How do you solve a problem like SEP-ia?

OK, that was possibly the worst headline I've ever written.

So, the insurance carriers have been getting the vapors of late because they're absolutely certain that uninsured off-season Special Enrollment Period (SEP) enrollees are gaming the system, taking advantage of the poor multi-billion dollar insurance carriers' hearts by tracking down the list of Qualifying Life Events (QLEs), shoehorning themselves in so that they're eligible for one of them (or simply lying about having one), signing up for coverage during the off-season, racking up gobs of expensive medical procedures and then kicking the policy to the curb without so much as a promise to call them the next day.

I've already discussed this issue ad nauseum, including point/counterpoints with folks like Richard Mayhew of Balloon Juice, Michael Hiltzik of the L.A. Times and Michael "King v. Burwell" Cannon of the CATO Institute.

My takeaway is that yes, people "gaming" the system probably is happening to some extent, but that there are some fairly simple steps which can be taken to prevent this (or at least to reduce it considerably).

My ideas were twofold: First, actually require people to verify that they're actually eligible by having them scan/upload a copy of the appropriate proof--a marriage certificate, an employment termination notice, a military discharge form, whatever.

During a lengthy Twitter discussion with Mayhew, Louise Norris, Josh Schultz and Larry Levitt earlier today (with cameos by Judy Solomon and Aaron Albright), it was pointed out that HHS has to walk a very fine line on stuff like this. Not everyone has a scanner, after all, and some types of QLEs may not provide an obvious, consistent form of verification anyway. I suggested that one obvious solution would be for CMS to develop a smartphone app similar to the apps many banks already provide for taking photos of checks; the app recognizes watermarks and scans for the routing number/etc, then electronically deposits the check. I have to imagine something similar could be arranged for here to improve the process.

Judy Solomon of Center on Budget and Policy Priorities also called my attention to an interesting article she just authored about improving data matching for the normal open enrollment process:

In 2015, about 470,000 people lost FFM coverage and over 1 million households lost some or all of their subsidies because they had problems proving their eligibility.

Health insurers have made claims,[2] which haven’t been substantiated, that too many people are inappropriately using special enrollment periods (SEPs) in the health insurance marketplaces to get care when they are sick and that this is weakening insurance markets and raising premiums.[3] Insurers have paid less attention to data-matching issues that present problems for both themselves and consumers.

The vast majority of people who lose coverage or subsidies because of documentation issues are likely eligible. This should raise concerns for insurers, because when people are faced with repeated requests for additional documentation, those who give up on the process at some point — and consequently lose coverage — are likely to be healthier-than-average people rather than sicker ones. The result is that insurers end up with a less healthy, costlier group of enrollees. HHS has made some improvements in the processes for verifying eligibility in order to reduce the number of people who have to provide follow-up documents, but more can be done to limit the number of eligible people who lose coverage — and thereby to help in stabilizing insurance markets.

This is an important point, when you think about it: The issue isn't just about sicker, more-expensive-to-treat people being more likely to enroll (whether on or off season); it's also about healthier, less-expensive-to-treat people not being as likely to enroll. If you put too many hoops in front of someone, those who are more desperate (ie, sicker) will be a lot more willing to go through the bother of jumping through them. In this sense, the problem of adverse selection is unintentionally baked into the ACA exchange enrollment process itself.

This presents a heck of a balancing act for the HHS Dept. and the ACA exchanges. They want to encourage as many people as possible to sign up whether or not they're sick or healthy, which means making the process as simple/easy as possible...but at the same time, they have to include verification steps to make sure those who do sign up are eligible to do so, to receive tax credits as appropriate and so on.

Solomon notes several potential improvements to the data matching issue:

The best approach to keeping people from losing coverage is to prevent data-matching issues from occurring in the first place by increasing the number of cases with successful data matches. The President’s 2017 budget would take a step in that direction by allowing the marketplaces to access the National Directory of New Hires, which includes more recent wage information than the IRS tax data.

The whole thing is worth a read.

As for myself, my other major suggestion was to simply shorten the various grace periods which are currently allowed for under the ACA:

  • Shorten the 90-day Payment grace period down to 30 or 60 days
  • Shorten the 95-day Residency Documentation data matching grace period down to 30 or 60 days
  • Shorten the 2-month Individual Mandate grace period to 1 month
  • Shorten the 60-day grace periods for having a baby, getting married, etc. down to 30 days

I've been informed that some of these may require Congressional action, while others are at the discretion of the HHS Dept/White House. Either way, I was curious to see how much of a difference this might make.

I have no idea how many people are "gaming" by stretching out the SEP grace periods, but I can make a somewhat educated guess about the 2-month Individual Mandate period, in which people technically only have to have ACA-compliant coverage for 10 months of the year instead of year-round:

You may be eligible for an exemption if you:

  • Had a gap in coverage for less than 3 consecutive months during the year

I had several short coverage gaps in a year – I was uninsured in March, then again in August. Since the total gap was less than 3 months, am I exempt from the penalty?

...The rule for short coverage gaps is that only the first short coverage gap in a year will be recognized. You wouldn’t be penalized for lacking coverage in March, but you may owe a penalty for your second gap in coverage in August if you don’t otherwise qualify for an exemption during that period.

This, by the way, is one of the main reasons why people are likely to wait until after the mid-December enrollment deadline and sign up in January. Their policy doesn't kick into effect until either February 1st or March 1st, but they still don't have to pay the penalty as long as you're covered for the rest of the year.

UPDATE: Louise Norris corrected me on this; it's less than 3 consecutive months...which means that it's actually only up to 2 months right now, not 3.

In addition, there's another 90-day grace period relating to making premium payments for your policy. As Michael Kolber and Hans Leida noted back in November 2014:

Under the Patient Protection and Affordable Care Act (ACA), individuals receiving a federal subsidy are entitled to a three-month premium nonpayment grace period. As long as such an individual has paid at least the first month’s premium of the year, in any subsequent month the individual has three months to make the premium payment before coverage is terminated.

The grace period has obvious benefits for consumers, yet as a recent Health Affairs Health Policy Brief describes, this provision of the law has created significant apprehension among doctors and other health care providers who worry they will go unpaid when coverage is retroactively terminated for their patients. Unfortunately, as we explain here, this provision could have even broader adverse implications for the health care system.

The grace period law could encourage subsidized individuals to regularly pay only nine months of premiums and receive, in effect, twelve months of coverage. Should this gaming become widespread it could increase premiums (perhaps by as much as several percentage points) for everyone who purchases coverage in the individual (non-group) exchanges.

Ultimately, it may lead to higher premiums for coverage offered through the exchanges than for coverage offered outside of the exchanges; though there is still much uncertainty about exactly how insurers will be allowed to build the cost of gaming into rates. New market entrants who have a larger proportion of subsidized individuals may be disproportionately affected. for 9 months, get 12 months of coverage. Ouch. Shortening both of these from 2 months / 90 days to just 1 month / 60 days should improve the situation tremendously (at least from the insurance carrier's POV, of course), but by how much?

Well, let's look at the just-concluded 2016 Open Enrollment Period as an example. Roughly 11.3 million people signed up nationally in time for January 1st, 2016 policy effectuation, and another 1.4 million or so signed up after that, for February or March coverage.

Let's suppose that HHS/Congress did knock the 2-month penalty grace period down to 1 month. I'm guessing that would spark perhaps 1/3 of those 1.4 million people into enrolling at least a month earlier (some who enrolled between 12/18 - 1/15 would enroll before the December deadline, while others who enrolled from 1/16 - 1/31 would enroll before January 15th). Let's call it 400,000 people enrolling for 1 extra month of coverage to avoid the mandate tax.

What about at the opposite end? Well, right now, a lot of people appear to be taking advantage of the "buy 9, get 12!" option to simply stop paying their premiums starting in October...but their carrier and/or care providers are still on the hook for the remainder of the year.

How many? Well, CMS reported around 9.3 million people were still enrolled in effectuated exchange policies as of 9/30/16...and while HHS spent most of the year expecting around 9.1 million to still be enrolled as of the end of December, the partial renewal/re-enrollment data suggests that it was probably mor elike 8.9 million in the end.

Now, not all of those 400,000 people who dropped from September through December did so deliberately; some legitimately found other coverage, etc. However, let's assume that 3/4 of them did so; that's around 300,000 people.

If you drop the "payment" grace period from 90 days to 60, that's another 300,000 monthly premiums paid up.

Add them together and that would be around 700,000 monthly premiums paid. The average premium this year is around $408 per month for the 38 states, so that's around $286 million in additional premium payments made to the various exchange carriers from these changes alone.

CMS already announced that they're implementing some changes to the SEP issue, including eliminating some of them while tightening up verification of the others, but I suspect shortening the grace periods would have a more dramatic effect.

Ironically, Laurel Lucia writes over at HealthAffairs that we should go completely the opposite direction:

Insurers have alleged that SEPs are being misused by consumers, but have not publicly provided data demonstrating misuse. Insurers concerns seem to stem in part from their surprise at the number of SEP enrollees, and those enrollees’ shorter duration of coverage and higher costs compared to those who enroll during Open Enrollment.

However, research indicatethat many more individuals are likely eligible for SEPs than have taken advantage of them, and that the individual market has always faced considerable turnover among enrollees. Marketplaces and insurers have been very aware that high enrollment during Open Enrollment is crucial for maintaining a balance of low- and high-risk enrollees, thereby helping to keep premiums sustainable. The same principle applies equally to SEP enrollment. The best way to reduce adverse selection and keep average premiums down is to achieve greater mid-year enrollment, including among healthier individuals.

Lucia's point is similar to that of Solomon: Off-season enrollment is most likely to be taken advantage of by those who are most in need of coverage in the first place. Healthier individuals only have a financial incentive to get covered during the off-season (ie, the mandate tax, and concerns about being hit with unexpected medical expenses). Those who are already ill, or prone to becoming so, have a much higher personal incentive to jump through those extra hoops necessary to verify that yes, they're eligible for off-season enrollment.

For that matter, they also have a much higher incentive to find out whether they're eligible or not in the first place. Remember, according to the Kaiser Family Foundation, only 15% of those who are still uninsured even knew when the OE3 deadline was; the odds are that even fewer of them are aware that they can enroll off-season under certain circumstances, much less what those circumstances are.

HealthCare.Gov and the other state-based exchanges have various tools/information for getting the word out about SEPs, but it's still a tricky balancing act. They have to be as transparent as possible about all of the options available to prospective enrollees (ie, the government's job here isn't to try and play "gotcha!!"), but they also do have a responsibility to prevent people from gaming the system...whether you think the insurance carriers as a whole "deserve" to be "protected" or not, given the long history of abuse by some of them.