This spring, the Supreme Court will again hear a case that could bring down the Affordable Care Act. The question this time is whether it’s legal for the IRS to issue tax credits to low- and middle-income people who buy health insurance in the 34 federally run exchanges under the ACA. To date, more than five million people have bought insurance through a federal exchange, and more than 80 percent of those people have received tax credits. Those credits are now in jeopardy.
The case, King v. Burwell, is an appeal from a decision by United States Court of Appeals for the Fourth Circuit, which ruled in favor of the government in July. On the same day, in a parallel case, two conservative judges on a three-judge panel of the D.C. Circuit ruled the opposite way, striking down tax credits for insurance bought on federal exchanges.
The Supreme Court’s choice to hear the case is striking because the D.C. Circuit had vacated the panel’s decision, announcing that it would take the unusual step of rehearing the case “en banc,” or as a full court. Pending the rehearing, then, there was no disagreement between circuits, which is usually a prerequisite for the Supreme Court to take a case. Many observers have taken the high court’s surprising decision to get involved so early as a sign that at least four justices (the number needed for the court to hear a case) are eager to overturn the Fourth Circuit and invalidate the subsidies.
Without tax credits, as the government argues, the whole system would fall apart. (Some states without exchanges might end up creating their own, but there’s no guarantee.) Millions would be unable to afford insurance, so they would be exempt from the law’s mandate to buy coverage—and without enough healthy people buying insurance, the requirement that insurers cover people with preexisting conditions without charging more would make premiums for everyone skyrocket. The tax credits, individual mandate, and preexisting condition rule make the health care law a “three-legged stool” that topples if any leg is removed.
Judge Thomas Griffith, writing for the D.C. Circuit panel, was unmoved by those concerns. He held that because the ACA explicitly refers to tax credits only for insurance “enrolled in through an Exchange established by the State,” the law “does not authorize the IRS to provide tax credits for insurance purchased on federal Exchanges.”
The very real chance that the Supreme Court will agree with Griffith reveals the dangerous potential of a misleadingly simple theory, championed by some conservative academics and judges, known as textualism.
Textualists—most notably, Supreme Court Justice Antonin Scalia—argue that when interpreting a statute, a judge must consider nothing but the words of the text, rather than the legislative intent behind it. As Scalia puts it in a 2012 treatise, Reading Law, co-written with lexicographer Bryan Garner, a judge’s sole task is to figure out “how a reasonable reader, fully competent in the language, would have understood the text at the time it was issued.” Considering anything but the meaning of the words themselves gives judges too much discretion to rule according to their own personal or political preferences. It’s okay, Scalia says, to consider the purpose of the law, but only to the extent that the purpose can be divined from the text. It is not okay to consider outside evidence of what the legislators actually meant when they passed the law—even evidence that the legislators intended for judges to use, such as committee reports.
At first glance, the plaintiffs’ case in King looks like an open-and-shut textualist victory. Under the ACA, the federal government has set up insurance exchanges in the 34 states that chose not to create their own. Section 36B of the law, which establishes tax credits for low- and middle-income people to buy insurance, calculates those credits in terms of premiums on insurance “enrolled in through an Exchange established by the State.” The federal government is not a state, so credits seemingly aren’t available on federal exchanges—which means the IRS regulation providing credits through those exchanges violates the statute.
According to the plaintiffs, there’s nothing strange about that outcome: Congress must have chosen to withhold tax credits from residents of states that have federally run exchanges as a way to push the states to set up their own. It just underestimated how many states would reject the deal. Otherwise, why would Congress have added the words “established by the State” to Section 36B, when it could have just written “enrolled in through an Exchange” and left it at that? The government can’t get out of the plain meaning of the law now, the argument goes, just because it lost a risky bet.
That argument looks great—but only from a very particular angle. Start looking at how Section 36B fits with the rest of the law, and it becomes clear that the plain wording isn’t so plain; the question of whether a federal exchange can be treated as a state exchange for tax credit purposes is at least ambiguous. That’s important, because under Supreme Court precedent, if the law is ambiguous—merely ambiguous—the court is supposed to defer to the IRS’s interpretation.
This gets a little complicated, but the basic point is that other parts of the law make no sense if a federal exchange can’t be treated as one established by a state. Crucially, the section explaining who can buy insurance on the exchanges defines a “qualified individual” as someone who “resides in the State that established the Exchange.” Taking that section literally would mean that Congress intended for the federal exchanges to have no customers. To avoid that absurd result, the phrase “resides in the State that established the Exchange” must be read as treating the federal government as equivalent to the state. It’s therefore sensible, and consistent, to interpret the tax credit section in the same way.
(Judge Griffith’s response to this argument was one for the ages: “Federal Exchanges might not have qualified individuals, but they would still have customers—namely, individuals who are not ‘qualified individuals.’” For someone who says he’s devoted to plain meaning, it’s quite a stretch to say that people who aren’t qualified are still, well, qualified.)
So, as some other observers have pointed out, even a strictly text-based approach should favor the government, because one need not go beyond the text of the law to show that allowing tax credits under the federal exchanges is reasonable. But the fact that we’re even talking about this is testament to the success textualists have had in shaping the terms of the debate. The plaintiffs’ theory should fail under a responsible textualist approach, but it wouldn’t even get consideration—and millions of Americans wouldn’t be at serious risk of losing their health insurance—if some judges weren’t willing to ignore the obvious fact that creating a federal exchange without tax credits would make no sense.
Textualism’s shortcomings have been well documented. There’s no indication that textualist judges are any less biased than non-textualists, and recent research suggests that textualists misunderstand how Congress actually operates. One of Scalia’s favorite arguments is that by consistently using his approved canons, or rules, judges teach Congress to draft laws more clearly. For example, applying the “canon against superfluity,” which says every word of a statute must be given effect, encourages Congress to avoid repetition. Yet a study by Abbe R. Gluck and Lisa Schultz Bressman, law professors at Yale and Vanderbilt, respectively, found that members of Congress and their staffs either are unaware of or consciously disregard many of the most prominent rules, including the superfluity canon. (One reason is that their target audience is usually administrative agencies, like the IRS—not federal judges.) That should give pause to anyone who argues that Congress simply must have included the words “established by the State” for a reason.
These critiques of textualism are powerful, but they don’t get to the root fallacy underlying the whole concept.
The theory’s animating principle is the idea that, as Justice Oliver Wendell Holmes put it more than a century ago, “We do not inquire what the legislature meant; we ask only what the statute means.” But that is a false dichotomy. To see why, consider Scalia and Garner’s treatment of a problem posed by H.L.A. Hart in a classic 1958 Harvard Law Review article: “A legal rule forbids you to take a vehicle into the public park. Plainly this forbids an automobile, but what about bicycles, roller skates, toy automobiles? What about airplanes? Are these, as we say, to be called ‘vehicles’ for the purpose of the rule or not?”
Hart’s puzzle is often held out as an example of the inherent ambiguity of statutory language, but Scalia and Garner think solving it is a piece of cake; all you need to do is figure out the common meaning of the word “vehicle.” First, they identify the purpose of the law as “excluding certain things from the park—presumably things that would otherwise commonly be introduced.” Next, they look at a few dictionaries (textualists are often mocked, to little effect, for cherry-picking dictionary definitions) until they find a definition they like: “A self-propelled conveyance that runs on tires; a motor vehicle.” But wait, they say: “if taken literally, this definition would embrace a remote-controlled, miniature model car—which does not seem right. The proper colloquial meaning in our view (not all of them are to be found in dictionaries) is simply a sizable wheeled conveyance … Taking the word to mean a sizable wheeled conveyance would exclude form the park only one of the five examples given by Hart: automobiles.”
There is a lot to take issue with in that passage, which Scalia and Garner hold out as a model of textualist analysis, but the two most important phrases are almost throwaways: “presumably things that would otherwise commonly be introduced,” and “which does not seem right.” What’s the basis for presuming that the law is addressed to things that would be commonly introduced into parks? Why doesn’t a ban on toy cars “seem right”? The answer has nothing to do with the “common usage” of the word vehicle. The answer, rather, is that of course the law is aimed at things that otherwise might be brought into a park—otherwise, why write the law? And of course it doesn’t ban toy cars; who would ban toys from a park?
But now, we’re no longer talking solely about the meaning of words; we’re talking, quite rightly, about the intent of whoever wrote them. Scalia and Garner claim that their inquiry is into “how a reasonable reader … would have understood the text,”; but that’s misleading. What they are really asking, without saying so, is how a reasonable reader would understand the intent behind the text. There is no way to resolve an ambiguous statute without implicitly asking that question. Most of the time, there is no way to decide whether a statute is clear or ambiguous in the first place without checking it against our sense of what a legislator would have probably intended by it. The question about how to read the language of a law smuggles in the supposedly forbidden question about intent, since any reasonable reader would take into account the fact that words in laws are intended to accomplish something specific, something different from other forms of communication.
We see examples of this in the D.C. Circuit opinion striking down the tax credits.
At one point, Griffith dismissed one of the government’s arguments by saying it would be “odd” for Congress to define a characteristic of exchanges in a section of the statute other than in the definition section. The implication is that judges should avoid interpretations that lead to “odd” results. Fine, but if that’s true, wouldn’t it be really “odd” for Congress to prohibit tax credits under the federal exchanges in two obscure clauses defining credit amounts and coverage months rather than in the provision specifying who is eligible for credits? Either way, it’s an argument about what Congress intended, not about what the words themselves mean.
If every judge, textualist and non-textualist alike, is ultimately trying to get to an objective reading of the legislative intent, then there’s no reason to ignore non-textual evidence of what that intent was. In King, that includes the knowledge, common at the time the law was passed, that no exchange would work without credits, mandatory coverage, and the requirement that preexisting conditions be covered. Even if Congress had wanted to encourage states to establish exchanges, it wouldn’t have done so by creating self-destructing federal exchanges without tax credits. If judges pay attention to that evidence, then the plaintiffs’ theory disintegrates. With all due respect for Justices Holmes and Scalia, the notion that a statute’s meaning can be separated from what its authors meant is nonsense.