Severability and Estoppel

By Mike Dorf


In my post yesterday on the severability ruling in Florida v. U.S. Dep't of HHS, I addressed Judge Vinson's determination that the entire Patient Protection and Affordable Health Care Act must fall following the conclusion that its individual mandate is unconstitutional.  As I explained, at least with respect to wholly unrelated provisions, this conclusion seems unwarranted.  Here I want to address the more plausible contention that the provisions of the law that are intertwined with the mandate are non-severable.  I conclude that this contention is weaker than it may at first appear.


The best argument for invalidating the mandate as outside the scope of the Commerce power begins with the proposition that the Court has never sustained the regulation of inactivity.  This is true but misleading; the Court has never faced a challenge to a law on the ground that it regulates inactivity; but Congress has regulated inactivity as commerce; consider the Taft-Hartley Act, which forbids certain kinds of strikes, i.e., which regulates the decision to withhold labor.  In other words, it regulates the inactivity of not working.  Now, if you say that a strike is not merely inactivity because it occurs in the broader context of economic activity, you're right, but the same is true, as the government has repeatedly argued, with respect to the decision not to purchase health insurance when healthy.  But I digress.


Suppose a court is persuaded, as Judge Vinson was, by the argument that Congress may not regulate inactivity, and thinks either that PPAHCA is appreciably different from Taft-Hartley or that they're both unconstitutional.  To reach that conclusion, one would have to reject the government's argument that the mandate is necessary and proper to effectuating the restrictions on insurers' ability to screen customers for pre-existing conditions.  The government says, in defense of the law's validity, that without the mandate, healthy, otherwise uninsured, people will wait to buy health insurance until they get sick, thus undermining the system's ability to function as insurance (because there will be inadequate premiums and reserves to cover the resulting costs).


How might a judge reject that argument?  One way would be to deny the facts.  Perhaps the judge thinks that the other aspects of the law can function perfectly well without the mandate.  After all, the mandate isn't really a mandate anyway, given that the tax penalty for failure to comply with it is relatively small.  And there are other things the government could do to address this problem.  For example, it could subsidize health care for the uninsured.


But a court that accepted this argument would be hoist by its own petard with respect to severability.  As I noted in an earlier post, if the mandate is not sufficiently connected to provisions like the prohibition on screening for pre-existing conditions to fall within the scope of Congressional power under the Commerce and Necessary and Proper Clauses, then it seems clear enough that the mandate can be severed from the Act without fatally undermining such other provisions.


There is another possibility, however.  As I read Judge Vinson's opinion, he says that neither the Commerce Clause nor the Necessary and Proper Clause authorizes Congress to regulate inactivity, full stop.  Even where the regulated inactivity is closely connected to the regulation of economic activity, there is no federal power, according to Judge Vinson.  He then turns around and says that the government, insofar as it argued that the mandate is inextricably connected with various other provisions of the law, has admitted that the mandate is not severable from those other provisions.   Although Judge Vinson does not use the term "estoppel," he essentially estops the government from contesting the mandate's severability from various other core provisions of the Act.  He says:
the defendants concede that [the individual mandate] is absolutely necessary for the Act’s insurance market reforms to work as intended. In fact, they refer to it as an ‘essential’ part of the Act at least fourteen times in their motion to dismiss.
Is that right?  Not having examined the record closely enough, I don't know whether the government literally conceded the severability point.  But I do think that the government did not have to concede the severability point.


Why not?  Because the relatedness threshold for satisfying the Necessary and Proper Clause is lower than the relatedness threshold for saying that a provision of law is non-severable from another provision.  In order to find that some measure satisfies the Necessary and Proper Clause, a court need only find that the provision is "convenient" or "useful" for accomplishing an end that is within an enumerated power such as the regulation of Commerce.  That language doesn't come from some post-New Deal/Warren Court expansion of the scope of federal power, but from the leading case on the scope of Congressional power, CJ John Marshall's 1819 opinion in McCulloch v. Maryland.


By contrast, to find that a provision is non-severable from otherwise valid provisions requires a court to conclude that "the balance of the legislation is incapable of functioning independently."  (That's a quote from Alaska Airlines v. Brock).  That is on its face a tougher test to satisfy than the Necessary and Proper test.  


Thus, the government can consistently argue: 1) that the mandate is sufficiently closely related to the other provisions of the Act that the mandate is necessary and proper to the regulation of the interstate commercial market in health insurance; but 2) that the mandate is not so closely related to those other provisions that they are incapable of functioning without it.


Perhaps the government did not leave itself room to distinguish between these two standards, but it certainly could have--and in the next stage of this litigation, it should.