Reportedly, Chief Justice John Roberts was initially inclined to strike down the individual mandate in Obamacare but switched sides after something changed his mind. Speculation has been that it was concern over the Supreme Court’s prestige and susceptibility to public criticism. But I think there was an irresolvable dispute over a crucial question of constitutional law, one that the Court never reached because Roberts switched sides in order to avoid it.
The issue of severability—what to do with the rest of the law when one of its provisions is found invalid—has bedeviled the Court since its earliest days. Had the Court reached that issue in this case, it would have been faced with the most complex and excruciating severability analysis of its entire history. The mandate is just a single provision in a huge law. But it is so interwoven with the health insurance reforms (guaranteed issue, for example) that those reforms cannot function properly without it, as both the Obama Administration and the challengers argued. The insurance reforms are in turn organic to the rest of the law—and the whole law is a fantastic tangle.
I suspect that Chief Justice Roberts was willing to strike down the mandate, and perhaps some of the insurance reforms, but not the whole law. He would have argued that judicial deference to the prerogatives of Congress required striking down only as much of the law as necessary to cure it. But Justice Scalia might have insisted that judicial deference to the prerogatives of Congress required striking down the whole law, not rewriting it, and that upholding part of the law would be even worse than upholding the whole thing.
If that is what happened, then the real explanation of the Court’s decision in Obamacare lies in a little-noticed problem with what is perhaps the most important decision in the history of the Court: Marbury v. Madison (1803). In a case that every law student remembers, Chief Justice John Marshall struck down a relatively insignificant section of the Judiciary Act of 1789, which created our system of federal courts, and left the rest of the law in place.
The case is remembered because it established the Court’s power of constitutional review: the power to strike down laws of Congress that violate the Constitution. That power does not appear anywhere in the Constitution, but everyone agrees that the Court must have it, otherwise the whole structure of the Constitution makes no sense.
So far so good. The problem is that the power to strike down only one part of a law and leave the rest of the law in place is not just a lesser-included element of the power to strike down whole laws. It is an entirely different kind of power. It is the power to rewrite laws of Congress through a sort of judicial line-item veto. When a law is only partially invalidated, the remaining law is one that by definition fails to meet the Constitution’s minimum requirement for a bill to become law: passage by both Houses of Congress, and presentment to the President for his signature.
What is ironic in the doctrine of partial invalidation is that the Court invented it out of deference to the prerogatives of Congress. At first glance, the ruling in Marbury v. Madison seems intuitively correct: Why strike down a whole complex law on account of one tiny invalid provision that nobody will miss?
Leaving to a side for just a moment the constitutional problem, there is an enormous practical problem in trying to apply any doctrine of partial invalidation: As legislation has become more complex, laws have become more organic, so that any provision of a law may be inseparably interwoven with a host of others in an impossible maze of statutory interactions. Over the last two centuries, an entire edifice of Court rulings has been erected to tackle the issue that Marshall did not face: What to do when the invalid provision is so interwoven with other valid provisions that the valid ones won’t function as intended without the invalid part?
Obamacare did not merely raise the issue: It promised the most complex and difficult severability decision in the history of the Court. It is not for nothing that the dissent devoted nearly twenty pages to a severability analysis that did not come up once in the majority opinion. It would have been surprising, actually, if five votes could have been mustered in favor of any outcome on the severability issue, precisely because the possible outcomes were so many.
The Court’s current formulation of how to resolve the severability problem is a two-part test: Unless (1) it is evident that Congress would not have enacted the valid provisions independently of the invalid part, the invalid part may be struck, if (2) what is left is fully operative as a law. Roberts himself wrote the last major articulation of that standard, Free Enterprise Fund v. Public Co. Accounting Oversight Bd. (2010). But in oral argument Justice Antonin Scalia stressed, as many Justices have, that this test draws the Court into pure speculation about what Congress might have intended in an imaginary bill that never became law.
That is the sort of argument that Supreme Court justices can spend many years having. And had the Court ever reached the severability analysis, it would likely have mired the judges in rehearsing that same argument over and over again with respect to countless different provisions of the law. The potential for an enormous waste of time was almost infinite. It is entirely conceivable that Roberts saw this quicksand ahead, and simply refused to go near it.
But that is the quicksand the Court steps into when it embraces any doctrine of partial invalidation. There are no two ways about it: When the Court partially invalidates a statute, it is proclaiming, by decree, an entirely new law that might never even have occurred to anyone in the legislature.
It is crucial to understand that the partial invalidation of a law is not a single action, but rather two separate ones. First of all, the Court exerts the power of constitutional review and invalidation. Okay. Still, that action must precede a second action, namely that of proclaiming what the valid law is. When the Court partially invalidates a law, and upholds the rest, it must first decide that the law is invalid as written, and then it must proclaim what the new law is. Nobody disputes that it the Court has the power to do the first of these; should it have the second? The Court is charged with interpreting acts of Congress, but that power should not extend to replacing acts of Congress with entirely new laws that no Congress ever acted on.
The text of the Constitution is absolutely clear that in order for a bill to become law it must be enacted by Congress and it must be presented to the President. Both Congress and the President are accountable to the people, and the basic responsibility of their office is to represent the voters. Their constitutional prerogative is clearly diminished when either is deprived of its power to consider a proposed law as a whole before deciding whether to enact it.
Who cares what Congress might have done with a bill that was never voted on? Whatever it might have done, it didn’t do it. And yet, out of supposed deference to Congress, Marbury v. Madison instructs the Court to saddle the country with laws that no Congress ever passed.
The reasonable legislator might object that it is virtually impossible to predict what the Court might find unconstitutional, and that he would rather the Court just delete any problematic provision and leave the rest of his handiwork in place.
But that is not convenient for the people; on the contrary, it is convenient only for the legislator and the special interests he’s mindful of, precisely at the expense of the people.
We hold our representatives accountable for the laws they vote on. That is why bills must be enacted by both houses of Congress before they can become law. The Constitution clearly entrusts lawmaking to the legislative and executive branches, and makes it no part of the Supreme Court’s competence. Out of deference to Congress, courts should vote up or down on the constitutionality of laws, and not take it upon themselves to rewrite them.
To the extent that Marbury v. Madison establishes a power of partial invalidation of laws, it should be reversed. That would preserve the prerogatives of Congress and of the people. And Congress would have to be more careful about the Constitution when it writes its laws.