More than (four) words

The viability of the Affordable Care Act’s subsidies in 36 states is going to be decided by the Supreme Court.  King v. Burwell, the Fourth Circuit iteration of a multi-pronged challenge to availability of health insurance tax credits across much of the country, has been granted cert by the Court to be heard this term.

I’ve written about these cases (focusing on the D.C. Circuit case, Halbig v. Sebelius Burwell) frequently (I, II, III, IV, and V). But as the case readies for it’s moment before the Highest Court, I wanted to take a moment to address a clever rhetorical deceit that is being propagated by those cheering on the subsidy challenge.

That’s the idea that this is a cut and dry case, a foretold outcome based on the unambiguous text of the statute. Take George Will: “Four words in the ACA could spell its doom.” “The four words that threaten disaster for the ACA,” Will writes, “say the subsidies shall be available to persons who purchase health insurance in an exchange ‘established by the state.’ But 34 states have chosen not to establish exchanges.”

Or take Patrick Wyrick, the solicitor general of Oklahoma, who is challenging the law’s subsidies in a separate suit.  “The phrase ‘Exchange established by a state under Section 1311′ ​leaves nothing to the IRS’s imagination​,” he argues at SCOTUSblog.

Indeed, the architect of these lawsuits, Michael Cannon of the Cato Institute, asserts much the same, contending that “the tax-credit eligibility rules ‘clearly say’ exchange subsidies are available only through state-established exchanges,” and that any ambiguity has been retrospectively manufactured by the government’s lawyers.

There’s a reason the challengers in King are so adamant that this is a simple, unambiguous case. If there is any ambiguity in the statute — any uncertainty whatsoever in whether the text permits subsidies in non-exchange states — the challengers likely lose. That’s because ambiguity triggers so-called Chevron deference to the IRS’s interpretation of the law, which favors making subsidies available in all states.

It could also trigger constitutional avoidance doctrine, since the challengers’ constitutionally troublesome and coercive reading of the law would have to compete with the government’s abjectly constitutional, non-coercive reading of how the subsidies operate.

But if what the law says is clear and unambiguous, then there’s no discretion for the IRS to employ, and no deference owed to its interpretation. And there’s no constitutional avoidance obstructing the challengers’ path to victory, because there would be no alternative reasonable interpretation at hand.

Now “established by the State” sounds pretty unambiguous, right? Pretty damning for the government and the law’s supporters, no?

Of course it does. But it’s also a neat sleight of hand, because those four dooming words aren’t the ones that this case hinges on.

Instead, the case turns on two words, not four. Those two words are “such Exchange.” See, Section 1311 of the law instructs that each state create a health exchange. But Congress can’t order the states to do anything, so it created a federal fallback. For states that don’t elect to run their own exchanges, Congress said, the federal government would step in to “establish and operate such Exchange within the State.”

What’s more, the term “Exchange” in the statute is defined as “an American Health Benefit Exchange established under [section 1311 of the ACA].” That’s the section that directs states to create exchanges.

So what does “such Exchange” mean, exactly? One eminently plausible interpretation is that, if you connect the daisy chain of definitions and provisions laid out above, the federal government creates the functional equivalent of an exchange established by the state in each non-compliant state. That is, for purposes of the statute, the federal government can create an “exchange established by the State.”

Counterintuitive, sure. But it’s clearly a reasonable interpretation, as I’ve explained before. And the meaning of “such” is what split the lower courts in the first place. Shouldn’t this alone be enough to suggest ambiguity, which in turn triggers deference to the IRS’s interpretation?

One would think so.  But the ObamaCare challengers evidently insist that the federal government can only create “an exchange,” not “such Exchange,” for it can create one with all the attributes of a state-created exchange except for subsidy availability. Where’s the support for this in the text?

The challengers’ case looks a lot weaker when we hone in on the text of the law that really matters to this case. And liberals like Paul Krugman ought to stop arguing that the law is facing “death by typo,” for it concedes that the law’s text as written can’t accommodate subsidies on the federal exchange.

That’s only the case if you accept the challengers’ false framing of the language that this case turns on. They’re trying to hide the ball behind the seemingly open-and-shut certainty of “established by the State.” But that’s not the language that really matters. And the fact that “such Exchange” is ambiguous is just enough to expose the challengers’ case to potentially crippling vulnerabilities that could spell their doom before the Supreme Court.

The data prove Boehner’s “idea” about the unemployed wrong, wrong, wrong

John Boehner thinks we’ve made it too easy for the unemployed to go without work. Addressing the conservative American Enterprise Institute last week, the Speaker of the House criticized “this idea” among the jobless that “I really don’t have to work. I don’t really want to do this. I think I’d rather just sit around.” To Boehner, our policy has corrosively enabled those inclined toward sloth to give way to that temptation.

This theorizing has been roundly blasted in liberal circles. Paul Krugman characterized it is a revealing gaffe, exposing conservative disdain for the unemployed. Simon Maloy at Salon said that Boehner had gone off the reservation, drifting off-message from efforts by those like Rep. Paul Ryan to move away from (or at least better conceal) makers/takers rhetoric and knee-jerk distaste for the jobless.

But few have discussed the absolute wrongheadedness of Boehner’s policy diagnosis here. He believes that it’s too easy to stay unemployed. The implication is that public policy should make it harder to be unemployed by, say, cutting off unemployment insurance benefits earlier and making benefits stingier to begin with. This, it’s thought, would give these slackers the kick they need to go out and work.

Never mind that unemployment benefits are already unavailable to the vast majority of the unemployed and already historically stingy. As Krugman explains, “Only 26 percent of jobless Americans are receiving any kind of unemployment benefit, the lowest level in many decades. The total value of unemployment benefits is less than 0.25 percent of G.D.P., half what it was in 2003, when the unemployment rate was roughly the same as it is now.”

Yet the idea that our lavish unemployment benefits drive up unemployment rates persists among conservatives. And it persists in the face of clear evidence to the contrary.

For one thing, unemployment benefits vary drastically from state to state. States aim to replace some portion of an unemployed person’s lost wages. Maximum benefits range from Mississippi’s $235 per week to Massachusetts’s $1,019.

If the Boehner Hypothesis had something to it, we might expect to see lower unemployment rates in states providing the stingiest unemployment benefits. But in fact, Mississippi has 7.9 percent unemployment, one of the highest in the country. And Massachusetts has 5.8 percent unemployment, lower than the national rate.

The Federal Reserve recently studied whether extending unemployment benefits had any effect on the unemployment rate. During the recession, the federal government extended funds for state unemployment benefits. These emergency measures made benefits available to those without jobs past the usual 26 weeks of unemployment to as many as 99 weeks. Conservatives regularly fretted that this weakened the incentive to find work, frequently obstructing extensions to provide more help to the jobless.

The Fed, however, found that these extensions had a minimal impact on the unemployment rate and virtually no impact on the number of people participating in the workforce. “[T]he overall effect of EEB [emergency extended benefits] on the unemployment rate is fairly modest; at its peak (in terms of the average number of benefit weeks provided) EEB boosted the unemployment rate by one-third percentage point.” Moreover, “the effect of EEB on the [labor force] participation rate is estimated to have been quite small.”

This is just the latest piece of evidence refuting the Boehner’s Bootstraps Hypothesis. James Pethokoukis at the American Enterprise Institute helpfully summarized a series of studies that found similar results. One study by the Boston Fed found that cutting off unemployment benefits didn’t nudge people to find work, but rather made them “more likely to drop out of the labor force; transitions to a job appear to be unaffected by UI benefit extensions.”

Another study compared the tellingly similar results in North and South Carolina. Last summer, North Carolina slashed its unemployment benefits so severely that it made itself ineligible for federal emergency unemployment extensions. Mirroring the Red State rejection of ObamaCare’s free Medicaid expansion, North Carolina effectively spurned free federal support for its own long-term unemployed.

Economist Justin Wolfers found that since North Carolina cut its benefits, its economy has fared no better than neighboring states with similar economies that didn’t slash benefits for the unemployed. South Carolina, for instance, has seen slightly faster employment growth that North Carolina has. Contrary to the conservative bootstraps theory, North Carolina doesn’t stand out at all from similar states despite its experiment in inflicting hardship on the long-term jobless. “The bottom line,” Wolfers concluded, “is that North Carolina looks quite similar to its peers, and certainly not better.”

The irony is that Pethokoukis’s work comes from the very think tank before which Boehner continued to cling to his disproven theory of what ails the unemployed. The lesson, I suppose, is that Boehner would do well to survey the work from the forums he speaks at.

And he’d do well to heed Pethokoukis’s conclusion that the “safety net supported American incomes during the recession and its aftermath” — jobless benefits included. It’s time we lay to rest the conservative notion that the jobless need constant work incentives, and maybe instead just need some financial support to help them afford things like Internet, new suits, dry-cleaning, resumes — you know, things that cost money that help people find a job.

Instead of moralizing against the unemployed, we should understand that unemployment insurance plays an important role in helping people through hard economic times. The unemployed don’t need a kick in the pants to help them find work, they need a stronger economy. And that comes when everyone — the jobless included — has a little more spending money in their pockets.

Hey, Congress actually did something! — and made American childcare a little better in the process

Stop the presses: Congress actually passed a bill last week. An overwhelming bipartisan majority in Congress voted to reauthorize the Child Care and Development Block Grant program, a $5 billion fund for the states to subsidize the cost of childcare for low-income families, which President Obama is expected to sign. And in the process, it patched some glaring holes in our childcare system to make it a little better and safer.

Calling what we have a childcare “system” is frankly too generous. What we have in much of the country is a lightly regulated mess. In a highly-regarded piece last year called “The Hell of American Daycare,” Jonathan Cohn of the New Republic exposed the neglectful patchwork of state under-regulation that led a tragic fire at a home daycare in Texas.

“Excellent day cares are available, of course, if you have the money to pay for them and the luck to secure a spot,” Cohn explained. “But the overall quality is wildly uneven and barely monitored, and at the lower end, it’s Dickensian.”

Some states require hardly any training for daycare providers in health, safety, and child development. As the Center for American Progress has noted, only 13 states even require background checks for workers at daycare centers—and four states don’t even screen for sex abuse history. Sixteen states permit teachers without a high school degree or G.E.D. to lead a daycare center.

Despite lackluster quality, daycare costs continue to soar. Average annual childcare costs are nearly $8,000. For families earning less than $18,000, the rising cost of childcare has consumes a prohibitive 39 percent of monthly income. Researchers believe this has led to a recent uptick in the number of stay-at-home moms, as the prospective cost of childcare leaves little net gain from working.

There are two federal programs that attempt to ease this financial burden. The Child and Dependent Care Tax Credit subsidizes some of the cost of childcare for families. It provides up to $1,050 per child for families earning below $15,000 a year, steadily decreasing to $600 for families earning over $43,000. However, it is nonrefundable, meaning any subsidy for low-income families’ childcare expenses is lost once their tax liability reaches zero.

The second program is a $5 billion annual block grant to the states to pay for childcare for low-income families. It was last authorized in 1996 as part of congressional welfare reform, but Congress had not previously attached strings to the grant to promote childcare quality.

Until now. When Congress reauthorized the CCDBG program, it attached important regulatory conditions addressing some of the hellish problems plaguing American daycare. If states want funding, their daycare centers must now run background checks on all employees. Providers must be trained in first aid, safe sleep practices, emergency preparedness, and avoiding shaken baby syndrome. And states must conduct annual random inspections of each daycare facility, publicizing the results online — a major improvement given that states like California and Iowa had been conducting inspections only once every five years.

These are significant reforms that ought to improve the quality of care across the country. They bring some uniformity to what had been a wildly disparate field of state regulations, bringing into the twenty-first century the places where 11 million children under the age of 5 spend most of their days.

But there’s still much more to be done. Though the reauthorization increases CCDBG funding somewhat over the next six years, it remains an underfunded program. It serves only one out of six eligible families, leaving long waiting lists for vouchers and subsidized daycare seats in many states.

Low-income families who aren’t lucky enough to get subsidized spots continue to pay out larger shares of their income for worse care. And a $600 tax credit for middle-class families is a relative drop in the bucket, enough to cover only about a month of childcare costs for the average family. It’s a drag on our economy when parents can’t justify working in the face of expensive childcare.

Still, it’s heartening to see Congress responding to an under-noticed national crisis. Reauthorizing the CCDBG with conditions to improve quality was the right thing to do. Perhaps going forward, the bipartisan coalition behind the reauthorization will push for more support for working families in order to expand childcare access and affordability.

What would conservatives do with Medicaid?

State-level Republican governors and legislatures are using the voluntary Medicaid expansion to leverage conservative policy concessions from the Obama administration. In states like Missouri and Pennsylvania, this has meant trying to enact work requirements for Medicaid. This would mean that, to receive Medicaid coverage, low-income Americans must be working, actively seeking out work, or participating in a job-training program.

This would be a significant departure from how Medicaid programs are currently constructed. No state has a work requirement for Medicaid — the original Medicaid statute prohibits conditioning coverage on work participation. And the Obama administration has flatly refused to grant a Medicaid waiver for proposed work requirement rules, so these state proposals are unlikely to become law anytime soon.

Nonetheless, the conservative instinct to condition health insurance for the poor on working is telling. And if a Republican wins the White House in 2016, red states would gain a new hearing to enact these ideas through federal waivers.

Interestingly, the attempt to tie Medicaid to working comes at the exact moment that health economists and policy wonks across the political spectrum are aiming to decouple of health insurance from employment for everyone else. Most economists consider traditional employer-sponsored insurance to be inefficient, as it locks workers into their current jobs for fear of losing coverage. They’ve therefore supported reforms to let individuals purchase insurance outside of their workplaces, like ObamaCare’s exchanges.

Some conservatives favor doing just the opposite when it comes to insurance for the poor. True, Medicaid coverage doesn’t have the same job-lock issues as employer-sponsored insurance. It’s a single-payer system, so it follows you from job to job.

But the urge to link Medicaid to employment flows from the conservative obsession with legislating work incentives for the poor. To some conservatives, any government benefit that reaches the poor must be conditioned on working, seemingly because poor people supposedly need constant shocks and prods to keep them employed.

Prominent conservative health economist Avik Roy showed the foolishness of this line of thinking in a recent interview. Roy, the author of How Medicaid Fails the Poor, recently released a conservative healthcare reform proposal detailing how we can “transcend” ObamaCare, using its existing structure to achieve conservative policy preferences.

In the interview, Roy explained — and justified — why his proposed revamp of Medicaid doesn’t include a work requirement:

 “[S]omebody asked me the other [d]ay, he said ‘Avik, is there a work requirement … ‘ (A conservative asked me this …) ‘Is there a work requirement in your plan for eligibility for these exchange subsidies for low-income people?’ and I said no.

“The guy said: ‘Well that’s a problem. We should have a work requirement.’

“I said to him: ‘Would you ask for a work requirement for a low-income unemployed parent to send his child to primary school?’ Of course he didn’t answer.”

Now of course, there may be conservatives for whom this point isn’t an argument-stopper — they may very well support work requirements for parents to send their kids to public school. After all, we’ve seen recent conservative discomfort with giving school kids government handouts in the form of free school lunches.

Though he defends public health insurance for the poor from shortsighted work requirements, Roy is hardly a fan of Medicaid. In Transcending ObamaCare, he argues that it generates little in the way of beneficial health outcomes for the poor in comparison to going uninsured.

Roy acknowledges that the root cause of this has been low Medicaid reimbursement rates to physicians. Doctors are reluctant to take Medicaid patients because Medicaid pays so little. Studies have found Medicaid patients are denied doctors appointments six times more often than those with private insurance — even when they tell the doctor that their child has a serious medical illness.

Why does Medicaid pay so much less? It’s because of the quirky and inefficient cooperative federal structure of Medicaid. “Medicaid is jointly funded by state governments and the federal government,” Roy explains. “Because neither party has full responsibility for the program, both parties have engaged in irresponsible behavior.” Federal regulations prohibit states from charging more to Medicaid patients, and states have responded to budgetary crises by slashing reimbursement rates, plunging as low as 29 percent of private reimbursement rates in New York.

(It should be noted that comparing Medicaid and private insurance reimbursement makes Medicaid look exceptionally bad. It fares a little better when compared to what Medicare pays, as calculated by the Kaiser Family Foundation. For instance, New York’s Medicaid program pays 55 percent of the Medicare rate, and the average state pays about two-thirds of what Medicare pays. Still underfunded, but a bit better.)

Roy’s solution is to blow up Medicaid and give the poor subsidized private insurance on health exchanges. Very well. It’s a totally legitimate proposal that would extend ObamaCare’s subsidy-exchange structure to those below 133 percent of the federal poverty line — the current cut-off point between expanded Medicaid and ObamaCare’s marketplaces.

And in the Obama era of liberal pragmatism, it’s a proposal in spirit with the technocratic amenability to using any practical means to achieve progressive goals. Obama wanted to enact affordable universal coverage in the United States — long a liberal goal — but didn’t mind using Mitt Romney’s private insurance-based marketplace structure to get us there, rather than insisting on liberal means like single-payer.

In the Medicaid expansion already, the administration has been willing to accept conservative ideas on how to extend health insurance to the poor. Indeed, Roy’s plan would largely mimic the so-called “private option” that Arkansas negotiated with the Department of Health and Human Services, using Medicaid funds to subsidize private insurance for those newly eligible for Medicaid coverage.

These are fine ideas, but are hardly the only way to heal Medicaid’s woes. As Roy explained, Medicaid’s problems stem from low reimbursement rates, which in turn stem from its inefficient joint federal-state financing structure. So why not simply federalize Medicaid? The federal government is better positioned to sustain social insurance programs than the stares are. It faces less budgetary pressure, particularly during recessions, because (unlike the states) it can run up deficit spending when social welfare program rolls expand in a weakened economy. This stabilizes financing for these sorts of programs, preventing them from being hollowed out by emergency state budget cuts.

The federal government could then raise reimbursement rates to something closer to what Medicare pays. And to get rid of any lingering discrimination against Medicaid patients, it could simply make such discrimination illegal, relying on testers in the same way housing discrimination laws are enforced.

Yes, this would cost the federal government more, but it would also relieve the states of a significant budgetary burden. It would also better serve the poor, as historically, the federal government has been a far better steward of low-income programs than the states have been. (See welfare reform / TANF; the 24 Medicaid expansion opt-out states; etc.)  And it might even save the government money in the longterm by letting the poor get more check-ups and preventive care, keeping the government from footing costly ER and advanced illness bills later on.

Still, Roy’s proposal is a refreshingly thoughtful constructive critique of ObamaCare — an all-too-rare conservative feat in the years since healthcare reform. If nothing else, maybe it will teach conservatives to love (or at least begrudgingly accept) ObamaCare by showing that it lays a foundation that they too can work with, finally breaking that fever that has plagued conservative politics since 2009. And though Roy would revamp Medicaid, at least doesn’t want to shackle the uninsured poor with work requirements.

Deterrence and deferred action

Last week, I wrote a piece at The Week arguing that conservative objections to the merits of a broader deferred action immigration order from President Obama (so-called “DACA 2.0″) must be either inhumane or outright dangerous.

The foundation for the argument was essentially this:

  1. Immigration officials lack the capacity to remove all undocumented immigrants
  2. They have therefore prioritized removing public safety and national security threats (Prosecutorial discretion)
  3. This prioritization and resource scarcity make it functionally impossible for peaceable, law-abiding immigrants to ever face deportation
  4. These peaceable, law-abiding immigrants are the same group that could benefit from expanded deferred action (we don’t know yet exactly who the President has in mind)
  5. A deferred action order would essentially notify this group that they are not at risk for deportation, thus allowing them to live and work freely
  6. Thus, his executive order likely amounts to little more than (1) reaffirming existing policy (the “Morton Memo”) + (2) notice

So conservative opposition must either be to (1) existing policy — which would be dangerous in that it would jeopardize public safety and national security (see point 2) — or to (2) notice — which would be cruel in that it would keep immigrants in the dark even though the immigration system will never deport them, driving them to exploitative work.

Danny Vinik has a post at The New Republic addressing the notice point. Vinik worries that the notice that deferred action provides crosses a legal line from prosecutorial discretion into policymaking. This is so because secret prosecutorial discretion (with no notice) preserves a deterrent against law-breaking (here, entering/remaining in the United States illegally).

Vinik relies on a law review article by law professor David Price. Under Price’s analysis, preserving the deterrent value of the law is key to furthering the will of Congress in the face of scarce prosecutorial resources. “[D]eclining to prioritize certain cases, as the executive branch might properly have done, may have very different effects from an announced, categorical policy like DACA,” Price asserts. “While the former preserves the deterrent effect of federal statutes by leaving all individuals covered by the statute in some jeopardy, the latter removes the risk of enforcement altogether. It thus contradicts the statutory policy to a degree that mere prioritization of enforcement resources does not.”

Vinik rightly notes that Congress legislates with many different goals in mind, and deterrence is just one of them. So how significantly should we weigh the deterrent goal here?

I’ve argued that the humanitarian costs outweigh the deterrent benefits of keeping prosecutorial discretion hazy and un-notified. For the sake of “supposedly discouraging migrant flows,” critics would have “millions of immigrants to needlessly live with the specter of deportation hanging over their heads. This would condemn them to living in the shadows and working in tenuous, often-exploitative conditions — even though immigration officials have no interest in deporting them.”

Even under closer scrutiny, I think this basic analysis holds up. To assess the deterrence benefits of going without deferred action, we must speculate a bit about who, exactly, might constitute the benefited group. One group of immigrants commonly mentioned is the parents of children granted deferred action under DACA. If so, can we really believe that the mere threat of potential deportation will cause any significant number of them to self-remove from the country? Would they really leave their (quasi-legalized) children behind, or even less likely, give up their children’s safe harbor from deportation?

In my assessment, the deterrent benefits for this group approaches naught, but the humanitarian costs remain huge. Though DACA children face no threat of deportation, their parents still struggle to earn a safe living and provide for a stable upbringing for their children under the threat of work penalties. This deepens the odds against DREAMer kids to do things like graduate high school or go to college.

We must also remember that any grant of deferred action will be (a) temporary, falling short of full legal status, and (b) require that its beneficiaries have lived in the United States for some years. Point (a) thus does preserve some deterrence value while removing the short-term humanitarian costs of the illusory deportation threat. Point (b) means that any deferred action will not benefit new arrivals, and therefore shouldn’t incentivize increased illegal immigration.

Skeptics might quibble with this. They might argue that deferred action shows that the U.S. is getting soft on immigration, and will eventually legalize everyone. A new arrival just needs to hang low and give it time. Moreover, they could point to the border child crisis, driven in part by misinformation throughout Central America over who is eligible for deferred action under DACA. Smugglers and coyotes may have lied to families and led them to believe that newly arrived children would be granted protection from deportation, even though this is plainly not the case under DACA.

But we can hardly afford to let our immigration policy be dictated by coyotes and deliberate misinformation. We would also face a diminished ethical quandary if a flood of adults arrived at the border, in comparison to that which we face from young children fleeing violence. Moreover, ICE already concentrates most of its removal forces on stopping new arrivals at the border. We therefore might already be equipped to blunt any unintended effects of expanding deferred action for immigrants already present.

Vinik and Price are right that there is surely some deterrence lost from expanding deferred action. But just how much deterrence is lost is a more pressing question. We must remember to consider the specifics of what the policy actually accomplishes and who it reaches to evaluate how much deterrence has really been forsaken. It might be the case that the deterrence cost is a price worthy paying.

Halbig’s hellscape

I’ve written (several times) about the Achilles’ heel in the subsidies-as-incentives theory of Obamacare – the cornerstone of the Halbig argument. In short, eliminating subsidies while retaining the ban on community rating and pre-existing condition exclusions will lead to insurance death spirals on Red State individual markets. This is because without subsidies, many more Americans would be exempt from the individual mandate on affordability grounds. They could then wait until becoming sick to purchase insurance, and insurers would be legally obligated to accept them.

The Kaiser Family Foundation crunches the numbers and figures out exactly how many more Americans would be exempt from the individual mandate if Halbig were the law of the land:

With subsidies available, less than 3% of uninsured people eligible for subsidies in the 36 federal marketplace states would be exempt. However, if the Halbig case prevails and the subsidies are invalidated in federal marketplace states, we estimate that 8.1 million (or 83%) of those formerly subsidy-eligible uninsured people would end up being exempt from the individual mandate. With the subsidies unavailable and the individual mandate rendered partially ineffective, it might be difficult to attract healthy people into the individual market and premiums could rise significantly in these states. The result could be what is commonly called a “death spiral,” as healthy people exit the market and premiums rise even more.

That’s a huge number. Of course, this would only be the first wave of Halbig‘s effects. As the healthiest people drop out of insurance pools, insurance premiums would rise further, making insurance unaffordable for even more people.

The legal import of this dynamic cannot go ignored. A full reading of Halbig means that Congress threatened states with a devastated insurance market if they didn’t create health exchanges. Such behavior can’t be constitutional — so the Halbig reading of the law cannot be right.

“Such” a mess

Wow. This morning was a doozy for ObamaCare, as two federal appellate courts issued opposite rulings on the validity of the law’s crucial insurance subsidies. Plaintiffs in each of these cases challenge whether the language of the Affordable Care Act gives the IRS authority to give subsidies to all Americans or to only those living in the 14 states that elected to create their own health exchanges. (I have previewed the D.C. Circuit iteration of this lawsuit previously: I, II, and III.)

How did the two courts reach different conclusions? After a quick read through of the two opinions, it looks like the divergence comes down to the meaning of the word “such.” Really.

First, consider the key statutory sections of the law at play:

  • Section 1311 of the law directs states to create health exchanges.
  • If states fail to do so, the law directs the federal government (through the Department of Health and Human Services) to “establish and operate such Exchange within the State.”
  • Section 36B gives subsidies to those who enroll in plans “through an Exchange established by the State under section 1311.”
  • But the law also defines the term “Exchange” as “an American Health Benefit Exchange established under [section 1311 of the ACA]” — the section that directs states to create exchanges.

How did the courts reconcile this mess? First, let’s start with the D.C. Circuit case, Halbig v. Burwell. The court there determined that the statute requires three elements to receive subsidies: “(1) an Exchange (2) established by the State (3) under section 1311.” “[F]ederal Exchanges satisfy only two: they are Exchanges established under section 1311. Nothing in section 1321 deems federally-established Exchanges to be ‘Exchange[s] established by the State.'”

It’s the court’s second element that dictates its resultant decision against the government and against broad subsidy availability. Does “such exchange” mean that an HHS-created exchange is (for statutory purposes) an “exchange established by the state”?

The Fourth Circuit Court of Appeals in King v. Burwell determined that it did. “Given that Congress defined ‘Exchange’ as an Exchange established by the state, it makes sense to read § 1321(c)’s directive that HHS establish ‘such Exchange’ to mean that the federal government acts on behalf of the state when it establishes its own Exchange.”

In other words, the Fourth Circuit thinks that the D.C. Circuit parsed the statute too finely. By authorizing HHS to create an exchange under Section 1311, the law lets the federal government fill in for the state. “Such” encompasses more than just a 1311 exchange — it means a 1311 exchange established by a state.

That’s the threshold disagreement here. And it’s a really, really important disagreement, for it determines whether health care reform will function in most of the country. I think the government has the better of the argument for reasons beyond the basic arguments over statutory text, which is frankly a mess to make sense of. But the consequences of that mess, to paraphrase Vice President Biden, are a big f***ing deal.

The submerged state common denominator

Sens. Kirsten Gillibrand (D-New York) and Rand Paul (R-Kentucky) have (improbably) teamed up to sponsor legislation that would expand the tax credit for child care. The bill would double the current child care tax credit from $3,000 to $6,000.

It’s a laudable legislative endeavor. As Gillibrand notes, the tax credit has only increased $600 since its 1981 inception, despite child care costs rising above the rate of inflation and consuming a growing share of working families’ incomes. The growing cost of childcare has not only strained family budgets, but may have begun nudging women out of the labor force.

The bill would take other worthy steps, like subsidizing employers to provide on-site childcare and aiding them in locating quality nearby childcare, along with encouraging more trained professionals to work in childcare. But the heart of the bill is the expanded tax credit. Importantly, this tax credit will be fully refundable, which provides full relief to low-income families who would already face minimal tax liability.

Paul’s support for the legislation is more than just a novelty, however. It is a convergence of the two parties on expanding what political scientist Suzanne Mettler calls the submerged state.

In her 2011 book of the same name, Mettler explains that the submerged state is the set of tax code subsidies that we’ve enacted to provide relief to working parents or to encourage home ownership. We’ve increasingly shifted our social welfare state toward a set of targeted carve-outs from tax liability, Mettler says, in ways that are often so subtle that beneficiaries don’t even realize they’ve received government aid.

Our polarized politics have spurred the proliferation of the submerged state. “In recent decades of conservative dominance and political polarization, the submerged state [. . .] became not a last resort but the template of choice for new policy initiatives,” Mettler writes. Targeted tax credits further conservative goals of reducing tax burdens generally. They also increasingly appealed to liberals as a way to provide relief to low-income and working-class Americans. During the ascendancy of conservatives during the Reagan era of the 1980s, liberals “increasingly acquiesced and supported policies such as tax expenditures because, to quote one member of Congress, they became ‘the only game in town,'” Mettler writes.

This interest convergence made submerged state tax credits something of a political least-common denominator in a polarized environment. It was a policy reform that conservatives and liberals could each support for ideological reasons, making them easier to muster through Congress.

We see this driving the Gillibrand-Paul alliance today. The traditional liberal ideal might be a universal childcare system of the type that Emily Badger described last month in the Washington Post: direct government provision of childcare centers across the country — a sort of public option for daycare — that we once had during World War II, nearly had again but for Richard Nixon, and would be unimaginable today.

But the pragmatic liberalism that developed during the 1980s (largely out of desperation) and has flourished during the Obama administration now embraces tax subsidies that allow Americans to better afford the services that the private market offers. This allows room for bipartisan agreement, particularly on relatively low-salience issues like childcare subsidies.

But joining conservatives in expanding the submerged state comes with distinct costs for liberals. Mettler describes how submerged state policymaking undermines democratic accountability, as taxpayers and subsidy beneficiaries barely realize that they’ve gained from public subsidies. This also poses asymmetric political harm for liberals, for when they propose bolder expansions of government assistance, few constituents realize that they have likely already benefited from tax subsidies. As Mettler notes, it’s the “keep your government hands off my Medicare” problem.

Expanding the subsidy we give to parents for childcare is a worthy policy goal. But liberals should nonetheless be wary of relying on submerged, low-visibility tax credits to provide relief to Americans in need. Admirable though the sentiment may be, subsidizing through tax credits undermines larger progressive goals of mobilizing government to alleviate the pressures on vulnerable Americans.

The inequality of our employer-provided welfare state

Over at The Week, I have an article on the unequal access to social insurance when we rely on employers to provide these programs as fringe benefits:

[L]eaving social insurance to the market . . . layers social insurance inequality on top of rising income inequality. Liberals want to alleviate this by giving all Americans access to the kinds of social insurance benefits that the highest earners enjoy now through their jobs. [. . .] In response to the liberal impulse to combat social insurance inequality, conservatives dwell on the potential disruptions to existing private contracts that might result from expanded access.

The proliferation of social insurance benefits among high earners and flush companies is telling. It’s a market signal that basic social insurance benefits like paid sick leave or paid maternity leave are life-improving benefits that workers are willing to pay for – whether through foregone salary or taxation.

Liberal policymakers ought to loosen the grip that employers have as exclusive providers of these social insurance programs just as they did during healthcare reform. A just society requires that these aspects of the good life be made available to all Americans regardless of earning power or employer clout.

Note that this does not require tearing up contracts or abolishing employer-provided fringe benefits. It only requires crafting public alternatives that supplement the employer-provided regime.

But the idea of an employer-provided welfare state and its corresponding inequities offers a powerful framing for liberal politicians and policy minds. Access-improving reforms to the welfare state immediately run into the conservative charge that they are socialist or would turn America into a coddled Nordic state. But liberals can point at the generous benefits offered by the Facebooks and Googles of the Fortune 500 and say, “Hey, the market has spoken. These benefits are goods that more people ought to have access to.” It’s not about giving low-income Americans what Sweden has. It’s about giving them what Google has.

The anatomy of a death spiral

We are still waiting for the D.C. Circuit Court of Appeals’ decision in the Obamacare subsidy case Halbig v. Sebelius. I’ve written before (here and here) on a significant glaring flaw in the challengers’ case. In short, their theory that Congress threatened states that refused to create exchanges with an inactive individual mandate (due to lost insurance subsidies) and an active ban on preexisting condition exclusions raises constitutional issues regarding improper coercion upon the states. This is because such a regulatory regime would plunge state insurance markets into adverse selection death spirals. Because the challengers’ theory raises constitutional problems, basic statutory interpretation principles will guide courts toward the government’s reading of the statute – a reading that permits subsidies to be offered on all exchanges. Therefore, the plaintiffs’ challenge to Obamacare will fail.

I’ve explained the contours of this argument, and I’ve justified why the government’s interpretation of the relevant Affordable Care Act provisions is reasonable. But I haven’t yet walked through the anatomy of an insurance market death spiral in much depth. Indeed, this is the linchpin of the theory of why the Halbig plaintiffs’ argument fails on its face.

An amicus brief filed by a group of health economists in Halbig provides a succinct outline of how a death spiral occurs:

Without premium subsidies, millions of people will be exempt from the mandate altogether or will choose to pay the tax penalty rather than purchase unaffordable insurance. Yet the sickest people will continue to sign up for insurance and insurers will have to cover them. The resulting higher premiums will threaten an adverse selection “death spiral”: as premiums increase, more and more healthy people will be exempt from the mandate or will choose to pay the tax penalty rather than buy insurance, leaving sicker people an ever greater portion of the risk pool, leading to escalating premiums, and even fewer enrollees.

The economists explain the death spirals of this kind have occurred in Massachusetts, New York, New Jersey, and the U.S. Virgin Islands.

Looking at two pertinent case studies is instructive. First, the individual insurance market in New York. New York experienced some of the sharpest post-Obamacare premium declines in the country. New Yorkers who were previously paying over $1,000 a month for individual insurance plans are now barely paying $300 – and that was even before Obamacare’s subsidies kicked in.

This rapid price decline was due to the fact that, prior to national health reform, New York had one of the most expensive and severely broken individual markets in the country. The reason, according to Sarah Kliff, was “a law passed in 1993, which required insurance plans to accept all applicants, regardless of how sick or healthy they were. That law did not, however, require everyone to sign up, as the Affordable Care Act does. [ ] New York has, for 20 years now, been a long-running experiment in what happens to universal coverage without an individual mandate.”

The result of the experiment was the highest health premiums in the nation. Insurers had to accept all comers, but there was no reciprocal obligation on all New Yorkers – both healthy and sick – to carry health coverage. This led to a sicker risk pool, leading to higher prices, leading healthier people to drop out of the risk pool in greater numbers, leading to a still sicker and more expensive risk pool. In short, a death spiral.

Second, consider the case of child-only insurance plans. In 2010, insurance companies started dropping child-only insurance plans. The reason? The Affordable Care Act was imposing a ban on preexisting conditions but no individual mandate (until 2014). A spokesman for America’s Health Insurance Plans explained the danger of this regulatory structure, fearing that it “provides a very powerful incentive for a parent to wait until their child becomes very sick before purchasing coverage.” Rather than face a market meltdown, many insurers simply stopped offering child-only policies under the new rules.

The meltdown in the child-only markets has not gone unnoticed, including by Halbig architect Michael Cannon. Importantly, Cannon concedes that the deliberate imposition of insurance market death spirals by Congress is on par with the coercive “gun to the head” threat of losing Medicaid funds in NFIB v. Sebelius:

Congress enacted even worse policy (community-rating price controls with zero protections against adverse selection) in both the CLASS Act and the markets for child-only health insurance, and enacted similarly bad policy (community rating with weak protections against adverse selection) in the non-Exchange individual market and in U.S. territories. Moreover, the potential adverse-selection effects amici describe are not out of character for a Congress that was trying to put “a gun to the head” of uncooperative states, which is what the Supreme Court found this Congress was trying to do.

Cannon fails to appreciate how this accusation of congressional coercion – the logical endpoint of the plaintiffs’ argument – undermines the anti-Obamacare case. If the plaintiffs’ theory depends on coercion by Congress, constitutional avoidance doctrine commands courts to adopt the government’s theory that the hastily-drafted Affordable Care Act sections at issue can be plausibly read to make subsidies available on all health exchanges.

The experiences of New York and the child-only insurance market shares the same characteristics as the Halbig plaintiffs’ understanding of congressional intent in Obamacare: the absence of an individual mandate, coupled with a requirement that insurers accept all customers regardless of their preexisting conditions. Under the Court’s coercion doctrine, Congress could not have constitutionally threatened states with becoming New York if they didn’t create health exchanges. It would have devastated their insurance markets.

So on the threshold question of how exactly we are to interpret the statute, and how Congress meant Obamacare’s subsidies to work, courts can’t buy what the plaintiffs are selling. In their quest to kill Obamacare and escape the individual mandate, the Halbig challengers prove too much by bringing on a death spiral.