Wednesday, February 05, 2014

The health insurance "who pays?" tug-of-war

If you're going to cut a string to a certain length, whichever end you cut, you'll still have the same length.

The first substantive "repeal and replace" legislative proposal offered by Republicans since the ACA passed, unveiled in outline on January 27 by Republican Sens. Tom Coburn (OK), Richard Burr (NC), and Orrin Hatch (UT), illustrates that basic principle on multiple fronts. That's because its core provisions are to greater or lesser degrees cousins of their counterparts in the ACA.

As Don Taylor has been pointing out for a week, the rollout of the Coburn-Burr-Hatch  proposal last Monday highlights the fact that any real attempt to replace the ACA while at least roughly matching its coverage goals will expose the GOP to the same kinds of attacks -- perhaps more from their right wing than from the Democrats -- as they've been leveling at Democrats since the ACA took shape in 2009. 

In some cases, the choices advanced in Coburn et al's so-called Patient CARE Act (following Taylor, I'll call it PCARE) may represent improvements over parallel provisions in the ACA. But all such choices involve tradeoffs -- alternate plans will have, to varying degree different winners and losers  (high taxpayers vs. subsidized insureds; young vs. old; employer-insured vs. individual market-insured; healthy vs. sick).  Many if not most of us will be on different sides of each of these divides at different points.

One feature of PCARE, or rather two interrelating features, throw the nature of these tradeoffs into relief. For years, Republicans have been screaming that the ACA forces young adults in the individual market to subsidize older participants. That's because the ACA limits the allowable extent to which the oldest participants can be charged more than the youngest to a 3-to-1 ratio, whereas pre-ACA, most states allowed ratios of  5-to-1 or more (42 states, according to AHIP). That alleged bilking of the young was at the heart of the case against the ACA's constitutionality argued in the Supreme Court in March 2012 and fueled all the cries of "rate shock" from conservative critics of the ACA in 2013. According to the Kaiser Family Foundation, the impact of the narrower age banding on young adults' premiums is pretty modest, but never mind.

Besides expanding the allowable  "age band" to 5-to-1, PCARE puts forward a simpler and less generous premium subsidy scheme than the ACA's. Under the ACA, the subsidies are calculated to make people at different income levels spend a fixed percentage of their income. Those percentages not only rise with each income band but scale up within each income band:
Income as % of FPL Cap % (Lower End) Cap % (Higher End) Up to 133% 2.0% 2.0% 133% - 150% 3.0% 4.0% 150% - 200% 4.0% 6.3% 200% - 250% 6.3% 8.05% 250% - 300% 8.05% 9.5% 300% - 400% 9.5% 9.5% - See more at:
Income as % of FPL Cap % (Lower End) Cap % (Higher End) Up to 133% 2.0% 2.0% 133% - 150% 3.0% 4.0% 150% - 200% 4.0% 6.3% 200% - 250% 6.3% 8.05% 250% - 300% 8.05% 9.5% 300% - 400% 9.5% 9.5% - See more at:

Income as % of FPL
Cap % (Lower End)
Cap % (Higher End)
Up to 133%
133% - 150%
150% - 200%
200% - 250%
250% - 300%
300% - 400%

                                         Source: ValuePenguin

PCARE puts forward a simpler and skimpier subsidy schedule, perhaps not fully worked out yet. Here are the envisioned subsidies for those earning under 200% FPL:


  50-64                     $3,720                    $8,810

Note that while a plan for 60-somethings may cost five times as much as a plan for 20-somethings, the subsidies are only 2.6 times as high for a family and 2.4 times as much for an individual. That means that older participants who are subsidy-eligible will payer a higher percentage of their income than younger participants with the same income. Older participants who are ineligible for subsidies will pay much more than their younger counterparts, of course, as they did in the individual market prior to full ACA implementation this year.

Let's consider a family of four, with adults aged 60 and 61, living in Beaufort County, North Carolina, with a family income that's 175% of poverty, or $41,212. Their tax credit for premium subsidies under the ACA is $14,772 under the ACA, vs. $8,810 under PCARE; they'll pay $94 per month for the cheapest silver plan under the ACA. Their out-of-pocket expenses are reduced to a $1,000 family deductible and a $1,400 per year maximum. The unsubsidized base price of the plan as a whole is $15,990.

If the couple were 27 and 25 with two children, the base price of that same plan would be $5916.  The 61 and 60 year-old couple pay 2.7 times as much, or 90% of the allowable age-banding. Under PCARE banding, they might pay 4.5 times as much, which would set the base price of the plan at $26,662. With the maximum PCARE family credit, they'd pay $17,812.

PCARE proponents will claim that the plan in question will be less expensive, because it will not be subject to various ACA coverage mandates. True, perhaps. But then it will also be worth less, because it will cover less. Some families would win under the skimpier coverage and some would lose.  Wherever you cut the string, you still have the same amount of string -- unless you manage to reform the system to deliver more effective healthcare for less money. e.g. by disincentivizing wasteful care and reducing the pricing power of hospitals and doctors.

I don't know what PCARE would do to the ACA requirement that health plans spend 80-85% of their premium income on patient care. If it removes that, the string is simply cut in favor of insurers at the expense of patients.

It may be argued that 5-to-1 age-banding is an actuarial norm, not an arbitrary one: the average 64 year-old is likely to cost the insurer five times as much as the average 21 year-old. The National Association of Insurance Commissioners safe harbor guidelines accept 5-to-1 as "reasonable," i.e., actuarially justifiable. To the extent that 5-to-1 is a "real" reflection of risk (and just because insurance companies say so and the generally insurer-friendly NAIC agrees doesn't make it so), the young are in fact subsidizing the older to a degree in the ACA.   That subsidy is offset in large part by the higher percentage of young adults who qualify for subsidies, as well as by the ACA provision enabling parents to carry children up to 26 on their own plans.

Another string sliced differently by different age rating and subsidy schemes is incentive to work vs. security. 

In characteristic Pavlovian fashion, Republicans pounced on the CBO's new projection that the Affordable Care Act would lead to a labor force reduction equivalent to 2.3 million jobs by 2021. That's in large part because a) subsidized ACA policies will free some people from dependence on their employer-sponsored plans, and b) some people will be reluctant to boost their income because doing so may reduce or eliminate their ACA subsidies.

The more generous a means-tested subsidy, the more a beneficiary has to lose by "earning out" of it.  The steeper a given "subsidy cliff" -- the size of a subsidy lost by earning out of eligibility -- the stronger the incentive to avoid earning too much, e.g., by working more hours.

The ACA's premium subsidy schedule is finely graded for young single workers, phasing out gradually, in increments ranging from about $8-$13 dollars per month per thousand dollars of income. For older adults, however, there's a serious "subsidy cliff,"  because a) premiums for the oldest eligible participants can be up to three times higher than for the youngest, and b) older participants' subsidies are bigger, because they're calculated to reduce the premium to a fixed percentage of the user's income. A 56 year-old earning $28k will pay the same premium as a 28 year old earning the same, though her unsubsidized premium may be 2.5 times more.   Our 60-something couple with two teen children described above qualify for a pretty enormous subsidy under the ACA.  If they earned 399% FPL that subsidy would be less, but still quite substantial.  In a prior post, I cited a similar example:
The problem [of a subsidy cliff] is exacerbated for a pair of 55 year-olds with a 23 year-old son or daughter. For them, the cheapest silver plan in New Jersey is $1,354 per month.  If the family income is $63k, the subsidy is $858, their monthly payment $496.  At $78k, the subsidy has dropped modestly, to $739. At $79k, it's -- zero.
Under PCARE, that couple would have less to lose if they hit the 300% FPL cap, because their subsidy would be less generous (the PCARE plan outline does not spell out the subsidies for 200-300% FPL).  The disincentive to hit the cap would still be there, but it would be somewhat weaker. On the other hand, the cost of insurance might be out of reach, with or without the subsidy.

It would seem optimal to me to amend the ACA to phase the subsidies out more gradually for older buyers, more in line with the phase-out for younger buyers. Doing so would reduce the disincentive to earn more, which really is quite high for the small sliver of late-middle-aged relatively high earners without access to employer-sponsored insurance.  But the more gradual phase-out would require more money from taxpayers of one kind or another (high earners, industry participants, everyone?) Another string. 

1 comment:

  1. Really appreciate your thorough and careful analysis!