Just How Indefensible Does an Argument in a Government Brief Have to Be to Cause All the Career Litigators in Federal Programs to Withdraw from the Case?

Late yesterday afternoon, something remarkable—perhaps unprecedented—happened at the Department of Justice.   In Texas v. United States, No. 18-167 (N.D. Tex.)—a suit brought by 18 states, two governors and two private individuals, seeking an injunction against the operation of the entire Affordable Care Act (ACA)—three of the four attorneys representing the United States from DOJ’s Federal Programs Section of the Civil Division, including an Assistant Branch Director (Joel McElvain) and a Senior Counsel (Eric Beckenhauer), moved to withdraw from the case, leaving from Federal Programs only the politically appointed head of the Section, Brett Shumate (who filed an appearance on Tuesday) and a line attorney who was hired only a few weeks ago (Daniel Mauler, who entered an appearance earlier yesterday).  As soon as this remarkable filing was made, close observers assumed that it might be a sign that the Administration could find virtually no career FedPrograms lawyers—none who have been with the Department since before May, anyway—who were willing to file the government’s substantive brief that was due at close of business yesterday.

Perhaps such a mass withdrawal of DOJ attorneys from a case has happened before.  If so, however, I am not aware of it.  (The 1981 Bob Jones litigation in the Supreme Court, in which the Acting Solicitor General noted his withdrawal in the government’s brief after the government switched positions, is probably the closest analogy.)  It’s important to understand that Federal Programs lawyers often put their signatures to briefs that make, shall we say, very aggressive and unlikely-to-prevail arguments in defense of federal programs and statutes.  This is not a timid litigating component.  Of course, lawyers in that section, like any other, occasionally request not to be assigned to a particular case when they have moral or other serious qualms about the government’s actions or arguments.  But such recusal is not common and, more to the point, once those lawyers do represent the government in a case they rarely seek the court’s permission to withdraw representation because they’ve concluded that an argument their supervisors choose to include in a brief is too embarrassing or indefensible to defend.  For three such respected DOJ attorneys to do so simultaneously—just hours before a major filing, and without replacement by any other career lawyers other than a rookie—is simply flabbergasting.  It did not portend well for the brief that was to follow.

And, sure enough, a couple of hours after the mass withdrawal motion, the government filed its brief in response to plaintiffs’ motion for a preliminary injunction, which made it clear why the FedPrograms attorneys, en masse, refused to be any part of this case:  The government’s brief includes not one but two arguments about the intent and effect of Congress’s amendment to ACA in December 2017 that are simply implausible—some might even say preposterous, in light of what everyone knows about the intent and design of the Republican Congress last December. 

The Government’s Implausible Reading of the 2017 Amendment to Section 5000A, the “Individual Mandate” Provision

In order to understand the extraordinary nature of the government’s arguments, it’s first necessary to recall how the so-called “individual mandate” provision of the ACA, 26 U.S.C. 5000A, operates and how the Supreme Court construed that provision when it upheld its constitutionality in 2012 in NFIB v. Sebelius. 

Subsection 5000A(a) provided (and still does) that most individuals “shall … ensure” that they and their dependents are “covered under minimum essential coverage,” i.e., that they maintain health insurance.  Subsection 5000A(b), in turn, provides that if (most) of those covered individuals covered by subsection (a) fail to maintain such insurance, they must pay a “shared responsibility payment” to the IRS “in the amount determined under subsection (c).”  (I say “most” because lower income persons and members of Indian tribes are exempt from the subsection (b) “penalty,” something I discuss below.)  When the Supreme Court reviewed the ACA, the amount of this “penalty” option was the greater of 2.5% of household income or $695.

A majority of the Court in NFIB held that Congress would lack the constitutional authority to actually impose upon individuals a legal duty to maintain insurance.  Nevertheless, it upheld Section 5000A because it construed the provision, read as a whole, to give individuals a choice between (i) maintaining insurance and (ii) making the payment to the IRS, which the Court deemed a tax that Congress could have independently imposed.  Because one of the two options (the latter) was within Congress’s constitutional authority to require, a fortiori Congress had the power to offer people a choice between the two.

On December 22, 2017, President Trump signed the Tax Cuts and Jobs Act of 2017, section 11081 of which amended subsection 5000A(c) to change the “amount” of the applicable tax to zero, effective on January 1, 2019.

Texas and the other plaintiffs argue, in the first instance, that this 2017 amendment effectively eliminated one of the two choices Congress had previously afforded to covered individuals, so that now they all must maintain health insurance—something that the Court has said would be an unconstitutional requirement. 

Yesterday, the Department of Justice somewhat shockingly agreed with this specious argument.

It should be self-evident why it’s mistaken.  Obviously the 2017 Republican Congress did not intend to diminish individuals’ choices and to require them to maintain health insurance beginning in 2019 whereas they previously did not have to do so.  (Indeed, I’d be willing to wager than no members of Congress who voted for the bill would favor such a result.)  Instead, the new Congress’s design was manifestly to lessen the burden on individuals by giving them an option they previously did not have—namely, to decline to either maintain insurance or pay a tax:  Come 2019, payment of “zero” will be a lawful option.

The government insists, however, that because of what the 2017 Congress did, any covered individual who does not maintain health insurance come January will be violating a legal duty, even though they were not doing so until then.  Again, I’m fairly confident that no member of the legislature who voted for the 2017 amendment thinks that is the case, or intended anything like it. 

A hypothetical and an illustrative application of the existing, pre-2017 ACA will help demonstrate the point:

First, imagine, for instance, that an individual who has not maintained insurance for the past few years voluntarily chooses to continue paying the IRS $695 a year, even after January 1, 2019, when the tax requirement will be reduced to zero.  That person, in other words, does not change her behavior in the slightest.  Will she then be violating the law because of the 2017 amendment?  Of course not.  But according to DOJ, she will be.

Alternatively, let’s look at what many (i) lower income individuals and (ii) members of Indian tribes are doing today, before the 2017 amendment goes into effect.  Under the ACA as it has existed since its enactment (Subsection 5000A(e), in particular), such persons are expressly exempt from the Shared Responsibility Payment (the “penalty”).  Accordingly, many of those persons have neither maintained insurance nor made any shared repsonsibility payments to the IRS.  Have they therefore been violating the ACA, by virtue of being exempt from the tax provision, i.e., from one of the two choices available to others?  Of course not.  Well, the effect of the December 2017 amendment will simply be to put everyone else in exactly the same legal position as lower-income individuals and tribal members have been for almost a decade:  They will have a lawful choice to do nothing—to make a “payment” of zero—and will not have a duty to purchase or maintain insurance.

Contrary to the implausible argument in the Justice Department brief, the amended version of Section 5000A does not and will not require anyone to maintain health insurance.  Therefore it is not unconstitutional—instead it is, at most, toothless.  And if Section 5000A is not unconstitutional, that’s the end of Texas’s case.  DOJ’s argument to the contrary is inexplicable.  And that’s true even before we apply the canon of constitutional avoidance.  Or before we recall that President Trump signed the bill just a few months ago without expressing any constitutional doubts about the amendment (and presumably without being advised by OLC that the amendment was unconstitutional).  [UPDATE:  In his letter to Congress yesterday, Attorney General Sessions states that President Trump "has concluded that the statute is unconstitutional."  That's interesting--I wonder why he and DOJ did not say so when he signed the law in December.]  Or before we consider the traditional practice of DOJ to defend the constitutionality of statutes when reasonable arguments are available:  None of the established reasons for deviating from that practice (which I discussed in this post, about why it was not necessarily problematic that Acting Solicitor General John Roberts attacked the constitutionality of statutes in Metro Broadcasting) are remotely applicable here.

But it gets worse . . . .

The Government’s Implausible Understanding of Congress’s Intent About Whether the Guaranteed Issue and Community Rating Provisions Should be Operative if There is No Mandate to Maintain Health Insurance

DOJ’s failure to defend the constittionality of Section 5000A as amended—of the statute that President Trump signed a few months ago—is indefensible, but standing alone it wouldn’t much matter, because (as I discuss above) whether or not 5000A is unconstitutional, it will no longer be doing much work once the tax is reduced to nothing in January.  The much more serious problem with the government’s brief is what comes next.

Let’s assume—contrary to what I’ve written above—that the amended version of Section 5000A does require people to maintain insurance and that it is therefore unconstitutional and will not be enforced.  Texas’s next argument—the one that’s truly at the heart of the case—is that if that’s so, then the entire remainder of the ACA must be declared inoperative because it’s not severable from Subsection 5000A, i.e., that Congress would not have intended any provisions of the ACA to remain in place if there were no enforceable individual “mandate” to maintain health insurance. 

DOJ disagrees with Texas on that broad nonseverability proposition, but nevertheless insists that two crucial provisions of the Act are inseverable from Section 5000A and therefore must be declared invalid beginning in January, when the 2017 amendent goes into effect.  The provisions in question are the well-known Guaranteed Issue and Community Rating provisions (GI and CR) at the heart of the ACA, which prohibit insurers from denying coverage or charging higher premiums, respectively, to any person because of that person’s preexisting medical conditions or medical history.  DOJ argues that Congress would not have wanted the GI and CR provisions to exist if there were no enforceable individual mandate.

In support of this argument, DOJ relies almost exclusively on a finding in the original ACA (Section 18091(2)(I)) that the individual mandate requirement was “essential” to the effective operation of the GI and CR provisions.  That 2010 finding states that “[t]he [Section 5000A] requirement is essential to creating effective health insurance markets in which improved health insurance products that are guaranteed issue and do not exclude coverage of pre-existing conditions can be sold.” 

The problem with relying upon this finding from 2010 is that the 2017 Congress deliberately both: (i) set the “penalty” for not maintaining insurance at zero, understanding that therefore many more individuals would neither make a payment nor maintain insurance; and (ii) left the CR and GI insurance reforms in place.  The 2017 Congress, in other words, concluded that the insurance reform provisions now could and should function sufficiently even without an effective individual insurance “mandate.”  Nor was such a conclusion irrational, even if the 2010 congressional finding was true at the time it was enacted.  In 2017 the Congressional Budget Office issued a report concluding that insurance markets would still continue to function effectively even if rates increased because of a zeroing out of the fee alternative to the 5000A insurance maintenance requirement.  The CBO found that in light of the situation in the markets as of today (rather than 2010, when fewer people had already purchased insurance policies), “[n]ongroup insurance markets would continue to be stable in almost all areas of the country throughout the coming decade.”  Whether or not others could take issue with that conclusion, it was certainly rational for Congress to rely upon it in 2017.  Which it did.

In this case, in other words, there’s no need to indulge the sort of “counterfactual” that the courts apply to ordinary severability questions about what Congress’s intent would have been for provisions B and C if provision A were rendered a dead letter, for we know what the 2017 Congress’s intent actually was in such a case:  provisions B and C are to remain intact.  As Ilya Somin sharply put the point when Texas filed its suit:

[T]here is a big difference between a court choosing to sever a part of a law, and Congress doing so itself.  And in this case, Congress has already effectively neutered the individual mandate, while leaving the rest of the ACA in place.  It was Congress that removed the monetary penalty imposed on violators of the individual mandate, thus rendering it ineffective.  And it was also Congress which chose to leave the rest of the law in place, nonetheless (largely because President Trump and the GOP leadership repeatedly failed to round up enough votes in the Senate to repeal any more of Obamacare). Unlike in NFIB, a court could not conclude that Congress’ design for the ACA would be fatally undermined without an effective individual mandate. … In this case, Congress itself has concluded that a mandate-less ACA is acceptable (or at least a lesser evil than the available alternatives).

To the same effect, see also Nick Bagley here or listen to Jonathan Adler here.

If this sounds like a slam-dunk point on severability, that’s because it is.  And so how does DOJ counter it?  It devotes all of three sentences to the argument about the 2017 Congress (which is three sentences more than what Attorney General Sessions offers in his letter to Congress yesterday!).  Here are those sentences (in bold), on which the government’s argument crucially depends:

That conclusion [that the GR and CI provisions are nonseverable] is not affected by the fact that the [2017] TCJA eliminated the mandate’s penalty.  It still remains the case that, in the complete absence of the mandate, retention of the guaranteed-issue and community-rating requirements would expose health insurers (and their customers) to unfettered adverse selection by individuals who can game the system by waiting until they are sick to purchase insurance, contrary to Congress’s express intent.  42 U.S.C. § 18091(2)(I).  Nor is this conclusion undermined by the fact that the TCJA did not itself eliminate the guaranteed-issue and community-rating requirements at the same time it eliminated the mandate’s penalty and thereby rendered the mandate unconstitutional.  The best evidence of Congress’s intent is found in the legislative findings [from 2010], which continue to remain part of the ACA after the TCJA.  These express findings continue to describe the mandate as “essential” to the operation of the guaranteed-issue and community-rating provisions. See EEOC v. Hernando Bank, Inc., 724 F.2d 1188, 1190–91 (5th Cir. 1984) (noting that in determining “whether Congress would have enacted the remainder of the statute in the absence of the invalid provision[,] … [c]ongressional intent and purpose are best determined by an analysis of the language of the statute in question”).  Those findings cannot be deemed to have been impliedly repealed by Congress’s mere elimination of the financial penalty.  See Nat’l Ass’n of Home Builders v. Defs. of Wildlife, 551 U.S. 644, 662 (2007) (explaining that “‘repeals by implication are not favored’ and will not be presumed unless the ‘intention of the legislature to repeal is clear and manifest’” (citation omitted)).

It’s easy to see why virtually no attorneys in the Federal Programs branch would agree to sign their names to a brief that depends upon those three sentences.  For even assuming that the “repeals by implication are disfavored” canon applies to a findings provision of a law that Congress subsequently amends in fundamental relevant respects, the “intention of the [2017] legislature” to abandon the thrust of the 2010 finding, in light of changed circumstances in the interim, could not be “clear[er]” or more “manifest.”  And DOJ doesn't offer a single word to explain why not.

DOJ attorneys, especially those from Federal Programs and elsewhere in the Civil Division, do not withdraw from a case simply because the government is making a weak argument or an argument the lawyer privately thinks the courts should or will reject—something that happens rather frequently.  The bar for a DOJ lawyer to withdraw from a case because of the implausibility or weakness of a government argument in support of a federal program is understandably set very high.  DOJ’s brief yesterday in Texas v. United States cleared that hurdle with room to spare.