By Mike Dorf
The summary judgment ruling in Garden State Equality v. Dow raises interesting substantive questions but as a federal jurisdiction scholar, I can't help but find myself fascinated by a seemingly arcane procedural aspect of the case: If it is affirmed by the NJ intermediate appellate court and the New Jersey Supreme Court, will SCOTUS review be possible? I think the answer is just barely yes, but for reasons that readers may find surprising. The case implicates a very nice question of federal procedural law: When does a state court discussion of federal law create a case that arises under federal law within the meaning of Article III and the relevant federal jurisdictional statute?
Let's begin with the NJ background. Back in 2006, in Lewis v. Harris, the NJ Supreme Court held that the NJ Constitution obligates the state either to recognize same-sex marriage or to create an institution that affords same-sex couples all of the legal benefits of marriage. The NJ legislature responded by enacting a marriage-in-all-but-name civil union statute. That was sufficient to satisfy the Lewis requirement until the Supreme Court's ruling this past June in United States v. Windsor, striking down the federal definition of marriage as only opposite-sex marriage. In light of Windsor, the Garden State Equality plaintiffs said, NJ civil unions no longer afford same-sex couples all of the benefits of marriage: In particular, Windsor only applies to couples who are "married" under state law--and not to same-sex couples in civil unions. Thus, said the plaintiffs, NJ's civil union law does not afford same-sex couples the opportunity to receive federal benefits as married couples. The state defendants countered with two related arguments: First, Lewis only requires the state to extend equal state benefits to same-sex couples, so if the federal government denies those couples the benefits of marriage, that's not the state's fault; and second, Windsor, properly understood, should extend federal benefits to same-sex couples in civil unions, so again, if federal agencies aren't implementing Windsor that way, the plaintiffs should sue the federal agencies, not the state. Judge Jacobson sided with the plaintiffs.
Under the applicable federal jurisdictional statute and longstanding precedent, the SCOTUS can only review state court rulings that decide federal questions where reversal on the federal question would affect the outcome of the case. If a state court decision rests on an "adequate and independent" non-federal ground, then there is no SCOTUS jurisdiction. So let's suppose that the NJ Supreme Court were to write an opinion that is identical to the opinion written by the trial court. Would that be reviewable or would it rest on an adequate and independent state law ground?
Our first task is to locate a federal question. The notion of a federal question for Supreme Court appellate review is somewhat broader than the notion of a federal question for original jurisdiction in the federal district courts (where the case must satisfy the "well-pleaded complaint" rule), but still, there are limits. Here, there is only one possible candidate for a decided federal question: Post-Windsor, do civil unions count as marriages for purposes of federal law? At first blush, it appears that the trial court resolved this question--saying they don't count--and that its resolution was essential to the outcome of the case: it is that not-counting that makes NJ civil unions no longer adequate to satisfy NJ equal protection requirements as articulated in Lewis.
In particular, there is language in the Garden State Equality opinion that indicates that Judge Jacobson thinks that Windsor doesn't apply to civil unions. For example, she quotes the following language from Windsor for the proposition: "This opinion and its holding are confined to those lawful marriages." She also explains that it makes sense for federal law not to treat state civil unions as marriages because state definitions of civil unions differ among themselves.
Yet there is other language in the opinion indicating that Judge Jacobson's view of the best reading of Windsor is irrelevant to the outcome of the case. This other language suggests that whether or not Windsor is properly construed to treat civil unions as marriages for federal purposes, many federal agencies have already concluded (and are acting on the conclusion) that civil unions do not count as marriages; thus, NJ same-sex couples in civil unions are as a practical matter, already being denied federal benefits in virtue of the fact that NJ fails to call their unions "marriages."
Is that a federal question? Although the matter is not entirely free from doubt, I think the right answer is no. An undisputed factual claim about how federal agencies are construing federal law is a proposition about federal law, but it is not a proposition of federal law. And so, if one reads the NJ ruling as resting on the proposition that federal agencies are, rightly or wrongly, treating Windsor as inapplicable to civil unions, then the ruling (if affirmed as is by the NJ Supreme Court) would not be reviewable by the SCOTUS. There would be no question of federal law that could be reversed by the Supreme Court to produce a different outcome.
Nevertheless, I think that as written (and if affirmed in terms by the NJ Supreme Court), the trial court ruling would be SCOTUS-reviewable, because it is not entirely clear to me that the opinion's factual analysis of what the federal agencies are doing is truly independent of its legal analysis of the federal question of whether Windsor properly applies to civil unions. The opinion intersperses discussion of the two points in a way that suggests that the trial judge might have drawn a different conclusion with regard to the factual question if she thought the legal question were less clear-cut.
Thus, we come to Michigan v. Long. In that 1983 ruling, the SCOTUS held that where a state court resolves a case in a way that is ambiguous as to whether non-federal grounds for the decision were independent of federal grounds, the SCOTUS will presume that the state court meant to rely on the federal grounds, and can take review. I think the best reading of the trial court ruling in Garden State Equality makes the non-federal ground independent, but I think there is enough ambiguity in the opinion to trigger the Michigan v. Long presumption.
Going forward, the lesson of the foregoing analysis for the NJ intermediate appellate and Supreme courts should be clear: To make the ruling bullet-proof, be crystal clear in any opinion affirming the basic ruling of the trial court that the factual assessment of what the federal agencies are doing is an entirely sufficient basis for the conclusion that civil unions no longer satisfy Lewis. Here is the language I would include at the end of any NJ Supreme Court opinion that ends up affirming the trial court ruling:
In summary, we hold that regardless of whether Windsor is best read to treat state civil unions as "marriages" for purposes of federal law, it is an undisputed fact that many federal agencies have already concluded that civil unions will not be treated as marriages for purposes of federal law. In light of that fact, the State of New Jersey no longer satisfies its state constitutional obligation of equal protection under Lewis by affording same-sex couples the right to enter civil unions. Only "marriage" now satisfies that obligation.
You're welcome, New Jersey.