Analysis of King v. Burwell -- The Obamacare Tax Subsidies Case

Here are my thoughts on the Supreme Court’s decision on June 25, 2015, in King v. Burwell.  In that case, the Court ruled that tax subsides are available to those purchasing health insurance on both (1) exchanges set up by a state itself, and (2) exchanges set up by the federal government for a state.  I agree with the Court’s ruling.  While I believe reasonable minds can differ in this case—something that is virtually always true when the decision is 6 to 3—I think the majority has the better argument, and by a good margin.

 

It will be difficult for me to write a summary of the case that digs deeply into the substance without almost rewriting the various opinions in full.  That is because complex statutory interpretation cases, like complex contract interpretation issues, come down to the weight of the evidence taken as a whole.  There are numerous relevant categories of evidence in these types of lawsuits.  And there are often multiple pieces of evidence from each category.  That is certainly true in King v. Burwell.  As a result, any true summary that I could write would leave out critical detail.  Given that, and for reasons of time, I am going to summarize the opinions in more general terms and include selected arguments for each side.

 

Before getting to those thoughts, however, let me note that statutory interpretation is not rocket science.  And it does not take a law degree to understand the basic concepts, even if it does take such a degree to capture some of the deeper nuances.  Accordingly, given that the case is not that long, those nonlawyers (and lawyers) deeply interested in this matter ought to consider reading the case in full rather than relying solely upon my assessment below.  The opinion is available here on the Supreme Court’s website.

 

Now, to business.  The Affordable Care Act says that tax subsidies are available to those people who buy health insurance on an exchange “established by the State.”  Given that the statute defines “State” to mean the 50 states and some territories, there is a good case that the four-word phrase “Established by the state” means that a person may only receive tax subsidies if the person buys insurance on a state exchange; those purchasing on a federal exchange may not receive the subsidies.  However, the single most important principle in statutory, contractual, and constitutional interpretation is that language must be read in context.  And the full context of the statute creates a powerful argument that “established by the State,” as used in the relevant provision, does not rule out subsidies on federal exchanges.  Indeed, Chief Justice Roberts marshals multiple types of arguments to make a compelling case that the tax subsidies apply to both state and federal exchanges, including (1) textual arguments (a close reading of the text), (2) structural arguments (analyzing the relationship of textual provisions throughout the statute), (3) purposive arguments (assessing the purposes or goals of the act, as reflected in both the language of the statute and material from outside the statute—called “extrinsic evidence”—such as legislative history), and (4) consequential arguments (focusing on the consequences of various interpretations and how those consequences match up with the language and purposes of the statute).

 

For an example of a close textual reading, the statute provides that if the state chooses not to set up an exchange, the federal government “shall . . . establish and operate such Exchange within the state.”  (Emphasis added.)  Roberts argues that this means that federal exchanges essentially stand in for state exchanges and should be treated in the same manner for many purposes.  In other words, Roberts is saying that the words “such exchange” support the conclusion that any federal exchange just is a state exchange (for many purposes).

 

For an example of a structural reading, Roberts points out that if tax subsidies are only available on state exchanges, then there will be no individuals who meet the tax subsidy eligibility standards in states with a federal exchange.  But, Roberts continues, the statute “clearly contemplates” that there will be qualified individuals for every exchange because all exchanges must make available health plans for qualified individuals.  How can an exchange make available health plans for qualified individuals if there are no qualified individuals for that exchange?  To put this argument in broad terms, Roberts is arguing that between the two readings of “established by the State,” one creates a conflict with other language in the statute and one doesn’t.  Consistent with long-established canons of statutory (and contractual) interpretation, it is better to chose the reading that avoids the conflict—the reading that allows subsidies for the federal exchanges.

 

Here is an example of an argument that combines structural, purposive, and consequentialist reasoning.  First, some basic principles.  The guaranteed issue provision requires that insurance companies provide insurance to anyone who requests to buy it regardless of preexisting conditions.  The community rating provision requires that everyone be charged largely the same price for health insurance regardless of their health condition.  And the individual mandate provision requires that everyone own insurance or pay a tax/penalty.  Roberts argues that these provisions will not work together as intended by the Affordable Care Act if the tax subsidies are not available on federal exchanges.  That is because, without the subsidies, many people cannot afford to buy insurance on the exchanges.  Most of those people will then be exempt from the individual mandate under one of the exceptions in the law and thus need not buy insurance.  Next, if the healthy people in that group do not buy insurance and the sick people do (a very likely occurrence), that will raise premiums, pricing more healthy people out of the market.  Finally, as more people are priced out of the market, insurance will become even more expensive, creating a feedback loop that Roberts calls a “death spiral.”  That feedback loop will defeat essential purposes of Obamacare reflected throughout the statute.

 

Roberts makes other powerful arguments.  In fact, I agreed with almost every point he made.  I do believe he overstated his case in a couple of places.  But overall, his arguments—and responses to the dissent’s arguments—were excellent.

 

Turning to the dissent, Justice Scalia presented a number of solid points.  But unlike Roberts, he also makes a number of unpersuasive arguments; and he overstates in multiple places.

 

Scalia makes more than a dozen distinct arguments in the dissent.  But I think three stand out from the others as strong bases for his position.

 

First, Scalia contends that the words “exchange established by the State” will have no meaning in the provision in question if that provision applies to both federal and state exchanges.  Such a result violates the canon of interpretation that all words in a legal text should be given meaning, if possible.  But I called this a “canon” for a reason.  It is not a rule.  It is just one, albeit important, factor to consider when interpreting legal language.  Moreover, the canon may not even apply here.  Remember that Roberts essentially argues that “established by the State” incorporates the federal exchanges due to other language (e.g., “such exchange” discussed above).  That would mean the three highlighted words do have meaning—just not their ordinary meaning.

 

Second, Congress chose to use the word “exchange” in some places and the phrase “exchange established by the State” in others.  That suggests that “exchange” and “exchange established by the State” have different meanings.  Point for Scalia.  This is one of Scalia’s strongest arguments, if not the best.

 

Third, Congress wrote that if a territory establishes an exchange (e.g., Puerto Rico, Guam, or the U.S. Virgin Islands), it shall be treated the same as if a state established the exchange.  Why didn’t Congress do the same with respect to exchanges established by the Federal Government?  Another good argument for Scalia.  However, Roberts essentially argues, in response, that Congress did not need to craft a similar provision for federal exchanges because the phrase “such Exchange,” discussed above, obviated the need to do so.

 

These three points, when mixed with a couple of other decent arguments Scalia offers, are simply not sufficient to overcome the much larger collection of persuasive arguments that Roberts presents.

 

As I said, Scalia also makes some unpersuasive arguments throughout his dissent.  Here are three examples. 

 

First, Scalia writes that it would be “hard to come up with a clearer way to limit tax credits to state Exchanges than to use the words ‘established by the State.’”  This is false.  In fact, it is actually quite easy to come up with stronger language.  Scalia fails to see this because he ignores a basic and recognized principle in both statutory and contract drafting: If your goal is to exclude something, it is always more effective to exclude that thing expressly than it is to implicitly exclude it by simply leaving that thing out of a list of what is included.  And failing to follow this rule has cost legislatures and contracting parties in more cases than I can count.  Given this principle, here are three examples of how the language could more clearly have prohibited subsidies on the federal exchanges:

 

1.         “. . . exchanges established by the State, excluding those created by the Secretary of Health and Human Services,”

 

2.         “. . . exchanges established by the State pursuant to [the section setting forth the precise manner in which state exchanges are created] and not pursuant to [the section setting forth the precise manner in which federal exchanges are created]

 

3.         “. . . exchanges, except those established by the Secretary of Health and Human Services.”

 

All of these are clearly superior to the current wording in the statute, if the goal is to establish that tax subsidies do not apply to federal exchanges.

 

Second, Scalia essentially argues that “established by the State” must have the same meaning in each provision in which it is used in the act.  And if the four words are ignored when it comes to tax subsidies, the same must be true elsewhere.  This is incorrect.  It is indisputable that the same language can, and often does, mean different things in different contexts, even within the same statute or contract.  The fact that the same phrase is used in two places in a law or contract is important evidence that the meaning is the same.  But it is not the stringent rule that Scalia suggests.  Moreover, Roberts is best read as not actually arguing that the words mean something different in the relevant provision.  Instead, he is arguing that the provision, when read in context, does not exclude tax subsidies from federal exchanges because federal exchanges are supposed to be treated the same as state exchanges (at least for this purpose).

 

Third, according to the majority, if the tax subsidies are not available, then certain other provisions in the law would make little sense.  Scalia responds by saying that this only shows “oddity, not ambiguity.”  But here Scalia misses the point.  Oddity is precisely one of the bases that counts in favor of rejecting an interpretation.  Scalia is correct that oddity and ambiguity are two different things.  But the odder the result of an interpretation, the weaker that interpretation is.  Scalia continues by saying that laws often have unusual or mismatched provisions.  But when an interpretation leads to an usual result or a mismatch of two sections of a statute, that counts against the interpretation.  This result flows from multiple canons of construction, including (1) the preference for reasonable interpretations, (2) the preference for interpretations that are consistent with the principal purpose of a law, and (3) the canon that provisions should be read in harmony if possible.

 

Let me end by noting that King v. Burwell is a rather run-of-the mill case on statutory interpretation.  The opinions reflected some of the basic divisions among the members of the Court (and among lawyers and law professors more generally) regarding the appropriate method for interpreting statutes (and contracts and constitutions).  But both Roberts’s majority opinions and Scalia’s dissent were perfectly normal Supreme Court opinions addressing a perfectly mundane case (from the perspective of statutory interpretation).  Yes, this case had high political salience.  Thus, the attention it will get and the rhetoric it will spawn is going to be abnormal.  Indeed, Scalia’s dissent is filled with sky-is-falling rhetoric.  But that is rather common in dissents in cases like this—cases with high political salience.  Had Roberts and Kennedy sided with the Conservatives, I am rather sure that one of the four Liberals would have used similar rhetoric in dissent, much like Ginsburg (wrongfully) did in the Hobby Lobby/religious accommodation case.   In short, do not believe any statements that this case has “fundamentally altered our constitutional order” or any other such hyperbolic nonsense.  This was a typical statutory construction case, quite similar to the statutory and contract interpretation disputes American courts handle every day.

The Same-Sex Marriage SCOTUS Oral Argument

A number of thoughts regarding the oral argument in the same-sex marriage case this week.  Before I begin, please note that for reasons of time, this posting is directed more towards lawyers or others with some legal training.  I am in the middle of final exams and so I do not have time to lay everything out in full.

 

First, among the commentary I have reviewed, I found this posting at the Originalism Blog to be quite interesting.

 

Equal Protection versus Due Process in Obergefell v. Hodges

Michael Ramsey

 

This is my favorite part of the oral argument in the same-sex marriage cases (from Lyle Denniston's analysis at SCOTUSblog):

 

The federal government’s view is that the Court should rule in favor of same-sex marriage, based upon the Fourteenth Amendment’s guarantee of legal equality.   He [Solicitor General Verrilli] had no answer when Justice Kennedy asked him for the government’s views on whether gays should have a right to enter marriage, under the Fourteenth Amendment’s Due Process Clause, because marriage is a fundamental right.   The government was not prepared to opine on that issue, Verrilli told Kennedy. 

 

I like the subtle message sent here.

 

Among other reasons, there's a practical (though originalist-based) argument for preferring the equal protection clause over the due process clause.  A problem with finding a fundamental right to marriage unmoored from the traditional understanding of marriage is that there is no obvious stopping point.  If same-sex couples have a fundamental right to marriage, why not polygamists? Or indeed people who want to marry their brother, or their dog, or their piano?  At some point the Court would have to say what is "really" a marriage and (critically) what is not.  There is simply no way to do this on the basis of anything other than the judge's moral intuition.  Like other arguments from moral intuition purporting to rest on the due process clause, the whole enterprise is corrosive of the rule of law.  No one will think of it as anything but the Court making it up.   (This concern has been raised by commentators generally sympathetic to same-sex marriage, for example here by Elizabeth Price Foley and here by Josh Blackman.) 

 

But this concern has force only if one thinks of the case as being about marriage (the due process perspective) instead of being about discrimination (the equal protection perspective).  Equal protection offers a more defensible grounding, and a more defensible stopping point.  As I've argued, the key is describing sexual orientation discrimination as a form of caste legislation.  (In modern doctrinal terms, to think of laws discriminating against people of same-sex sexual orientation as targeting a "suspect class" and thus requiring a "compelling" justification.)  It seems fairly straightforward to say that sexual orientation -- but not a preference for polygamy, or for other nontraditional versions of "marriage" -- constitutes such a suspect class.  Put another way, sexual orientation discrimination can be analogized to race discrimination (clearly proscribed by the Fourteenth Amendment) in a way that laws against other kinds of non-traditional marriage cannot.  And that provides a stopping point that can be explained on legal grounds connected to the Amendment's original meaning.  

 

Let me begin by saying that I actually disagree that there is no obvious stopping point under the due process analysis.  When addressing individual rights issues under the Constitution, the Court has generally balanced the importance of the right against the importance of the state’s regulatory interest.  On the state interest side of the balance, I think it is actually rather easy to show a rational basis (the lowest level of protection for individual rights), an important interest (for intermediate scrutiny, the middle level of protection for individual rights), and even a compelling state interest (for strict scrutiny, the highest level of protection for individual rights) when it comes to polygamy.  Research shows, for example, that polygamy raises consent issues far more often than two-person marriage.  And, more importantly, polygamy has clear impacts on third parties since the research shows that men of privilege tend to take extra wives, depriving other men of the opportunity to marry.  In addition, I think there are at least important interests and often compelling state interests with respect to the other hypothetical extensions of marriage rights.  On the rights side of the balance, I think there are historical, doctrinal, and conceptual bases that can be used to draw a line short of polygamy and the other hypotheticals Ramsey offered.  Thus, I’m not concerned with most slippery slope arguments here.  With that said, I think the equal protection argument is more straightforward and works better for the general reasons Ramsey articulates.  Even if the due process/fundamental rights analysis is not as problematic as Ramsey contends, he is right that it is more troublesome than the equal protection analysis, where slippery slope arguments are of virtually no concern.

 

Second, I was somewhat disappointed by the oral advocacy of the four lawyers who argued the case.  In particular, I was hoping that respondents’ counsel, John Bursch, the former solicitor general of Michigan, would do a better job.  He was outstanding in the affirmative action case out of Michigan last year.  Thus, I had very high expectations for him.  While his performance in the marriage case was the best of the four advocates by a good margin, I thought he was considerably below the standard he set last year when his argument was one of the best I’ve heard in some time.   In fairness, perhaps part of the drop off is because his legal arguments are much weaker this time—in the affirmative action suit, his position was close to a slam dunk.  Bursch’s principal argument here was that (1) same-sex marriage will cause heterosexuals to value marriage less; (2) that will decrease opposite sex marriage, which will decrease the number of children raised in the state-preferred manner; and (3) even if you don’t believe these points, the state rationally could believe it.  Thus, since rational basis applies here (again, the lowest standard of individual rights protection), the same-sex marriage bans are constitutional.  It took him until near the end of his argument to finally get this point out, though he tried to state it multiple times before then.  Part of the problem was that he kept getting interrupted by the justices, but part of it was that he kept going back to the beginning and did not follow through until near the end of his time.  In sum, given the importance of the issue, I wish the advocates on both sides had been more effective.

 

Third, I found it interesting that Chief Justice Roberts was the one who brought up the sex discrimination argument.  Sex discrimination is governed by intermediate scrutiny.  And if intermediate scrutiny (or something higher) is the appropriate standard, there is no way the same-sex marriage bans cans survive.  Assuming the traditional doctrinal structure applies here, the defenders of the bans must establish two points in order to win: (1) rational basis governs the constitutionality of the same-sex marriage bans; and (2) there is in fact a rational basis for the bans.  I think the defenders have strong legal arguments on both of these issues (for an example on the latter point, see Bursch’s argument discussed in the previous paragraph).  But I think that the correct legal result is the defenders of the bans lose on both points (though not by a large margin).  Thus, I would strike down the bans on same-sex marriage under the equal protection clause of the Fourteenth Amendment.  My reasoning is as follows: (1) Roberts is right; this is sex discrimination, so intermediate scrutiny applies.  (2) Sexual orientation should at least be a quasi-suspect class given the history of discrimination against gays, lesbians, and bisexuals.  Once again, intermediate scrutiny applies.  (3) Even if the proper standard is rational basis, the argument that same-sex marriage will harm children is just too attenuated to meet even that extremely deferential standard.  And none of the other rational basis arguments work at all.  Note that because I am an exceptionally strong supporter of same sex marriage on moral grounds (indeed, I think it is arguably the last major example of formal inequality in American law), I must acknowledge that my legal analysis here could be biased by my moral views.  However, please also note that I believed that same-sex marriage bans were constitutional from Goodrich (the Mass. case from 2003—the first time I really thought this issue through legally) until Windsor (in 2013).  Windsor pushed me over me over the edge—as it seemed to do for the vast majority of both Liberal and Conservative federal and state judges who have considered same-sex marriage over the the last two years.  With all of this said, Justice Kennedy has been willing to jettison the traditional doctrinal framework (i.e., the tiers of scrutiny) in past sexual orientation cases.  In fact, he did so in both Lawrence v. Texas (in 2003, the case that upheld the right to engage in sodomy) and in Windsor.  Given this, and given the willingness of the other justices to go along with Kennedy’s alternative modes of constitutional analysis, my discussion in this paragraph might all end up being entirely besides the point.

 

Fourth, oral argument is seldom a sure sign of things to come, and thus I remain largely agnostic on what is going to happen.  I lean slightly towards it being 5-4 striking down the bans, but I will not be surprised at all if Kennedy joins the Conservatives and it is 5-4 in favor of the bans.  I also won’t be surprised if both Kennedy and Roberts join the Liberals, making it 6-3.  Anything other than those three options would surprise me.  This has been my view since shortly after Windsor and the oral argument this week did nothing to change my very modest prediction.  (Note that I have no predictions re. the full faith and credit analysis, just the 14^th Amendment).

 

Thanks for reading.  As always, all comments are welcome.

 

Josh