The New York eviction moratorium decision and the problems of the shadow docket

Mark Tushnet

Maybe the Court’s decision in the New York eviction moratorium case was right. But the legal problem it presented was way more complicated than the two sentences the Court devoted to its substantive analysis. Here are those sentences:  “If a tenant self-certifies financial hardship, Part A of CEEFPA generally precludes a landlord from contesting that certification and denies the landlord a hearing. This scheme violates the Court’s longstanding teaching that ordinarily ‘no man can be a judge in his own case’ consistent with the Due Process Clause.” What follows is my analysis of the analytic difficulties bound up with those sentences. (A caution: I’m on vacation and haven’t done research specific to this problem – I’m drawing on my general knowledge of constitutional law. And at some points I will make assertions in which my confidence level varies from “quite sure” to “pretty sure” to “reasonably sure.” That means, though, that some of the assertions might be mistaken.)

 1. Framing the statutory context:  The Court accurately said that the statute before it provided no opportunity for a landlord to challenge a tenant’s self-certification. Justice Breyer replied, also accurately, that if you looked at New York’s statute books generally, the landlord would have an opportunity to challenge the self-certification on or after September 1, because the statute before the Court was expressly temporary; its provisions would expire on September 1, at which point the rest of New York’s statutory scheme would spring back into action – and under that scheme landlords could challenge the self-certification.

So, the framing question: In analyzing the specific statute’s constitutionality, do you look only at it or can you look at what I’ll call the entire corpus juris? I know that in other settings the Court has held that an apparent deficiency in a challenged statute can be remedied or alleviated enough to eliminate unconstitutionality by provisions in other statutes. What’s the basis for distinguishing this case from those? (I’m not saying that no distinctions are possible, only that a serious analysis would seek to explain why it was appropriate in this case to focus solely on the statute being challenged.)

2. Self-certifications seen more generally: Let’s assume that you can look to the entire corpus juris, so that Justice Breyer was correct to refer to the rest of New York’s landlord-tenant law. That means that – at some point – landlords can challenge the self-certification. Still, for some period (until September 1), the statute allows the tenant to alter the legal relations between him/her and the landlord. There are two substantive constitutional issues lurking here: a regulatory takings challenge (a three-week removal of one of the sticks in the landlord’s bundle of right) or a contracts clause challenge. That’s not what the Court dealt with, though.

Self-certifications that alter existing legal relations, pending a later assessment of the accuracy of the self-certification, are quite common. The federal income tax law is shot through with them (you self-certify that you are married filing jointly or separately, which affects the bottom-line taxes you’ll pay). Closer to the current problem: A criminal defendant fills out a form in which s/he self-certifies an inability to afford a private lawyer, at which point a public defender (or equivalent) is assigned and provides initial legal defense, pending an evaluation of the accuracy of the self-certification.

Now, these are different because they change the legal relations between the self-certifier and the government (or, put another way, the self-certification affects the distribution of a government benefit). And maybe there’s a difference (relevant to procedural due process) between that and a self-certification that alters the legal relations between private parties. (Again, in the present context the most natural clause relevant here would be the contracts clause – though I would note that contract clause doctrine as it stands today probably would not find unconstitutional the imposition by statute of a term postponing the effective date of a landlord’s legal entitlement by three weeks.)

But I’m reasonably certain that “private rights” self-certifications do exist. One place to look would be the bankruptcy code. In conversation a friend said that “stand your ground” laws were a form of self-certification: A defendant charged with assault and asserting self-defense “self-certifies” that he feared for his life (or whatever the stand-your-ground law provides), and as a result gains a right to use force, not otherwise authorized, against the (private) assailant. (I’m not entirely sure that that’s a good analogy, but it’s worth thinking about.)

Again, it might be that there are relevant distinctions between the “private law” self-certifications that exist and the self-certification in the New York eviction moratorium case. Or maybe, having learned something from this case, the justices would be willing to consider the constitutionality of other self-certifications.

3. The state action problem. When a warehouse holding repossessed property sells that property after a statutorily prescribed period has passed, the Court held, the sale didn’t involve state action, and for that reason the warehouse owner (and the state) didn’t have to provide the “owner” any procedural due process rights.. How is that different from the tenant’s self-certification? That is, why is the self-certification state action that requires procedural due process?

The obvious answer is that it’s not the self-certification that is state action, but the New York statute. But, in the warehouse case, the “owner” said that there was state action in the statute authorizing the warehouse to sell material in its possession, and the Court said that that didn’t make action authorized by the statute statute state action.

Once again, maybe there are relevant distinctions. (The one that comes first to mind is a distinction between material property, in the warehouse case, and intangible rights as in the New York eviction moratorium case.) But, once again, it would be helpful to have some explicit discussion of this issue.

4. The corpus juris again. Assume that the problem is that New York’s statute might alter the pre-existing legal rights of landlords and tenants. What are those rights? At first glance, they are the rights contained in the lease. We know, though, that those rights are subject to alteration if the state exercises its police powers in a manner consistent with the contracts or takings clauses. So the preexisting rights are the ones in the lease, contingent upon later exercises of the police power that are consistent with the contracts and takings clauses. The New York eviction moratorium doesn’t alter that (complex) set of pre-existing rights if it’s consistent with the contracts or takings clauses. Maybe, though, in thinking about the question of what are the pre-existing rights we’re not allowed to “build in” the contingent role of the police power. I confess that I don’t see any obvious reasons why not, but maybe there are some.

5. Another version of the substantive law issue. Consider statute A, which provides: “Tenants in place are entitled to remain until September 1.” There’s no procedural due process question here, only substantive contracts and takings clause ones. Now consider statute B, which provides: “Tenants in place who suffer from financial hardship due to COVID are entitled to remain until September 1.” Maybe landlords have a due process right to contest a tenant’s claim that s/he will suffer financial hardship. The Court thought that statute B was the one before it. But consider statutes C and D. Statute C is, “Tenants in place who suffer from financial hardship, etc., are entitled to stay in place, etc. A tenant who self-certifies that s/he will suffer financial hardship is conclusively presumed to have done so.” And statute D provides, “Tenants in place who self-certify, etc., are entitled to stay in place, etc.” Now: The law of conclusive presumptions is that statutes C and D are the same – merely different ways of saying the same thing. And statute D is the same as statute A, only with the beneficiary class narrowed. So, the only constitutional questions about statute D are substantive ones. And, finally, it’s not difficult to interpret statute B to mean what statute C explicitly says.

Again, there are responses. The first is that the asserted interpretation of statute B – the actual moratorium statute – isn’t available. I’d want to hear why not (it looks to me like a question about the limits upon interpretation imposed by state law, with no obvious federal law component). The second is that the move from a conclusive presumption to a substantive statute isn’t allowed, which leaves space for the procedural due process argument By saying that all we really have is statute B. Two old cases do say that such a move isn’t allowed, but they are widely criticized and haven’t been followed. (Both involve holdings of unconstitutionality for denial of a right to a hearing in situations where, had the move to a substantive statute been made, the statute would have been constitutionally permissible.) The parallel cases in the criminal context involve presumptions that shift the burden of proof from the prosecution to the defendant, challenged as inconsistent with the “beyond a reasonable doubt” requirement.  After toying with the idea that you can’t move from the presumption to a substantive rule, the Court sensibly concluded that you could, and analyzes the problems as involving definitions of crimes challenged on substantive (typically, cruel and unusual punishment) grounds.

So, where does all this leave us? It looks as if the Supreme Court resolved a single constitutional claim, but in doing so it (necessarily) came to holdings on four or five additional claims. Its resolution of those additional claims might have been right, or – when in subsequent cases litigants say, “You know that you’ve already held that you can’t assess this statute in light of the entire corpus juris,” or, “You know that you’ve already held that there’s state action here,” – the Court will say, “Why yes, you’re right, and we’ll follow that holding in this case (or distinguish the New York moratorium case’s holding on that issue).”

But making these additional holdings in the context of a shadow docket case, where (I’m pretty sure) the papers available to the justices and the creativity of their law clerks are unlikely to have flushed the issues into the open, thereby enabling considered and deliberate reflection upon them and the analytic issues associated them, seems to me a terrible way to make constitutional law. Perhaps the Court should add to its criteria for deciding a shadow docket case on the merits something like, “This problem is way too complicated to solve in the time we have available. So we’ll grant plenary review, knowing that there’s a decent chance that, on standard views, it’s going to become moot. But maybe we can think – then – about how we want to apply or develop mootness doctrine when cases of this sort arise on the shadow docket.”