Ryan v. ICE

Opinion

United States Court of Appeals For the First Circuit

No. 19-1838

MARIAN RYAN, in her official capacity as Middlesex County District Attorney, ET AL.,

Plaintiffs, Appellees,

v.

U.S. IMMIGRATION AND CUSTOMS ENFORCEMENT ET AL., Defendants, Appellants.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF MASSACHUSETTS

[Hon. Indira Talwani, U.S. District Judge]

Before

Torruella, Selya, and Kayatta, Circuit Judges.

Francesca M. Genova, Trial Attorney, Office of Immigration Litigation, U.S. Department of Justice, with whom Joseph H. Hunt, Assistant Attorney General, Civil Division, William C. Peachey, Director, Office of Immigration Litigation, and Erez R. Reuveni, Assistant Director, were on brief, for appellants.

Michael M. Hethmon, Christopher J. Hajec, and Ralph L. Casale on brief for Immigration Reform Law Institute, amicus curiae.

David J. Zimmer, Special Assistant Attorney General of Massachusetts, with whom Daryl L. Wiesen, Alicia Rubio-Spring, and Goodwin Procter LLP were on brief, for appellees Ryan and Rollins.

Wendy S. Wayne on brief for appellee Committee for Public Counsel Services.

Oren N. Nimni, Lawyers for Civil Rights, David J. Zimmer, Daryl L. Wiesen, Alicia Rubio-Spring, and Goodwin Procter LLP on brief for appellee Chelsea Collaborative, Inc.

Dayna J. Zolle, Elizabeth B. Wydra, Brianne J. Gorod, and Ashwin Phatak on brief for Constitutional Accountability Center, amicus curiae.

Nikolas Bowie, Sabrineh Ardalan, Philip L. Torrey, and Norah Rast on brief for Nikolas Bowie and Harvard Immigration and Refugee Clinical Program, amici curiae.

Ari J. Savitzky, Assistant Solicitor General of New York, Letitia James, Attorney General, Barbara D. Underwood, Solicitor General, Steven C. Wu, Deputy Solicitor General, William Tong, Attorney General of Connecticut, Karl A. Racine, Attorney General for the District of Columbia, Kwame Raoul, Attorney General of Illinois, Brian E. Frosh, Attorney General of Maryland, Keith Ellison, Attorney General of Minnesota, Gurbir S. Grewal, Attorney General of New Jersey, Hector Balderas, Attorney General of New Mexico, Ellen F. Rosenblum, Attorney General of Oregon, Josh Shapiro, Attorney General of Pennsylvania, Peter F. Neronha, Attorney General of Rhode Island, Robert W. Ferguson, Attorney General of Washington, Thomas J. Donovan, Jr., Attorney General of Vermont, and Mark R. Herring, Attorney General of Virginia, on brief for states of New York, Connecticut, Illinois, Maryland, Minnesota, New Jersey, New Mexico, Oregon, Pennsylvania, Rhode Island, Vermont, Virginia, and Washington and the District of Columbia, amici curiae.

Douglas E. Keith, Alicia L. Bannon, and Brennan Center for Justice on brief for 19 Former Massachusetts Judges, amici curiae.

Thomas J. Carey, Jr., Martin W. Healy, and Christopher N.

Lasch on brief for Massachusetts Bar Association, Boston Bar Association, Massachusetts Academy of Trial Attorneys, Women's Bar Association of Massachusetts, and South Asian Bar Association of Greater Boston, amici curiae.

Maria T. Davis, Howard M. Cooper, and Todd & Weld, LLP on brief for Massachusetts Association of Criminal Defense Lawyers, amicus curiae.

Lauren Godles Milgroom, Joel A. Fleming, Amanda R. Crawford, and Block & Leviton LLP on brief for 27 Domestic and Sexual Violence Advocacy Organizations, amici curiae.

September 1, 2020

SELYA, Circuit Judge. United States Immigration and Customs Enforcement (ICE) is the arm of the federal government charged with the apprehension and detention of noncitizens who are subject to removal. Believing state courthouses to be appropriate locations in which to conduct civil enforcement actions, ICE increased its efforts to arrest allegedly removable noncitizens in and around state courthouses when they appeared for judicial proceedings. In January of 2018, ICE issued Directive 11072.1 (the Directive), formalizing its policy regarding civil enforcement actions in such courthouses.

ICE's growing presence in Massachusetts courthouses concerned a number of persons and organizations, including Marian Ryan and Rachael Rollins (the District Attorneys of Middlesex County and Suffolk County, respectively), the Committee for Public Counsel Services (the main public defender agency for the Commonwealth of Massachusetts), and Chelsea Collaborative, Inc. (a nonprofit that provides services to the immigrant community in Chelsea, Massachusetts). Fearing the effects of ICE's activities on the proper functioning of both the state judicial system and access to justice in immigrant communities, they sued ICE, the United States Department of Homeland Security (DHS), and three DHS officials (collectively, the defendants), specifically challenging

4

the Directive and generally challenging ICE's policy of civilly arresting individuals attending court on official business.1 When the plaintiffs moved for a preliminary injunction, they argued primarily that ICE lacked statutory authority under the Immigration and Nationality Act (INA), 8 U.S.C. §§ 1101-1537, to conduct such arrests because the INA implicitly incorporates a hoary common law privilege against civil arrests for parties and witnesses attending court proceedings. The district court determined that the plaintiffs were likely to succeed on the merits of this argument and preliminarily enjoined ICE from implementing the Directive or otherwise civilly arresting individuals attending court on official business anywhere in Massachusetts. See Ryan v. U.S. Immigr. & Customs Enf't, 382 F. Supp. 3d 142, 159, 161 (D. Mass. 2019).2 On this interlocutory appeal, we have carefully considered the district court's rescript and the compendious

1 At present, the three individual defendants, named in their official capacities, are Chad Wolf, Acting Secretary of DHS; Matthew T. Albence, Acting Deputy Director of ICE (who, as the senior official currently in place, performs the duties of the Director); and Todd M. Lyons, ICE's Acting Boston Field Office Director.

2 Throughout this litigation, the plaintiffs have described

the individuals whom they believe ICE officers may not civilly arrest in and around courthouses as those attending court "on official business." The plaintiffs have not clearly defined the contours of this phrase, but they seem to mean parties, witnesses, and victims at a bare minimum. Notwithstanding this potential lack of clarity, the district court adopted the terminology. See Ryan, 382 F. Supp. 3d at 146, 161. We, too, use it as a convenient shorthand.

5

briefing furnished by both the parties and an array of helpful amici. We conclude that the district court abused its discretion in finding that the plaintiffs were likely to succeed on the merits of their argument that the INA implicitly incorporates a common law privilege against civil arrests for individuals attending court on official business. Turning to the plaintiffs' backup argument, we likewise conclude that, on the underdeveloped record before us, the plaintiffs have so far failed to show that they are likely to succeed in arguing that ICE lacks statutory authority to conduct such arrests in Massachusetts because Congress has not clearly stated its intent to permit arrests that violate state law. Consequently, we vacate the preliminary injunction and remand for further proceedings consistent with this opinion. I. BACKGROUND We briefly rehearse the relevant facts and travel of the case. For some time, ICE has conducted civil enforcement actions designed to take removable noncitizens into custody in courthouses across the country. During the Obama administration, ICE imposed certain restrictions on the ability of its officers to conduct such actions in courthouses. In March of 2014, ICE issued guidance directing that "[e]nforcement actions at or near courthouses will only be undertaken against Priority 1 aliens." An earlier ICE policy, which remained in effect, defined "Priority 1 aliens" as those posing a threat to national security or public safety. The

6

2014 guidance also forbade ICE officers from arresting "collateral" noncitizens whom they encounter during an enforcement action against a Priority 1 target, such as family members or friends accompanying the target to a court appearance.

ICE's enforcement priorities changed when the political winds shifted. Shortly after taking office, President Trump issued an executive order (the EO) on January 25, 2017, declaring the federal government's intention to "[e]nsure the faithful execution of the immigration laws . . . against all removable aliens." Exec. Order No. 13,768, 3 C.F.R., 2017 Comp., p. 268, reprinted in 8 U.S.C. § 1103 app. at 647-49 (2018). To this end, the EO expanded the classes of noncitizens prioritized for removal. See id. at 269. A month later, the Secretary of DHS handed down a memorandum implementing the EO and rescinding any conflicting directives or guidance. This memorandum reiterated the broader enforcement priorities delineated in the EO.

Neither the EO nor the implementing memorandum directly addressed courthouse arrests. It nonetheless appears that ICE officers began to conduct more civil arrests in and around state courthouses, including those in Massachusetts. ICE attributes this change to a newfound unwillingness on the part of many state and local governments to honor civil immigration detainers, which ask law enforcement agencies to hold allegedly removable noncitizens beyond their scheduled release from criminal custody

7

so that federal immigration officers may detain them. See City of Providence v. Barr, 954 F.3d 23, 29 (1st Cir. 2020). Such an unwillingness was antithetic to ICE's claim that courthouses were the safest place to arrest such noncitizens because courthouse visitors are customarily screened for weapons upon their arrival. In Massachusetts, the situation took on a new dimension when the Massachusetts Supreme Judicial Court (SJC) held that state-court functionaries could not detain noncitizens based solely on civil immigration detainers. See Lunn v. Commonwealth, 78 N.E.3d 1143, 1146 (Mass. 2017) (per curiam).

The Chief Justice of the Massachusetts Trial Court, in response to Lunn and ICE's more pervasive presence in Massachusetts courthouses, promulgated a policy for state-court personnel regarding civil immigration enforcement actions in state courthouses. This policy took effect in November of 2017. Under it, ICE officers "may enter a courthouse and perform their official duties provided that their conduct in no way disrupts or delays court operations, or compromises court safety or decorum." The policy directs state-court personnel to ask any armed ICE officer seeking entry into a courthouse to state his law-enforcement purpose and to describe the enforcement action that he proposes to undertake. If an ICE officer attempts to effect a civil arrest of a noncitizen who is not in the court's custody, the policy instructs state-court personnel neither to impede nor to assist

8

with the arrest. ICE officers may not conduct civil arrests either in nonpublic spaces within a courthouse or (absent permission in advance) in courtrooms.

In January of 2018, ICE issued the Directive, codifying its policy anent civil enforcement actions in courthouses. The Directive enlarged the categories of noncitizens subject to civil arrest in courthouses beyond those specified in the agency's 2014 guidance. Specifically, ICE officers were authorized to target noncitizens "who have been ordered removed from the United States but have failed to depart" and those "who have re-entered the country illegally after being removed." What is more, the Directive relaxed the restriction on arresting "collateral" noncitizens present during an enforcement action: it authorized ICE officers to arrest such an individual under "special circumstances, such as where [he or she] poses a threat to public safety or interferes with ICE's enforcement actions." Although the Directive instructed ICE officers to "generally avoid enforcement actions in courthouses, or areas within courthouses that are dedicated to non-criminal . . . proceedings," it allowed such actions when "operationally necessary."3

3 ICE's implementation of the Directive was debated extensively during the preliminary injunction hearing. The defendants represented that ICE applies the Directive in conjunction with a prior policy discouraging the removal of victims, witnesses, and other noncitizens attempting to protect their noncriminal legal rights. In line with this policy, ICE

9

As ICE started to conduct more civil enforcement actions in courthouses across Massachusetts, the plaintiffs' concerns mounted. In their view, the specter of courthouse arrest deters noncitizens from appearing in court (whether as criminal defendants, victims of crimes, parties to civil litigation, or witnesses). This deterrent effect, they say, interferes with the ability of prosecutors and defense counsel to resolve criminal charges and the ability of noncitizens to enforce their legal rights in noncriminal areas ranging from domestic violence to employment.

Spurred by these concerns, the plaintiffs filed suit in the United States District Court for the District of Massachusetts. At the same time, they moved for preliminary injunctive relief with respect to the first count of their complaint, which alleged that the Directive exceeds ICE's statutory authority and, thus, violates the Administrative Procedure Act (APA). See 5 U.S.C. § 706(2)(C). In support, they argued chiefly that the Directive exceeds ICE's statutory authority because the civil arrest power

"generally" conducts courthouse arrests only of criminal defendants and not of victims, witnesses, and civil litigants. Ryan, 382 F. Supp. 3d at 161. The plaintiffs did not dispute this account of how ICE was implementing the Directive but observed that the Directive's language appears to authorize a broader range of civil arrests. At any rate, the precise manner in which ICE is implementing the Directive is largely irrelevant to the resolution of this appeal. To the extent that we rely on facts relative to ICE's implementation of the Directive, those facts are uncontroverted.

10

in the INA implicitly incorporates a common law privilege protecting against the civil arrest of individuals attending court on official business.

After a two-day hearing at which no witnesses were called, the district court granted the plaintiffs' motion. See Ryan, 382 F. Supp. 3d at 161. To begin, the court found that the plaintiffs had both constitutional and prudential standing to bring their APA challenge to the legality of the Directive.4 See id. at 152-55. Next, the court found that the plaintiffs were likely to succeed in showing that the Directive exceeds ICE's statutory authority. See id. at 155-59. Finally, the court found that the remaining factors in the preliminary injunction calculus favored the plaintiffs. See id. at 159-61. Summing up, the court preliminarily enjoined the defendants "from implementing [the Directive] in Massachusetts and from civilly arresting parties, witnesses, and others attending Massachusetts courthouses on official business while they are going to, attending, or leaving the courthouse." See id. at 161.

4The defendants did not renew their attack on standing in their appellate briefing and, at oral argument, they disclaimed any intention of pressing this offensive. Even so, we have an independent duty to ensure that constitutional standing exists before proceeding to the merits. See Merrimon v. Unum Life Ins. Co. of Am., 758 F.3d 46, 52 (1st Cir. 2014). We have reviewed the parties' submissions in the court below and, for substantially the reasons given by the district court, see Ryan, 382 F. Supp. 3d at 152-55, we conclude that the plaintiffs have constitutional standing to pursue their APA claim.

11

This interlocutory appeal followed. We have jurisdiction under 28 U.S.C. § 1292(a)(1). II. ANALYSIS We have erected a four-part framework for district courts to use in determining whether to grant or deny a preliminary injunction. Under this framework, a district court is tasked with considering the movant's likelihood of success on the merits; whether and to what extent the movant will suffer irreparable harm in the absence of preliminary injunctive relief; the balance of relative hardships, that is, the hardship to the nonmovant if enjoined as opposed to the hardship to the movant if no injunction issues; and the effect, if any, that either a preliminary injunction or the absence of one will have on the public interest. See Corp. Techs., Inc. v. Harnett, 731 F.3d 6, 9 (1st Cir. 2013); Ross-Simons of Warwick, Inc. v. Baccarat, Inc., 102 F.3d 12, 15 (1st Cir. 1996).

The movant's likelihood of success on the merits weighs most heavily in the preliminary injunction calculus. See Ross- Simons, 102 F.3d at 16. Indeed, we have described likelihood of success as the "sine qua non" of preliminary injunctive relief. New Comm Wireless Servs., Inc. v. SprintCom, Inc., 287 F.3d 1, 9 (1st Cir. 2002). If the movant "cannot demonstrate that he is likely to succeed in his quest, the remaining factors become matters of idle curiosity." Id.

12

There is, of course, an important distinction between a decision at the preliminary injunction stage and a final decision on the merits. Battles over preliminary injunctions normally are waged early in a case. At such an embryonic stage in the litigation, an inquiring court need not conclusively determine the merits of the movant's claim; it is enough for the court simply to evaluate the likelihood vel non that the movant ultimately will prevail on the merits. See Ross-Simons, 102 F.3d at 16.

We assay a district court's decision to grant or deny a preliminary injunction for abuse of discretion. See Corp. Techs., 731 F.3d at 10. Within this rubric, we examine answers to abstract legal questions de novo, findings of fact for clear error, and any judgment calls concerning the balancing of the four factors with significant deference to the district court. See id.; Ross-Simons, 102 F.3d at 16. A district court may be held to have abused its discretion by, say, making a material error of law, "ignor[ing] pertinent elements deserving significant weight, consider[ing] improper criteria, or, though assessing all appropriate and no inappropriate factors, plainly err[ing] in balancing them." Ross- Simons, 102 F.3d at 16; see Corp. Techs., 731 F.3d at 10.

The parties' briefs leave no doubt that they hold diametrically opposite positions on whether it is good public policy for ICE to arrest noncitizens in courthouses. In the plaintiffs' view, these arrests undermine access to justice in

13

immigrant communities and interfere with the timely resolution of criminal prosecutions. In the defendants' view, the game is worth the candle: they insist that courthouses provide a safe location for arresting noncitizens who pose a threat to public safety.

It is not for us to say whether ICE's strategy is sound public policy or, conversely, whether that strategy is antithetic to sound public policy. That question lies within the domain of the politically accountable branches of the federal and state governments. Our task is simply to decide the pertinent legal issues and determine whether the district court abused its discretion in granting the preliminary injunction.

A.

Our starting point is the plaintiffs' likelihood of success on the merits. In support of their assertion that they are likely to succeed on their APA claim, the plaintiffs renew the principal argument that they advanced below: that the Directive and ICE's policy of civilly arresting individuals attending court on official business exceed ICE's statutory authority under the INA. Importantly, they do not challenge the power of ICE officers to conduct criminal arrests in and around courthouses. See 8 U.S.C. § 1357(a)(4)-(5) (permitting ICE officers to make warrantless arrests for certain federal crimes). Nor do they challenge ICE's authority to make civil arrests of noncitizens brought to courthouses in either state or federal custody.

14

Instead, their argument focuses exclusively on ICE's lack of any authority to civilly arrest individuals who travel to courthouses on their own to attend court on official business. For ease in exposition, we henceforth refer to this type of arrest as a "courthouse arrest."

The plaintiffs' argument presents a pure question of law, which we review de novo. See Corp. Techs., 731 F.3d at 10. It is a bedrock principle that the power of an executive agency administering a federal statute "is 'authoritatively prescribed by Congress.'" City of Providence, 954 F.3d at 31 (quoting City of Arlington v. FCC, 569 U.S. 290, 297 (2013)). When an agency acts in a manner not authorized by statute, its action is ultra vires and a violation of the APA. See id.; see also 5 U.S.C. § 706(2)(C).

So viewed, the plaintiffs' argument concerning the scope of ICE's civil arrest authority under the INA reduces to a question of statutory construction. As with any effort to decipher the meaning of a federal statute, the touchstone of this inquiry is congressional intent. See City of Providence, 954 F.3d at 31. And it is nose-on-the-face plain that "the quest to determine this intent must start with the text of the statute itself." Id.

The text of the INA confers broad authority upon ICE to conduct civil arrests. Under 8 U.S.C. § 1226(a), an ICE officer may arrest a noncitizen pursuant to an administrative warrant and may detain him during the pendency of removal proceedings. An

15

interlocking INA provision, 8 U.S.C. § 1357(a)(2), authorizes the warrantless arrest of a noncitizen if an ICE officer has reason to believe that the noncitizen is in the United States unlawfully and is likely to escape before he can obtain a warrant. On their face, neither of these provisions bars ICE officers from exercising their civil arrest power either in courthouses or against individuals attending court on official business. Nor do the plaintiffs identify an explicit limitation to this effect anywhere else in the text of the INA.

Recognizing that the INA does not prohibit courthouse arrests in haec verba, the plaintiffs' primary theory invokes what has been called the "nonderogation canon" of statutory construction. Pasquantino v. United States, 544 U.S. 349, 359 (2005). They contend that there is a long-standing common law privilege against civil courthouse arrests. Given this privilege, they add, we must presume that Congress intended not to permit courthouse arrests when it authorized civil arrests in the INA. To cap the matter, they submit that nothing in the text of the INA rebuts the presumption that Congress intended to incorporate this common law privilege, albeit sub silentio, into the statute.

The nonderogation canon instructs that "[s]tatutes which invade the common law . . . are to be read with a presumption favoring the retention of long-established and familiar principles, except when a statutory purpose to the contrary is

16

evident." Id. (alterations in original) (quoting United States v. Texas, 507 U.S. 529, 534 (1993)). This instruction is rooted in the notion that when Congress legislates in an area permeated by such principles, it "does not write upon a clean slate." Texas, 507 U.S. at 534. Consequently, a court must assume that Congress is aware of such long-established and familiar principles and — in the absence of an evident statutory purpose to the contrary — intends to retain them. See Lexmark Int'l, Inc. v. Static Control Components, Inc., 572 U.S. 118, 132 (2014); Samantar v. Yousuf, 560 U.S. 305, 320 n.13 (2010). Properly envisioned, the nonderogation canon is just a tool to assist in discerning congressional intent, which remains the lodestar of the judicial inquiry into statutory meaning. See Pasquantino, 544 U.S. at 360; Astoria Fed. Sav. & Loan Ass'n v. Solimino, 501 U.S. 104, 108 (1991); cf. Scheidler v. Nat'l Org. for Women, Inc., 547 U.S. 9, 23 (2006) (explaining that "canons of interpretation . . . are tools designed to help courts better determine what Congress intended").

Although we do not question the continuing vitality of the nonderogation canon, we are less sanguine about its applicability in the circumstances of this case. After all, the nonderogation canon does not give courts carte blanche to read a grab bag of common law rules into federal statutes simply to effectuate what those courts may perceive as good policy. See

17

Astoria, 501 U.S. at 108. It follows that a court should apply the presumption that Congress intended to retain a common law rule only if that rule was "long-established and familiar" at the time of the statute's enactment. Pasquantino, 544 U.S. at 359-60 (quoting Texas, 507 U.S. at 534). In other words, the nonderogation canon comes into play only if the terms of a statute appear to disregard a common law rule that was both long- established and familiar when Congress enacted the statute. See id.; United States v. Craft, 535 U.S. 274, 288 (2002). This requirement ensures that the assumption that Congress was aware of a particular rule and, through its silence, intended to retain it is a reasonable one. See Pasquantino, 544 U.S. at 359; Astoria, 501 U.S. at 108.

Against this backdrop, we train the lens of our inquiry on the INA. Congress enacted the provisions that authorize civil immigration arrests — 8 U.S.C. § 1226(a) and 8 U.S.C. § 1357(a)(2) — in 1952. See Immigration and Nationality Act, Pub. L. No. 82- 414, §§ 242(a), 287(a)(2), 66 Stat. 163, 208-09, 233 (1952). Although Congress revised the wording of these provisions in 1996, it did not alter the substantive authority that they conferred. See Illegal Immigration Reform and Immigrant Responsibility Act of 1996, Pub. L. No. 104-208, §§ 303(a), 308(d)(4)(L)(i), 110 Stat. 3009-546, 3009-585, 3009-618. Because the substance of these statutory provisions dates back to 1952, we must examine the state

18

of the common law as of that time to determine whether there was a long-established and familiar common law rule relating to courthouse arrests, which a court could presume Congress meant to incorporate into the INA. See Pasquantino, 544 U.S. at 360.

The plaintiffs purport to locate such a rule in the common law privilege against courthouse arrests for parties and witnesses to a civil suit. For the most part, the origins of this privilege are uncontroversial. At common law, a plaintiff in a civil action obtained personal jurisdiction over a defendant by means of a writ of capias ad respondendum. See Murphy Bros. v. Michetti Pipe Stringing, Inc., 526 U.S. 344, 350 (1999). This writ "directed the sheriff to secure the defendant's appearance by taking him into custody." Id. The common law also recognized other types of arrests in civil suits. The writ of capias ad satisfaciendum, for example, was a method of executing on a civil judgment that directed the arrest of the judgment debtor. See Magniac v. Thomson, 56 U.S. (15 How.) 281, 300 (1853).

English courts in the eighteenth and early nineteenth centuries protected parties and witnesses attending court from at least some of these forms of civil arrest. See, e.g., Ex parte Byne, (1813) 35 Eng. Rep. 123, 123, 126 (Ch.); Walpole v. Alexander, (1782) 99 Eng. Rep. 530, 531 (K.B.). This protection — which the plaintiffs refer to as a "privilege" — was designed both to remove a disincentive for inhibiting parties and witnesses

19

from coming forward (especially the risk of arrest in connection with another matter) and to ensure that arrests did not disrupt the orderly operation of the courts. See Orchard's Case, (1828) 38 Eng. Rep. 987, 987 (Ch.); Walpole, 99 Eng. Rep. at 531. To effectuate these dual purposes, the protection extended not only to residents of England but also to foreigners entering the country for the purpose of attending court. See Walpole, 99 Eng. Rep. at 531.

The protective carapace sheltered such individuals while they were physically present in the courthouse and while traveling to and from court. See Spence v. Stuart, (1802) 102 Eng. Rep. 530, 531 (K.B.); Meekins v. Smith, (1791) 126 Eng. Rep. 363, 363 (C.P.); Walpole, 99 Eng. Rep. at 531. And it applied regardless of whether the individual attended court voluntarily or under subpoena. See Meekins, 126 Eng. Rep. at 363. As William Blackstone — "whose works constituted the preeminent authority on English law for the founding generation," Alden v. Maine, 527 U.S. 706, 715 (1999) — summarized in his Commentaries on the Laws of England,

Suitors, witnesses, and other persons, necessarily attending any courts of record upon business, are not to be arrested during their actual attendance, which includes their necessary coming and returning. And no arrest can be made in the king's presence, nor within the verge of his royal palace, nor in any place where the king's justices are actually sitting.

20

3 William Blackstone, Commentaries *289 (emphasis in original).

This privilege survived an ocean crossing. "[A]rrests in civil suits were still common in America" at the time of the founding, Long v. Ansell, 293 U.S. 76, 83 (1934), and there is no legitimate doubt that courts in the newly independent United States engrafted this privilege against courthouse arrests onto American common law. In 1804, for instance, Justice Washington, riding circuit in Pennsylvania, invoked the privilege to authorize the release of a New York resident arrested on a writ of capias ad satisfaciendum while in the state to testify at trial. See Hurst's Case, 4 U.S. (4 Dall.) 387, 387-89, 12 F. Cas. 1019, 1019-20 (C.C.D. Pa. 1804). State courts in the early nineteenth century widely agreed that parties and witnesses attending court were not subject to arrest in civil suits. See, e.g., Norris v. Beach, 2 Johns. 294, 294 (N.Y. Sup. Ct. 1807) (per curiam); Fletcher v. Baxter, 2 Aik. 224, 228-29 (Vt. 1827); Richards v. Goodson, 4 Va. (2 Va. Cas.) 381, 381-82 (Gen. Ct. 1823); cf. In re M'Neil, 3 Mass. (2 Tyng) 288, 288 (1807) (recognizing privilege but not specifying nature of arrest). Questions concerning the privilege often arose in cases involving parties and witnesses crossing state lines to attend court, but the scope of the privilege was not expressly limited to nonresidents. See, e.g., Hurst's Case, 4 U.S. (4 Dall.) at 387-89, 12 F. Cas. at 1019-20; Norris, 2 Johns. at 294; cf.

21

Thompson's Case, 122 Mass. 428, 429 (1877) (explaining that privilege against arrest applied to parties and witnesses "whether they are residents of this state or come from abroad").

The practice of arresting parties as a means of securing personal jurisdiction in civil suits appears to have persisted to some degree into the late nineteenth and early twentieth centuries, and courts in such cases continued to recognize the vitality of the common law privilege against courthouse arrests. See, e.g., Larned v. Griffin, 12 F. 590, 590 (C.C.D. Mass. 1882); Dickinson v. Farwell, 51 A. 624, 625 (N.H. 1902); Ellis v. De Garmo, 24 A. 579, 579-80 (R.I. 1892); see also Monroe v. St. Clair Cir. Judge, 84 N.W. 305, 306 (Mich. 1900) (confirming existence of privilege but declining to discharge arrestee based on circumstances of arrest). Over time, though, personal service of a summons generally supplanted the writ of capias ad respondendum as the method for securing personal jurisdiction over a defendant in a civil action, see Murphy Bros., 526 U.S. at 350, and arrests in civil suits fell largely out of fashion.

As this shift took place, some courts determined, early on, that the privilege against arrest pursuant to a writ of capias ad respondendum should not extend to service of a summons. See Christian v. Williams, 35 Mo. App. 297, 303 (Ct. App. 1889) (collecting cases); see also Blight v. Fisher, 3 F. Cas. 704, 704- 05 (C.C.D.N.J. 1809) (holding that privilege of parties and

22

witnesses "extend[ed] only to an exemption from arrest," not service of summons). But this view did not prevail: recognizing that the threat of service of a summons still risked chilling court attendance, the majority of courts eventually ruled that a similar privilege against service of a summons should extend to at least some parties and witnesses. See Stewart v. Ramsay, 242 U.S. 128, 129-31 (1916) (collecting cases). Courts usually framed this privilege as protecting nonresidents who had to enter a jurisdiction to attend court proceedings. See, e.g., Diamond v. Earle, 105 N.E. 363, 363 (Mass. 1914); Richardson v. Smith, 65 A. 162, 163 (N.J. 1906); Parker v. Marco, 32 N.E. 989, 989 (N.Y. 1893). The Supreme Court adopted such a framing of the privilege as a matter of federal common law in 1916, holding that "suitors, as well as witnesses, coming from another State or jurisdiction, are exempt from the service of civil process while in attendance upon court, and during a reasonable time in coming and going." Stewart, 242 U.S. at 129; see Page Co. v. Macdonald, 261 U.S. 446, 448 (1923). Nevertheless, the Court restated the rule twice in the early 1930s without explicitly limiting its scope to nonresidents. See Long, 293 U.S. at 83; Lamb v. Schmitt, 285 U.S. 222, 225 (1932).

Relying on this history, the defendants argue that the common law privilege against courthouse arrests in civil suits evolved well before 1952 into a privilege against personal service

23

of a summons — a privilege that they say was available only to nonresidents entering a jurisdiction for the purpose of attending court. We should not presume that Congress intended to incorporate the common law privilege against courthouse arrests into the INA, their thesis runs, because the privilege (at least insofar as it applied to arrests) was a dead letter when Congress passed the statute. The plaintiffs rejoin that the privilege against courthouse arrests expanded to safeguard against service of a summons but did not die off in its original form.

The short of it is that the parties draw radically different conclusions from essentially the same nucleus of historical facts. Although the defendants' argument about the evolution of the common law privilege seems plausible, we need not make a definitive ruling in this regard because there is a clearer path leading to the conclusion that the plaintiffs have failed to demonstrate a likelihood of success with respect to their nonderogation theory.5 Even if (as the plaintiffs posit) the privilege against courthouse arrests retained some vitality in 1952, the plaintiffs' reliance on the nonderogation canon is

5 For the same reason, we have no need to explore the defendants' related argument that the privilege as it applied to service of process narrowed significantly after the Supreme Court held in International Shoe Co. v. Washington, 326 U.S. 310, 316 (1945), that due process permits the exercise of personal jurisdiction over certain parties who reside outside of the territorial boundaries of the state.

24

misplaced. They have not demonstrated that there was a long- established and familiar common law rule protecting against civil arrests on behalf of the sovereign. Put another way, we cannot presume that Congress intended to incorporate the privilege into the INA because the plaintiffs have not shown that it was clear in 1952 that the privilege had any application to the type of arrest authorized by the INA. We explain briefly.

To recapitulate, the nonderogation canon only informs the construction of a federal statute when a relevant common law rule was long-established and familiar at the time of the statute's enactment. See Pasquantino, 544 U.S. at 359-60; Texas, 507 U.S. at 534. The contours of the relevant common law rule are of decretory significance. See Pasquantino, 544 U.S. at 362. Those contours must be delineated clearly and precisely. See id. at 360, 362.

Pasquantino illustrates the point. There, the Supreme Court addressed whether a scheme to defraud a foreign nation of tax revenue came within the ambit of the federal wire fraud statute. See id. at 352-53. The petitioners contended that the Court should refrain from construing the statute in this manner in order to avoid derogating from the common law revenue rule, which prohibited the collection of a foreign tax liability. See id. at 359, 361. The Court rejected this contention, holding that "[t]he wire fraud statute derogates from no well-established revenue rule

25

principle." Id. at 360. In this wise, the Court explained that no case decided before the statute's enactment "held or clearly implied that the revenue rule barred the United States from prosecuting a fraudulent scheme to evade foreign taxes" and that its purposes did not "plainly suggest that it swept so broadly." Id. Along the way, the Court declined the plaintiffs' invitation to analogize the type of prosecution under consideration to factually distinct circumstances in which the revenue rule had been applied, such as civil suits indirectly seeking to enforce a foreign tax obligation. See id. at 362-68.

Pasquantino teaches that the party seeking to read a common law rule into the statute can only do so if the factual circumstances are closely analogous to a case of which Congress would have been aware. See id. at 364-65 (explaining that absence of such a case prevented Court from "say[ing] with any reasonable certainty whether Congress in 1952 would have considered this prosecution within the revenue rule"). In other words, the proponent of the rule cannot rely exclusively on attenuated analogies and speculative inferences to show that the rule was long-established and familiar at the time of the statute's enactment. See id. While the nonderogation canon allows certain commonsense and obvious comparisons, see Bank of Am. Corp. v. City of Miami, 137 S. Ct. 1296, 1305 (2017) (reading proximate cause requirement into cause of action for damages under Fair Housing

26

Act, which is "akin" to tort claim), we only read a common law rule into the statute if it generally qualifies as long-established and familiar and if it is widely regarded as applicable to the specific subject matter covered by the statute.

In the case at hand, the plaintiffs have failed to demonstrate that they are likely to succeed in showing that the common law privilege against courthouse arrests clearly applied to civil immigration arrests. First and foremost, the plaintiffs have not offered any pre-1952 case law from an American court that directly addresses the applicability of the privilege either to civil immigration arrests or to fairly comparable forms of civil arrest, that is, civil arrests on behalf of a sovereign. Nor (to the extent that they might be probative of how Congress would have viewed the common law in 1952) do we find direct guidance on this score in the English sources proffered by the parties. None expressly states whether the privilege protected against arrests in some or all crown-initiated civil suits. The same lack of clarity is characteristic of the literature: the plaintiffs do not identify a single treatise or article directly stating that the common law privilege extended to civil arrests on behalf of the sovereign.

What skimpy authority there is tends to suggest that crown-initiated suits may well have been exempted from the operation of the privileges against arrest in court under English

27

common law. For example, cases and treatises indicate that certain litigation-related privileges protecting attorneys on account of their necessary attendance in court did not apply in the context of such suits. See, e.g., Wheely v. Richam, (1734) 92 Eng. Rep. 882, 882 (K.B.); 1 William Tidd, The Practice of the Court of King's Bench in Personal Actions 264, 268 (William P. Farrand 1807) (1804). At the end of the day, these authorities may strengthen the defendants' hand — the two sets of privileges are intimately related, as both of them developed in order to ensure that necessary persons attended court proceedings, see 6 Matthew Bacon, A New Abridgement of the Law 530 (7th ed. 1832) (explaining that "[t]he law not only allows privileges to the officers of the court, but also protects all those whose attendance is necessary in courts," including parties to a civil suit) — but they are not dispositive because they concern the privilege enjoyed by attorneys, not the privilege at issue here.

For their part, the plaintiffs — like Rumpelstiltskin —

try to convert dross into gold. They strive to persuade us to treat the fact that various English treatises did not mention an exception for crown-initiated civil suits to the privilege for parties and witnesses, see, e.g., Tidd, supra, at 174-75, as conclusive evidence that no such exception existed. We are not convinced: the absence of any mention of arrests on behalf of the sovereign, when coupled with the utter dearth of any case law

28

holding such arrests to be within the ambit of the privilege, undercuts the claim that such an application of the privilege should be regarded as long-established and familiar.6 Where does this leave us? The absence of clear precedent involving fairly comparable forms of civil arrest throws considerable shade on the plaintiffs' effort to show that, in 1952, there was a long-established and familiar common law rule privileging individuals attending court on official business from civil immigration arrests, but it does not drive a final nail into the coffin of their nonderogation theory. It remains a possibility that we might be able to presume that Congress meant to incorporate such a common law privilege into the INA if the case law as of 1952 "clearly implied" that the privilege would extend to such arrests. Pasquantino, 544 U.S. at 360. On further perscrutation, though, this possibility dissipates: our canvass of the case law discloses nothing resembling a clear implication to this effect.

The centerpiece of the plaintiffs' forecast that the common law privilege would have applied to civil immigration

6 For similar reasons, we do not find dispositive that Blackstone's treatise described the common law privilege in a chapter on initiating process for suits involving "private wrongs," a category that apparently included certain "injuries proceeding from, or affecting, the crown." 3 William Blackstone, Commentaries *254, *289. Notably, Blackstone's description of the attorney privileges appears on the same page, but he did not mention the well-settled exception for crown-initiated civil suits applicable to those privileges. See id. at *289.

29

arrests is their claim that, historically, the privilege protected against any and all forms of civil arrest. This claim, if supportable, would take them a long way toward their goal, given that immigration arrests are undeniably civil in nature. See, e.g., United States v. Encarnacion, 239 F.3d 395, 400 (1st Cir. 2001) (referring to "civil deportation arrests and detentions under 8 U.S.C. § 1357(a)(2)"). Such arrests aim to facilitate the removal of noncitizens from the country, see United States v. Vasquez-Benitez, 919 F.3d 546, 553 (D.C. Cir. 2019); cf. Demore v. Kim, 538 U.S. 510, 528 (2003) (explaining that detention during removal proceedings of noncitizens who have certain criminal convictions "increas[es] the chance that, if ordered removed, [they] will be successfully removed"), which is a civil — not criminal — sanction, see Padilla v. Kentucky, 559 U.S. 356, 365 (2010); Harisiades v. Shaughnessy, 342 U.S. 580, 594 (1952). So, too, courts discussing the scope of the common law privilege sometimes used language suggesting that the privilege had broad application to many types of civil arrests. See, e.g., Underwood v. Fosha, 85 P. 564, 565 (Kan. 1906) (describing privilege for parties and witnesses "from civil arrest"); Fisher v. Bouchelle, 61 S.E.2d 305, 306 (W. Va. 1950) (referring to "long existing rule . . . that courts will not permit their proceedings to be disturbed by the arrest in a civil case" of parties and witnesses).

30

Digging deeper, though, discloses that the plaintiffs'

claim is insupportable. The authorities that the plaintiffs muster in their attempt to demonstrate a well-established and familiar common law rule mostly involved civil arrests in private litigation, and the exceptions, though silent on the nature of the underlying lawsuits, offer no reason to think that they arose out of actions brought on behalf of a sovereign. The plaintiffs track the origins of the privilege to cases involving arrests on common law writs in suits between private parties. See, e.g., Norris, 2 Johns. at 294; Fletcher, 2 Aik. at 224-25, 228-29; Richards, 4 Va. (2 Va. Cas.) at 381-82; Walpole, 99 Eng. Rep. at 530-31; Ex parte Byne, 35 Eng. Rep. at 124. Similarly, cases that recognized the continuing vitality of the privilege in the late nineteenth and early twentieth centuries almost always involved the arrest of a party to a private civil suit. See, e.g., Larned, 12 F. at 590; Monroe, 84 N.W. at 306; Dickinson, 51 A. at 625; Ellis, 24 A. at 579-80. As far we can tell, none involved an arrest on behalf of a sovereign.

Because the entirety of the pre-1952 case law pertaining to the common law privilege appears to have involved private civil suits, we think that any language in the case law suggesting a broader rule that the privilege applied to all forms of civil arrest can best be read as shorthand for a statement that the privilege applied to a wide swath of arrests in private civil

31

suits. Had Congress inspected the case law in 1952 with any degree of care, we are confident that it would have concluded — as we do — that there was no clear historical precedent for extending the privilege to arrests on behalf of the sovereign. Thus, we reject the plaintiffs' assertion that Congress would have reflexively inferred that the privilege protected against any and all forms of civil arrest, including the civil immigration arrests that it was authorizing in the INA.

Of course, the fact that it was well-established and widely understood in 1952 only that the privilege applied to arrests in private civil suits does not necessarily doom the plaintiffs' proposed application of the nonderogation theory. After all, Congress might have had no reason to believe that the common law would have treated arrests on behalf of the sovereign as a breed apart. Cf. Pasquantino, 544 U.S. at 364 (rejecting nonderogation argument when factual differences between common law cases and application of statute were "significant"). Here, however, this possibility is more theoretical than real: it is luminously clear to us (and it would have been luminously clear to Congress in 1952) that civil immigration arrests differ from arrests in private civil suits in a key respect. Civil immigration arrests are initiated by the sovereign in order to vindicate uniquely sovereign interests rather than private or proprietary interests. Controlling immigration and the presence of

32

noncitizens within the country are duties and powers vested exclusively in the sovereign. See Dep't of Homeland Sec. v. Thuraissigiam, 140 S. Ct. 1959, 1982 (2020) ("[T]he power to admit or exclude aliens is a sovereign prerogative." (quoting Landon v. Plasencia, 459 U.S. 21, 32 (1982))); Galvan v. Press, 347 U.S. 522, 531 (1954) ("Policies pertaining to the entry of aliens and their right to remain here are peculiarly concerned with the political conduct of government."). It is too abecedarian a proposition to warrant citation of authority that a private party cannot initiate proceedings to remove a noncitizen from the country. Nor does removal remotely resemble any type of civil remedy available to private litigants.

The fact that civil immigration arrests are initiated by the sovereign to vindicate uniquely sovereign interests is crucial to our analysis. This fact affords a powerful reason to believe that courts would have treated such arrests more like criminal arrests than like the types of civil arrest at issue in the cases to which plaintiffs advert. And the plaintiffs — whose skillful lawyers vigorously contest every arguable point — do not dispute that the privilege has never been thought to protect against criminal arrests or other forms of criminal process. See, e.g., Ex parte Levi, 28 F. 651, 652-53 (W.D.S.C. 1886); Cooper v. United States, 48 A.2d 771, 773 (D.C. 1946); State v. Gillmore, 129 P. 1123, 1125 (Kan. 1913); Schwartz v. Dutro, 298 S.W. 769, 771 (Mo.

33

1927); see also 2 Joseph Story, Commentaries on the Constitution of the United States 325 (Hilliard, Gray, and Co. 1833) (explaining that constitutional privilege of members of Congress against arrest, which excludes arrests for crimes, mirrors common law privilege against courthouse arrests for parties and witnesses). This is no mere happenstance: although criminal arrests in courthouses risk deterring parties and witnesses from coming forward and also risk disrupting ongoing proceedings, courts have refrained from extending the privilege to criminal arrests due to the overriding sovereign interests in enforcing the penal laws and protecting the public. See United States v. Conley, 80 F. Supp. 700, 702-03 (D. Mass. 1948); cf. Bacon, supra, at 532-33 (explaining that related English common law privilege for attorneys did not apply to "indictments, informations, or suits, in which the king alone is concerned" because courts should not protect "those who offend against the public peace of the community and the king's interest").

We add, moreover, that the analogy between criminal arrests and civil immigration arrests is close enough to preclude us from saying with sufficient confidence that immigration arrests would have fit within the privilege from civil arrest. Just as criminal arrests implicate the uniquely sovereign interests in enforcing the penal laws and protecting the public, so too do civil immigration arrests seek to vindicate similar kinds of interests

34

in controlling immigration and the presence of noncitizens in the country. And just as the common law privilege was not applied to criminal arrests because of these overriding sovereign interests, one would think (for the same reason) that the privilege would not shield civil immigration arrests. Especially in light of this plausible argument for treating civil immigration arrests like criminal arrests vis-à-vis the privilege, there is no principled way to find that the case law in 1952 clearly implied that the privilege would have afforded a shield against civil immigration arrests.

The plaintiffs' remaining arguments regarding the privilege are unavailing. They emphasize, for example, that the dual purposes undergirding the privilege in the context of arrests in private civil suits apply in much the same way to civil immigration arrests. This argument has a patina of plausibility: ICE's policy of conducting civil courthouse arrests may inhibit parties and witnesses from attending court proceedings and, in the bargain, it may in certain circumstances disrupt orderly court operations. But even though such considerations may have held dispositive weight when courts determined whether to privilege parties and witnesses from arrests in private civil suits, that reasoning carries much less weight with respect to civil immigration arrests. As we already have made pellucid, civil immigration arrests implicate uniquely sovereign interests.

35

Consequently, the purposes underlying the common law privilege comprise too frail a foundation for an assumption that the privilege would have applied to civil immigration arrests.

We find equally unconvincing the plaintiffs' attempt to analogize civil immigration arrests and arrests on a writ of capias ad respondendum. It is true that, viewed from ten thousand feet, these types of arrest bear a faint resemblance: each involves a government officer taking someone into custody to ensure his appearance at a civil proceeding. Apart from the "civil" label, though, removal proceedings have little in common with a typical private lawsuit. No less an authority than the Supreme Court has referred to removal as a "unique" civil penalty in light of its particularly harsh consequences and its close connection to the criminal process. Padilla, 559 U.S. at 365-66. To assume that a privilege that protected against civil arrests pursuant to a writ of capias ad respondendum would translate to immigration arrests because of a few superficial similarities would be to accept exactly the type of attenuated analogy that the Pasquantino Court deemed insufficient to warrant the application of the nonderogation canon. Simply put, the two types of arrest are not fair congeners.

We summarize succinctly. The case law is wholly devoid of any clear precedent on whether the common law in 1952 would have applied the privilege against courthouse arrests to civil

36

immigration arrests. The literature is equally nebulous. But even in this shadowy corner of the law, some things are manifest. Although the privilege protected against arrests in private civil suits, it did not apply to criminal arrests — and the fact that civil immigration arrests aim to vindicate uniquely sovereign interests supplies a strong reason to think that the common law would have treated them like criminal arrests for purposes of this privilege. The unique nature of removal further undermines any analogy between civil immigration arrests and arrests in private civil suits. Given these disparities, we cannot say with a reasonable degree of assurance that, in 1952, Congress would have considered the civil immigration arrests authorized in the INA to come within the scope of the common law privilege. See Pasquantino, 544 U.S. at 364-65. We conclude, therefore, that the plaintiffs are unlikely to succeed in demonstrating a long- established and familiar common law rule barring courthouse arrests that can be presumed to have been incorporated into the INA's civil arrest authority.

B.

This conclusion does not end our odyssey. The plaintiffs have a fallback position: they offer a different reason why, in Massachusetts, ICE's implementation of the Directive and its policy of conducting civil courthouse arrests exceed its statutory authority. Their rationale is that even if Congress has not

37

incorporated the common law privilege against courthouse arrests into the INA's civil arrest authority, the statute still should not be construed to allow ICE officers to conduct civil courthouse arrests that transgress state law. In support, they assert that the privilege against courthouse arrests for persons attending court on official business is firmly established in Massachusetts common law; that this privilege is emblematic of an exercise of the Commonwealth's sovereign power to operate its judiciary; and that the INA does not contain a clear statement of Congress's intent to permit civil immigration arrests that violate such a core state-law privilege.

Before tackling this argument, we pause to examine the framework that guides our inquiry. The plaintiffs dress their argument in the raiment of preemption, maintaining that the text of the INA fails to rebut the presumption that Congress would not have intended to preempt the Massachusetts common law privilege against courthouse arrests. Federal preemption, though, is typically a defense to an alleged violation of state law, see Lawless v. Steward Health Care Sys., LLC, 894 F.3d 9, 17 (1st Cir. 2018), and the plaintiffs have sought preliminary injunctive relief only on their APA claim, not on the ground that ICE's implementation of the Directive and its policy of conducting civil courthouse arrests violate Massachusetts law. Refined to bare essence, their argument suggests that we should construe the INA

38

in a way that would not authorize civil arrests in a manner that interferes with a state's chosen method of maintaining order and liberty within the confines of the state's courthouses. This suggestion fits more neatly under the federalism canon of statutory construction described in Gregory v. Ashcroft, 501 U.S. 452 (1991). Mindful that the relevant portion of the plaintiffs' brief relies heavily on Gregory, we examine their argument through that lens.7 The federalism canon of statutory construction flows from the elementary principle that "the States retain substantial sovereign powers under our [federal] constitutional scheme." Id. at 461. When Congress acts within the bounds of its enumerated powers, the Supremacy Clause permits it to "impose its will on the States" and to "legislate in areas traditionally regulated by the States." Id. at 460. Given the "extraordinary" nature of this power, we assume that "Congress does not exercise [it] lightly." Id. It follows that before construing a federal statute in a way that "would upset the usual constitutional balance of federal and

7It is worth noting that the federalism canon of statutory construction and the presumption against preemption of state law are intimately related doctrines. Both reflect a reluctance to ascribe to Congress an intent to interfere with a state's exercise of its sovereign powers, see Raygor v. Regents of the Univ. of Minn., 534 U.S. 533, 543-44 (2002), and Congress must clearly state such an intent if the presumption created by either doctrine is to be rebutted, compare Wyeth v. Levine, 555 U.S. 555, 565 (2009) (preemption), with Gregory, 501 U.S. at 460-61 (federalism canon). Consequently, our analysis would proceed in much the same manner were we to examine the plaintiffs' argument under the rubric of preemption.

39

state powers," courts must search for a clear statement indicating that such a result represents Congress's intent. Id. at 460-61; see Bond v. United States, 572 U.S. 844, 858 (2014).

The Supreme Court first described this canon of statutory construction in Gregory, which addressed whether Missouri's constitutional mandate that state judges must retire at age seventy violated the federal Age Discrimination in Employment Act of 1967 (ADEA). See 501 U.S. at 455. The question before the Court boiled down to whether a state judge qualifies as "an appointee on the policymaking level" — a category of persons excluded from the ADEA's sweep. Id. at 464-65, 467 (quoting 29 U.S.C. § 630(f)). The Court recognized that a state's choice of qualifications for its judges "is a decision of the most fundamental sort for a sovereign entity." Id. at 460. Because "[c]ongressional interference" with this type of state decisionmaking "would upset the usual constitutional balance of federal and state powers," the Court asked whether the ADEA contained a clear statement that state judges were included within the statute's scope. Id. at 460-61, 467. Answering this question, the Court concluded that it was "at least ambiguous whether a state judge is an 'appointee on the policymaking level'" and, therefore, found that the ADEA did not contain the requisite clear statement indicating that Congress intended to cover state judges. Id. at 467.

40

Invoking the reasoning of Gregory, the plaintiffs contend that the INA's generic civil arrest authority does not authorize ICE to conduct civil courthouse arrests in Massachusetts because the statute does not contain a clear statement of Congress's intent to interfere with Massachusetts's sovereign decision to protect individuals attending court on official business from civil arrests. The district court did not address this contention because it found that the plaintiffs had shown a likelihood of success on the merits of their nonderogation theory. See Ryan, 382 F. Supp. 3d at 155-59. This gap in the record presents a daunting obstacle. While we generally may affirm a district court's decree on any ground made manifest by the record, see Kando v. R.I. State Bd. of Elections, 880 F.3d 53, 58 (1st Cir. 2018); Mason v. Telefunken Semiconductors Am., LLC, 797 F.3d 33, 37-38 (1st Cir. 2015), this tenet does not permit us to affirm on grounds that are premised on factual determinations that the district court did not make and that the parties dispute. After all, except in rare circumstances (not present here), an appellate court may not engage in its own factfinding where the record contains evidence on a particular point that could lead reasonable factfinders to competing conclusions. See Candelario-Del-Moral v. UBS Fin. Servs. Inc. of P.R. (In re Efron), 746 F.3d 30, 38 (1st Cir. 2014); Dedham Water Co. v. Cumberland Farms Dairy, Inc., 972 F.2d 453, 463 (1st Cir. 1992).

41

This is such a case. The absence of pertinent factfinding by the district court, coupled with the conflicting data points in the record on Massachusetts's policy on courthouse arrests, leads inexorably to a conclusion that we cannot affirm the entry of the preliminary injunction based on the plaintiffs' clear-statement argument.

To be sure, the plaintiffs' argument rests on an uncontroversial premise: the operation of a functioning judiciary is unmistakably a fundamental exercise of state sovereignty. See Atl. Coast Line R.R. Co. v. Brotherhood of Locomotive Eng'rs, 398 U.S. 281, 285, 287 (1970) (explaining that states reserved power to maintain "judicial systems for the decision of legal controversies" and referring to "the fundamental constitutional independence of the States and their courts"). Inasmuch as congressional interference with a state's ability to manage the core functions of its judiciary would tilt the constitutional balance between state and federal power, courts must be certain of Congress's intent before interpreting a federal statute to authorize such interference. Against this backdrop, we assume — solely for purposes of this appeal — that a decision by Massachusetts to prohibit at least some courthouse arrests would represent an exercise of the Commonwealth's sovereign power to operate its judiciary, one with which Congress would not readily interfere.

42

Here, however, the train goes off the tracks when we move from the theoretical state of affairs to the actual state of affairs. Because the plaintiffs' argument is premised on the notion that construing the INA to authorize civil courthouse arrests would clash with a sovereign state decision, we must know the scope of Massachusetts's policy on courthouse arrests in order to evaluate the argument. Without such knowledge, we cannot determine whether and to what extent the INA's civil arrest authority and ICE's actions pursuant to that authority may (or may not) interfere with Massachusetts's exercise of its sovereign power. And because the parties seeking a preliminary injunction bear the burden of demonstrating a likelihood of success on the merits of their claim, see New Comm Wireless Servs., 287 F.3d at 9, the plaintiffs bear the burden of establishing the contours of the relevant Massachusetts policy (which is an essential factual predicate to the success of their argument). They have so far failed to carry this burden.

The plaintiffs demur, contending that the Commonwealth's policy is embodied exclusively in the relevant Massachusetts case law on the common law privilege against courthouse arrests. To buttress this contention, they point out that throughout the nineteenth century, the SJC recognized that parties and witnesses were privileged from civil arrests while attending court proceedings. See Thompson's Case, 122 Mass. at 429; May v.

43

Shumway, 82 Mass. (16 Gray) 86, 86-87 (1860) (per curiam); Wood v. Neale, 71 Mass. (5 Gray) 538, 538 (1855); In re M'Neil, 3 Mass. (2 Tyng) at 288. They note, as well, that the SJC referred to this rule again in both 1914 and 1968 (albeit in cases involving the related privilege protecting against service of a summons). See Valley Bank & Tr. Co. v. Marrewa, 237 N.E.2d 677, 680 (Mass. 1968); Diamond, 105 N.E. at 363. Although the SJC has never addressed whether the privilege protects against civil immigration arrests, Justice Cypher, in a single justice opinion, recently described her view "that there exists a common law privilege against civil arrest in Massachusetts and that the privilege, as a matter of State law, is broad enough to include arrests by Federal officers." Matter of C. Doe, No. SJ-2018-119, slip op. at 12 (Mass. Sept. 18, 2018). The plaintiffs urge us to take Justice Cypher's analysis to mean that all civil immigration arrests of noncitizens attending court on official business violate Massachusetts law and policy.

But this case law on the common law privilege is far from the only data point in the record regarding Massachusetts's policy on courthouse arrests. In November of 2017, the Chief Justice of the Massachusetts Trial Court promulgated a set of rules governing how "staff shall respond when [ICE officers] enter a Massachusetts courthouse with the intent of taking custody of an individual." These rules specify that ICE officers may not civilly arrest noncitizens in nonpublic areas of state courthouses or

44

(absent advance permission) in courtrooms. Withal, they recognize that ICE officers may otherwise conduct civil immigration arrests in courthouses as long as they follow certain procedures and do not disrupt court operations. This arrangement, as the defendants emphasize, is seemingly at odds with the plaintiffs' interpretation of the Massachusetts common law privilege; it indicates that Massachusetts courts do not object to all civil immigration arrests of noncitizens attending court on official business.

The bottom line is that we are confronted with an unsettled record concerning Massachusetts's policy on courthouse arrests. Since the district court grounded the preliminary injunction solely on its determination that the plaintiffs were likely to succeed on their nonderogation theory — a determination that turned on its mistaken assessment of long-established common law throughout the country rather than the particulars of Massachusetts law or policy, see Ryan, 382 F. Supp. 3d at 155-59 — it did not perform any factfinding that would assist in clarifying this unsettled area of the record. Nor did the court attempt to reconcile the plaintiffs' interpretation of the Massachusetts common law privilege against courthouse arrests with the Trial Court's rules, thus appearing to accept as a given ICE's authority to arrest certain noncitizens purportedly protected by the privilege.

45

This gap in the record is material in the sense that the manner in which it is filled could affect the outcome of the case. What is more, there is no avenue (short of a remand) through which we can bridge it. Undertaking the missing factfinding in the first instance not only would usurp the district court's prerogative but also would exceed the limits of our appellate role. See In re Efron, 746 F.3d at 38; Dedham Water Co., 972 F.2d at 463. Such an act of judicial hubris, problematic at any time, would be especially unwise where, as here, the record on appeal is glaringly underdeveloped.

An example serves to illustrate this point. The plaintiffs attempt to diminish the significance of the Trial Court's rules as evidence of Massachusetts's policy on courthouse arrests by arguing that a broader ban on civil immigration arrests in courthouses would be futile because ICE has made clear that it would not comply with such a ban. Relatedly, the plaintiffs argue that the Trial Court enacted these rules in order to restrict the involvement of court staff in ICE's enforcement actions rather than to cooperate with ICE. But without more information about both ICE's intent and the nature and purpose of the Trial Court's rules, we are left with a predicate that is manifestly inadequate for answering the factual questions that must be answered to resolve these arguments.

46

That ends this aspect of the matter. Given the unsettled nature of the record and the absence of pertinent factfinding by the district court, we cannot conclude that the plaintiffs have made a sufficient showing of Massachusetts's policy on courthouse arrests to carry their burden. And without such a showing, we have no baseline from which we can assess whether the INA's civil arrest authority may (or may not) interfere with the state's sovereign power to operate its judiciary.8 Here, moreover, a myriad of other factors beyond the underdeveloped record must be factored into the equation. For one thing, the parties have not fully explored the ramifications of the Gregory-based argument in their briefing.9 For another thing, the argument raises complex and novel legal issues, and we do not have the benefit of the district court's insights on those issues. To add yet another complication, it is unclear whether a policy

8 Once again, an example helps to illustrate the point. The defendants represented at oral argument that ICE has exercised its civil arrest authority in compliance with the Trial Court's rules. If so, and if those rules comport with Massachusetts's policy, we would be hard-pressed to hold that ICE's exercise of its civil arrest authority clashes with the Commonwealth's sovereign power in a way that would engender federalism concerns. Under such a scenario, it is questionable whether Gregory's clear-statement rule would be triggered at all.

9 Although there are many such ramifications, one is particularly striking. If taken to its logical end point, the plaintiffs' Gregory-based argument leads to a strange result: ICE's authority to conduct civil courthouse arrests might vary on a state-by-state basis, depending on each state's policy on the subject.

47

forbidding the arrest of all individuals attending court proceedings would constitute the type of exercise of state sovereignty that triggers Gregory's clear-statement rule. See 501 U.S. at 460 (emphasizing that state constitutional provision setting qualifications for judges "is a decision of the most fundamental sort for a sovereign entity"); cf. EEOC v. Massachusetts, 987 F.2d 64, 66, 69 (1st Cir. 1993) (finding Gregory inapplicable in deciding whether state law requiring public employees to pass medical examination at age seventy violated ADEA). It is possible that only a narrower range of courthouse arrests (say, those occurring in the courtroom itself) would fall within this taxonomy. And this is just the tip of the iceberg: taken together, the uncertainties that pervade this area of the record counsel in favor of judicial restraint. See Levinsky's, Inc. v. Wal-Mart Stores, Inc., 127 F.3d 122, 134 (1st Cir. 1997) (refusing to resolve complex legal issue for first time on appeal in light of deficiencies in the record, absence of district court opinion, and lack of full briefing from parties).

Let us be perfectly clear. We take no view on the role, if any, that Gregory and the federalism canon may play in construing the scope of the INA's civil arrest authority. Nor do we make any determination about the law or policy of Massachusetts vis-à-vis courthouse arrests. With respect to this issue, we conclude only that, without additional factfinding, the lack of

48

clarity in the record about Massachusetts's policy on courthouse arrests prevents us from determining whether or not the plaintiffs are likely to succeed on the merits of their Gregory-based argument. Following remand, the district court may evaluate this argument, either on a renewed motion for preliminary injunction or on the merits, with the help of both a better-developed record and more exegetic briefing. As part of its evaluation, the district court may consider whether, on a better-developed record, it would be appropriate to certify a question or questions to the SJC in order to pin down Massachusetts's policy on courthouse arrests. See Mass. S.J.C. Rule 1:03; see also Matter of C. Doe, No. SJ-2018-119, slip op. at 12 n.14 (Mass. Sept. 18, 2018) ("If a Federal Court were at all in doubt about the continued existence of the privilege in Massachusetts, or its applicability in given circumstances, it would of course be free to certify to this court any questions it has about the existence and applicability of the privilege."). III. CONCLUSION We need go no further. For the reasons elucidated above, we conclude that the plaintiffs thus far have failed to show a likelihood of success on the merits of their nonderogation theory, that is, that the Directive and its authorization of civil courthouse arrests by the federal government to enforce the immigration laws, exceed ICE's statutory authority because the INA

49

implicitly incorporates a common law privilege against courthouse arrests. The district court's contrary ruling was based on a material error of law and, thus, we hold that it constitutes an abuse of discretion. See Corp. Techs., 731 F.3d at 10. So, too, we hold that the plaintiffs have not to this point shown a likelihood of success on their APA claim based on the argument that, in Massachusetts, ICE's implementation of the Directive and its policy of conducting civil courthouse arrests exceed its statutory authority because Congress has not made clear its intent to permit ICE to conduct arrests in violation of state law.10 As a movant may not secure a preliminary injunction without demonstrating a likelihood of success on the merits, we have no need to consider the parties' arguments concerning either the remainder of the preliminary injunction calculus or the scope of the district court's injunction. See Wine & Spirits Retailers, Inc. v. Rhode Island, 418 F.3d 36, 54 (1st Cir. 2005).

We vacate the preliminary injunction and remand to the district court for further proceedings consistent with this opinion. All parties shall bear their own costs.

Vacated and remanded.

10 Certain amici supporting the plaintiffs asseverate that ICE's policy of arresting noncitizens in courthouses is unlawful for a variety of other reasons. We decline to address these asseverations. The customary praxis in this circuit is to eschew arguments raised only by amici and not by the parties. See, e.g., In re Sony BMG Music Ent., 564 F.3d 1, 3 (1st Cir. 2009); Lane v. First Nat'l Bank of Bos., 871 F.2d 166, 175 (1st Cir. 1989). We see no reason to depart from this praxis today.