Apr 19, 2018

      In an article entitled Jurisdiction and Its Effects, I argued that jurisdiction has inherent descriptive meaning but mutable effects. In response, Professor John Preis challenges my framework on a number of grounds and offers his own presumption-based approach. In this reply, I defend my original framework and register my own skepticism of his alternative approach.

Introduction

I.  Jurisdiction and Its Effects

II.  Responses to Criticisms
   A.    Getting the Argument Right
   B.    Defending Against Positivism
   C.    The Continuing Vitality of Jurisdiction

III.  The Presumption-Based Alternative

Conclusion

Introduction

In an article entitled Jurisdiction and Its Effects, I made two arguments that turned jurisdictional orthodoxy on its head.^[1] First, I argued that jurisdiction has inherent descriptive meaning: jurisdiction refers to legal boundaries that “determine[] forum in a multiforum system.”^[2] Second, I argued that jurisdiction’s effects are mutable and subject to positive law.^[3] I then explained how this recalibration of jurisdiction’s identity offered a variety of salutary benefits for existing doctrine, precedent, and litigation practice.^[4]

Professor John Preis has responded to my article by challenging my framework and offering an alternative of his own.^[5] He denies that jurisdiction has an immutable definition and argues instead that Congress can call whatever it wishes “jurisdictional” or “nonjurisdictional.”^[6] He also objects that my framework leaves little work for the term jurisdiction; the game is all in the effects.^[7] Finally, he offers a different approach, which is to recognize a presumption in favor of my definition that is nevertheless defeasible by a clear statement from Congress.^[8]

In this reply, I take on his critique and defend my original framework as the best approach. In Part I, I summarize the basic arguments and merits of my framework. In Part II, I defend my framework against Professor Preis’s criticisms. First, I correct several mistakes Professor Preis makes in interrogating my framework. Second, I explain why an immutable definition of jurisdiction is more defensible than a positivist conception of jurisdiction. Third, I show that the term “jurisdiction” has continued utility and should be retained. In Part III, I lodge my own objections to Professor Preis’s alternative approach.

I.  Jurisdiction and Its Effects

It is a familiar lament that jurisdiction has “many, too many, meanings.”^[9] To remedy the problem, the Supreme Court has attempted to bring discipline to the use of the term by erecting a set of guideposts for determining when something is jurisdictional. The primary guidepost is a recognition that jurisdiction is a positivist concept: Congress gets to say when something is jurisdictional,^[10] and, when it clearly demarcates a boundary as jurisdictional, so be it.^[11] Jurisdiction’s definition, therefore, is elusive, for Congress can use it to describe whatever kind of limit it wishes.

At the same time, current doctrine makes jurisdiction’s effects immutable. Those effects are rote for any law student: the parties cannot waive or forfeit jurisdictional defects; parties cannot consent to jurisdiction; jurisdictional defects cannot be excused for reasons of equity or judicial discretion cannot excuse jurisdictional defects; the court must verify jurisdiction sua sponte, and jurisdictional defects can be raised at any time, by any party, before final judgment; and judgments lacking jurisdiction are void.^[12] In conventional wisdom, jurisdiction’s definition may be mutable, but its effects are not.

I have offered an approach that inverts current doctrine. Under my framework, jurisdiction has definitional meaning: “it determines forum in a multiforum system.”^[13] Any such boundary—subject matter jurisdiction, appellate jurisdiction, personal jurisdiction—is jurisdictional by definition.^[14] All other limits—statutes of limitations, immunity, limits on remedies, statutory coverage limits—are nonjurisdictional by definition.^[15] This definition is inherent and immutable; it is not subject to the whims of Congress. Congress can no more call something nonjurisdictional that is, or something jurisdictional that is not, any more than one could try to call a sheep a wolf. A sheep is a sheep.^[16] And a law that determines forum in a multiforum system is a jurisdictional law.

By contrast, the effects of a particular jurisdictional law may be set by positive law.^[17] If the jurisdictional law is statutory, then Congress gets to prescribe whether the jurisdictional limits can be waived or forfeited, are subject to judicial discretion, and the like.^[18] If the jurisdictional law is a judicial creation—like abstention—then the courts can prescribe its effects in the absence of statutory override.^[19] If the jurisdictional law is constitutional, then the Constitution, as a source of positive law, may supply its effects.^[20] In this way, the term “jurisdiction” actually means something—it defines which courts have authority to hear a case and which do not—but positive law can mold the effects of a particular jurisdictional line to fit its context and litigation realities.

These two moves upend current doctrine. Right now, jurisdiction’s definition is subject to positive law while its effects are static and immutable.^[21] I mean to switch those around, with some dramatic and unorthodox results, and I am not surprised that my proposal has generated some resistance, including from Professor Preis. I now turn to his criticisms.

II.  Responses to Criticisms

In this Part, I respond directly to Professor Preis’s criticisms. I begin by correcting his misunderstanding of my approach. Then, I defend my approach against his positivist challenge. Finally, I answer his objection that jurisdiction lacks meaning under my framework.

A. Getting the Argument Right

The first step is to understand my approach, and Professor Preis makes at least three mistakes in recounting and applying it.

First, Professor Preis asserts that my definition “falters” because he thinks that I “overlook[] Congress’s long-standing power to define federal jurisdiction”^[22] and that I “fail[] to acknowledge that Congress has a wide-ranging power to draw boundaries.”^[23] Assuming he means “define [the scope of] federal jurisdiction,” then I do nothing of the sort. I accept that Congress’s constitutional power to “ordain and establish” inferior federal courts^[24] includes the power to set the scope of their adjudicative authority (within constitutional bounds), including the power to draw both jurisdictional and nonjurisdictional boundaries.^[25] If Congress wishes to impose an amount-in-controversy requirement on the diversity jurisdiction of the federal district courts, it of course may do so, at whatever amount it wishes. Similarly, if Congress wishes to set certain forms of service of process in the federal courts, it may do so, consistent with due process. The Constitution grants Congress the authority to draw both nonjurisdictional and jurisdictional lines and to supply the content of those lines, and nothing in my framework suggests otherwise.

But I do deny that Congress has the power to define which lines are jurisdictional and which are nonjurisdictional. Once Congress draws lines, those lines’ characterizations arise directly from them. The amount-in-controversy requirement is jurisdictional because it determines which cases can be heard in federal court and which cases can be heard in state court.^[26] Service of process is nonjurisdictional because it speaks only to the limits of the federal courts and not to alternate forums.^[27] Congress may, when consistent with the Constitution, draw the lines in any place, and with whatever content, that it wishes.^[28] But Congress may not call the amount-in-controversy requirement nonjurisdictional and the service rule jurisdictional, and nothing in the Constitution says otherwise.

Second, Professor Preis misapplies my framework to the amount-in-controversy requirement of diversity jurisdiction^[29] and the employee-numerosity requirement of Title VII.^[30] Because I assert that the employee-numerosity requirement is nonjurisdictional,^[31] Professor Preis believes I must also think the amount-in-controversy requirement nonjurisdictional.^[32]

But this is a fundamental misunderstanding of the operation of the two requirements. The employee-numerosity requirement sets a limit on the scope of Title VII’s substantive obligations irrespective of the forum; it applies equally to limit adjudicatory authority in both federal court and state court.^[33] Professor Preis is wrong to say that “plaintiffs suing employers with fewer than fifteen employees can simply take their cases to state courts.”^[34] The employee-numerosity requirement limits the claim in state court, too. The claim is either viable (because the requirement is met) or not (because it is not met), and viability has no bearing on forum. Accordingly, the requirement does not determine forum in a multiforum system, and, as a result, it is nonjurisdictional.^[35]

The amount-in-controversy requirement is very different. It does speak to forum. A tort claim between diverse parties may have any number of restrictions on it akin to the employee-numerosity requirement of Title VII—limitations periods, damages limits, et cetera—but the amount-in-controversy requirement is not one of them. Instead, the amount-in-controversy requirement determines in which forum a litigant may adjudicate the claim. If the requirement is met, then either federal or state court may hear the claim.^[36] If not, then only a state court may hear the claim.^[37] The requirement determines forum in a multiforum system and is therefore jurisdictional by definition.  

Professor Preis further confuses this distinction by hypothesizing an employee-numerosity requirement phrased as, “federal district courts shall only have jurisdiction to consider claims under this Act if the employer accused of violating the Act has fifteen or more employees,”^[38] and he seems to believe that my view would characterize this requirement as nonjurisdictional.^[39] But the result here is unclear because his hypothetical is incomplete. What matters is not Congress’s deployment of the term “jurisdiction,” but what the law does. If his hypothetical requirement limits the kinds of employers covered by Title VII regardless of where the case is heard, then it is indeed nonjurisdictional (just like the real employee-numerosity requirement), and Congress’s misuse of the term “jurisdiction” should be ignored. But the hypothetical language seems phrased much more in the way of a forum determinant; in other words, the language could leave Title VII applicable to all employers regardless of how many employees they have, restrict federal courts to adjudicating Title VII claims only if pursued against employers with fifteen or more employees, and leave state court as the sole forum for employers with fewer employees.^[40]

More information about the substantive scope of this hypothetical Title VII would be required to confirm it, but if that is what Professor Preis means, then the hypothetical language is jurisdictional—not because Congress denoted it as such, but because the language determines which adjudicative forum (federal or state court) can hear such claims.

Third, Professor Preis mischaracterizes the source of authority for my approach as a “Platonic ideal” of jurisdiction^[41] that must manifest itself as some kind of natural law.^[42] But I do not purport to derive some “brooding omnipresence” of jurisdiction from natural law.^[43] Rather, jurisdiction is definitional. It is no different in kind from other definitional terms in our legal lexicon that have static meaning, such as “agency,” “relief,” “remedy,” “appeal,” “adjudication,” and a host of others found throughout the pages of Black’s Law Dictionary. A sheep is not a sheep because of natural law. A sheep is a sheep because that is what we have chosen to call that kind of animal, and, once chosen, its meaning is static.^[44] Likewise, some terminology must describe a law that determines forum in a multiforum system. The appropriate term is jurisdiction. And, once that is settled, jurisdiction’s definition is as static as “sheep,” not because of natural law, but as a matter of definitional law.

B. Defending Against Positivism

Having set the record straight, I now turn to Professor Preis’s direct attacks on my framework. The first attack is a press for positivism in jurisdiction’s definition. In Professor Preis’s view, jurisdiction is positive law, and lawmakers are free to call a wolf a sheep and force the rest of us to agree.^[45] I do not dispute that positive law gets to set where jurisdictional boundaries lie and what the content of those boundaries are. But I reject Professor Preis’s assertion that positive law also may affix the jurisdictional label to a boundary that is not jurisdictional.

A principal reason why my definitional account of jurisdiction is superior to Professor Preis’s positivist account of jurisdiction is that my definitional account better aligns like doctrines and separates unlike doctrines. Under my framework, all doctrines that determine forum in a multiforum system are jurisdictional, while all doctrines that do not are nonjurisdictional.^[46] This definition consistently labels as jurisdictional the usual suspects—personal jurisdiction, original jurisdiction, and appellate jurisdiction—while augmenting that group with similar kinds of forum-determining doctrines like venue, forum non conveniens, and abstention.^[47] At the same time, this definition excludes doctrines that limit a court in isolation from other courts, such as court-specific procedures, claim-specific requirements, and separation of powers limits.^[48] A few doctrines that I acknowledge, including state sovereign immunity and the political question doctrine, are admittedly more difficult to classify,^[49] but, on the whole, my definitional account makes a great deal of sense.

A positivist approach to defining jurisdiction, by contrast, results in some very odd groupings. A limitations period in one statute might be jurisdictional, while in another it is not.^[50] A deadline to file a notice of appeal in a civil case might be jurisdictional, while a deadline to file a notice of appeal in a criminal case is not^.[51] A claim-limiting element that naturally appears to be a merits question might be denoted as jurisdictional.^[52] One portion of an appellate rule might be jurisdictional while another is not.^[53] Presumably, Congress could denote the amount-in-controversy requirement as nonjurisdictional. These distinctions have nothing to do with common sense or the differing functions of the laws at issue. They are simply the product of congressional whim. A positivist approach—with its invitation to give “too many[] meanings” to jurisdiction—results in the very mess the Court has lamented.^[54]

A further demerit of a positivist approach is that it relies upon an interpretive process that, at times, can be difficult—even fictional. Congress rarely takes the time to clarify whether a limit is jurisdictional, and, in many instances, Congress likely never even considered the question. Thus, recent cases attempting to divine congressional intent have produced some unconvincing and confusing decisions.^[55] If avoiding complexity and uncertainty in jurisdictional doctrine are primary virtues,^[56] there’s no better approach than my definitional account, a point I will return to in Part III.

C. The Continuing Vitality of Jurisdiction

Professor Preis’s second main criticism of my framework is that, by settling jurisdiction but giving Congress control of its effects, I have “render[ed] jurisdiction functionally irrelevant” because “Congress can get exactly what it wants simply by speaking in terms of effects,” such that “the law of jurisdiction will be replaced with the law of effects.”^[57] I agree that Congress can legislate rules that elide a jurisdictional characterization and speak solely in terms of effects. Indeed, Congress should. After all, under my framework, Congress has no control over what is or is not jurisdictional but has full control over the effects of a rule. And focusing on the effects puts the legislative thought and emphasis where it matters for the litigants.^[58] In my view, that is entirely appropriate; parties should focus on and litigate the effects of a rule, not its jurisdictional character.

But just because the effects have overriding importance to litigants does not mean the jurisdictional character of a limit is irrelevant. As I noted, “Jurisdictional lines are important for ... identifying where the boundaries between forums are and how the various forums relate to each other in the context of a particular case.”^[59] In other words, jurisdiction retains doctrinal, relational, linguistic, and precedential importance.

Doctrinally, jurisdiction is a useful organizing principle for categorizing adjudicatory limits. As explained above, it helps group together doctrines that determine forum in a multiforum system, and it distinguishes that group from doctrines that do not.^[60] It serves descriptively to align personal jurisdiction and diversity jurisdiction without being undermined by their differing effects. At the same time, it differentiates diversity jurisdiction and standing, despite the perhaps similar effects of those doctrines.

Jurisdiction is also a relational concept. It “resolves or encourages territorial disputes within a community of forums” by “erect[ing] both the fences that separate forums and the gates that cases may pass through.”^[61] Importantly, jurisdiction always involves more than one forum: it either groups them together as legitimate competitors or alternatives for the same adjudication, or it allocates the adjudication to one forum over another.^[62] Jurisdiction thus manages the relationships among forums with competing claims to a particular adjudication.

Jurisdiction also has linguistic value. It enhances effective and productive communication by representing a term with consistent and immutable meaning. Under a positivist conception of jurisdiction, the term cannot have generalizable meaning because it means one thing for one law and another thing for another law.^[63] Under my framework, by contrast, the assertion “this requirement is jurisdictional” has a settled and useful meaning: the requirement determines which forums can hear the case and which cannot.^[64]

My approach to jurisdiction would also affect precedent. Admittedly, it would require revisiting some precedent.^[65] But it would also legitimize others, especially the discredited mantra from United States v. Robinson that appellate deadlines are “mandatory and jurisdictional.”^[66] Under current doctrine, the word “jurisdictional” does all the work, for “jurisdictional” necessarily includes “mandatory.”^[67] But under my framework, both terms do independent work: “jurisdictional” describes the boundary line and imbues it with structural meaning, while “mandatory” prescribes one of its effects.^[68]

I close this Part with a note of irony. Professor Preis chides me for “render[ing] jurisdiction ... irrelevant.”^[69] As I have shown, I have not done so: jurisdiction retains vitality under my framework. But under Professor Preis’s approach, jurisdiction is, indeed, irrelevant. Under his approach, both jurisdiction’s meaning and its effects are subject to positive law.^[70] As a result, Professor Preis’s jurisdiction has neither inherent meaning nor a defined set of effects; it is simply whatever Congress says it is at the particular time and place. It offers no definitional, doctrinal, relational, linguistic, functional, or legal relevance.^[71] What is the point?

III.  The Presumption-Based Alternative

Professor Preis nevertheless presses forward with a proposal: use my definition of jurisdiction as a presumption rebuttable by a clear statement from Congress, which courts would assess by resort to “various ‘jurisdictional indicators.’”^[72] He illustrates his approach by applying it to the requirement of the Administrative Procedure Act (APA) that a court may review only a “final agency action.”^[73] This application itself illustrates several difficulties with Professor Preis’s approach to the jurisdictionality question.

At the outset, his illustration shows just how complicated, and potentially unhelpful, his approach can be. In his application, Professor Preis spills a good deal of ink on background understandings of sovereign immunity, the lack of the term “jurisdiction” in the statute, a close parsing of the textual provision and related provisions, how courts have characterized the language in the past, and a consideration of the likely effects of the provision in litigation.^[74] Yet he concludes that most of these features do not clearly resolve the question,^[75] and it is unclear (to me, at least) how his presumption in favor of a jurisdictional ideal operates within his analysis.

Professor Preis ultimately relies primarily on precedent characterizing federal sovereign immunity as jurisdictional and on the presumed effects of a jurisdictional characterization,^[76] but there are problems here, too. One problem with relying on precedent is that such reliance is a bootstrap: Shouldn’t Professor Preis instead use his own framework to interrogate the Court’s characterization of federal sovereign immunity as jurisdictional rather than relying on the Court’s characterization as part of the framework? Another problem with relying on precedent is that the precedent here is not as clear as Professor Preis makes it out to be. True, the Supreme Court has called federal sovereign immunity “jurisdictional in nature,”^[77] but the characterization predominantly comes from the era in which the term “jurisdiction” was used in profligate and unthinking ways; today’s federal sovereign immunity opinions tend to be cagier about its jurisdictional status.^[78] Further, the APA’s language is a limitation on a waiver of federal sovereign immunity, rather than an independent grant of federal subject matter jurisdiction,^[79] and the Supreme Court has been clear that not all conditions on federal sovereign immunity waivers are themselves jurisdictional.^[80]

Professor Preis relies somewhat more heavily on the supposed effects of a jurisdictional characterization, including its “strict application.”^[81] But as I have shown, and as Professor Preis agrees, laws can have strict application without being jurisdictional (and jurisdictional laws can have flexible application).^[82] Perhaps the APA should be strictly applied, but that has no bearing on its jurisdictional character. And even were Professor Preis correct that a strict application supports a jurisdictional characterization, the fact that federal sovereign immunity is subject to waiver would seem to undermine his conclusion that the APA provision is jurisdictional. On the whole, Professor Preis’s resort to “common jurisdictional characteristics”^[83] seems unhelpful to his characterization inquiry.

The far easier approach is to rely on my framework: because the APA’s language sets a boundary between agency adjudication and judicial adjudication, it is jurisdictional. Simple. In addition, Congress is free to attach whatever effects it wishes to the finality requirement, including whether it is waivable, subject to equitable exceptions, or deserves a “strict application.”

Conclusion

Professor Preis’s alternative approach to characterization questions lacks the organizational power and simplicity of my framework while simultaneously diluting the vitality of jurisdiction. I repeat my call for a framework based on an inherent and definitional meaning of jurisdiction, coupled with a positivist approach to its effects in specific contexts.

* John Edgar Hervey Chair in Litigation and Associate Dean for Research, UC Hastings College of the Law. Thanks to Rich Freer, Liz McCuskey, Michael Morley, Phil Pucillo, Brad Shannon, Joan Shaughnessy, Howard Wasserman, and others at the SEALS Annual Conference Jurisdiction Roundtable for discussions related to this reply.