What if exchange enrollment never catches up?

December 9, 2013
Back to TopCommentsE-mailPrintBookmark and Share

By now, everyone knows that if enrollment in the Obamacare exchanges doesn’t mount a roaring comeback, health insurers could end up with far more sick people than they counted on and sustain serious losses.

But until last week, I had never seen anyone try to calculate—in public—exactly how far short of projections the exchange enrollment might come. Or what that might mean for the premiums health insurers charge next year.

Seth Chandler, a law professor at the University of Houston, started a blog called ACADeathSpiral in mid-November, because he believes the exchanges will either fail as insurance marketplaces or will require far more financial propping up before they succeed.

Some readers will dismiss his (and my post) at this point as hopelessly biased. But I would encourage you not to. The conclusions to come are not good for Obamacare, but they’re not quite as bad as you may think.

Chandler’s basic methodology is to take what we know about enrollment via Healthcare.gov and extrapolate it over the entire open enrollment period, to see how many people are likely to actually get insurance coverage before March 23, 2014—the last day individuals can sign up without triggering Obamacare’s individual mandate tax.

Nationally, the 36 states that are relying on Healthcare.gov for exchange enrollment—which includes Indiana—saw 127,000 enrollments in October and November, according to data reported by Reuters. November enrollment was about 100,000, or four times as much as the 26,000 that enrolled in October.

November enrollment hasn’t been reported for Indiana, but let’s assume that it too was four times as much as October’s enrollment of 701. That would mean that as of Nov. 30, 3,505 Hoosiers had enrolled in private insurance via Healthcare.gov.

We also know that, on Dec. 1 and 2, a rush of 29,000 people enrolled via Healthcare.gov. Let’s assume that Indiana got its proportional share of those enrollments (1/36th of the total). That would mean Indiana saw 806 enrollments on those two days, for a pace of 403 enrollments per day.

That pace could very well be higher than normal, since there was pent up demand caused by Healthcare.gov’s technical problems in October and November. But let’s assume that Indiana maintains that pace of enrollment from Dec. 1 through March 23. That’s a period of 113 days. So enrollment during that time would total:

113 days x 403 enrollments per day = 45,539 enrollments.

Add to that total the 3,505 figure we calculated before, and we get total exchange enrollment of 49,044 Hoosiers.

But I’m not entirely satisfied with that number. Because there will likely be two additional rushes of enrollment, before the year-end deadline of Dec. 23 and then again before the March 23 deadline.

Chandler noted that as many as 25 percent of Americans wait to file their taxes each year until two weeks before the deadline. So let’s add that same assumption to Indiana’s total.

To do this, we must subtract four weeks (two in December, two in March) from our previous period of 113 days. That leaves us with 85 days, during which enrollment averages 403 per day. That gives us enrollment during that period of 34,255. We add back in the 3,505 from October and November, to get 37,760.

Then we assume that that number is only half the actual total, because 25 percent of enrollees will sign up before the Dec. 23 deadline and another 25 percent will sign up before March 23. So we double our number, to get total enrollment of 75,520.

How does enrollment of 75,500 stack up against insurers’ expectations? Not well.

The four insurers selling exchange policies in Indiana told the Indiana Department of Insurance that they expect about 200,000 total enrollees.

If enrollment, instead, totals 75,500, it will equal just 38 percent of that total. That’s just more than one-third—even though I used fairly generous assumptions about enrollment.

Chandler asserts that such low enrollment will inevitably lead to higher premiums for 2015. He noted that enrollment in the temporary high-risk pool that Obamacare established, called the Pre-Existing Condition Insurance Plan, totaled just one-third of expectations. That low enrollment led to costs more than double projections.

The same thing occurred here with the Healthy Indiana Plan. Enrollment in that program has been about one-third of initial projections, while costs per member have been double what was expected.

So let’s assume expenses per member insured in the exchanges are double what insurers were expecting. Using Anthem Blue Cross and Blue Shield’s rate filing, that means the average member would generate claims next year of $745 per month, instead of the expected $372.50 per month.

A spike like that, assuming everything else remains the same, would cause Anthem to suffer losses next year of $314.25 per member per month in its exchange plans. Ouch!

The good news for Anthem is that Obamacare includes provisions—known as risk corridors—that will absorb 80 percent of those losses, leaving Anthem will losses of just $65.85 per member per month.

Still, Anthem will certainly want to recover those losses. To do so, it would need to raise average premiums for 2015 by 15.4 percent. And if medical costs keep trending up at the same rate they have been, Anthem will have to pass on an average rate hike next year of about 28 percent.

That would push Anthem’s average premium to $520 per member per month, or $6,240 per adult, before tax subsidies kick in.

Is that large enough of an increase to cause a death spiral, as Chandler expects in some states? I don’t know.

As pointed out by former Indiana Insurance Commissioner Sally McCarty in her comments below, any increase in exchange losses will likely be mitigated because customers buying off-exchange policies will also be included in insurers' risk pools.

Also, because insurers are supposed to file their 2015 rates with regulators in April 2014, they will have very limited data on which to base any rate hike request. So there may be more time for enrollment to catch up before it directly affects premiums.

Even if low enrollment leads to larger losses, some in Washington are discussing ways in which to cover more than 80 percent of insurer’s losses, which might mitigate 2015 rate increases.

Lastly, Obamacare also raises its tax on individuals that don’t buy insurance in 2015 to $395 per adult, or 2 percent of household income, up from taxes next year of $95 per adult or 1 percent of household income. So that will put more inducement on uninsured individuals to buy, even with rising prices.

But it’s likely that the cost of insurance in 2015, even after subsidies, will continue to be far higher than the taxes for a large chunk of Hoosier families.

Chandler also proposed an alternate scenario, in which member expenses rise only 25 percent due to under-enrollment, rather than the 100 percent rise assumed in my first scenario. That lower projection should account for some of the mitigating factors I listed above.

If losses exceeded projections by 25 percent, then Anthem would lose only $7 per member per month. After factoring in the risk corridor payments from the federal government, Anthem would need to raise rates just 1.7 percent. Again, assuming rising costs of medical care, that would lead Anthem to seek a 2015 premium increase of 14 percent. Not good, but not out of the ordinary in the individual insurance marketplace.

For low-income Hoosiers, most if not all of those higher premiums would be absorbed by Obamacare's tax subsidies. For those who don't qualify for a subsidy, or only get a small one, that alternate scenario may cause them to start scoffing at Obamacare’s formal name, the Affordable Care Act. But it might not induce a death spiral, in which only sick people buy through the exchanges and premiums spike ever higher.

We’ll just have to wait and see.

  • Fantastic adaptation of my work
    Great job. You have written a sophisticated analysis and engaged in independent thinking. I'm proud that my work helped catalyze your analysis.
    • To Seth Chandler
      Thanks for your feedback. And thanks for your work. I'll keep reading.
    • Non-grandfathered risk pool
      Keep in mind the new individual market risk pool includes both the exchange and off-exchange products. Would be interesting to know how much off-exchange enrollment insurers are experiencing.
      • To Paul Houchens
        That's a good point, Paul. I'll try to dig up what the insurers were expecting for off-exchange enrollment. Getting them to disclose their actual enrollment so far will be difficult, for competitive reasons. But I'll see what I can come up with.
      • With All Due Respect
        The biggest drivers of rate increases are claims costs, utilization, and medical trend. Since there's a 3-month lag between the time claims are incurred and claims payments show up on "the books," those numbers (except trend) will be pretty sparse and won't have much validity by the time 2015 Exchange policy rates are due to be filed for review. It's not likely that counting enrollees and guessing about their health status is an accurate method for predicting 2015 rates. The rumblings among health policy folks say that there won't be much change in rates for 2015 because there will not be enough data available to justify increases. At least that's what's expected among the more consumer-friendly states. The person commenting here who pointed out that risk will be pooled by statewide claims experience per market (individual and small group) both on and off the Exchange, and not just new business, provides additional support for refuting the projections made here.
        • Great Point
          This is a crucial point. Even if only the sickest people enroll in the Exchange, which isn't likely, their experience will be pooled with the off-Exchange non-grandfathered business, which should be balanced between low and high risk folks. As evidenced by the number of policy cancellations that some elements have been screaming about, there isn't much non-grandfathered business left in the off-Exchange market, so that business will, no doubt, mitigate the effects of the higher cost insureds bot in and out of the Exchang.
          • ACA
            A look at the cost structure of the ACA, for those who are interested. Any feedback would be much appreciated. Good luck. https://www.youtube.com/watch?v=j7Y-5rjsaJY&list=TLbTo_19wIo4co4QkBulVc6pxUEbNOIOUH
          • To Sally McCarty
            Commissioner McCarty, Thank you for your two excellent points: 1) that my assumption that higher claims experience will be factored into 2015 rates may be wrong because not enough time will have elapsed for insurers to have a good sense of claims experience and 2) that my projections for exchange claims experience will be mitigated by the pooling of the experience of consumers buying off-exchange policies. I am adding those items to my list of mitigating factors in my main text. I also have two questions for you: 1) Do you think my alternate projection of losses 25% above expectations would more than account for the mitigating factors you mention? and 2) Since I doubt you would include Indiana under your label of "consumer-friendly states," how unreasonable is it to expect that insurers in Indiana will, indeed, be allowed to raise rates to reflect even incomplete data showing higher-than-expected claims experience?
            • Still Don't Agree
              JK, I'll try to be responsive: 1) The assumptions being used to project premiums are not valid on two levels: a) The assumption that people buying on the Exchange will all be high utilizers is baseless and any regulator worth her salt would not accept that as a valid justification for rate increases without hard data to back it up. Comparing enrollment shortfalls in the Exchange to those in the Federal high risk pool is especially invalid because there will be a much more varied mix of high and low risk enrollees in the Exchange plans. b)I can't emphasize enough that whatever the utilization experience of Exchange plans, it will be pooled with the off-Exchange market risk, which means that it's not likely that experience will be much different than it is currently. 2) Commissioner Robertson has taken some very powerful actions to protect consumers from unreasonable rate increases in the past. If he is able to (or allowed to) put his strong negative feelings toward the ACA aside, he would insist on more credible experience than what will be available this summer before approving any rate increases for 2015 Exchange policies. So, I'm not necessarily elminating Indiana from the consumer-friendly category.
              • To Sally McCarty
                That's helpful. Are you saying that the national concern with predominantly unhealthy people signing up for the exchanges is simply misplaced, because it's almost certain to be counter-balanced by healthy people buying off-exchange? You may be right. It's just not a view that I've heard before. But the views we hear a lot are often incorrect. So thanks for offering some balance.
              • Incorrect Rate Increase Calculations
                Your rate increase calculations are understated by 80%. If costs are double what was anticipated in the premium rates, plans will need to double premiums to hit profit totals. If they only raise them 20%, they will still have an 80% loss prior to risk corridors - which will still be a 16% loss after the corridors. These calculations ignore medical trend. They also presume that no effort will be made to recoup initial losses.
              • Not provable <> baseless
                While we cannot definitively conclude that lower enrollment would mean a disproportionate share of high utilizers, it is the most logical assumption. It's also the assumption underlying the CBO projections of budget neutrality. They assumed that the losses associated with the high risk members that were previously unable to get health insurance would be offset by the money made on "young invincibles". These young people get a rotten deal since there is no longer medical underwriting to lower their premium rates, and age band restrictions force them to pay higher rates to subsidize higher age bands. The mechanism for forcing these people was the (woefully inadequate) individual mandate fee (or tax according to SCOTUS). Lower than expected enrollment rates are indicative that voluntary enrollment - or people who did not feel they needed the insurance - were enrolling at lower rates than expected. This is a reasonable assumption for forecasting. Now, you mention than any examiner worth his/her salt would challenge this assumption. I expect you're right - since health insurance examiners, in my experience, don't approach rate review from an unbiased perspective. Unless they are worried about the solvency of the plan, they are typically far more concerned with member rate increases than plan profitability. It's always good politics to say you held health insurance premium rates down. They'll have a better time with this argument due to limited information availability - this being said, at minimum, plans will be able to point to differences in age/sex and industry factors from rating assumptions; so it's unlikely to be purely an argument of conjecture.
                • Why?
                  For plan 2014 financials, sure the experience may be pooled. But why would plans voluntarily sell business that is detrimental to their bottom line?
                  • Incredibly impressed
                    At the caliber of the dialog on this site. Pretty impressive to have the Indiana Insurance Commissioner commenting on a blog. I want to think about what she has to say; I regard it as a novel perspective. Also, if anyone wants an update on the national picture in light of today's news, I've posted a new blog entry available at http://acadeathspiral.org/2013/12/11/the-new-exchange-enrollment-numbers-are-bad/
                    • Responses
                      J.K.: My perspective is not so novel in the circles I travel in now, i.e., national health policy folks. However, we are "glass half full" people who are very supportive of the ACA, just as the perspectives floating around Indiana and Texas are most likely to be "glass half empty." Many people who are uninsured have been waiting for a better deal, waiting to retire early with insurance, etc. and are not necessarily high utiliziers. With the absurd list of conditions insurers reject applicants for on the current market (acne, bunions, etc.) coverage options available on the federal exchange will be welcome for a lot of healthy or nearly-healthy people. Seth Chandler: I'm a FORMER Indiana Insurance Commissioner (It's been 9 years!) who is currently working as a Senior Research Fellow at the Georgetown University Health Policy Institute. J.K. is kind enough to still address me as "Commissioner."
                    • Health Actuary
                      It sure would be nice if you would identify yourself, so we know your credentials and experience. Regarding the solvency issue: In my 11+ years as a state and federal insurance regulator I did not see ONE instance where denying a rate increase would threaten the solvency of an insurance company -- not ONE. A lower increase might result in lower profits for shareholders, but that should not be a concern for insurance regulators. The "solvency" argument is one that's often raised by insurers, so I suspect you develop rates for a health insurer.
                      • Solvency
                        I would prefer to remain anonymous as I don't want my opinions to come up in google searches of me. I'm a health actuary with 11 years of experience specializing in Medicare Advantage. That being said - I'm not asking that credentials be used or not used to add credibility to my arguments. I think you misunderstood my statement on solvency. I was saying that examiners are biased in favor of the consumer, and would prefer lower rates in the absence of evidence - note that I don't think that this is necessarily inappropriate. The solvency comment was the exception to this rule - IE, I imagine that an examiner concerned with the solvency of an insurance company might actually force things the other way. I'm claiming bias not to suggest that your opinion is incorrect of invalid. Rather, that the comment that an examiner would challenge an argument with the purpose of rejecting or reducing a rate increase is not indicative that the requested increase was not the appropriate one.
                      • One more thing
                        I'll also gladly concede that someone submitting rates may be equally biased. However, rate submitted can be biased in either direction. I've been involved in pricing products with conservative rate assumptions as well aggressive ones.
                        • To Sally, Paul and Health Actuary
                          Thanks to each of you for your comments. Clearly, we don't all agree in our outlooks for 2015 individual insurance rates in Indiana. But the three of you certainly gave readers the full range of other ways to look at the issue. You make this blog far more interesting to read than my meager posts do on their own. Thanks again.

                        Post a comment to this blog

                        COMMENTS POLICY
                        We reserve the right to remove any post that we feel is obscene, profane, vulgar, racist, sexually explicit, abusive, or hateful.
                        You are legally responsible for what you post and your anonymity is not guaranteed.
                        Posts that insult, defame, threaten, harass or abuse other readers or people mentioned in IBJ editorial content are also subject to removal. Please respect the privacy of individuals and refrain from posting personal information.
                        No solicitations, spamming or advertisements are allowed. Readers may post links to other informational websites that are relevant to the topic at hand, but please do not link to objectionable material.
                        We may remove messages that are unrelated to the topic, encourage illegal activity, use all capital letters or are unreadable.

                        Messages that are flagged by readers as objectionable will be reviewed and may or may not be removed. Please do not flag a post simply because you disagree with it.

                        Sponsored by