521 U.S. 424 (1997)

METRO-NORTH COMMUTER RAILROAD CO.
v.
BUCKLEY

No. 96-320.

United States Supreme Court.

Argued February 18, 1997. Decided June 23, 1997. CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT

*426 *426 Breyer, J., delivered the opinion of the Court, in which Rehnquist, C. J., and O'Connor, Scalia, Kennedy, Souter, and Thomas, JJ., joined. Ginsburg, J., filed an opinion concurring in the judgment in part and dissenting in part, in which Stevens, J., joined, post, p. 444.

Sheila L. Birnbaum argued the cause for petitioner. With her on the briefs were Barbara Wrubel, Douglas W. Dunham, Ellen P. Quackenbos, and Richard K. Bernard.

Charles C. Goetsch argued the cause for respondent. With him on the brief were George J. Cahill, Jr., and John G. DiPersia.^[*]

Justice Breyer, delivered the opinion of the Court.

The basic question in this case is whether a railroad worker negligently exposed to a carcinogen (here, asbestos) but without symptoms of any disease can recover under the *427 Federal Employers' Liability Act (FELA or Act), 35 Stat. 65, as amended, 45 U. S. C. § 51 et seq., for negligently inflicted emotional distress. We conclude that the worker before us here cannot recover unless, and until, he manifests symptoms of a disease. We also consider a related claim for medical monitoring costs, and we hold, for reasons set out below, that the respondent in this case has not shown that he is legally entitled to recover those costs.

I

Respondent, Michael Buckley, works as a pipefitter for Metro-North, a railroad. For three years (1985-1988) his job exposed him to asbestos for about one hour per working day. During that time Buckley would remove insulation from pipes, often covering himself with insulation dust that contained asbestos. Since 1987, when he attended an "asbestos awareness" class, Buckley has feared that he would develop cancer—and with some cause, for his two expert witnesses testified that, even after taking account of his nowdiscarded 15-year habit of smoking up to a pack of cigarettes per day, the exposure created an added risk of death due to cancer, or to other asbestos-related diseases, of either 1% to 5% (in the view of one of plaintiff's experts), or 1% to 3% (in the view of another). Since 1989, Buckley has received periodic medical checkups for cancer and asbestosis. So far, those checkups have not revealed any evidence of cancer or any other asbestos-related disease.

Buckley sued Metro-North under the FELA, a statute that permits a railroad worker to recover for an "injury . . . resulting . . . from" his employer's "negligence." 45 U. S. C. § 51. He sought damages for his emotional distress and to cover the cost of future medical checkups. His employer conceded negligence, but it did not concede that Buckley had actually suffered emotional distress, and it argued that the FELA did not permit a worker like Buckley, who had suffered no physical harm, to recover for injuries of either sort. *428 After hearing Buckley's case, the District Court dismissed the action. The court found that Buckley did not "offer sufficient evidence to allow a jury to find that he suffered a real emotional injury." App. 623. And, in any event, Buckley suffered no "physical impact"; hence any emotional injury fell outside the limited set of circumstances in which, according to this Court, the FELA permits recovery. Id., at 620; see Consolidated Rail Corporation v. Gottshall, 512 U. S. 532 (1994). The District Court did not discuss Buckley's further claim for the costs of medical monitoring.

Buckley appealed, and the Second Circuit reversed. 79 F. 3d 1337 (1996). Buckley's evidence, it said, showed that his contact with the insulation dust (containing asbestos) was "massive, lengthy, and tangible," id., at 1345, and that the contact "would cause fear in a reasonable person," id., at 1344. Under these circumstances, the court held, the contact was what this Court in Gottshall had called a "physical impact"—a "physical impact" that, when present, permits a FELA plaintiff to recover for accompanying emotional distress. The Second Circuit also found in certain of Buckley's work place statements sufficient expression of worry to permit sending his emotional distress claim to a jury. Finally, the court held that Buckley could recover for the costs of medical checkups because the FELA permits recovery of all reasonably incurred extra medical monitoring costs whenever a "reasonable physician would prescribe . . . a monitoring regime different than the one that would have been prescribed in the absence of" a particular negligently caused exposure to a toxic substance. 79 F. 3d, at 1347 (internal quotation marks omitted).

We granted certiorari to review the Second Circuit's holdings in light of Gottshall.

II

The critical question before us in respect to Buckley's "emotional distress" claim is whether the physical contact with insulation dust that accompanied his emotional distress *429 amounts to a "physical impact" as this Court used that term in Gottshall. In Gottshall, an emotional distress case, the Court interpreted the word "injury" in FELA § 1, a provision that makes "[e]very common carrier by railroad . . . liable in damages to any person suffering injury while . . . employed" by the carrier if the "injury" results from carrier "negligence." 45 U. S. C. § 51. In doing so, it initially set forth several general legal principles applicable here. Gottshall described FELA's purposes as basically "humanitarian." Gottshall, supra, at 542; see also, e. g., Urie v. Thompson, 337 U. S. 163 (1949). It pointed out that the Act expressly abolishes or modifies a host of common-law doctrines that previously had limited recovery. See, e. g., 45 U. S. C. §§ 51, 53, and 54. It added that this Court has interpreted the Act's language "liberally" in light of its humanitarian purposes. Gottshall, supra, at 543. But, at the same time, the Court noted that liability under the Act rests upon "negligence" and that the Act does not make the railroad "`the insurer' " for all employee injuries. 512 U. S., at 543 (quoting Ellis v. Union Pacific R. Co., 329 U. S. 649, 653 (1947)). The Court stated that "common-law principles," where not rejected in the text of the statute, "are entitled to great weight" in interpreting the Act, and that those principles "play a significant role" in determining whether, or when, an employee can recover damages for "negligent infliction of emotional distress." 512 U. S., at 544. See also id., at 558 (Souter, J., concurring) (Court's duty "in interpreting FELA . . . is to develop a federal common law of negligence . . . informed by reference to the evolving common law"); Atchison, T. & S. F. R. Co. v. Buell, 480 U. S. 557 (1987).

The Court also set forth several more specific legal propositions. It recognized that the common law of torts does not permit recovery for negligently inflicted emotional distress unless the distress falls within certain specific categories that amount to recovery-permitting exceptions. The law, *430 for example, does permit recovery for emotional distress where that distress accompanies a physical injury, see, e. g., Simmons v. Pacor, Inc., 543 Pa. 664, 678, 674 A. 2d 232, 239 (1996); Restatement (Second) of Torts § 924(a) (1977), and it often permits recovery for distress suffered by a close relative who witnesses the physical injury of a negligence victim, e. g., Dillon v. Legg, 68 Cal. 2d 728, 441 P. 2d 912 (1968); Gottshall, 512 U. S., at 549, n. 10 (citing cases). The Court then held that FELA § 1, mirroring the law of many States, sometimes permitted recovery "for damages for negligent infliction of emotional distress," id., at 550, and, in particular, it does so where a plaintiff seeking such damages satisfies the common law's "zone of danger" test. It defined that test by stating that the law permits "recovery for emotional injury" by

"those plaintiffs who sustain a physical impact as a result of a defendant's negligent conduct, or who are placed in immediate risk of physical harm by that conduct." Id., at 547-548 (emphasis added).

The case before us, as we have said, focuses on the italicized words "physical impact." The Second Circuit interpreted those words as including a simple physical contact with a substance that might cause a disease at a future time, so long as the contact was of a kind that would "cause fear in a reasonable person." 79 F. 3d, at 1344. In our view, however, the "physical impact" to which Gottshall referred does not include a simple physical contact with a substance that might cause a disease at a substantially later time— where that substance, or related circumstance, threatens no harm other than that disease-related risk.

First, Gottshall cited many state cases in support of its adoption of the "zone of danger" test quoted above. And in each case where recovery for emotional distress was permitted, the case involved a threatened physical contact that caused, or might have caused, immediate traumatic harm. *431 Keck v. Jackson, 122 Ariz. 114, 593 P. 2d 668 (1979) (car accident); Towns v. Anderson, 195 Colo. 517, 579 P. 2d 1163 (1978) (gas explosion); Robb v. Pennsylvania R. Co., 58 Del. 454, 210 A. 2d 709 (1965) (train struck car); Rickey v. Chicago Transit Authority, 98 Ill. 2d 546, 457 N. E. 2d 1 (1983) (clothing caught in escalator choked victim); Shuamber v. Henderson, 579 N. E. 2d 452 (Ind. 1991) (car accident); Watson v. Dilts, 116 Iowa 249, 89 N. W. 1068 (1902) (intruder assaulted plaintiff's husband); Stewart v. Arkansas Southern R. Co., 112 La. 764, 36 So. 676 (1904) (train accident); Purcell v. St. Paul City R. Co., 48 Minn. 134, 50 N. W. 1034 (1892) (near streetcar collision); Bovsun v. Sanperi, 61 N. Y. 2d 219, 461 N. E. 2d 843 (1984) (car accident); Kimberly v. Howland, 143 N. C. 398, 55 S. E. 778 (1906) (rock from blasting crashed through plaintiffs' residence); Simone v. Rhode Island Co., 28 R. I. 186, 66 A. 202 (1907) (streetcar collision); Mack v. South-Bound R. Co., 52 S. C. 323, 29 S. E. 905 (1898) (train narrowly missed plaintiff); Gulf, C. & S. F. R. Co. v. Hayter, 93 Tex. 239, 54 S. W. 944 (1900) (train collision); Pankopf v. Hinkley, 141 Wis. 146, 123 N. W. 625 (1909) (automobile struck carriage); Garrett v. New Berlin, 122 Wis. 2d 223, 362 N. W. 2d 137 (1985) (car accident). Cf. Deutsch v. Shein, 597 S. W. 2d 141 (Ky. 1980) (holding that exposure to X rays was "physical contact" supporting recovery for emotional suffering where immediate physical harm to fetus was suspected).

Second, Gottshall' s language, read in light of this precedent, seems similarly limited. 512 U. S., at 555 ("zone of danger test . . . is consistent with FELA's central focus on physical perils"); id., at 556 (quoting Lancaster v. Norfolk & Western R. Co., 773 F. 2d 807, 813 (CA7 1985)) (FELA seeks to protect workers "`from physical invasions or menaces' "), cert. denied, 480 U. S. 945 (1987); 512 U. S., at 556 (employer should be liable for "emotional injury caused by the apprehension of physical impact"); id., at 547-548 (quoting Pearson, Liability to Bystanders for Negligently Inflicted Emotional Harm—A Comment on the Nature of Arbitrary Rules, *432 34 U. Fla. L. Rev. 477, 488-489 (1982)) ("`[T]hose within the zone of danger of physical impact' " should be able to "`recover for fright' " because "`a near miss may be as frightening as a direct hit' ").

Taken together, language and cited precedent indicate that the words "physical impact" do not encompass every form of "physical contact." And, in particular, they do not include a contact that amounts to no more than an exposure—an exposure, such as that before us, to a substance that poses some future risk of disease and which contact causes emotional distress only because the worker learns that he may become ill after a substantial period of time.

Third, common-law precedent does not favor the plaintiff. Common-law courts do permit a plaintiff who suffers from a disease to recover for related negligently caused emotional distress, see supra, at 429, and some courts permit a plaintiff who exhibits a physical symptom of exposure to recover, see, e. g., Herber v. Johns-Manville Corp., 785 F. 2d 79, 85 (CA3 1986); Mauro v. Owens-Corning Fiberglas Corp., 225 N. J. Super. 196, 542 A. 2d 16 (App. Div. 1988). But with only a few exceptions, common-law courts have denied recovery to those who, like Buckley, are disease and symptom free. E. g., Burns v. Jacquays Mining Corp., 156 Ariz. 375, 752 P. 2d 28 (Ct. App. 1987), review dism'd, 162 Ariz. 186, 781 P. 2d 1373 (1989); Mergenthaler v. Asbestos Corp. of Am., 480 A. 2d 647 (Del. 1984); Eagle-Picher Industries, Inc. v. Cox, 481 So. 2d 517 (Fla. App. 1985), review denied, 492 So. 2d 1331 (Fla. 1986); Capital Holding Corp. v. Bailey, 873 S. W. 2d 187 (Ky. 1994); Payton v. Abbott Labs, 386 Mass. 540, 437 N. E. 2d 171 (1982); Simmons v. Pacor, Inc., 543 Pa. 664, 674 A. 2d 232 (1996); Ball v. Joy Technologies, Inc., 958 F. 2d 36 (CA4 1991); Deleski v. Raymark Industries, Inc., 819 F. 2d 377 (CA3 1987) (Pennsylvania and New Jersey law); Adams v. Johns-Manville Sales Corp., 783 F. 2d 589 (CA5 1986) (Louisiana law); Wisniewski v. Johns-Manville Corp., 759 F. 2d 271 (CA3 1985) (Pennsylvania law); In re Hawaii Federal *433 Asbestos Cases, 734 F. Supp. 1563 (Haw. 1990) (Hawaii law); Amendola v. Kansas City So. R. Co., 699 F. Supp. 1401 (WD Mo. 1988) (FELA); see also Potter v. Firestone Tire & Rubber Co., 6 Cal. 4th 965, 863 P. 2d 795 (1993) (in banc) (no recovery for fear of cancer in a negligence action unless plaintiff is "more likely than not" to develop cancer).

Fourth, the general policy reasons to which Gottshall referred—in its explanation of why common-law courts have restricted recovery for emotional harm to cases falling within rather narrowly defined categories—militate against an expansive definition of "physical impact" here. Those reasons include: (a) special "difficult[y] for judges and juries" in separating valid, important claims from those that are invalid or "trivial," Gottshall, 512 U. S., at 557; (b) a threat of "unlimited and unpredictable liability," ibid.; and (c) the "potential for a flood" of comparatively unimportant, or "trivial," claims, ibid.

To separate meritorious and important claims from invalid or trivial claims does not seem easier here than in other cases in which a plaintiff might seek recovery for typical negligently caused emotional distress. The facts before us illustrate the problem. The District Court, when concluding that Buckley had failed to present "sufficient evidence to allow a jury to find . . . a real emotional injury," pointed out that, apart from Buckley's own testimony, there was virtually no evidence of distress. App. 623-625. Indeed, Buckley continued to work with insulating material "even though. . . he could have transferred" elsewhere, he "continued to smoke cigarettes" despite doctors' warnings, and his doctor did not refer him "either to a psychologist or to a social worker." Id., at 624. The Court of Appeals reversed because it found certain objective corroborating evidence, namely, "workers' complaints to supervisors and investigative bodies." 79 F. 3d, at 1346. Both kinds of "objective" evidence—the confirming and disconfirming evidence—seem only indirectly related to the question at issue, the existence *434 and seriousness of Buckley's claimed emotional distress. Yet, given the difficulty of separating valid from invalid emotional injury claims, the evidence before us may typify the kind of evidence to which parties and the courts would have to look.

The Court in Gottshall made a similar point: "[T]esting for the `genuineness' of an injury alone . .. would be bound to lead to haphazard results. Judges would be forced to make highly subjective determinations concerning the authenticity of claims for emotional injury, which are far less susceptible to objective medical proof than are their physical counterparts. To the extent the genuineness test could limit potential liability, it could do so only inconsistently." 512 U. S., at 552.

And Justice Ginsburg, too, in her opinion concurring in the judgment in part and dissenting in part, seems to recognize this problem, for she would limit recovery in emotional injury cases to those who can show more objective evidence than simply having expressed fear and concern to supervisors. See post, at 445.

More important, the physical contact at issue here—a simple (though extensive) contact with a carcinogenic substance—does not seem to offer much help in separating valid from invalid emotional distress claims. That is because contacts, even extensive contacts, with serious carcinogens are common. See, e. g., Nicholson, Perkel, & Selikoff, Occupational Exposure to Asbestos: Population at Risk and Projected Mortality—1980-2030, 3 Am. J. Indust. Med. 259 (1982) (estimating that 21 million Americans have been exposed to work-related asbestos); U. S. Dept. of Health and Human Services, 1 Seventh Annual Report on Carcinogens 71 (1994) (3 million workers exposed to benzene, a majority of Americans exposed outside the work place); Pirkle, et al., Exposure of the U S Population to Environmental Tobacco Smoke, 275 JAMA 1233, 1237 (1996) (reporting that 43% of *435 American children lived in a home with at least one smoker, and 37% of adult nonsmokers lived in a home with at least one smoker or reported environmental tobacco smoke at work). They may occur without causing serious emotional distress, but sometimes they do cause distress, and reasonably so, for cancer is both an unusually threatening and unusually frightening disease. See Statistical Abstract of United States 94 (1996) (23.5% of Americans who died in 1994 died of cancer); American Cancer Society, Cancer Facts & Figures—1997, p. 1 (half of all men and one-third of all women will develop cancer). The relevant problem, however, remains one of evaluating a claimed emotional reaction to an increased risk of dying. An external circumstance—exposure—makes some emotional distress more likely. But how can one determine from the external circumstance of exposure whether, or when, a claimed strong emotional reaction to an increased mortality risk (say, from 23% to 28%) is reasonable and genuine, rather than overstated—particularly when the relevant statistics themselves are controversial and uncertain (as is usually the case), and particularly since neither those exposed nor judges or juries are experts in statistics? The evaluation problem seems a serious one.

The large number of those exposed and the uncertainties that may surround recovery also suggest what Gottshall called the problem of "unlimited and unpredictable liability." Does such liability mean, for example, that the costs associated with a rule of liability would become so great that, given the nature of the harm, it would seem unreasonable to require the public to pay the higher prices that may result? Cf. Priest, The Current Insurance Crisis and Modern Tort Law, 96 Yale L. J. 1521, 1585-1587 (1987). The same characteristics further suggest what Gottshall called the problem of a "flood" of cases that, if not "trivial," are comparatively less important. In a world of limited resources, would a rule permitting immediate large-scale recoveries for widespread emotional distress caused by fear of future disease *436 diminish the likelihood of recovery by those who later suffer from the disease? Cf. J. Weinstein, Individual Justice in Mass Tort Litigation 10-11, 141 (1995); Schuck, The Worst Should Go First: Deferral Registries in Asbestos Litigation, 15 Harv. J. L. & Pub. Pol'y 541 (1992).

We do not raise these questions to answer them (for we do not have the answers), but rather to show that general policy concerns of a kind that have led common-law courts to deny recovery for certain classes of negligently caused harms are present in this case as well. That being so, we cannot find in Gottshall' s underlying rationale any basis for departing from Gottshall' s language and precedent or from the current common-law consensus. That is to say, we cannot find in Gottshall' s language, cited precedent, other common-law precedent, or related concerns of policy a legal basis for adopting the emotional distress recovery rule adopted by the Court of Appeals.

Buckley raises several important arguments in reply. He points out, for example, that common-law courts do permit recovery for emotional distress where a plaintiff has physical symptoms; and he argues that his evidence of exposure and enhanced mortality risk is as strong a proof as an accompanying physical symptom that his emotional distress is genuine.

This argument, however, while important, overlooks the fact that the common law in this area does not examine the genuineness of emotional harm case by case. Rather, it has developed recovery-permitting categories the contours of which more distantly reflect this, and other, abstract general policy concerns. The point of such categorization is to deny courts the authority to undertake a case-by-case examination. The common law permits emotional distress recovery for that category of plaintiffs who suffer from a disease (or exhibit a physical symptom), for example, thereby finding a special effort to evaluate emotional symptoms warranted in that category of cases—perhaps from a desire to make a *437 physically injured victim whole or because the parties are likely to be in court in any event. In other cases, however, falling outside the special recovery-permitting categories, it has reached a different conclusion. The relevant question here concerns the validity of a rule that seeks to redefine such a category. It would not be easy to redefine "physical impact" in terms of a rule that turned on, say, the "massive, lengthy, [or] tangible" nature of a contact that amounted to an exposure, whether to contaminated water, or to germladen air, or to carcinogen-containing substances, such as insulation dust containing asbestos. But, in any event, for the reasons we have stated, supra, at 430-436, we cannot find that the common law has done so.

Buckley also points to a series of common-law cases that he believes offer him support. Many of these cases, however, find that the plaintiff at issue fell within a category where the law already permitted recovery for emotional distress. E. g., Marchica v. Long Island R. Co., 31 F. 3d 1197 (CA2 1994) (traumatic injury); Clark v. Taylor, 710 F. 2d 4 (CA1 1983) (intentional infliction of harm); Laxton v. Orkin Exterminating Co., 639 S. W. 2d 431, 433-434 (Tenn. 1982) (nuisance claim); Lavelle v. Owens-Corning Fiberglas Corp., 30 Ohio Misc. 2d 11, 507 N. E. 2d 476 (Ct. Common Pleas, Cayahoga Cty. 1987) (emotional distress damages sought by asbestosis-afflicted plaintiff). We have found only three asbestos-related cases, all involving state law, that support Buckley directly. Watkins v. Fibreboard Corp., 994 F. 2d 253, 259 (CA5 1993) (Texas law) (recognizing cause of action for emotional distress based on exposures to asbestos in the absence of physical symptoms); In re Moorenovich, 634 F. Supp. 634 (Me. 1986) (Maine law) (same); Gerardi v. Nuclear Utility Services, Inc., 149 Misc. 2d 657, 566 N. Y. S. 2d 1002 (Westchester Cty. 1991) (same). None of them was decided by the highest court of the relevant State. And we do not find that minority view a sufficient basis for reaching Buckley's proposed conclusion.

*438 Finally, Buckley argues that the "humanitarian" nature of the FELA warrants a holding in his favor. We do not doubt that the FELA's purpose militates in favor of recovery for a serious and negligently caused emotional harm. Cf. Gottshall, 512 U. S., at 550. But just as courts must interpret that law to take proper account of the harms suffered by a sympathetic individual plaintiff, so they must consider the general impact, on workers as well as employers, of the general liability rules they would thereby create. Here the relevant question concerns not simply recovery in an individual case, but the consequences and effects of a rule of law that would permit that recovery. And if the common law concludes that a legal rule permitting recovery here, from a tort law perspective, and despite benefits in some individual cases, would on balance cause more harm than good, and if we find that judgment reasonable, we cannot find that conclusion inconsistent with the FELA's humanitarian purpose.

III

Buckley also sought recovery for a different kind of "injury," namely, the economic cost of the extra medical checkups that he expects to incur as a result of his exposure to asbestos-laden insulation dust. The District Court, when it dismissed the action, did not discuss this aspect of Buckley's case. But the Second Circuit, when reversing the District Court, held that "a reasonable jury could award" Buckley the "costs" of "medical monitoring" in this case. 79 F. 3d, at 1347. We agreed to decide whether the court correctly found that the FELA permitted a plaintiff without symptoms or disease to recover this economic loss.

The parties do not dispute—and we assume—that an exposed plaintiff can recover related reasonable medical monitoring costs if and when he develops symptoms. As the Second Circuit pointed out, a plaintiff injured through negligence can recover related reasonable medical expenses as an element of damages. Ibid. (citing C. McCormick, Law of *439 Damages § 90 (1935)); see also Restatement (Second) of Torts § 924(c) (1977); J. Stein, Stein on Personal Injury Damages § 5.18 (2d ed. 1991). No one has argued that any different principle would apply in the case of a plaintiff whose "injury" consists of a disease, a symptom, or those sorts of emotional distress that fall within the FELA's definition of "injury." See Part II, supra. Much of the Second Circuit's opinion suggests it intended only to apply this basic principle of the law of damages. See, e. g., 79 F. 3d, at 1342 ("[T]his case turns upon whether . . . emotional harm . . . is an injury compensable under FELA"); id., at 1347 (monitoring costs are a "traditional element of tort damages"). Insofar as that is so, Part II of our opinion, holding that the emotional distress at issue here is not a compensable "injury," requires reversal on this point as well.

Other portions of the Second Circuit's opinion, however, indicate that it may have rested this portion of its decision upon a broader ground, namely, that medical monitoring costs themselves represent a separate negligently caused economic "injury," 45 U. S. C. § 51, for which a negligently exposed FELA plaintiff (including a plaintiff without disease or symptoms) may recover to the extent that the medical monitoring costs that a reasonable physician would prescribe for the plaintiff exceed the medical monitoring costs that "would have been prescribed in the absence of [the] exposure." 79 F. 3d, at 1347 (citation omitted). This portion of the opinion, when viewed in light of Buckley's straightforward claim for an "amount of money" sufficient to "compensate" him for "future medical monitoring expenses," Plaintiff's Proposed Charges to the Jury 25, Record, Doc. 33, suggests the existence of an ordinary, but separate, tort law cause of action permitting (as tort law ordinarily permits) the recovery of medical cost damages in the form of a lump sum, see Stein, supra, at §§ 5.1 and 5.18, and irrespective of insurance, Restatement (Second) of Torts, supra, § 920A(2). As so characterized, the Second Circuit's holding, in our *440 view, went beyond the bounds of currently "evolving common law." Gottshall, supra, at 558 (Souter, J., concurring).

Guided by the parties' briefs,