Health exchange design has a clear impact on metal level choices

[New data from Connecticut: please read updated version of this post here.]

Why was there such huge variation among states in the proportion of ACA private plan buyers who bought bronze plans -- the plans with the lowest premiums and highest deductibles and copays? In Hawaii, 41% of ACA shoppers bought bronze; in Mississippi, 8% did.

Part of the answer, as I've noted before, lies in a state's relative levels of wealth and health. Lower income buyers are eligible for generous Cost Sharing Reduction subsidies that reduce deductibles, co-pays and yearly out-of-pocket (OOP) maximums -- but only if they buy silver plans. Fortunately, most did. If you're sick and poor, a $6,000 deductible is likely to give you pause -- even if the plan is all but free and you don't come in knowing what a deductible is. On Healthcare.gov, only 15% of buyers eligible for any kind of subsidy bought bronze. The percentage is probably considerably lower among those eligible for strong CSR subsidies -- that is, buyers with incomes under 200% of the Federal Poverty Level.

Another factor plainly has a strong impact, though, and accounts for some wealth/health anomalies. That's website design. In Connecticut, which has a median household income of $67,8k, second highest in the nation, just 16% of all buyers selected bronze. In Colorado, with a median income of $63.4k, 40% bought bronze.

That's doubtless because the Connecticut site shows CSR-eligible applicants silver plans first; that is, the search results default to silver (at least they do in the pre-application shop-around feature; I've been trying to confirm that they do in the actual application process). [UPDATE: an Access Health CT spokesperson has confirmed that the actual application also defaults to silver for CSR-eligible users.] The Colorado site, in contrast, does next to nothing to steer CSR-eligible buyers toward silver. The shop-around is extremely cumbersome; the filter by metal level is hard to find (at the bottom of the screen, and you have to scroll back up to activate it); and unlike on healthcare.gov, applicants who qualify for CSR and make a move to buy a bronze plan receive no warning that they're leaving benefits on the table.

Healthcare.gov outperformed most state exchanges on this front. Search results on the federal exchange do default to ranking plans by premium, cheapest first. But the metal-level filter is easy to find and easy to use. In the shop-around, CSR-eligible buyers see a notice, prior to seeing plans and prices, to the effect that CSR benefit is available, and only with silver plans -- though the mention of "silver' is easy to miss, subordinated like this:

In addition this household may also be eligible for a cost-sharing reduction on a Silver plan that reduces the out-of-pocket expenses paid for deductibles, copayments, and coinsurance.

Perhaps most decisive, though, is the pop-up warning for CSR-eligible would-be bronze buyers, to the effect that they're eligible for CSR and that the plan in question doesn't have it. This warning does have a baffling design flaw -- not mentioning silver, and not defaulting results to silver for those who, asked whether they want to complete the bronze purchase, click on "no, I'll  select a new plan." If not optimal, though, it seems to have been effective.

The clearest proof that site design matters comes from New York (a middle-income state), where 89% of buyers with incomes under 200% FPL bought silver plans. New York had no shop-around in the first open season; you had to apply before you saw available plans and prices. But a state Health Department official recently confirmed for me that  CSR-eligible applicants are shown silver plans first when they begin shopping. He also said that 99.5% of applicants under 150% FPL -- those eligible for the strongest CSR, which raises the actuarial value of a silver plan to 94% -- bought silver.* That's as compared to 77% in Colorado. (Connecticut might provide a better case study than New York, if its enrollment reporting broke out buyers' metal level choices by income level.)

Cheap silver also helps

A third factor in metal level selection stems from the vagaries of plan pricing in different regions. ACA premium subsidies are designed so that a subsidy-eligible buyer pays a fixed percentage of income for the second-cheapest silver plan in her region. This year, for example, every solo buyer of any age with an income of $23,000 will pay $118 for the second-cheapest silver plan.  If the cheapest silver plan has a significantly lower premium than the benchmark, that's a golden opportunity to access CSR for less. In 2014, this happened in Southeast Pennsylvania, which includes Philadelphia. There, the cheapest silver plan for a 44 year-old with an income of $23k was $73 per month, vs. $118 for the second-cheapest, and $53 for the cheapest bronze. The deductible for the cheapest silver was $0 -- vs. $6,000 for the cheapest bronze. It was a no-brainer for almost every low income buyer in the region.  The Pittsburgh area also had just a $20 price spread between cheapest silver and cheapest bronze. In Pennsylvania, accordingly, just 5% of buyers eligible for any kind of subsidy bought bronze -- lowest in the nation. Among subsidized buyers, Pennsylvania, a middle income state (median household income $53.9k) edged out low-income Mississippi ($40.8k) and Alabama ($41.4k).

So there you have it. Three factors clearly have an impact on metal level selection: the wealth/health of the state's population, the effectiveness of communication on its exchange, and the particular features of the plan offerings through out the state.

This year should provide a couple of natural experiments on this front. First, Maryland outsourced its exchange redesign to Connecticut's, and it now shares Connecticut's "default to silver" for CSR-eligibles. Let's see if its bronze takeup rate declines from last year's 31%. Second, California (bronze takeup 25%) and Idaho (15%)  have introduced decision support tools that ask visitors how heavy their medical usage is likely to be and ranks plans accordingly. I suspect that these tools will steer fewer CSR-eligible buyers to silver than a simple default. Some might argue that that's good, if healthy low-income buyers forecast their medical usage accurately. I disagree. Insurance is about risk management. I don't think a $5,000 or $6,000 deductible makes sense for anyone who's not wealthy. Unless, perhaps, their employer fully funds an HSA instead of offering a health plan. More on that in an upcoming piece.
---

* There's an apparent anomaly here, as the New York enrollment report seems to suggest that the number of buyers of plans with the highest-level CSR (dubbed "CSR3") -- expressed as a percentage of all private plan buyers -- exceeds the number of buyers under 150% FPL -- expressed as a percentage of all subsidized buyers. That is, 12% of 273,888 subsidized buyers were under 150% FPL -- that would be 32,867, if 12% were a precise figure. But another chart claims that  9% of all buyers (370,604) bought CSR3 silver -- that comes out to 33,354. I'm waiting for an explanation of this.

Related:
Don't add copper plans to ACA -- enhance bronze
Connecticut's exchange steers low-income buyers toward silver plans
Rational choice in the ACA market
In Mississippi ACA rollout, one large disaster, one small success
News from New York: Most low-income ACA plan buyers chose wisely