Brookings bombshell: ACA lowered premiums, subsidies aside. Some caveats....

In a rigorous study of health insurance premiums in the individual market pre- and post-ACA, the Brookings Institution's Loren Adler and Paul Ginsburg come to a startling conclusion: in 2014, the year the ACA marketplace launched, average premiums were 10-21% lower than in 2013 -- leaving aside the subsidies accessed by over 80% of marketplace enrollees. Further: even accounting for the steep increases coming into effect in 2017, premiums will remain lower than they would have been if the ACA had not become law.

Based on CBO estimates and  projections of premiums from 2009 and later years, and data from the Medical Expenditure Panel Survey (MEPS), Adler and Ginsburg conclude:

According to our analysis, average premiums for the second-lowest cost silver-level (SLS) marketplace plan in 2014, which serves as a benchmark for ACA subsidies, were between 10 and 21 percent lower than average individual market premiums in 2013, before the ACA, even while providing enrollees with significantly richer coverage and a broader set of benefits. Silver-level ACA plans cover roughly 17 percent more of an enrollee’s health expenses than pre-ACA plans did, on average. In essence, then, consumers received more coverage at a lower price.

Moreover, ACA marketplace SLS plan premiums are still lower in 2016 than individual market premiums were in 2013, on average, and a full 20 percent below where the Congressional Budget Office (CBO) originally projected they would be when they first estimated the impacts of the ACA in 2009.

And with regard to this year's increases:

this analysis indicates that had premiums grown at 5 percent annually after 2013 in the absence of the ACA, average individual market premiums in 2017 for a 70 percent actuarial value plan (equivalent to “silver” level under the ACA) would have been between 30 and 50 percent higher on average than actual ACA premiums will be in 2017 for the second-lowest cost silver plan — even if marketplace premiums increase by 10 percent next year. Put another way, ACA premiums would have to grow by more than 44 percent in 2017 to approach where individual market premiums would have likely been in the absence of the ACA, even under conservative assumptions.

We've all heard tales of people whose premiums rose dramatically when the ACA was implemented. Notwithstanding a lot of politically-motivated distortions, many of these stories were credible: many individual market enrollees did see steep increases -- often in deductible as well as premium, and often with a narrower provider network available. Therefore some caveats to the broad picture are in order.

Mean vs. median premium

First, as Loren Adler tweeted in response to a query:

@xpostfactoid Well there's a mean vs. median question here too, so just cause mean went down, median almost certainly went up.

— Loren Adler (@LorenAdler) July 21, 2016

Ken Kelly fleshed that out with a snip from this 2009 America's Health Insuance Plans (AHIP) survey of the individual market in that year:

Roughly two-thirds of individuals (66 percent) were offered the coverage they had requested at standard or lower (preferred) premium rates.. The remaining third (34 percent) of applicants were offered rates that were higher than standard rates.

Six (6) percent of applicants were offered coverage with “condition waivers,” which stipulate exceptions from coverage for specified health conditions.

The 34% of those who had to pay above-standard rates drove up the average. In many states where insurers could not terminate enrollees who became sick, they would stop selling new policies in plans that had a high proportion of high-use members, then jack up the premiums on enrollees remaining in those plans, who could not obtain a new policy elsewhere. A minority thus drove up average premiums. In a conversation with me, Adler noted that according to MEPS data, mean premiums were about 30% higher than median premiums.

It should be added that according to AHIP 12.7% of applicants in 2009 were denied coverage because of pre-existing conditions, including 29% of those aged 60-64 and 24% of those aged 55-59.  "Everyone thinks of those who couldn't get insurance at all, who had cancer, for example," says Adler. But some of them did get insurance - they just paid a whole lot of money for it."

Adler further points out that the overall average contains huge regional differences, though the most expensive states, such as New York and Washington, which had community rating with no individual mandate, had too few enrollees to strongly affect the average.

Measuring 2017 increases

The Brookings study anticipates a premium hike in 2017 that looks modest compared to some estimates: about 10%.  That's in keeping with estimates by Kaiser and Avalere Health of increases  in cheapest silver and (second-cheapest) benchmark silver plans (in selected markets for which data is available). A different measure, the average weighted premium increase in all plans offered in 2016, is much higher: Charles Gaba pegs the weighted average of requested rate hikes so far at 22.7%,with data in for states representing 80% of the population. The weighted average, however, obscures the fact that actual buyers avoid the plans that raise prices disproportionately.   Adler points out that the cheapest and second cheapest silver plans are far and away the most popular. And in fact, in the 38 states using HealthCare.gov in 2016, the average increase in benchmark silver, 7.2%, was fairly close to the average overall premium increase among plans actually sold, 8.4%. Still, even based on that precedent, average premiums will probably rise somewhat more than 10% in 2017.

Apples and oranges

The Brookings report cites a CBO estimate that the average actuarial value of an individual market plan pre-ACA (in 2013) was 60%. In the ACA marketplace, about 69% of enrollees select silver plans, which have an AV of 70% if not enhanced by the Cost Sharing Reduction (CSR) subsidies that lower income buyers are eligible for. About 80% of silver plan buyers do qualify for CSR, which raises AV to 73%, 87% or 94% depending on income, with a large majority at 94% or 87%. My own estimate is that the average AV for the entire individual market in 2016 is about 75%.

The Brookings report does note that its comparisons of pre- and post-ACA premiums are even more favorable to the post-ACA world when the difference in AV is taken into account.  My caveat is in the other direction. For the unsubsidized, deductibles and out-of-pocket costs are higher in the post-ACA world -- though I don't have means to calculate how much higher they would have been by now had the ACA not passed.  I do think, however, that the AV difference obscures what some people feel they have lost.  If your pre-ACA plan did not cover pregnancy, or mental health, or other benefits now mandatory that you didn't feel you needed, or if it didn't cap out-of-pocket costs*and you didn't get seriously ill, you would be satisfied with less comprehensive coverage and lower out-of-pocket costs.

The broadest caveat is that a good number of healthy unsubsidized enrollees in the current individual market might have preferred the pre-ACA market, at least in the short term. That is, if they weren't in among the 34% who paid above-market rates for medical reasons, or the 13% who couldn't get coverage at all (many of whom may still be among the satisfied unsubsidized). At present, a bit less than half of the individual market is unsubsidized. A bit more than half of them might have paid less in a non-ACA market for coverage that met their needs. At a very gross estimate, then, perhaps a quarter of the individual market are net losers from the ACA restructuring, at least in the short term.

---
* According to the AHIP 2009 survey cited above, most plans in that year did have out-of-pocket limits, but 25% of HMO plans did not.