (partially) silver loaded market in 2018 and years following.

Subscribe to xpostfactoid