Appellate Jurisdiction in Sanchez-Gomez: A Hard Case That Should Be Easy


     In 2013, the Southern District of California adopted a rule requiring all criminal defendants to appear in full restraints—hands cuffed and connected to a waist chain, feet shackled and chained together—for most non-jury proceedings.1 There were a few exceptions, including appearances at sentencing. And a defendant could ask the district court to have the shackles removed. Four defendants, represented by federal public defenders, challenged the shackling rule, contending that it violated their due process rights. After the district court rejected these objections, the defendants sought immediate appellate review. The Ninth Circuit, sitting en banc, held that the shackling rule violated due process. In United States v. Sanchez-Gomez, the Supreme Court will decide whether the Ninth Circuit had jurisdiction to reach that holding.2

     As the Court and the parties have approached it, Sanchez-Gomez is a hard case that raises difficult and unresolved questions about the meaning and application of two major aspects of federal appellate jurisdiction: the collateral-order doctrine and appellate mandamus. But it should be an easy case. Sanchez-Gomez is a “one-shot” interlocutory appeal—once decided, it will not necessarily produce any future interlocutory appeals. It thus presents few of the costs normally associated with early appeals. And there are substantial benefits to allowing immediate review of the shackling rule, which might otherwise evade appellate scrutiny.

     Sanchez-Gomez is thus a case in which the current system of federal appellate jurisdiction gets in the way of reaching a practically sound result. In this brief essay, I begin with an explanation of why Sanchez-Gomez, as the Court and parties have approached it, is a hard case, and I then explain why it should be an easy one. I end with both a solution for Sanchez-Gomez itself—recognizing the one-shot nature of the appeal—and a suggestion for a rule of procedure governing discretionary appeals of any district court decision, which might avoid turning easy cases like Sanchez-Gomez into hard ones.

     One final note before proceeding. Appellate jurisdiction is not the only issue at play in Sanchez-Gomez. There is also a question about whether the case is moot, as all of the defendants’ charges have been resolved via pleas or dismissal as well as the question of whether the shackling rule is constitutional. The Supreme Court has granted certiorari on only the appellate jurisdiction and mootness issues. In this essay, I limit myself to appellate jurisdiction and do not address mootness or the underlying merits of the shackling rule.3

I. Why Sanchez-Gomez is a Hard Case

     First, some background.4 Most appeals in federal court must wait until the end of district court proceedings, when all issues have been decided and all that remains is enforcing the judgment.5 This general rule comes from 28 U.S.C. § 1291’s grant of appellate jurisdiction over only “final decisions” of the district courts, and it is commonly called the “final-judgment rule.”

     The final-judgment rule strikes the general balance between the interests that underlie the timing of appeals: efficiency, error correction, and development of the law. Delaying appeals has several benefits. For example, consolidating all issues in a case into a single appeal reduces appellate caseloads, requiring only a single panel of appellate judges to learn the case once, and allowing the panel to resolve all issues in a single opinion. Litigants are also saved the expense, inconvenience, and delay—as well as potential harassment by better-resourced litigants—of multiple appeals. And interlocutory appeals that might eventually become unnecessary—say, because the aggrieved party ultimately prevailed at trial—are avoided.

     But the final-judgment rule also has costs. Appellate decisions can correct errors and develop unclear areas of the law. Cases that end in settlement—a common outcome in federal court—do not produce appealable decisions, leaving errors or issues unexamined. In some cases, appellate intervention might also speed along district court proceedings or cut short what would later be deemed unnecessary litigation. And the delay between an erroneous district court decision and vindication on appeal can cause substantial, sometimes irreparable, harms.

     The final-judgment rule strikes a balance between these interests, reflecting the conclusion that the benefits of delaying appeals generally outweigh the costs. But it strikes that balance only generally. In some cases, the costs of an immediate appeal will be particularly low, or the benefits of an immediate appeal will be particularly high (or both). So the final-judgment rule has exceptions. In fact, it has many exceptions. Some come from statutes, some come from rules of procedure, and still others come from judicial interpretations of § 1291.6

     The decisions at issue in Sanchez-Gomez did not end proceedings in the district court. They addressed only the constitutionality of the shackling rule, leaving other issues—including the merits of the criminal charges—for later resolution. For the Ninth Circuit to have jurisdiction to review the shackling rule, one of the exceptions to the final-judgment rule must apply. The defendants have invoked two: the collateral-order doctrine and appellate mandamus.

     Neither avenue to appellate review is an easy one. The appellate-jurisdiction literature regularly bemoans the complex, unpredictable, and fluid nature of the existing system of federal appellate jurisdiction. There’s no need to repeat those criticisms here. All that’s necessary is a glance into the questions that Sanchez-Gomez raises.

     Let’s start with the collateral-order doctrine. That doctrine is a judicial gloss on § 1291’s grant of jurisdiction over “final decisions”. For a decision to be “final”, and thus appealable under the collateral-order doctrine, four requirements must be satisfied: (1) the district court must have conclusively resolved the issue; (2) the appealed issue must be completely separate from the merits; (3) the district court’s decision must be effectively unreviewable in an appeal from a final judgment; and (4) the issue must be sufficiently important to merit a deviation from the normal rule of delayed appeals.7

     In Sanchez-Gomez, everyone appears to agree that the first and second requirements are satisfied. The district court conclusively rejected the constitutional challenge to the shackling rule, which had nothing to do with the underlying criminal charges. But the parties dispute the third and fourth requirements—whether the district court’s decisions are both effectively unreviewable and important.

     At first blush, those final two requirements might seem to be easily satisfied in this case. After all, review of the shackling rule after conviction and sentencing doesn’t seem terribly effective for defendants that had to appear throughout their proceedings in shackles; the court of appeals cannot go back in time and un-shackle them. And the issue is likely to evade appellate review. Most criminal cases end in pleas, which are unlikely to allow review of the shackling rule. Even if the issue did make it to an appeal—say, in an appeal after conviction and sentencing, or in an appeal from a conditional guilty plea that reserved the option of appealing a suppression motion—there’s no guaranty that a court of appeals would actually review it. Shackling at non-jury proceedings is unlikely to affect a conviction or sentence, rendering any error harmless. That harmlessness would allow a court of appeals to avoid addressing the issue.

     There also seem to be some important interests at stake. The shackling rule implicates defendants’ liberty interests and their interests in being treated as individuals instead of as part of a general group of “dangerous” defendants. It might even interfere with their ability to assist their counsel, as the shackling itself might hinder defendants from getting their counsel’s attention or writing notes to counsel. The rule also implicates the safety of district court proceedings. The Southern District of California purportedly adopted the rule because it was concerned that the lack of background information on some defendants made individualized determinations of dangerousness too difficult.8 Rather than risk violence at the hands of some defendants, the district decided to shackle them all.

     The thing is, the Supreme Court often treats the collateral-order doctrine as far narrower than its stated requirements might make it seem. After all, all sorts of district court decisions can be only imperfectly remedied in an appeal after a final judgment. Lots of district court decisions evade appellate review. And many district court decisions implicate important interests. Were it enough for a decision to be effectively unreviewable and important in some sense, oodles of district court decisions would be immediately appealable collateral orders. But that can’t be the case; otherwise the exception would swallow the rule. Indeed, the Supreme Court rarely mentions the collateral-order doctrine without also mentioning its narrow scope.9

     Something more is required for an order to be immediately appealable under the collateral-order doctrine. Often that something amounts to a judgment call about whether the interests at stake are sufficiently important, and whether delayed appellate review would be sufficiently ineffective, to justify an immediate appeal.10 That’s not always a simple question. In making similar calls in the past, the Supreme Court has drawn some fine lines. It has held, for example, that criminal defendants can immediately appeal the denial of claims that the district court has set bail excessively high.11 They cannot, however, appeal the denial of a motion to dismiss on prosecutorial-vindictiveness grounds.12 Defendants can immediately appeal denials of a speech-and-debate-clause defense and denials of a claim that a prosecution violates the prohibition on double jeopardy,13 but they cannot appeal the denial of a motion to dismiss on speedy-trial grounds.14 And although orders requiring that defendants be involuntarily medicated so as to render them competent to stand trial can be immediately appealed, orders disqualifying defense counsel cannot.15

     Review via mandamus poses its own difficulties. Technically, mandamus is not an appeal. It’s an independent proceeding in the court of appeals, begun with a petition for a writ ordering the district court to take (or cease taking) some particular action. For a writ of mandamus to be proper, three requirements must be met: (1) the petitioner must have no other adequate means for relief; (2) the petitioner’s entitlement to relief must be clear and indisputable; and (3) the court of appeals, in its discretion, must deem the writ appropriate.16

     It’s not clear that the circumstances in Sanchez-Gomez satisfy any of these requirements. The dispute about whether the defendants have other adequate means for relief largely mirrors the dispute over the collateral-order requirement that the decision be effectively unreviewable. One additional wrinkle comes from the United States’ suggestion that the detainees could challenge the shackling rule via a class action against the Marshals Service under Ex parte Young. The alternative of a civil suit has not played a role in the collateral-order context, but its possibility raises questions about the appropriateness of mandamus. The likelihood of a civil suit challenging the shackling rule is also questionable, as the federal public defenders currently representing the defendants in Sanchez-Gomez probably could not represent them (or any other defendants) in a class action.

     The second requirement—that the right to the writ be clear and indisputable—is more unique to the mandamus issue. And no one is quite sure what that requirement means.17 At first glance, it seems to be something akin to plain error—that is, the district court’s error must be obvious. The rationale for such a strict requirement would be that mandamus is not a substitute for appeal, but is instead an extraordinary and rare remedy for instances when district courts have gone completely out of bounds. Indeed, at oral argument, Justice Kagan suggested that mandamus was usually reserved for correcting obvious errors, not for resolving important issues.18

     But that’s not entirely true. The Supreme Court has approved the use of mandamus to decide difficult issues that had no clear answer. In Schlagenhauf v. Holder, the Court affirmed the issuance of a writ of mandamus to decide an issue of first impression.19 And in Mallard v. U.S. District Court, it endorsed the use of the writ to decide an issue on which the courts of appeals had split.20 These and other instances demonstrate the “advisory” or “supervisory” use of mandamus, which is fairly well established in the courts of appeals.21

     That being said, it’s not clear that the current Supreme Court is ready to openly embrace the advisory and supervisory uses of mandamus. Just as the Court regularly describes the collateral-order doctrine as a narrow avenue for appeals, it also emphasizes the extraordinary nature of mandamus review.22

II. Why Sanchez-Gomez Should Be Easy

     Thus, as the Court and the parties have approached the case, Sanchez-Gomez poses some difficult questions about the meaning and application of the collateral-order doctrine and appellate mandamus. These questions are not un-answerable; indeed, they’re no more difficult than many of the other issues the Supreme Court regularly encounters.

     It’s just that these issues—again, as the Court and the parties have approached them—are much harder than they need to be. Sanchez-Gomez should be an easy case. It presents few of the costs normally associated with interlocutory appeals, as it poses no real risk of increasing appellate caseloads. The benefits of immediate review are also greater than normal, given the chance that the shackling rule might otherwise evade appellate review.

     As for costs, Sanchez-Gomez is relatively unique in its likely impact on appellate caseloads. Most disputes over appellate jurisdiction raise the specter of increased appellate caseloads, which courts of appeals are often eager to avoid. In many cases, allowing an immediate appeal would open the door for numerous subsequent litigants to appeal the same issue. Consider, for example, appeals from a denial of qualified immunity. Qualified immunity is a defense in civil rights actions that shields government officials from damages unless their actions violated clearly established law. In Mitchell v. Forsyth, the Supreme Court held that denials of qualified immunity are immediately appealable via the collateral-order doctrine.23 Mitchell has led to thousands of subsequent appeals, as government officials seek immediate review of whether they—under the particular circumstances of their cases—are entitled to qualified immunity.

     To avoid this influx of appeals, the federal courts try to keep the avenues to immediate appellate review narrow and extraordinary. And for good reason. Increases in the number of actual appeals can be costly and burdensome for resource-strapped courts. So, too, are increases in attempts to immediately appeal district court decisions, even if those attempts are ultimately unsuccessful. The courts of appeals must expend effort beating back all of these attempted appeals. These efforts can be just as burdensome and costly as actual appeals, as the courts must consider and reject the proffered grounds for jurisdiction. Strict doctrines and requirements mitigate some of that burden by discouraging at least some litigants from trying for immediate review. When litigants do seek immediate review, the courts of appeals are well equipped to reject those attempts by pointing to these strict requirements.

     But not all interlocutory appeals necessarily open the door for subsequent appeals. Consider Cohen v. Beneficial Loan Corp., the case in which the Supreme Court created the collateral-order doctrine.24 Cohen involved an Erie question: must a federal court apply a state law requiring the posting of a litigation bond? That question was relatively abstract, as the appellant was seeking a decision only on whether the bond requirement applied in federal court. There were no case-specific issues, such as the amount of the required bond. Nor was the anyone arguing that something unique about the case affected the applicability of the bond requirement. Once the Supreme Court decided the issue, it was settled once and for all. No other litigants would need to appeal the question of whether that particular bond requirement applied in federal court. There was no real risk of subsequent appeals raising the same issue. Cohen thus involved a relatively abstract issue of law that, once resolved, need not be addressed in subsequent appeals. It was a “one-shot” interlocutory appeal.

     Sanchez-Gomez is, too. The Ninth Circuit addressed only the constitutionality of the district-wide shackling rule. Once that issue is resolved, there will be no need for subsequent appeals of it. Importantly, immediate review of the constitutionality of the shackling rule need not open the door to immediate review of individual shackling determinations. Appeals from individual shackling determinations raise different considerations than an attack on the shackling rule itself. This isn’t to say that individual shackling determinations should not be appealable. It’s just that the appealability of those determinations is a different question, one that Sanchez-Gomez need not resolve.

     Allowing the appeal in Sanchez-Gomez thus does not pose a real risk of increased appellate caseloads. Other issues that often accompany immediate appeals—such as potential delay of district court proceedings and the possibility of duplicative appeals—are also relatively unlikely to arise due to the abstract nature of the question Sanchez-Gomez presents. And, as discussed above in relation to the collateral-order doctrine, there is a real risk that without immediate appeal the shackling rule will continue to evade appellate review. If there is to be any appellate review of the rule, it might need to be immediate.

     Framing Sanchez-Gomez in terms of the collateral-order doctrine or appellate mandamus thus turns an easy case into a hard one. It might be too late to treat Sanchez-Gomez like the easy case that it is. Absent a decision that radically changes the law of appellate jurisdiction, the Court is stuck with the current rules for the collateral-order doctrine and appellate mandamus. But the Court might make Sanchez-Gomez an easier case by acknowledging the one-shot nature of the case. In many circumstances, immediate appeals are more appropriate when they raise relatively abstract issues of law, the resolution of which would resolve the issue for all litigants and would not require any future appeals on case-specific applications of the law. Granted, recognition of one-shot interlocutory appeals via either the collateral-order doctrine or mandamus might invite litigants to appeal other serious, unsettled, and relatively abstract issues of law. But it would not drastically increase appellate caseloads by opening the door to an influx of case-specific appeals.

     There is support for this approach in both of the avenues for immediate appellate review discussed above. Cohen is one of the few collateral-order decisions that can be characterized as a one-shot interlocutory appeal. Most open the door to subsequent, case-specific appeals of the same issue. One-shots are more common in the mandamus context. In Thermtron Products, Inc. v. Hermansdorfer, for example, the Supreme Court held that mandamus was the proper vehicle to decide that a district court could not remand a removed case due to the district court’s crowded docket.25 And in Mallard v. U.S. District Court, the Court held that mandamus was appropriate to determine that district courts could not force attorneys to represent indigent defendants in civil proceedings.26 To be sure, not all mandamus cases are one-shots. Schagenhauf v. Holder, for example, first addressed the abstract issue of whether a district court could order the mental and physical examination of a defendant under Civil Rule 35 (it could), and then the case-specific question of whether an examination of a particular defendant was appropriate.27 But mandamus has provided a valuable vehicle for one-shot appeals.

III. The Rulemaking Alternative

     Going forward, a better option for resolving issues of federal appellate jurisdiction might be rulemaking. In other work, I have advocated rules-based reform in this area.28 Much of that reform would consist of categorical rules providing either mandatory or discretionary jurisdiction over particular kinds of district court decisions. These rules could bring some much-needed transparency and accessibility to an unnecessarily obscure area of law. Explicit exceptions might also bring some simplicity, clarity, and predictability to appellate jurisdiction. And reform need not consist entirely of categorical rules—it could also be supplemented with a catchall giving the courts of appeals discretion to review any district court decision.29

     A discretionary catchall could avoid turning easy cases like Sanchez-Gomez into hard ones. Rather than try to fit a unique case into the framework of an existing exception to the final-judgment rule, courts could candidly address the costs and benefits of allowing an immediate appeal. The details of the discretionary catchall are unimportant at this point. It would be part of a wholesale reform, and its contours would depend in part on the other rules governing the timing of appeals. And there are real risks that a rule of discretion—whatever its form—would increase appellate caseloads.30

     But this essay is not the appropriate place for a full-throated defense of discretion and flexibility in appellate jurisdiction. I’ll simply note that the current system is one that already includes discretionary appeals.31 Those appeals are just hidden behind doctrines like the collateral-order doctrine and appellate mandamus.

     Obscuring this discretion has some benefits. Because some litigants simply are not aware of the possibilities for immediate appeals—and thus never attempt them—it reduces appellate caseloads. Wrapping this discretion in vague, unhelpful, and occasionally fluid doctrinal requirements also discourages some attempts to appeal. But it’s odd policy to keep appellate caseloads manageable by keeping lawyers ignorant of opportunities to appeal and unnecessarily complicating matters of appellate jurisdiction.

     What’s important for now is to recognize that absent some candid and transparent discretion and flexibility, the current system of appellate jurisdiction will keep turning easy cases like Sanchez-Gomez into hard ones.

IV. Conclusion

     Sanchez-Gomez—as the Supreme Court and the parties have approached it—is a hard case that should be easy. The difficulty is an inevitable byproduct of our current system of federal appellate jurisdiction. To avoid turning easy issues of appellate jurisdiction into hard ones, that system must change. And rulemaking—including a rule for discretionary appeals—is probably the best way to achieve that change.


1. See United States v. Sanchez-Gomez, 859 F.3d 649, 653–54 (9th Cir. 2017) (en banc).

2. See generally United States v. Sanchez Gomez, No. 17-312 (U.S. argued March 26, 2018).

3. For a discussion of the merits, see Recent Case, Ninth Circuit Deems Unconstitutional Routine Shackling in Pretrial Proceedings: United States v. Sanchez-Gomez, 131 Harv. L. Rev. 1163 (2018).

4. I’ve adapted some of this background material from Bryan Lammon, Dizzying Gillespie: The Exaggerated Death of the Balancing Approach and the Inescapable Allure of Flexibility in Appellate Jurisdiction, 51 U. Rich. L. Rev. 371, 374–76 (2017).

5. See Gelboim v. Bank of Am. Corp., 135 S. Ct. 897, 902 (2015); Catlin v. United States, 324 U.S. 229, 233 (1945).

6. Others have extensively discussed the various exceptions to the final-judgment rule. See, e.g., Thomas E. Baker, A Primer on the Jurisdiction of the U.S. Courts of Appeals 42–71 (2d ed. 2009); Aaron R. Petty, The Hidden Harmony of Appellate Jurisdiction, 62 S.C. L. Rev. 353, 360–93 (2010); Andrew S. Pollis, The Need for Non-Discretionary Interlocutory Appellate Review in Multidistrict Litigation, 79 Fordham L. Rev. 1643, 1652–59 (2011); Adam N. Steinman, Reinventing Appellate Jurisdiction, 48 B.C. L. Rev. 1237, 1244–72 (2007).

7. See Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 106 (2009); Coopers & Lybrand v. Livesay, 437 U.S. 463, 468 (1978). Courts often present these four requirements as three, combining the second and fourth. Again, others have extensively discussed the collateral-order doctrine. See, e.g., Baker, supra note 6, at 42–45; 15A Charles Alan Wright, Arthur R. Miller & Edward H. Cooper, Federal Practice and Procedure §§ 3911–3911.5 (2d ed. 1992); Lloyd C. Anderson, The Collateral Order Doctrine: A New “Serbonian Bog” and Four Proposals for Reform, 46 Drake L. Rev. 539 (1998); Bryan Lammon, Rules, Standards & Experimentation in Appellate Jurisdiction, 74 Ohio St. L.J. 423, 447–59 (2013); Petty, supra note 6, at 377–86; Steinman, supra note 6, at 1247–57.

8. See United States v. Sanchez-Gomez, 859 F.3d 649, 666–67 (9th Cir. 2017) (en banc) (Ikuta, J., dissenting).

9. See, e.g., Will v. Hallock, 546 U.S. 345, 350 (2006).

10. See Pollis, supra note 6, at 1652–53; Steinman, supra note 6, at 1255–56.

11. See Stack v. Boyle, 342 U.S. 1 (1951).

12. See United States v. Hollywood Motor Car Co., 458 U.S. 263 (1982).

13. See Helstoski v. Meanor, 442 U.S. 500 (1979) (speech-and-debate clause); Abney v. United States, 431 U.S. 651 (1977) (double jeopardy).

14. See United States v. MacDonald, 435 U.S. 850 (1978).

15. See Sell v. United States, 539 U.S. 166 (2003) (involuntary medication); Flanagan v. United States, 465 U.S. 259 (1984) (disqualification).

16. See Cheney v. U.S. Dist. Court for D.C., 542 U.S. 367, 380–81 (2004). For in-depth discussions of appellate mandamus, see 16 Wright et al., supra note 7, §§ 3932–3936.3; Robert S. Berger, The Mandamus Power of the United States Courts of Appeals: A Complex and Confused Means of Appellate Control, 31 Buff. L. Rev. 37 (1982); Paul R. Gugliuzza, The New Federal Circuit Mandamus, 45 Ind. L. Rev. 343, 351–61 (2012); Steinman, supra note 6, at 1257–72; Note, Supervisory and Advisory Mandamus Under the All Writs Act, 86 Harv. L. Rev. 595 (1973).

17. See supra citations in note 16.

18. See Transcript of Oral Argument at 39, United States v. Sanchez-Gomez, No. 17-312 (U.S. March 19, 2018),

19. See Schlagenhauf v. Holder, 379 U.S. 104, 110 (1964).

20. See Mallard v. U.S. Dist. Court for S. Dist. of Iowa, 490 U.S. 296, 309–10 (1989).

21. See generally Note, supra note 16.

22. See, e.g., Cheney v. U.S. Dist. Court for D.C., 542 U.S. 367, 403 (2004).

23. Mitchell v. Forsyth, 472 U.S. 511, 530 (1985).

24. See Cohen v. Beneficial Indus. Loan Corp., 337 U.S. 541 (1949).

25. See Thermtron Prods., Inc. v. Hermansdorfer, 423 U.S. 336, 351 (1976), abrogated on other grounds by Quackenbush v. Allstate Ins. Co., 517 U.S. 706 (1996).

26. See Mallard v. U.S. Dist. Court for S. Dist. of Iowa, 490 U.S. 296, 309 (1989).

27. See Schlagenhauf v. Holder, 379 U.S. 104, 115 (1964).

28. See Bryan Lammon, Hall v. Hall: A Lose-Lose Case for Appellate Jurisdiction, 68 Emory L.J. Online (forthcoming 2018); Bryan Lammon, Finality, Appealability, and the Scope of Interlocutory Review, 93 Wash. L. Rev. (forthcoming 2018); Bryan Lammon, Cumulative Finality, 52 Ga. L. Rev. (forthcoming 2018); Lammon, supra note 4.

29. See Lammon, supra note 4, at 415–18.

30. See Lammon, supra note 7, at 434.

31. See Lammon, supra note 4, at 411; Steinman, supra note 6, at 1276–77.

%d bloggers like this: