Unsafe Defaults in Statutory Severability Analysis
(I’m not a lawyer, and I did precious little research in drafting this. It should be understood as a thought experiment and intellectual provocation. It’s also possible, if not likely, that this is a well-known observation and adds nothing novel.)
From time to time, federal courts find that part of a statute is incompatible with the constitution. When this happens, the court must decide what remedy follows from this – which frequently means choosing which portions of a statute to hold invalid. This is called severability analysis, because the court is choosing whether the portion of the statute that it’s holding unconstitutional can be severed from the remainder of the statute. In some cases, this analysis is straightforward because the statute contains instructions to courts on how to perform severability analysis. Unfortunately, in many statutes no such guidance exists, and when considering provisions of statutes concerning constraints on the executive branch, such as in INS v. Chadha (legislative veto), courts have consistently reached conclusions that accrete executive power and are atextual.
In cases like these, courts uniformly find that the portions of a statute which impose constraints on the executive branch are severable from the portions which grant the executive branch new authorities. This is the crux of courts' mistakes. From a separation of powers perspective, the problem is that this is a one way ratchet, analogous to Justice Gorsuch’s observation in Learning Resources v. Trump. Once a court removes a constraint but leaves some authority in place, unless a statute has a sunset clause, it will require a veto-proof majority for Congress to re-assert some limit on the authority granted to the president (we assume that any president will veto nearly any bill that constrains their power). In contrast, if a court finds that the constraints are not severable, and thus both the constraint and the grant of authority must go, then Congress has the ability to craft a new compromise, striking the balance between executive authority and constraints they find appropriate, through the normal legislative process.
In addition to this separation of powers concern, there’s also a congressional intent problem. Statutes that contain both grants of authority and constraints are better understood as atomic packages which can’t be sliced and diced without entering a policy making role. Anyone who has participated in a contract negotiation, settlement process, or legislative drafting understands that there’s an incredible give and take to what each side is willing to accept. It’s extremely common for two, seemingly unrelated, provisions to in fact reflect a balance in what the parties were willing to accept. The court’s test of whether the provisions can function independently disrespects this legislative horse-trading process. And it puts the court in a role of guessing what Congress might have done, rather than calling balls and strikes with what it actually did.
Lastly, when courts pick and choose which parts of a statute to retain, they leave portions of the statute functioning as they had been. This likely sounds like a feature of severability analysis, but it has a serious flaw. While it keeps a portion of the statute working, it does so with a one way ratchet in the balance of separation of powers, and it further entrenches congressional dysfunction. If every statute containing a legislative veto had been struck down in Chadha, that would have been an extreme result. So extreme that it would have forced Congress to find a new compromise. Instead what happened is that numerous statutes continued to drift, over time substantially upsetting the balance between the branches.
In a world where many are concerned that it’s already too hard to legislate because of the risk of the Supreme Court striking statutes down, it may seem strange to advocate for courts to be even more sweeping in doing so. However, I believe that this may get cause and effect backwards. Because courts are presently able to take a scalpel to statutes, it provides the illusion that courts are being judicious; were courts forced to reckon with the proper choice of invalidating statutes as a whole or not, it would force them to have a greater respect for the legislative process.
If I’m wrong, and Congress really would prefer a partially operational statute, they have a clear remedy available to them: a severability clause. And importantly, this is available to them through the standard legislative process – either in the original drafting or after invalidation. This is in contrast with finding a constraint on the executive severable, which will require a veto-proof majority to correct.
In cases where Congress has not drafted a severability clause, courts are asked to decide whether one impermissible portion of a statute should cause the entire thing to fall. In cases concerning constraints on the executive, courts uniformly find that these are severable, which has the effect of accreting power to the executive in a way never contemplated by the Congress which passed the bill. Courts must not substitute their judgement for what Congress might have done for the reality of what Congress actually did. Though it may seem extreme, in separation of powers cases, declining to sever ultimately best serves the interests in the balance of the branches.