Urban Institute: At least half* of ACA plans cost LESS than employer-sponsored insurance?

*(study is based on benchmark exchange plans only, which I estimate make up about half of total individual market; see below)

There's obviously a lot of people getting upset and concerned about ACA-compliant individual market policy rate hikes shooting up significantly next year. Heck, I'm probably responsible for some of this concern due to my 24-26% national weighted average estimate being referenced in every healthcare story from Forbes to the New York Times and even The Economist (my estimate has been pretty much confirmed by Barclays, FWIW, although for all I know, they might've even cribbed their data from me as well).

Anyway, with all of this as backdrop, a new study by the Urban Institute has some news which surprised even me. As Dan Mangan reports:

Insuring people through Obamacare — which was crafted in part to cover people who can't get health insurance through their jobs — may be costing less money than if they had employer-based coverage, a new study suggests.

The study, by the Urban Institute, comes as premium rates for 2017 Obamacare plans are being finalized. Those premiums are expected to rise more sharply, on average, than in recent years.

But the report found that certain key Obamacare plans, on average, cost 10 percent less in premiums than average employer-based coverage, when adjusting for how much the plans cover for medical services, as well as for adjustments for health-care usage and age distribution.

I've been making a big deal over the past few weeks about the HHS Dept's irritating tendency to ignore the off-exchange market when using the "85% of exchange enrollees receive tax credits" talking point to downplay premium rate hikes on the individual market. This is important because that figure drops to around 50% when you include the 7 million or so off-exchange enrollees who are paying full price.

HOWEVER, just as it's not reasonable to ignore the fact that half the individual market has to pay full price for their premiums, it's equally unreasonable to ignore the fact that the 150 million or so people enrolled in employer-sponsored policies aren't paying full price for their premiums:

The average monthly Obamacare premium for the types of plans examined was $464 per month for an individual, nationally. That compared to $515 per month for an average job-based premium for an individual, according to the report, entitled "Are Nongroup Marketplace Premiums Really High? Not in Comparison with Employer Insurance."

"We find that in most states and in most metropolitan areas, [Obamacare] premiums are lower than for those for employer-sponsored plans," the report's authors wrote.

...The Urban Institute's report looked at the total premiums paid for both job-based plans and Obamacare plans.

In the case of job-based coverage, that includes what the employer pays toward the premium and the employee's share. For Obamacare plans, that includes what a customer personally pays, plus any subsidy they receive from the government in the form of tax credits.

In other words, the portion paid by APTC assistance for about half of all individual ACA-compliant enrollees plays the equivalent role of the portion paid by employers for pretty much all large/small group enrollees.

Now, it's important to note that this Urban Institute study only looked at the benchmark plans (2nd lowest-cost Silver plans), so Bronze, Gold, Platinum and other Silver plans aren't included in the comparison:

The report focused on premiums for the second-lowest-cost "silver" Obamacare plans against employer-sponsored plan premiums in the same geographic area. Silver plans, the most popular type of Obamacare coverage, cover about 70 percent of their customers' health services' costs, with customers owing the rest personally.

However, it's also worth noting that about 70% of all exchange enrollees have Silver plans, and I'm guessing that most of those are the benchmark plans. I don't know what the metal level breakout is for off-exchange policies, but assuming it's similar (around 70% Silver), it's safe to assume that a good 50% or so of all individual market enrollees have the benchmark Silver plan...the exact type that the Urban Institute study is comparing here.

It's also worth taking a look at the official ASPE report on 2016's open enrollment period, where it states:

For the 38 states using the HealthCare.gov eligibility and enrollment platform:

More than 8 in 10 individuals (more than 8.1 million, or 85 percent) who selected or were automatically enrolled in a 2016 Marketplace plan qualify for an advance premium tax credit with an average value of $290 per person per month.

The average advance premium tax credit covers about 73 percent of the gross premium for individuals who qualify for an average advance premium tax credit.

The average net premium after advance premium tax credit is $106 per month among individuals with 2016 plan selections through the Marketplaces in the HealthCare.gov states who qualify for an advance premium tax credit.

OK, so that means that the average full-price premium for exchange plans this year is around $396/month. It make sense that this is lower than the $464/month average noted in the Urban Institute study, since that $396 includes lower-priced Bronze and Catastrophic plans (around 22% of the total) as well as higher-priced Gold/Platinum plans (the remaining 8%). This is further complicated by the fact that the $396/mo average only applies to the 85% who receive APTC at all. By an amazing coincidence, if you assume that the remaining 15% of full-price exchange enrollees have similar metal level proportions, the average full-priced premium being paid for all exchange-based plans just happens to work out to:

  • $396 / 85% = $466/month

...or almost exactly the same as the Urban Institute's $464 average. This is almost certainly a coincidence, especially since the Urban Institute study gets that number by adjusting for things like age distribution, deductibles/actuarial value and so forth, but it's still interesting.

Again, I have no idea what the metal level market share breakout is for off-exchange enrollees, but let's assume that it's roughly the same. If so, then the actual situation is:

  • Around 9.4 million people on the indy market are paying 27% of their $396/mo average premiums
  • Around 8.7 million people on the indy market (1.7 million on exchange, 7.0 million off exchange) are paying 100% of their $466/mo (?) average premiums

This compares with around 150 million people in the group market who are having an average of 73% of their​ premiums paid for by their employer, according to the Kaiser Family Foundation ($4,710/year from the employees vs. $12,612/year for the employers).

By another amazing coincidence, that's the same 73% average that APTC assistance is paying for qualifying ACA exchange enrollees.

It seems to me that expanding APTC assistance beyond the 400% FPL threshold is one key to stabilizing the individual market...and judging from the evidence above, it sounds like this isn't unreasonable even from a "run government like a business" perspective.

Sidenote: As you'd imagine, Michael "King v. Burwell" Cannon has a slightly different perspective:

A new Urban Institute report claims premiums for plans sold in ObamaCare’s health-insurance Exchanges are lower than comparable premiums for employer-sponsored health plans. Unfortunately, Urban scholars used sleight of hand to hide the full premiums for ACA plans. Incorporating the full premiums shows Exchange plans are more expensive.

Cannon's main arguments are a) the impact of the "Reinsurance" program; b) the fact that overall, exchange carriers lost a bunch of money the first couple of years of the program (and presumably are set to do so again this year); and c) the impact of Cost Sharing Reduction (CSR) assistance on the "total" price.

I haven't read the UI study thoroughly enough to comment on how legitimate Cannon's criticism is, but I'm sure plenty of actuaries/etc. will chime in in the comments below.

For what it's worth, here's what Loren Adler of the Brookings Institute Center for Health Policy has to say:

@mfcannon (1/2) The S&P data confuses me as it disagrees with everything else out there, CSRs shouldn't matter in an AV-adjusted comparison

— Loren Adler (@LorenAdler) September 20, 2016

.@mfcannon (2/2) But yes, the existence of reinsurance + likely losses still in 2016 means the two markets likely close to on par.

— Loren Adler (@LorenAdler) September 20, 2016

The idea that ACA plans would cost less overall than ESI plans sounds rather odd to me, I admit; the difference in the size of the risk pools alone (150 million vs. perhaps 18 million) makes it seem unlikely...but even if they've missed a few factors, this study certainly makes it sound like indy market QHPs aren't excessively more expensive (at full price) than ESI once you factor in everything.