Editor’s note: From February 15 until March 5, Tim will be in Africa. We wish him well on his trip, even though limited internet connectivity and well-deserved time with family and friends will likely restrict his posts here during this period.
On February 12, 2014, the Department of Health and Human Services released an exchange enrollment report covering the period of October 1, 2013 to February 1, 2014. The headline is the 3.3 million people have now selected exchange plans, 1.4 million through state-based exchanges and 1.9 million through the federally administered exchanges. Exchanges have determined or assessed an additional 3.2 million individuals eligible for Medicaid or CHIP.
The 3.3 million individuals who have selected plans include 1.15 million who have done so in January, down from the 1.79 million who selected plans in December, but up considerably from the 365,000 who selected a plan during the troubled first two months of the program’s operation. It is not surprising that the number of new enrollees dropped in January, as December was the first month during which the exchanges were really functioning and the last month in which individuals could enroll and be eligible for coverage and premium tax credits when the exchanges opened. It is very likely that enrollment will climb sharply over the next two months before open enrollment ends on March 31.
The report contains exhaustively cross-referenced data on enrollees by age, gender, selected metal level, state, and type of exchange. The number of “young invincibles” (aged 18-34) who have selected plans in the exchanges increased somewhat in January, making up 27 percent of enrollees as compared to 24 percent in the first three months of enrollment (although the overall 18-34 enrollment for the first four months moved up only to 25 percent). Thirty-one percent of those who selected plans over the first four months were 55 to 64, compared to 33 percent for the first three months.
Women selecting exchange plans continue to outnumber men, 55 to 45 percent, compared to 54 to 46 percent for the first three months. The gender difference grows with age: women aged 18 to 25 in the federal exchange outnumber men 54 to 46 percent, but women aged 55 to 64 outnumber men 58 to 42 percent. Perhaps women are more risk averse or perhaps they are more likely to head single parent families and thus be looking out for their children in addition to themselves. As Larry Levitt pointed out to me, younger women cost more than younger men, but older men more than older women, so the gender difference may balance out.
The percentage of individuals who have selected a plan who will receive premium tax credits continues to grow and now is at 82 percent, up from 79 percent for the first three months. Over the first four months, 43 percent of those who were determined eligible for marketplace enrollment were determined to be ineligible for financial assistance. Presumably many of them found coverage elsewhere in the nongroup market or are uninsured.
Most individuals who have selected a plan — 62 percent — continue to choose the silver level plan. Nineteen percent selected a bronze plan, 12 percent gold, 7 percent platinum, and only 1 percent a catastrophic plan (4 percent of individuals ages 18 to 34 in the federal exchange, who account for 97 percent of catastrophic plan enrollees). Seventy-two percent of federal exchange enrollees who have selected a plan with financial assistance have chosen a silver plan, compared to 25 percent of those without financial assistance. Individuals who have selected a plan in the federal exchange without financial assistance are more likely to choose a bronze (30 percent) or gold (27 percent) plan than a silver plan. Stand-alone dental plans have been purchased by 422 thousand individuals in the federal exchange. Nearly 30 percent of these were between the ages of 18 and 34.
State variation. Enrollment numbers and patterns continue to vary from state to state. California leads the pack with 728,000 individuals having selected a plan; New York comes in second at 211,000. Only 3,614 individuals have selected a plan through the Hawaii exchange and 4,696 (presumably many of them members of Congress and their staff) have done so in the District of Columbia. In the federal exchange, Florida leads the way with 297,000 individuals who have selected a plan; Texas is next at 208,000. Enrollment seems high in North Carolina and Georgia, low in New Jersey, Ohio, and Arizona.
Metal level selections seem to vary considerably by state, no doubt reflecting plan competition and pricing. Forty percent of the residents of North Dakota who have selected a plan have a gold plan; only 11 percent of residents of Alaska or Wisconsin do. Conversely, 39 percent of Hawaiians who have selected a plan have a bronze plan, compared to 7 percent of Pennsylvanians. Ninety-two percent of Mississippi residents who have selected a plan have financial assistance, compared to only 11 percent of those who selected a plan through the District of Columbia exchange.
So how are the exchanges doing after four months? In 2010, at the time the ACA was adopted, the Congressional Budget Office projected that the exchanges would enroll 8 million individuals in 2014, with 23 million covered by 2017. By 2013, it had revised down its estimate of 2014 exchange coverage to 7 million, which it revised down further in February of 2014 to 6 million.
In June of 2013, HHS Secretary Sebelius accepted the CBO’s then estimate of 7 million enrollment projection as a “realistic target.” In an internal memo from September, 2013, CMS’s Acting Assistant Secretary for Planning and Evaluation adopted the 7 million number and projected monthly targets for the 6 month open enrollment period (with 4.4 million signed up by the end of January). In a Health Affairs article in the fall of 2013, however, the CMS actuary projected that private insurance enrollment would only increase by 2.9 million in 2014, with many new enrollees in the exchanges coming from the individual or employer market.
It is obvious that the numbers who have signed up for plans so far are below these projections. There are many reasons for this. First, and most importantly, the website didn’t work for October and November, and continues to have problems with complex applications. This has not only meant that fewer people than expected were able to sign up during the fall, it has also meant that the big enrollment push expected for the fall was largely postponed.
Second, opposition to the Affordable Car Act remains intense, and this affects enrollment. Seventeen states have passed laws limiting navigators and assisters. The George Washington University Public Health School recently published a study of consumer assistance in community health centers that found that assisters in states with restrictive laws were significantly less likely to help consumers enroll. ACA opponents have also been very vocal in claiming that the exchange websites are not secure, certainly deterring some Americans from enrolling.
Third, a host of other factors have impeded enrollment to date. First, there was Christmas. Who wants to spend Christmas enrolling in a health plan, and who has disposable income for premiums during the holiday season? Second, there was the administration’s equivocation on cancelled policies. A significant number of individuals who might have enrolled in the exchange either renewed non-compliant 2013 plans or decided to go uninsured in 2014 once they were excused from the mandate. Third, surveys show that many Americans still don’t know the premium tax credits and assume they cannot afford health insurance. As noted earlier, a significant percentage of people who have visited the website have apparently found that they are not eligible for premium tax credits and either purchased plans outside of the exchange or remained uninsured.
Much negative media coverage has focused on the fact that many of those who are signing up were not previously uninsured. Exchange enrollees presumably come from four groups. First, there are individuals who were previously insured in the nongroup market but whose prior insurance plan is no longer available — the cancelled policy group. Almost all of these were presumably offered replacement insurance by their insurer, but many went to the exchange and found a better deal on premiums, and perhaps qualified for premium tax credits. Second, some enrollees have lost employer coverage, either because they lost their job, left COBRA coverage for a better deal in the exchange, were dropped from coverage on their spouse’s plan, or were dropped by an employer who concluded that its workers could get better coverage for less in the exchange.
Third, there are individuals who lost other forms of coverage, in particular coverage through the federal pre-existing condition high risk pool or a state high risk pool. Finally, there are the uninsured, some of whom have been uninsured for years, and others who have become uninsured in recent months because of loss of individual, employer, or other coverage. At the present time, this fourth group probably accounts for a minority of enrollees, which is a concern.
How much difference is there, however, between an individual who was insured on January 1, but might have been uninsured 3 months later, and a person who was uninsured three months earlier but was insured at the time of enrollment uninsured? The individual insurance market is extremely volatile — fewer than half of individuals insured on January 1 of a given year will be continuously insured for the next 12 months. Ultimately, the goal of the ACA is to reduce the number of the uninsured, but this can happen by enrolling individuals who would otherwise have become uninsured as well as enrolling individuals who are currently uninsured. The January report does not tell us how many enrollees fit in each category. But a Gallup poll released on February 12 found that the rate of uninsured in the U.S. had dropped sharply from 17.1 in the fourth quarter of 2013 to 16 percent in the first quarter of 2014, while the rate of uninsured 26- to 34-year olds dropped from 30.2 to 25.7 percent.
There has also been a great deal of media coverage on the percent of enrollees who are “young invincibles” between 18 and 34 and the January report focuses on enrollment of this group. The one principle of health economics every reporter seems to have imbibed is that you need young invincibles to balance an insurance risk pool and avoid a death spiral. In fact, healthy enrollees are necessary to balance an insurance risk pool. Young people tend to be healthier than older people, but young people pay lower premiums than older people, 1/3 as high, because of this. Moreover, the threat of a death spiral is greatly overblown; a study by the Kaiser Family Foundation concluded that if only a quarter of enrollees are young invincibles, it will only raise premiums by 2.4 percent, hardly a death spiral. (I can find no evidence, by the way, to support the oft-made claim that HHS projected at some point that 40 percent of enrollees in the exchanges would be young invincibles. The KFF study concludes that 40 percent young adult enrollment would be optimal based on their share of the uninsured population, apparently taking into account that many young adults qualify for Medicaid or are on their parents’ plans.)
The exchange risk pool, insofar as we understand it, tends to reflect the individual market before the ACA, which is to say it has a disproportionate number of people in the 45 to 64 age range. This is an age where people become more risk averse — they are more prone to illness and often have accumulated assets they have to protect. These are early retirees without employer coverage, self-employed entrepreneurs, people hanging on to a low-wage job without health insurance until they can get Social Security and Medicare. It is not surprising that they are now looking to the exchange for coverage. Insurers understood this, and priced their plans accordingly.
Looking ahead. It is possible that if the insurers do not a good risk pool mix, premiums will go up for 2015, some small insurers may drop out of the market or fail, and some of the big insurers will stay on the sideline. But the reinsurance and risk corridor programs, which are being beefed up for 2014, are available as shock absorbers. And in the short run, increases in premiums will be covered by the premium tax credits, which cover premium costs above a fixed percentage of income, not by enrollees, so a death spiral is unlikely.
In any event, there is reason to believe that exchanges may still enroll 6 or 7 million and that the risk pool will be tolerable. Enrollment is just now ramping up, and we can expect a major enrollment push in most states between now and March 31. We are now getting into tax filing season, and millions of Americans are being told about the premium tax credits and the individual mandate penalty by their tax preparation service, which is recommending that they use their refund to purchase health insurance. Although the $95 penalty for the first year is pretty meager, higher-income individuals face a penalty of 1 percent of their income above the filing limit, which may get their attention. CMS is getting the kinks worked out of direct insurer and broker enrollment, and there are a lot of hungry brokers out there eager to sign people up.
And as March 31 draws near, we are likely to see a flood of last minute enrollees. Experience with Medicare Part D, the Massachusetts health plan, the Federal Employees health benefits program, and private employer coverage reaffirms what we already know about humans — they tend to procrastinate.
Finally, although March 31 is the end of open enrollment, we can expect millions more to qualify for special enrollment periods for the remainder of 2014. Everyone who loses Medicaid or employer coverage, has a baby, gets married, or otherwise qualifies for a special enrollment period will have an opportunity after March 31 to enroll. (Medicaid has rolling rather than open enrollment and will continue to enroll individuals through 2014.) No one knows whether we will hit 6 or 7 million, and if I had to guess right now, I would say that it is unlikely, but it is clearly not out of the range of possibility.