Friday, November 09, 2018

Gold in Texas's broad empty spaces

A county is a flexible unit of demarcation, especially in the U.S. Los Angeles County (population 10 million) is 100,000 times larger than Loving County, Texas (pop. 95).

Insurance premiums in the individual market also vary wildly by geography, with county lines often (but not always) marking a rating area boundary. This may make sense from a business standpoint for commercial insurers, but it's a ridiculous principle for social policy. Living on the wrong side of a street can cost a person thousands of dollars.

Assessing the national ACA marketplace by county can create mistaken impressions. In the runup to Open Enrollment 2017, Republicans, ever eager to denigrate the marketplace, were crowing that one third of counties had only one insurer. That was true, but only 19% of enrollees (estimated prior to open enrollment) had a choice of just one insurer. That reflected market deterioration for sure, but 19% is not 31%.

Last year, when Trump cut off direct federal reimbursement for Cost Sharing Reduction (CSR) subsidies, most states and insurers coped  by loading the cost of  CSR into silver plans only, since CSR is available only in silver plans. Since premium subsidies vary by income and are keyed to a silver benchmark, this had the effect in many states and rating areas of creating large discounts for subsidized enrollees in bronze and gold plans. In some cases, gold plans were cheaper than benchmark silver for subsidized buyers. This windfall boosted enrollment at the upper income range of subsidy eligibility, where CSR is weak to nonexistent. Gold plan enrollment quadrupled in some states, and bronze enrollment rose from 23% of total enrollment in 2017 to 29% in 2018.

Plan offerings were finalized last fall shortly after Trump cut off CSR reimbursement in October. David Anderson swiftly and very usefully mapped out the effects of silver loading by county nationally,and he has done so again this year, when more states (almost all) have adopted silver loading.  The map shows broad areas where the gold plans are available for less than benchmark silver. That's the case in 1,136 counties in 2019, compared to 595 this year. On the map, counties with such anomalous gold pricing are marked in green -- the darker the green, the cheaper the gold relative to the benchmark.

There are counties and counties, however. David's map shows a sea of green in Texas, so I've been meaning for a while to check out the offerings.

In Loving County (pop. 95, just under the southeast border of New Mexico), a great gold deal is indeed available. I checked out pricing for a 40 year-old with an income of $25,000. That's slightly above the cutoff (200% FPL) for strong CSR, which gives silver plans a higher actuarial value than gold; at $25k income, silver has an AV of 73%, compared to 80% for gold. At this income level, the benchmark silver premium is $144 per month, nationwide. And in Loving, a Blue Cross/BlueShield  gold plan with a deductible of $350 costs this person (40 years old, income $25k) just $84 per month.  Bronze is also available at a considerable discount -- in fact, it's free

In 2019, maybe someone in Loving County will enroll.  This year, no one did.

In Harris County, TXx, meanwhile -- population 4.65 million, and home to Houston -- the pickings are not so rich.  That same gold BCBS plan will set our 40 year-old $25k earner back $237 per month -- almost $100/month above the benchmark. A bronze plan will cost her $72.

I checked out premiums for a 40 year old at income $25k  in the ten Texas counties with the highest enrollment in 2018, together accounting for 61% of total enrollment in the state as of the end of Open Enrollment for 2018.

 In only one of these counties, Hidalgo (home of McAllen), was the premium for the cheapest goldplan within $50/month of the premium for benchmark silver -- and there just barely, at $192 per month. In all others, cheapest gold ranged from $213 to $253 per month.

Bronze plans in most of these counties were available at low cost, $18-23 per month -- though not in Harris, where cheapest bronze was $72/month. In some but not all counties, cheap bronze plans, with their very high deductibles, offered limited services not subject to the deductible -- generic drugs, and sometimes doctor visits, or a limited number of them. On balance, the bronze pricing shows some effect of silver loading in these counties, though the gold plans don't.

Gold shines brightly, in contrast (or rather dark green), in four counties in the far west of the state: Presidio, Hudspeth, Jeff Davis and Culberson. Here, for our 40 year-old at $25k, the cheapest bronze plan is free, cheapest gold is $55 per month ($89/month below the benchmark), and cheapest silver is also heavily discounted relative to the benchmark (second cheapest silver), at $74 per month. That low silver pricing is a bonanza for people with income below $200% FPL, where CSR renders silver plans more valuable (premiums being equal) than gold.

Combined enrollment in these four counties in 2018 was 833. BCBS Texas is the only insurer, and so free to set the pricing spreads to optimize discounts if they so choose -- as they did. BCBS is the sole insurer in many Texas counties -- e.g., Wichita, where it's structured even more extreme bargains. And Wichita has more substantial enrollment than the far west counties -- 3,196 in 2018.

The Kaiser Family Foundation could crunch numbers for the whole state and tell us what percentage of last year's enrollees (or of the population) have access to strong gold plan discounts, or strong discounts for cheapest silver. I'm limited to something of  a spot check, though one "spot" does account for 61% of state enrollment. In any case, as in our electoral maps, geographic space does not equal population density.

Update, 11/10: David Anderson tweets that 25% of uninsured (as of 2016) are in counties where a gold plan is cheaper than benchmark silver.  That's impressive. I wonder whether a comparable percentage would hold for 2018 enrollees.

No comments:

Post a Comment