When a federal court (especially a U.S. Court of Appeals or the U.S. Supreme Court) faces an unresolved issue of state law—either in a so-called diversity case or in a case where the meaning and scope of federal law may turn on the meaning and scope of state law in a particular state—the federal court often can certify the issue for resolution by the highest court of the state. The federal court essentially asks the state high court how state law would apply to specific circumstances. And after the state court answers the questions, the federal court can resolve the case before it, confident that federal judges will not be misapplying or mucking up state law.
Certification is a device that has been around for over 50 years. It first received significant attention by the U.S. Supreme Court in in 1960 in Clay v. Sun Insurance Office Ltd. Clay was a seemingly straightforward diversity action by a policyholder to recover on an insurance policy. The insurance contract was made in Illinois and contained a provision barring lawsuits to recover that commenced more than 12 months after discovery of the loss. Such a provision was valid under the laws of Illinois. A complexity arose because the lawsuit was brought in federal court in Florida (the jurisdictional basis being that the parties were citizens of different states). Florida by statute prohibited enforcement of contractual provisions that limited the time for bringing suit, instead forcing parties to make use of the statutory limitations embodied in Florida law for different kinds of claims.
When the case was appealed to the Fifth Circuit (Florida was then in the Fifth Circuit; the Eleventh Circuit did not come into being until 1981), that court did not rule on the key question whether the Florida statute prohibiting contractual limitations on the time for bringing claims actually applied, as a matter of Florida law, to a contract made in Illinois. Instead, it held that, as a matter of federal due process, the Florida statute could not constitutionally be applied to a contract made elsewhere, whether Florida wanted it to or not.
The case proceeded to the Supreme Court, where Justice Felix Frankfurter’s majority opinion questioned the wisdom of the Court of Appeals having first resolved the constitutional (due process) question without having first decided the statutory one; if as a matter of Florida law the Florida statute did not govern out-of-state statutes, there would be no need to address any due process questions. And as a general rule, federal courts are supposed to avoid constitutional questions in resolving cases if that is reasonably possible.
One doctrine that might have applied in this situation is so-called Pullman-type abstention, named after the famous Railroad Commission of Texas v Pullman case. Under Pullman abstention, federal courts are supposed to stay their hand – that is, abstain from moving forward, when: (1) an uncertain issue of state law underlies a federal constitutional claim; (2) the state issues are amenable to interpretation and such an interpretation obviates the need for or substantially narrows the scope of the federal claim; and (3) an incorrect decision of state law would hinder important state law policies. Such abstention would then relegate the parties to initiate state-court proceedings in the hope of getting state courts to resolve the underlying uncertainty of state law. These prerequisites existed in Clay, and it would thus not have been surprising if the Court had directed the lower courts to invoke Pullman abstention.
But it seems that about a year before Clay, a former Frankfurter clerk had delivered a speech in which he mentioned in passing a little-known and never-implemented Florida statute that empowered the Florida Supreme Court to set out a process by which it could directly receive questions of state law certified to it by a federal appellate court. The Clay opinion lauded the Florida legislature’s foresight in passing such a law, and in vacating the Fifth Circuit’s ruling indicated that such certification was the way to go (even though at that time the Florida Supreme Court had never pursued the invitation from the legislature to set up a certification process). Three Justices dissented, with Justice Douglas offering the following lamentation concerning certification: “Shuttling the parties between state and federal tribunals is a sure way of defeating the ends of justice. The pursuit of justice is not an academic exercise. There are no foundations to finance the resolution of nice state law questions involved in federal court litigation.”
On remand the Fifth Circuit dutifully certified to the Florida Supreme Court (which had in the meantime, seemingly in response to the U.S. Supreme Court’s praise of the certification route, adopted a rule allowing for certified questions from federal courts) questions about the meaning of the Florida statute. The Florida Supreme Court, although it agonized a bit over whether the Florida constitution permitted it to address questions of law posed in the somewhat abstract way that certified questions do, obliged the Fifth Circuit, and responded that the Florida statute did in fact apply to a case like Clay. The Fifth Circuit, armed with this definitive ruling on the applicability of Florida law, reiterated its prior holding that the Florida statute could not constitutionally apply to a contract in situations where the only contacts with Florida were that the insured had moved to the state and that the loss occurred there.
The case went back to the U.S. Supreme Court and this time, seven years after suit was filed and four years after the Court had first directed the use of certification, the Court unanimously disagreed with the Fifth Circuit on whether the Florida statute could constitutionally apply. The Court said Florida law could constitutionally apply on these facts, and that therefore the one-year limitation period provided for in the insurance contract was not enforceable, such that the original judgment of $6,800 in favor of the insured that the district court had entered way back when should stand.
It may seem that this first prominent use of certification was not particularly efficient, and that the experiment would not generate enthusiastic use of the device going forward. But in spite of its challenges, certification, as the Court has pointed out more recently, enjoys certain advantages over abstention: Certification “allows a federal court faced with a novel state-law question to put the question directly to the State’s highest court,” and has thus the benefit of “reducing the delay, cutting the cost, and increasing the assurance of gaining an authoritative response” from the state court as compared to abstention, in which the parties have to begin anew in state court and cannot cut in line, so to speak, straight to the highest state judicial tribunal. And certainly, when compared to a federal court charging ahead with its own interpretation of unclear or unsettled state law and risking major mistakes, the certification procedure promotes cooperative judicial federalism.
Clay was decided in 1960, and it wasn’t until about 15 years later that certification picked up steam. In the late 60s the Commissioners on Uniform State Laws had approved a Uniform Certification of Questions of Law Act, and the American Law Institute then proposed a statute explicitly empowering federal courts to certify questions to state courts if the states in question had procedures permitting the answer of certified questions. Yet only a handful of states had, following Florida’s lead, adopted a certification procedure. And in the absence of such state receptivity there is no way a federal court can require or compel a state court to answer state-law questions
But in 1974 in Lehman Brothers v. Schein, the Supreme Court again extolled the virtues of certification mechanisms. The Court made clear that a federal court has discretion to use an available certification procedure even when the unclear issue of state law arises in a routine diversity case and raises no constitutional issue. The Court enthusiastically discussed how certification saves “time, energy, and resources and helps build a cooperative judicial federalism.” The Court echoed this praise of certification in Lehman Brothers two years later in Bellotti v. Baird, interestingly enough in the abortion setting.
Bellotti involved Massachusetts’s new (and as of yet unconstrued) state law regulating the consent, including parental consent for unmarried minors, that was required before an abortion could be performed. The law was challenged, not long after Roe, as violating the due process rights of minors. Justice Blackmun, writing for the Court, thought it possible that the statute could be understood in a way that would avoid or at least reshape the federal constitutional issue presented. Pullman-type abstention was an option, but the Court doubled back to its earlier pronouncements that abstention is limited by considerations of delay and expense, and that certification saves time. In directing certification to the state court, the Court added:
The importance of speed in resolution of the instant case is manifest. Each day the statute is in effect, irretrievable events, with substantial personal consequences, occur. Although we do not mean to intimate that abstention would be improper in this case were certification not possible, the availability of certification greatly simplifies the analysis.
To be sure, certification has proven not to be a panacea. Sometimes it is improperly invoked where state law is not unclear, but the federal court for other reasons seems to want to avoid making a difficult ruling. And sometimes state courts do not fully comply in answering certified questions. The fact that a question of state law is unresolved may not be accidental; some questions of law are unresolved because courts are reluctant to resolve them. That may be especially true for state court judges and justices who are removable by the voters or who have to stand for retention elections periodically. If a state supreme court had semi-purposefully avoided resolving some thorny question of state law in the past, it will not be excited about answering that question in response to a certified question. Certification might also be a victim of its success: the more common certification becomes, the greater the burden on state courts.
Still other complexities arise because some state constitutions may impose constraints on state courts akin to the justiciability hurdles (standing, mootness, ripeness, etc.) that Article III is said to create for federal courts. So the willingness and ability of some state courts to answer certified questions will vary by state. On top of all this, answers can be only as good as the questions that are posed. Federal courts that certify questions must pay special attention to craft the questions in a way that do not violate a state court’s justiciability constraints (that is, in ways that are not too rarified, abstract, and divorced from actual facts) but that also will yield the information the federal courts need to resolve their cases.
As noted throughout this discussion, state law determines when and how state courts receive and respond to certified state law questions. And, again, the process varies by state; some states even permit federal district courts (and not just the Courts of Appeals and the Supreme Court) to certify questions. But despite the variations among states, one trend has been clear: the number of states that authorize some kind of certification procedure has grown steadily over the decades since Clay. Indeed, currently we are aware of just a single state, North Carolina, that provides no mechanism at all for federal court certification of state law issues. As we will explain in Part Two of this series, a case currently pending in the U.S. Supreme Court, Berger v. North Carolina State Conference of the NAACP, shows the downsides to North Carolina’s unique refusal to accept certified questions—and, we hope, will lead to the state providing for a certification process.