Nadine Pellegrino v. TSANadine Pellegrino v. TSA
(Opinion filed August 30, 2019)
Paul M. Thompson (Argued)
Sarah Hogarth
McDermott Will & Emery
500 North Capitol Street, N.W.
Washington, DC 20001
Matthew L. Knowles
McDermott Will & Emery
28 State Street, Suite 3400
Boston, MA 02109
Counsel for Appellants
Mark J. Sherer (Argued)
Office of the United States Attorney
615 Chestnut Street, Suite 1250
Philadelphia, PA 19106
Jonathan H. Feinberg
David Rudovsky
Kairys Rudovsky Messing & Feinberg
718 Arch Street, Suite 501 South
Philadelphia, PA 19106
Hugh Handeyside
Hina Shamsi
American Civil Liberties Union
125 Broad Street, 18th Floor
New York, NY 10004
Molly M. Tack-Hooper
American Civil Liberties Union of Pennsylvania
P.O. Box 60173
Philadelphia, PA 19106
Counsel for Amicus Appellants
American Civil Liberties Union;
American Civil Liberties Union of Pennsylvania;
Cato Institute; Rutherford Institute.
Mahesha P. Subbaraman
222 South 9th Street, Suite 1600
Minneapolis, MN 55402
Counsel for Amicus Appellants
Freedom to Travel USA; Restore the Fourth Inc.
AMBRO, Circuit Judge, with whom Chief Judge Smith and Judges McKee, Chagares, Greenaway, Jr., Shwartz, Restrepo, Bibas, and Porter join.
The Federal Government is typically immune from suit. The Federal Tort Claims Act waives the Government‘s immunity for certain torts committed by its employees.
Nadine Pellegrino relies on
Background
A. Factual Background
Pellegrino and her husband arrived at the Philadelphia International Airport to board a flight home to Florida. This meant passing through the security checkpoint maintained by the Transportation Security Administration (TSA) with TSOs. Congress created the TSA after the terrorist attacks of September 11, 2001, with the enactment of the Aviation and Transportation Security Act, Pub. L. No. 107-71, 115 Stat. 597 (2001). Under that Act, TSOs perform screenings at TSA checkpoints in airports in the United States. See
As Pellegrino passed through the security checkpoint, she was randomly selected for additional screening. A TSO began examining her bags, but she stopped him and requested a more discreet screening. In a private room, several TSOs combed through Pellegrino‘s luggage, papers, and other effects. One allegedly counted her coins and currency, examined her cell phone data, read the front and back of her membership and credit cards, and opened and smelled her cosmetics, mints, and hand sanitizer. Per Pellegrino, the TSO also spilled the contents of several containers and was so rough with her belongings that her jewelry and eyeglasses were damaged. Frustrated, she told the TSOs that she would report their conduct to a supervisor.
The screening ended, but the TSOs’ alleged torment did not. Pellegrino was left to clean up the mess created by
As a result of the TSOs’ allegations, the Philadelphia District Attorney‘s Office charged Pellegrino with ten crimes, including aggravated assault, possession of an instrument of a crime (her luggage), and making terroristic threats. At a preliminary hearing, the presiding judge dismissed many of the charges and the District Attorney abandoned others. The remaining charges came to naught when the TSA failed to produce surveillance video from the incident, one TSO failed to appear in court, and another TSO‘s testimony was self-contradictory on key points.
B. Procedural History
After her ordeal at the airport and victory in the courtroom, Pellegrino and her husband brought numerous constitutional and statutory claims (including under the Administrative Procedure Act and the Freedom of Information Act) against the TSA and several TSOs. The District Court winnowed them down to claims for property damage, false arrest, false imprisonment, and malicious prosecution under the Tort Claims Act and implied rights of action under Bivens v. Six Unknown Named Agents, 403 U.S. 388 (1971), for malicious prosecution in violation of the First and Fourth Amendments. The claim for property damage
As for the claims under the Tort Claims Act for false arrest, false imprisonment, and malicious prosecution, the Court granted summary judgment for the defendants on the ground that TSOs are not “investigative or law enforcement officer[s]” whose intentional torts expose the United States to liability. See Pellegrino v. U.S. Transp. Sec. Admin., No. 09-cv-5505, 2014 WL 1489939, at *7 (E.D. Pa. Apr. 16, 2014). In particular, the Court stated that it was “ambiguous” whether TSOs perform the requisite “searches . . . for violations of Federal law,” id. at *5, and turned to the legislative history of the proviso at
On appeal we appointed amicus counsel to argue Pellegrino‘s side on, inter alia, the Tort Claims Act issue. A divided panel of our Court affirmed the District Court in full (including as to summary judgment on the non–Tort Claims Act claims). See Pellegrino v. U.S. Transp. Sec. Admin., 896 F.3d 207, 209 (3d Cir. 2018). We then granted rehearing en banc to consider whether TSOs are “investigative or law enforcement officer[s]” as defined in the Tort Claims Act.
Jurisdiction and Standard of Review
The District Court had jurisdiction under
As noted, the United States enjoys baseline immunity from suit. See Millbrook v. United States, 569 U.S. 50, 51–52 (2013). Congress has overridden this rule with the Tort Claims Act‘s general waiver of immunity for injuries “caused by . . . any employee of the Government.” See
But even the intentional-tort exception has its limits. Under the proviso, the exception does not apply to (and thus the United States may still be sued for) six of the eleven torts — “assault, battery, false imprisonment, false arrest, abuse of process, [and] malicious prosecution” — committed by “investigative or law enforcement officers.”
A. Text of the Proviso, 28 U.S.C. § 2680(h)
Are TSOs (1) “officer[s] of the United States” who are (2) “empowered by law” to (3) “execute searches” for (4) “violations of Federal law“? To begin, we track the text.
“Ordinarily, a word‘s usage accords with its dictionary definition.” Yates v. United States, 135 S. Ct. 1074, 1082 (2015). Under one prominent dictionary definition shortly before 1974, the year of the proviso‘s enactment, an officer “serve[s] in a position of trust” or “authority,” especially as “provided for by law.” Officer, Webster‘s Third New International Dictionary (1971); see also Officer, Black‘s Law Dictionary (4th ed. rev. 1968) (“[A]n officer is one holding a position of trust and authority . . . .“). TSOs satisfy this definition, as they are “tasked with assisting in a critical aspect of national security — securing our nation‘s airports and air traffic.” Vanderklok v. United States, 868 F.3d 189, 207 (3d Cir. 2017). To take another definition from the time, officers are “charged” by the Government “with the power and duty of exercising certain functions . . . to be exercised for the public benefit.” Officer, Black‘s Law Dictionary, supra. TSOs qualify under this definition as well, as they perform “the screening of all passengers and property,”
But are they “officer[s]” under the Tort Claims Act? A distinction between “employee[s]” and “officer[s]” appears
We think not.2 Aside from the single shared word “officer,” there is no textual indication that only a specialized “law enforcement officer” in the Aviation Security Act,
And neither Act‘s statutory distinction between “officer[s]” and “employee[s]” is airtight. Instead, both statutes include “officers” within the meaning of the term “employee.” See
Even if there were uncertainty about the reach of the term “officer of the United States,” it would be resolved in favor of a broad scope. To begin, disputes over the breadth of the Tort Claims Act “do[] not implicate the general rule that ‘a waiver of the Government‘s sovereign immunity will be strictly construed . . . in favor of the sovereign.‘” Dolan v. U.S. Postal Serv., 546 U.S. 481, 491 (2006) (quoting Lane v. Peña, 518 U.S. 187, 192 (1996)). And here the statutory reference to “any officer” — as opposed to, say, criminal officer — supports an expansive reading. See Boyle v. United States, 556 U.S. 938, 944 (2009) (“The term ‘any’ ensures that the definition has a wide reach[.]” (citation omitted)). Furthermore, as recently as 2013 the Supreme Court clamped down on a cramped reading of the proviso. See Millbrook, 569 U.S. at 56–57. As the Fifth Circuit recently put it, “[t]he [Millbrook] Court held there to be no implicit limits on the
The Supreme Court‘s expansive reading also set the tone for the Seventh Circuit‘s sweeping view of the proviso last year. See Bunch v. United States, 880 F.3d 938, 945 (7th Cir. 2018) (Wood, C.J.) (concluding that an ATF chemist could qualify under the proviso, and explaining that “[w]e are also influenced by the broad reading of the law-enforcement proviso that the Court adopted in Millbrook“). Our decision today that TSOs are “officer[s] of the United States” is consistent with the broad constructions announced in Millbrook and Bunch.
2. ” . . . Empowered by Law . . .”
To repeat, the complete proviso definition for an “investigative or law enforcement officer” is “any officer of the United States who is empowered by law to execute searches, to seize evidence, or to make arrests for violations of Federal law.”
Turning, then, to the statutory authority of TSOs, they are empowered by law to conduct “the screening of all passengers and property.”
TSO screenings are “searches” (i) as a matter of ordinary meaning, (ii) under the Fourth Amendment, and (iii) under the definition provided in Terry v. Ohio, 392 U.S. 1 (1968). Attempts to distinguish (iv) between administrative and criminal “searches” are divorced from the plain text, and any distinction, if one must be made, should account for (v) the fact that TSA searches extend to the general public and involve examinations of an individual‘s physical person and her property.
(i) Ordinary Meaning. — TSOs perform “searches” as understood in ordinary parlance. Of the many dictionary definitions that bear this out, to search is “to examine (a person) thoroughly to check on whatever articles are carried or concealed.” Search, Webster‘s Third New International Dictionary (1971); see also Search, Black‘s Law Dictionary (4th ed. rev. 1968) (“an examination or inspection . . . with [a] view to discovery of stolen, contraband, or illicit property“). Dictionaries aside, one could simply ask any passenger at any airport. Indeed, the very TSOs who screened Pellegrino called their procedure a search: “While [a TSO] was doing the searches, [Pellegrino] continued to be verbally abusive. When the search was complete, the passenger asked that she repack her own bags . . . .” J.A. 215 (incident report).
The Aviation Security Act‘s statutory and regulatory regime reflects this ordinary usage. TSOs perform “screening[s] of all passengers and property,”
(ii) Fourth Amendment. — Setting aside the ordinary meaning of “search,” airport screenings are searches as well under the Fourth Amendment. George v. Rehiel, 738 F.3d 562, 577 (3d Cir. 2013). No warrant is required, and no individualized suspicion need exist. Id.; see also Nat‘l Treasury Emps. Union v. Von Raab, 489 U.S. 656, 675 n.3 (1989) (noting “the Federal Government‘s practice of requiring the search of all passengers seeking to board commercial airliners, as well as the search of their carry-on luggage, without any basis for suspecting any particular passenger of an untoward motive“) (emphases added).
The Government does not dispute that holding. Instead, it contends that consent by passengers cancels the Fourth Amendment‘s effect. But the presence or absence of consent does not determine whether a search has occurred for purposes of the Fourth Amendment. See George, 738 F.3d at 575 (“The constitutionality of an airport screening search . . . does not depend on consent . . . . [A]ll that is required is the passenger‘s election to attempt entry into the secured area.” (internal quotation marks omitted) (quoting United States v. Aukai, 497 F.3d 955, 961 (9th Cir. 2007) (en banc))). In any event, TSO screenings are not consensual. As noted, per TSA regulations any individual who does not consent to a “search or inspection” may not board a flight.
Terry provided a vivid definition of “search“: “[I]t is nothing less than sheer torture of the English language to suggest that a careful exploration of the outer surfaces of a person‘s clothing all over his or her body in an attempt to find weapons is not a ‘search.‘” Terry, 392 U.S. at 16 (emphasis added). This is an apt description of the duties of a TSO, who by statute may “thoroughly conduct” an exploration “over an individual‘s entire body.”
To be sure, Terry typically requires reasonable suspicion for a search. But that is not the point here. Under Terry, the existence of reasonable suspicion determines whether a search was justified, not whether it occurred in the first place. An inspection may meet the definition of “search” under Terry yet involve no reasonable suspicion. In that situation, the incident is still a search; it is simply an unlawful search.
The only textual support for this distinction comes from the interpretive canon noscitur a sociis (to know something by its accompanying words). At three points in the proviso — “execute searches,” “seize evidence,” and “make arrests” — neighboring words arguably carry criminal connotations that possibly color the meaning of “searches.” In reverse order, “mak[ing] arrests” to curb federal law violations no doubt has a criminal color, and TSOs (unless specially designated, see
But Congress chose not to include the terms “warrant” or “search warrant” in
Moreover, we are doubly slow to apply the noscitur canon here. Not only is the term “searches” clear, see Russell Motor Car Co. v. United States, 261 U.S. 514, 520 (1923), but the three duties in the proviso are listed in the disjunctive (“to execute searches, to seize evidence, or to make arrests“). “When Congress has separated terms with the conjunction ‘or,‘” the canon often “is of little help.” In re Continental Airlines, Inc., 932 F.2d 282, 288 (3d Cir. 1991) (Scirica, J.) (citations omitted). Each of the three duties independently suffices to define “investigative or law enforcement officer.” See Bunch, 880 F.3d at 943. As even the counsel for the Government stated at oral argument, satisfying the proviso “would depend . . . on the individual statutory authority” measured against the three listed duties. Tr. of En Banc Oral Arg. at 35:10–11. We agree; the three statutory duties in the proviso begin and end the inquiry. No resort to amorphous criminal connotations is warranted.
No surprise, then, that every decision on the scope of the proviso tests whether any single duty is statutorily present. Some federal officers qualify because they perform “searches.” See Bunch, 880 F.3d at 943 (ATF chemists); cf. Caban, 671 F.2d at 1234 n.4 (immigration agents). Others make arrests, and therefore qualify even if they don‘t play a traditional law enforcement role. See Campos, 888 F.3d at 737 (Customs and Border Protection officers); Nurse v.
Indeed, we could apply the same analysis to both TSOs and TSA “law enforcement officers” per
Another conceivable way the noscitur canon might arrive at a distinction between criminal and administrative searches is by parsing the particular intentional torts against which the proviso waives immunity: assault, battery, false imprisonment, false arrest, abuse of process, or malicious prosecution. To be sure, these torts are commonly claimed against criminal law enforcement officers performing criminal law functions. But as our case demonstrates, that these torts are typically brought against criminal law
Nor does a lack of training on the constitutional doctrines underpinning these torts absolve TSOs of liability. Congress knows how to define “law enforcement officers” by reference to training. See, e.g.,
At bottom, Congress chose to re-waive sovereign immunity only for certain torts to cabin the Government‘s liability, not to provide an indirect textual clue about the meaning of “investigative or law enforcement officer.” See Sami v. United States, 617 F.2d 755, 764–65 (D.C. Cir. 1979) (“[B]y limiting the wrongs covered in the § 2680(h) exception . . . , Congress set finite boundaries around the kind of law enforcement abuses for which it wished to make the government liable.“).
(v) Distinctions From Typical Administrative Searches. — If we must draw distinctions between “searches” in the proviso, the possible distinction between criminal and administrative searches is incomplete. A further distinction within administrative searches may be needed — one that accounts for the physically intrusive and ubiquitous nature of TSA searches.
The intimate physical nature of TSA searches also harmonizes our decision today with Matsko v. United States, 372 F.3d 556 (3d Cir. 2004), in which we held that an inspector of the Mine Safety and Health Administration, who had the “authority to inspect mines and investigate possible violations,” was not covered by the proviso. Id. at 560 (citation omitted). Our Court also asserted in a dictum that “employees of administrative agencies, no matter what investigative conduct they are involved in, do not come within the
Next, the risk of abuse is greater for TSO screenings than for most other administrative searches. Because TSA searches affect the public directly, the potential for widespread harm is elevated. This potential for abuse in borne out by Pellegrino’s own experience. There is a reason
In sum, we hold only that TSO screenings are “searches” under the proviso because they are more personal than traditional administrative inspections — they extend to the general public and involve examinations, often intrusive, of an individual’s physical person along with her property.
Under the proviso, investigative or law enforcement officers must be authorized “to execute searches, to seize evidence, or to make arrests for violations of Federal law.”
But we need not decide whether that rule applies here, as TSOs do execute searches “for violations of Federal law.” See, e.g.,
Finally, that airport contraband is legal in some non-flight contexts does not change this conclusion. As long as TSOs screen for items federal law bars on airplanes, they are searching for “violations of Federal law.”
B. Usages of “Law Enforcement Officer” Beyond the Proviso
The textual analysis above is enough to satisfy the proviso’s four-part definition of “investigative or law enforcement officer.” For three reasons, other statutory usages of “law enforcement officer” beyond the proviso do not change that outcome.
First, Congress’s use of the bare term “law enforcement officer” says nothing about the term “investigative or law enforcement officer” in the proviso. “[O]r” is “disjunctive,” and “terms connected by a disjunctive [should] be given separate meanings unless the context dictates otherwise.” United States v. Urban, 140 F.3d 229, 232 (3d Cir. 1998) (internal quotation marks omitted); see
Second, it is unnecessary to explore the entire U.S. Code to discern the contours of the term “investigative or law enforcement officer,” because Congress provided an expressly local definition in the proviso. “When a statute includes an explicit definition, we must follow that definition, even if it varies from that term’s ordinary meaning.” Stenberg v. Carhart, 530 U.S. 914, 942 (2000) (citation omitted); see also Bond v. United States, 572 U.S. 844, 871 (2014) (Scalia, J., concurring in the judgment) (observing that we may not resolve any perceived “‘dissonance’ between ordinary meaning and the unambiguous words of a definition . . . in favor of ordinary meaning” because, “[i]f that were the
Third, Congress knows how to give an explicitly “criminal” meaning to the term “law enforcement officer,” and it chose not to do so here. See, e.g.,
We make no mention of legislative history. Where a statute is unclear on its face, good arguments exist that materials making known Congress’s purpose “should be respected, lest the integrity of legislation be undermined.” Robert A. Katzmann, Judging Statutes 4 (2014). Accordingly, our Court “has declined to employ legislative history if a statute is clear on its face,” but “we have allowed recourse to legislative history in the face of ambiguity.” Bruesewitz v. Wyeth, Inc., 561 F.3d 233, 244 (3d Cir. 2009) (Smith, J.), aff‘d sub nom. Bruesewitz v. Wyeth LLC, 562 U.S. 223, 242 (2011) (Scalia, J.) (noting that “legislative history is persuasive to some because it is thought to shed light on what legislators understood an ambiguous statutory text to mean” (citation omitted)).
Here, however, the text tells the tale. Cf. United States v. A.M., 927 F.3d 718, 719 (3d Cir. 2019) (“The text of a law governs its reach. We will neither read in new limits nor read out existing limits on its application.”). Congress could have chosen to insert “criminal” into the proviso. It did not, and thus we follow suit.4
Our following Congress’s lead rests on our view of the proper relationship between Congress and the courts. As we recently put it,
[t]he critical question is who should decide whether to provide for a damages remedy, Congress or the courts? Most often, the answer is Congress. Because, when an issue involves a host of considerations that must be weighed and appraised, it should be committed to those who write the laws rather than those who interpret them.
D. Consequences of Our Ruling
Before concluding, we note the implications of the choice before us. If TSOs are not “investigative or law enforcement officers” under the proviso, then plaintiffs like Pellegrino are left with no avenue for redress. We have already held (and correctly so) that TSOs are not susceptible to an implied right of action under Bivens for alleged constitutional violations, see Vanderklok, 868 F.3d at 209, so a Tort Claims Act action is the only remaining route to recovery. Without recourse under that Act, plaintiffs like Pellegrino will have no remedy when TSOs assault them, wrongfully detain them, or even fabricate criminal charges against them.
If, on the other hand, TSOs are “investigative or law enforcement officers,” we discern no risk of sweeping liability and certainly no concomitant threat to the public fisc. In 2015, for example, fewer than 200 people (out of over 700 million screened) filed complaints with the TSA alleging harm that would fall within the scope of the proviso. Corrected Tr. of Panel Oral Arg. at 26:8–17. In 2017, only one out of every 100,000 passengers lodged a complaint about the “courtesy” of a TSO, see Gary S. Becker, TSA Complaint Data Reveals Airport Screening Trends, Security Debrief (Mar. 16, 2018), http://securitydebrief.com/2018/03/16/tsa-complaint-data-airport-screening/, a statistic beyond suits alleging harm that
Nor is our ruling meant to draw every administrative search into the ambit of the proviso. As explained above, TSO screenings fall within the proviso because they are more personal than traditional administrative inspections: They extend to the general public and involve searches of an individual’s physical person and her property. See supra pp. 20–22.
Conclusion
Words matter. This core tenet of statutory interpretation channels our conclusion today: TSOs are “investigative or law enforcement officers” as defined in the Tort Claims Act at
As nearly all of us can attest who have flown on an aircraft in the United States, the overwhelming majority of TSOs perform their jobs professionally despite far more grumbling than appreciation. Their professionalism is commensurate with the seriousness of their role in keeping our skies safe. The life-and-death duties entrusted to them fall naturally within the ambit of the proviso.
The Majority and I agree that words matter, that our role is to interpret Congress’s statute and not to rewrite it, and that the United States retains sovereign immunity absent a clear and unambiguous waiver. But our statutory analyses of the so-called “law enforcement proviso” lead us to very different conclusions and demonstrate definitively, in my view, that TSA screeners do not qualify as “investigative or law enforcement officers.” At the very minimum, however, these two thoughtful opinions demonstrate that the proviso is susceptible to divergent yet “plausible interpretation[s].” FAA v. Cooper, 566 U.S. 284, 290–91 (2012). That conclusion, in and of itself, requires us to affirm the District Court’s dismissal of Pellegrino’s FTCA claims because “a waiver of sovereign immunity must be ‘unequivocally expressed’ in statutory text,” and “[a]ny ambiguities in the statutory language are to be construed in favor of immunity.” Id. at 290 (quoting Lane v. Pena, 518 U.S. 187, 192 (1996)).
From its earliest days, the Supreme Court has counseled that the “meaning of a word, and consequently, the intention of the legislature,” must be “ascertained by reference to the context” in which that word appears. Neal v. Clark, 95 U.S. 704, 709 (1878). I therefore read the words at issue in the context of the statute as a whole and of settled Fourth Amendment doctrine to conclude that the law enforcement proviso is limited by its terms to officers empowered to exercise traditional police powers—including investigatory searches for law enforcement purposes, but not administrative searches for programmatic purposes.
That breathtaking expansion of the proviso is textually unsound, departs from other circuits, and contravenes the rule that waivers of sovereign immunity must be strictly construed in favor of the Government. Because we should be reading Congress’s words together to give them the meaning that Congress intended and because we should not subject the United States Treasury to vast tort liability where Congress has not done so clearly and unambiguously, I respectfully dissent.
I. The Plain Language of the Proviso Excludes Administrative Employees, Like TSA Screeners, Who Conduct Routine, Suspicionless Searches
As with all cases involving statutory interpretation, the text must guide our analysis. But text cannot be interpreted in a vacuum. In law as in life, the meaning that we ascribe to words depends on the words that surround them, considering both “the specific context in which that language is used, and
Applying this principle to interpret the words of the law enforcement proviso in their statutory context, TSA screeners are neither “empowered by law to execute searches, to seize evidence, or to make arrests for violations of Federal law,” nor empowered by law as “officer[s] of the United States.”
A. Screeners Are Not “Empowered by Law to Conduct Searches . . . for Violations of Federal Law”
The law enforcement proviso waives sovereign immunity only if the alleged tortfeasor “is empowered by law to execute searches, to seize evidence, or to make arrests for violations of Federal law.”
1. The Law Distinguishes Between Investigatory and Administrative Searches, with TSA Screeners Conducting Only the Latter
To interpret the meaning of “execute searches . . . for violations of Federal law,” I begin with contemporaneous dictionaries and Fourth Amendment jurisprudence. While leaning on Black’s Law Dictionary to pick a favored definition of “officer,” see Maj. Op. 9, the Majority omits that the same edition defined “search” as “[a]n examination of a man’s house or other buildings or premises, or of his person, with a view to the discovery of . . . some evidence of guilt to be used in the prosecution of a criminal action.”1 Search, Black’s Law Dictionary 1518 (4th ed. 1968) (emphasis added). As this definition reflects, when Congress passed the law enforcement proviso in 1974, “execute search,” just like “seize evidence” and “make arrests,” referred in the law primarily to the exercise
At that time, “search” had only recently entered the legal lexicon to refer to examinations “for non-law-enforcement purposes such as employee drug screenings, building inspections, health inspections, and other administrative inspections.” Special-Needs Doctrine, Black’s Law Dictionary (11th ed. 2019); see Administrative Search, Black’s Law Dictionary (11th ed. 2019) (origin date of 1960); Eve Brensike Primus, Disentangling Administrative Searches, 111 Colum. L. Rev. 254, 260 (2011) (noting that “the concept of administrative searches first entered the law in the 1960s”). To distinguish these searches from investigatory ones, they were (and often still are) called “inspections,” “inspection searches,” “regulatory searches,” or “administrative searches.” Administrative Search, Black’s Law Dictionary (11th ed. 2019); see Inspection Searches, Black’s Law Dictionary 717 (5th ed. 1979).
Then, as now, the distinction between these two types of searches undergirds much of Fourth Amendment doctrine. In its foray into administrative searches, the Supreme Court held that the Fourth Amendment posed no barrier, because they were not “searches for evidence to be used in criminal prosecutions.” Frank v. Maryland, 359 U.S. 360, 365 (1959); see 5 Wayne R. LaFave, Search & Seizure § 10.1(a) (5th ed. 2018) (observing that under Frank the Fourth Amendment’s applicability “depended upon whether the search was a part of a criminal investigation which might lead to prosecution”). Even when later overruling Frank, the Court deemed administrative searches a distinct category of “search” that did not require individualized suspicion because a routine inspection presented “a less hostile intrusion than the typical policeman’s search for the fruits and instrumentalities of
Investigatory searches, which pertain “to criminal investigations, not routine, noncriminal procedures,” Colorado v. Bertine, 479 U.S. 367, 371 (1987) (citation omitted), have as their “primary purpose . . . to detect evidence of ordinary criminal wrongdoing,” and serve “the general purpose of investigating crime,” Edmond, 531 U.S. at 38, 39; see Ferguson, 532 U.S. at 73, 77, 83 (distinguishing searches with “criminal investigatory purposes” and “the immediate objective . . . to generate evidence for law enforcement purposes” from “constitutionally permissible suspicionless searches” (emphasis omitted)); Whren v. United States, 517 U.S. 806, 811–12 (1996) (distinguishing searches for “violation of law” from administrative searches).
Administrative searches, on the other hand, require neither individualized suspicion nor a warrant, but only because “the ‘primary purpose’ of the search[] is ‘[d]istinguishable from the general interest in crime control’” and is “other than conducting criminal investigations.”2
Of these two types of searches, the screenings “now routine at airports and at entrances to courts and other official buildings,” Chandler v. Miller, 520 U.S. 305, 323 (1997), fall squarely within the realm of “administrative searches,” Nat’l Treasury Emps. Union v. Von Raab, 489 U.S. 656, 675 n.3 (1989). At airports, such suspicionless screenings are not implemented to gather evidence of a crime with an eye toward criminal prosecution,3 but rather to effect “an administrative
Unsurprisingly, then, TSA policy directs that screenings “be tailored to the transportation security purpose for which they are conducted” and forbids “[a]dministrative and special needs searches . . . to detect evidence of crimes unrelated to transportation security.” TSA Mgmt. Directive No. 100.4 ¶¶ 6.B(1), C(1). If a screener’s preventative screening happens to uncover evidence of a crime, she must “refer it to a supervisor or a law enforcement official for appropriate action”; she cannot seize the item, continue searching, or make an arrest. Id. ¶ 6.C(1); see also id. ¶ 6.G(2). The “only TSA personnel who [can] engage in law enforcement activities”—such as detentions, arrests, seizures, and investigatory searches—are TSA “law enforcement officers.” Id. ¶ 6.G(3); see TSA Mgmt. Directive No. 1100.88-1 ¶ 4.A.
Thus, properly framed, the question presented today is
2. The Law Enforcement Proviso Covers Only Investigatory Searches
In determining whether Congress intended to cover administrative searches, we need not look beyond the proviso’s inclusion of “execute,” “for violations of Federal law,” the other duties (“seize evidence” and “make arrests”), “investigatory or law enforcement officer,” and a narrowly defined list of intentional torts. Each phrase suggests that Congress intended to refer only to investigatory searches, and together, they convey Congress’s unmistakable intent.
Execute Searches. The proviso does not use the word “search” alone; it refers to the power to “execute searches”—a term of art. Without exception, every other statute in the United States Code that uses this phrase refers to investigatory searches. E.g.,
Implying that it has some relevance to the plain meaning of “execute searches,” the Majority observes that Congress sometimes uses “milder” verbs than “execute” in the criminal context and that the proviso does not include the terms “warrant” or “search warrant.” Maj. Op. 18. But those observations are neither here nor there: The point is not that “execute” is a “harsh” word (as opposed to a “mild” one), but that the words “execute” and “search” must be read together, not in isolation. When Congress uses the phrase “execute searches,” it invariably refers to traditional investigatory searches. And, of course, “execut[ing] searches” in the criminal context is not limited to the execution of warrants. Investigatory searches also include brief stops if an officer reasonably suspects that “criminal activity may be afoot,” Terry v. Ohio, 392 U.S. 1, 30 (1968); searches incident to arrest to protect officers and “to prevent the concealment or destruction” of evidence, Arizona v. Gant, 556 U.S. 332, 339 (2009) (internal alterations, emphasis, and citation omitted); and protective sweeps “to ensure [officers‘] safety after, and
For Violations of Federal Law. Beyond the word “execute,” the term “searches” is further qualified by the phrase “for violations of Federal law.”5
Given the criminal connotation of “for violations of Federal law,” i.e., “to generate evidence for law enforcement purposes,” Ferguson, 532 U.S. at 83, instead of for “programmatic purpose[s],” id. at 81, the searches TSA screeners conduct simply are not “search[es] . . . for violations of Federal law.”
The Majority argues that because carrying weapons or explosives on an aircraft is a criminal offense and screeners are authorized to search for those items, screeners are searching “for violations of Federal law,” even if that does have criminal connotations. But aside from the constitutional cloud that would place over the entire TSA screening program, this
In sum, screeners can conduct routine, suspicionless searches only for the programmatic purpose of removing prohibited items, which is designed to prevent “violations of Federal law” from occurring; they do not search, and may not constitutionally search, “for violations of Federal law.”
“Seize Evidence” and “Make Arrests.” The other terms in the list (“seize evidence” and “make arrests“) provide important context. As the Majority concedes, “make arrests” necessarily carries criminal connotations. See Maj. Op. 17. So does “seize evidence.”7 Under the canon noscitur a sociis, Congress‘s listing of “execute searches” alongside “seize evidence” and “make arrests“—three actions routinely listed in tandem to describe police powers, see, e.g.,
In a cautionary tale, albeit not heeded by the Majority, the Supreme Court held that a nearby provision in the FTCA barring claims arising out of the “loss, miscarriage, or negligent transmission of letters or postal matter,”
So too in § 2680(h)—just a few subsections later—where the terms “seize evidence” and “make arrests” are properly read in the proviso to “limit the reach,” id., of “execute searches.”
Investigative or Law Enforcement Officer. The very term being defined here—“investigative or law enforcement
Intentional Torts Covered. Finally, the law enforcement proviso waives immunity only for the types of tort claims typically asserted against traditional investigative or law enforcement officers—assault, battery, false imprisonment, false arrest, abuse of process, and malicious prosecution; it preserves the United States‘s immunity for libel, slander, misrepresentation, deceit, and interference with contract rights. See
Congress‘s intentional selection of torts premised on use of excessive force and lack of probable cause cannot be waved aside by the observation that screeners, like any other federal employee, can commit “assault, battery, false imprisonment, false arrest, malicious prosecution, abuse of process, libel, slander, misrepresentation, deceit, or interference with contract rights.”
For these reasons, the law enforcement proviso evinces no intent to waive sovereign immunity for administrative searches. And on that basis alone, Pellegrino‘s FTCA claims do not fall within the proviso.
B. TSA Screeners Are Not “Officers” Under the Proviso
The law enforcement proviso further limits the waiver of sovereign immunity to alleged torts committed by “officer[s].” See
Without explanation, the Majority adopts the broadest possible dictionary entries, defining “officer” as someone “‘charged’ by the Government ‘with the power and duty of exercising certain functions,‘” Maj. Op. 9 (quoting Officer, Black‘s Law Dictionary 1235 (4th ed. 1968)), or who “serve[s] in a position of trust [or] authority” id. (quoting Officer, Webster‘s Third New International Dictionary 1567 (1971)). Such expansive definitions of “officer” could theoretically pass muster if they did no violence to neighboring words in the statute. See, e.g., United States v. Locke, 529 U.S. 89, 105 (2000); King v. St. Vincent‘s Hosp., 502 U.S. 215, 221 (1991). But here they do.
1. The Majority‘s Reading Gives No Meaning to Congress‘s Choice of “Officer,” Rather Than “Employee,” in the Law Enforcement Proviso
Several provisions in the Federal Tort Claims Act use the term “employee” to identify whose acts or omissions are covered. For example, the FTCA grants federal district courts exclusive jurisdiction over torts “caused by the negligent or wrongful act or omission of any employee.”
Where, as here, Congress uses certain language in one part of a statute but distinct terminology elsewhere, courts should “‘presume[]’ that Congress intended a difference in meaning.” Loughrin v. United States, 573 U.S. 351, 358 (2014) (quoting Russello v. United States, 464 U.S. 16, 23 (1983)); see, e.g., Sosa v. Alvarez-Machain, 542 U.S. 692, 711 n.9 (2004). If Congress wanted the proviso to sweep broadly, it could have—just as in the discretionary function exception a few subsections above, see
The Majority disregards this presumption by choosing a dictionary definition of “officer” so broad as to render the term coextensive with “employee.” Indeed, I am hard pressed to conceive of any employee of an agency who conducts administrative searches and is not, as the Majority defines it, in “a position of trust and authority,” or, for that matter, any federal employee at all who has not been “‘charged’ by the Government ‘with the power and duty of exercising certain functions.‘” Maj. Op. 9 (citations omitted).
Instead of conflating “officer” with “employee,” I read Congress‘s markedly different language in the very same statutory section to signal an intent to limit the proviso to a specific class of federal government personnel: those “charged with police duties.” Officer, Webster‘s New Collegiate Dictionary 797 (1976). After all, the FTCA itself defines “[e]mployee of the government” as “officers or employees of any federal agency.”
2. The ATSA Distinguishes Between Screeners and Officers Empowered with Investigative and Law Enforcement Powers
The ATSA specifies that “screening . . . shall be carried out by a Federal Government employee (as defined in section 2105 of title 5).”
Only TSA employees designated as “officers” are empowered by law to “carry a firearm,” “make an arrest,” and “seek and execute warrants for arrest or seizure of evidence,”
And were there any doubt about whether Congress intended that mapping, the ATSA dispels it: Congress expressly cross-referenced the FTCA when it distinguished the liability of state and local law enforcement officers—who may be deputized by the Administrator to supplement “Federal law enforcement officers” at airports,
3. The Majority‘s Contrary Arguments Are Unavailing
In the face of the plain text to the contrary, the Majority offers several reasons why its broad interpretation of “officer,” which would encompass a screener, should prevail. None is persuasive.
First, the Majority tells us that we should not rely on the distinction between an “officer” and “employee” in the FTCA and ATSA because both statutes define “employee” to include “employees” and “officers,” and there can be no “distinction between two terms that are not themselves mutually exclusive.” Maj. Op. 11 (citing
Second, the Majority argues against mapping the ATSA‘s definition of “employee” and “officer” onto the FTCA. But its syllogism is flawed. According to the Majority, (a) because the ATSA defines the term “employee” by reference to the general civil-service laws, see
But both premises are wrong. Congress did not define “officer” in the proviso by reference to civil-service laws; it told us in no uncertain terms to look to the particular “law” that “empower[s]” employees of that agency—here, the ATSA, to act as officers; see also
With that reading, moreover, the FTCA‘s reference to “officer” is not “underinclusive.” The laws that empower employees of various agencies consistently demarcate those who carry police powers from regular employees. Examples include the Internal Revenue Code,
Third, the Majority relies on the fact that screeners “are officers by name” and “wear uniforms with badges noting that title.” Maj. Op. 10. It is surprising indeed that such a superficial gloss is deemed relevant to understanding a waiver of federal sovereign immunity—particularly as these employees were originally called “screeners” (and remain so in the statute and regulations), and TSA (not Congress) changed their name to Transportation Security Officer (TSO) in 2005 only in an effort to improve morale and foster “upward
Fourth, the Majority seizes on the modifier “any” before “officer” to garner support for its expansive interpretation. But, again, this is text without context. The argument would carry appeal “only if we stopped reading right there,” but “we do not stop there; we do not read statutes in little bites.” Kircher v. Putnam Funds Tr., 547 U.S. 633, 643 (2006). Whether “use of the word ‘any’ . . . indicate[s] that Congress intended particular statutory text to sweep broadly . . . necessarily depends on the statutory context.” See Nat‘l Ass‘n of Mfrs. v. Dep‘t of Def., 138 S. Ct. 617, 629 (2018). The definition of “investigative or law enforcement officer” does not end after the words “any officer of the United States“; it goes on to delineate those officers “empowered by law to” perform the traditional police powers that follow.
In sum, screeners are not “officers” and for that reason, too, they are not “investigative or law enforcement officers.”
II. The Legislative History Confirms That Congress Did Not Intend to Cover Administrative Searches
While legislative history cannot manufacture ambiguity where none exists, “for those of us who use legislative history to help interpret statutes, the history . . . supports our reading,” Obduskey v. McCarthy & Holthus LLP, 139 S. Ct. 1029, 1037 (2019), confirming the textual cues on which I rely. The Supreme Court and our Circuit have similarly considered legislative history as a useful “cross-check.” United States ex rel. Greenfield v. Medco Health Sols., Inc., 880 F.3d 89, 95 (3d Cir. 2018) (Ambro, J.); see, e.g., Sturgeon v. Frost, 139 S. Ct. 1066, 1085 (2019).
Congress did not come to use the word “officer” rather than “employee” in the proviso by accident. Responding to two appalling “no-knock” raids by federal narcotics officers, Congress considered three bills to amend the broad immunity
In its deliberations, policymakers discussed the fact that the Proviso Bill, unlike the other bills, would not cover federal employees who perform administrative searches. Some observed that the Proviso Bill “only applies to law enforcement officers. It does not apply to any other Federal employees that might violate the rights of an individual.” 120 Cong. Rec. 5287 (statements of Reps. Donohue and Wiggins). Others, urging passage of the alternative bills that waived immunity for all federal employees, lamented that the Proviso Bill would provide no remedy for assaults committed by those who perform administrative searches: “We have Department of Agriculture investigators who go [in to] look at books and records. We have Defense Department auditors to look at books and records. . . . They are not law enforcement officers even under this definition. They don‘t qualify.”15 But the
The legislative history concerning the particular torts selected for the proviso also confirms my plain text reading: Congress‘s intention, in excepting “assault, battery, false imprisonment, false arrest, abuse of process, [and] malicious prosecution,”
This history simply corroborates what the text itself conveys: After debating competing options, Congress decided to afford a remedy only to “victims of Federal law enforcement abuses.” Caban v. United States, 671 F.2d 1230, 1235 (2d Cir. 1982) (quoting S. Rep. No. 93-588, at 4 (1973)); see Carlson v. Green, 446 U.S. 14, 19 (1980) (“Congress amended FTCA in 1974 to create a cause of action against the United States for intentional torts committed by federal law enforcement officers.“).
(testimony of Irving Jaffe) (“It should be noted that . . . H.R. 8245 is confined in its applicability to Federal investigative or law enforcement officers, while . . . H.R. 10439 would waive the sovereign immunity of the United States as to the same acts or omissions on the part of all Government employees.“).
After dismissing “[a]ttempts to distinguish . . . between administrative and criminal ‘searches‘” as purportedly “divorced from the plain text,” the Majority offers up an atextual reading, positing that a distinction could be drawn between physical and nonphysical searches. Maj. Op. 14, 20. But the Majority provides no principled basis for that distinction. That is because there is not one: Its reading sweeps in all administrative searches.
A. The Majority Offers No Principled Basis for Limiting Its Reading to Physical Searches
Uneasy with the breadth of its holding, the Majority posits that TSA screenings can be distinguished from other administrative searches because they may include pat-downs. For an opinion premised on adhering to the text‘s plain meaning, this marks a striking shift. Nothing in the proviso even remotely hints at a distinction between administrative searches that include pat-downs and administrative searches that do not. It does not use the term “physical searches,” but simply “searches.” My colleagues cannot in the same breath proclaim fidelity to the text and devise an atextual line between “physical searches” and “non-physical searches” to attempt to cabin the proviso‘s reach: Congress either intended the proviso to waive sovereign immunity for those conducting both categories of
Not only is this reading of the proviso as limited to “physical searches” atextual, it is made out of whole cloth. The
The Majority also suggests that its expansion of the proviso today is limited only to TSA screenings because “they extend to the general public and involve examinations, often intrusive, of an individual‘s physical person along with her property.” Maj. Op. 22. But those features are not unique to airport searches. Searches to which the general public is subjected involving examinations of persons and property are “now routine . . . at entrances to courts and other official buildings.” Chandler, 520 U.S. at 323. They are “used widely at state and local levels to enforce laws regarding drivers’ licenses, safety requirements, weight limits, and similar matters.” Martinez-Fuerte, 428 U.S. at 560 n.14. All federal agencies “may, at their discretion, inspect packages, briefcases and other containers in the immediate possession of . . . persons arriving on, working at, visiting, or departing from Federal property.”
Finally, the Majority errs in conflating airport pat-downs with Terry stops. Airport pat-downs serve a programmatic purpose; Terry stops require individualized suspicion. We made that point clearly in Hartwell, where we upheld a frisk “without individualized suspicion” of an airport passenger as “permissible under the administrative search doctrine.” 436 F.3d at 181. Hartwell further observed that, unlike Terry stops, airport screenings are “well-tailored to protect personal privacy,” lack virtually any stigma, provide passengers with advance notice, and are “made under supervision and not far from the scrutiny of the traveling public.” Id. at 180–81 (citation omitted). Terry stops are on the other side of that
In the end, the Majority succumbs to the siren call that we need only concern ourselves today with “hold[ing] . . . that TSO screenings are ‘searches’ under the proviso,” leaving future panels to fend off the consequences. Maj. Op. 22. But we should not undertake even a purportedly narrow holding—and the Majority‘s holding is far from narrow—without having both a principled basis and a considered view of the repercussions. The Majority‘s inability to identify any sustainable distinction between TSA screenings and other administrative searches does not bode well for either.
Without a limiting principle, the Majority‘s interpretation of the law enforcement proviso works a staggering expansion of the Government‘s waiver of sovereign immunity. Much of what administrative agencies and their employees are empowered to do qualifies as a “search” under the
Here, the significance of the Supreme Court‘s decision in Millbrook v. United States, 569 U.S. 50 (2013), warrants emphasis. Under Millbrook, if an employee has the authority to perform any of these “searches“—and thus, under the Majority‘s interpretation, qualifies as an “investigative or law
The potential scale of that liability is why Congress sought to limit the proviso to “investigative or law enforcement officers” and the specific subset of intentional torts they are carefully trained to avoid. Law enforcement officers “are expected to ‘schoo[l] themselves in the niceties‘” of
In short, instead of drawing the principled and constitutionally grounded line between investigatory and administrative searches, my colleagues today open the United States Treasury to liability for the intentional torts of every federal employee with the authority to conduct any
IV. The Majority Creates, and Takes the Wrong Side of, a Circuit Split
If, as the Majority claims, its reading reflected an intuitive reading of the law enforcement proviso, presumably it would find some support in our precedent or that of other circuits. But instead it marks a dramatic departure.
Among our precedents, the Majority‘s approach is in sharp tension with Matsko v. United States, 372 F.3d 556 (3d Cir. 2004), where we held that a Mine Safety and Health Administration employee with “authority to inspect mines and investigate possible violations” was not covered by the proviso, because “employees of administrative agencies, no matter what investigative conduct they are involved in, do not
A unanimous panel of the Eleventh Circuit squarely rejected the Majority‘s interpretation in a persuasive and well-reasoned, albeit non-precedential,17 opinion. See Corbett v. TSA, 568 F. App‘x 690 (11th Cir. 2014) (per curiam). The court there considered the pertinent statutory language and concluded that TSA screeners are not covered by the proviso for the “simple[]” reason that they are “employees,” not “officers.” Id. at 701. The court therefore relied on the same textual distinction that the Majority in this case elides.18
Consistent with these decisions, the Seventh Circuit held in Bunch v. United States, 880 F.3d 938 (7th Cir. 2018), that the limited record “d[id] not foreclose the possibility” that the proviso could apply to an ATF forensic chemist whose duties may have included “the identification of relevant
On the other hand, other circuits have held that the proviso does not cover positions that lack a criminal law component. In EEOC v. First National Bank of Jackson, for example, the Fifth Circuit concluded that EEOC agents fell outside the proviso, distinguishing between federal employees who “have access to, for the purpose of examination, and the right to copy any evidence of any person being investigated or proceeded against that relates to unlawful employment practices,” and “investigative or law enforcement officers” who have the power to “execute searches.” 614 F.2d 1004, 1007–08 (5th Cir. 1980) (citation omitted); see also Wilson v. United States, 959 F.2d 12, 15 (2d Cir. 1992) (per curiam) (parole officers); Moore, 213 F.3d at 710 (federal prosecutors); Solomon v. United States, 559 F.2d 309, 310 (5th Cir. 1977) (per curiam) (security guards); Johnson v. United States, 547 F.2d 688, 691 (D.C. Cir. 1976) (per curiam) (VA hospital physicians).
In short, with no exception until today, the Courts of
V. Where, as Here, At Least Two Plausible Interpretations Exist, We Must Construe the Law Enforcement Proviso in Favor of the Sovereign
By departing from precedent to expose the United States to enormous liability, the Majority‘s interpretation runs afoul of another principle of statutory interpretation: that waivers of sovereign immunity must be construed narrowly in favor of the United States.
Because courts do not casually infer that the United States has waived its sovereign immunity, a waiver must be “strictly construed, in terms of its scope, in favor of the sovereign.” Lane v. Pena, 518 U.S. 187, 192 (1996). Thus, where “a plausible interpretation of the statute” exists that would preserve the United States‘s sovereign immunity, a court must adopt it. Cooper, 566 U.S. at 290–91. Our Circuit, just like every other, has applied these principles to the FTCA‘s waiver of immunity. Lightfoot v. United States, 564 F.3d 625, 628 (3d Cir. 2009); see also Evans v. United States, 876 F.3d 375, 380 (1st Cir. 2017); Tsolmon v. United States, 841 F.3d 378, 382 (5th Cir. 2016); McGowan v. United States, 825 F.3d 118, 126 (2d Cir. 2016); Lopez v. United States, 823 F.3d 970, 976 (10th Cir. 2016); Jackson v. United States, 751 F.3d 712, 717 (6th Cir. 2014).
While Dolan held that the general rule does not adhere
* * *
Like my colleagues, I am sympathetic to the concern that the current legal regime provides no obvious remedy for
But courts “do not sit as councils of revision, empowered to rewrite legislation in accord with their own conceptions of prudent public policy.” United States v. Rutherford, 442 U.S. 544, 555 (1979). Congress to date has limited the proviso to “investigative or law enforcement officers“—a term that covers only officers with traditional police powers. The wisdom of this policy, especially as it implicates the public fisc, lies beyond our purview. I therefore respectfully dissent.