Hey, HHS, you're boasting about the wrong metric

If you read the inimitable Elisabeth Rosenthal's account of unhappy ACA plan holders this weekend, you know that high deductibles and out-of-pocket costs are one serious weakness of the coverage offered to many buyers on ACA exchanges.*

One mitigating factor is that most buyers with incomes under 200% of the Federal Poverty Level get Cost Sharing Reduction (CSR) subsidies that strongly reduce deductibles and maximum out-of-pocket costs, to levels superior to those offered by most employer-sponsored plans.  CSR is only available with silver plans, but as I've labored to deduce, best evidence suggests that over 80%** of buyers with incomes under 200% FPL do buy silver, resisting the temptation of lower-premium bronze plans that would leave them on the hook for huge out-of-pocket (OOP) costs.

Given this partial success, I find it baffling that all of HHS' reports of enrollment data are written as if nothing matters but premium -- the lower the better. The report released today boasts:

Nearly 8 in 10 Individuals Could Select a Plan with a Premium of $100 or Less after Applying the Advance Premium Tax Credit in the HealthCare.gov States 

Across all consumers plan selections with or without advance premium tax credits,79 percent have available an option with a net premium of less than $100 after the advance premium tax credit given the available plans in their rating areas.6

Based on actual plan choices and reenrollments to date, 53 percent of individuals have selected or re-enrolled in a plan with a net premium of $100 or less after advance premium tax credit. Similarly, 66 percent of individuals could select a plan with a premium of $50 or less after the advance premium tax credit, but based on plan selections and re-enrollment to date, only 31 percent of individuals have selected or re-enrolled in plans with a net premium of $50 or less after the advance premium tax credit.7 (my italics)

No buts, HHS -- that's good news! There's more to life than premiums. That gap represents the buyers who are balancing premium against OOP costs. Again, that includes most buyers under 200% FPL, who generally paid tens of dollars per month more to buy silver plans and thus get a deductible of anywhere (typically) from $250 to $1000 instead of $5000-6,600 for bronze.

A footnote provides more (wistful?) detail about the many who could have bought plans with super-low premiums but didn't:

7 Among new plan selections and re-enrollments qualifying for an APTC, 89 percent of individuals could have selected a plan with a premium after tax credit of $100 or less, but among actual plan selections or re-enrollments, 61 percent of individuals have selected a plan with premium after tax credit of $100 or less. Similarly 76 percent of individuals qualifying for an APTC could have selected a premium after credit of $50 or less, but among actual plan selections or re-enrollments, 35 percent of individuals qualifying for an APTC have selected a plan with a premium after advance premium tax credit of $50 or less

This data does not give as good a read on the percentage of low-income buyers who accessed CSR as the comparisons of HHS and single-state data on metal-level selection that I have parsed elsewhere. Many of those who could buy plans for $100 or less, or even $50 or less, earn more than 200% FPL -- especially many older buyers. That's because the base price of a plan increases with the buyer's age, but the subsidy increases along with it so that people of different ages with the same income pay the same price -- a 21 year-old and a 64 year-old with an income of $23k will both pay about $118 for the benchmark second-cheapest silver plan.  Since bronze plans are cheaper than the benchmark, the older person's larger subsidy often covers the spread between a silver and bronze premium. For older buyers who earn too much to qualify for the strong CSR available under 200% FPL, bronze can be a rational choice.  And most buyers did choose rationally. Silver plan selection falls off a cliff at 201% FPL, because at that threshold CSR becomes so weak as to be almost negligible.

In any case, HHS is sending the wrong message by focusing all its boasts on the premiums that ACA buyers pay.  In the first open season, only 20% of all buyers chose bronze. HHS would do better to boast about the average actuarial value obtained by ACA plan buyers than the average premium paid. Or better, brag about both. Neither number is too shabby.

--
*Rosenthal also focuses on the narrow networks (constantly shifting and poorly documented) that characterize many plans, which can waylay plan holders with out-of-network costs.

** Stat and link updated, 11/15/2015. As the article at the link explains, I had previously somewhat overestimated CSR takeup.