Safeco Insurance Co. of America v. Burr

(Slip Opinion) OCTOBER TERM, 2006 1 Syllabus NOTE: Where it is feasible, a syllabus (headnote) will be released, as is being done in connection with this case, at the time the opinion is issued. The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U. S. 321, 337. SUPREME COURT OF THE UNITED STATES Syllabus SAFECO INSURANCE CO. OF AMERICA ET AL. v. BURR ET AL. CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT No. 06–84. Argued January 16, 2007 —Decided June 4, 2007* The Fair Credit Reporting Act (FCRA) requires notice to a consumer subjected to “adverse action . . . based in whole or in part on any in formation contained in a consumer [credit] report.” 15 U. S. C. §1681m(a). As applied to insurance companies, “adverse action” is “a denial or cancellation of, an increase in any charge for, or a reduction or other adverse or unfavorable change in the terms of coverage or amount of, any insurance, existing or applied for.” §1681a(k)(1)(B)(i). FCRA provides a private right of action against businesses that use consumer reports but fail to comply. A negligent violation entitles a consumer to actual damages, §1681o(a), and a willful one entitles the consumer to actual, statutory, and even punitive damages, §1681n(a). Petitioners in No. 06–100 (GEICO) use an applicant’s credit score to select the appropriate subsidiary insurance company and the par ticular rate at which a policy may be issued. GEICO sends an ad verse action notice only if a neutral credit score would have put the applicant in a lower priced tier or company; the applicant is not oth erwise told if he would have gotten better terms with a better credit score. Respondent Edo’s credit score was taken into account when GEICO issued him a policy, but GEICO sent no adverse action notice because his company and tier placement would have been the same with a neutral score. Edo filed a proposed class action, alleging will ful violation of §1681m(a) and seeking statutory and punitive dam ages under §1681n(a). The District Court granted GEICO summary judgment, finding no adverse action because the premium would —————— * Together with No. 06–100, GEICO General Insurance Co. et al. v. Edo, also on certiorari to the same court. 2 SAFECO INS. CO. OF AMERICA v. BURR Syllabus have been the same had Edo’s credit history not been considered. Pe titioners in No. 06–100 (Safeco) also rely on credit reports to set ini tial insurance premiums. Respondents Burr and Massey—whom Safeco offered higher than the best rates possible without sending adverse action notices—joined a proposed class action, alleging will ful violation of §1681m(a) and seeking statutory and punitive dam ages under §1681n(a). The District Court granted Safeco summary judgment on the ground that offering a single, initial rate for insur ance cannot be “adverse action.” The Ninth Circuit reversed both judgments. In GEICO’s case, it held that an adverse action occurs whenever a consumer would have received a lower rate had his con sumer report contained more favorable information. Since that would have happened to Edo, GEICO’s failure to give notice was an adverse action. The court also held that an insurer willfully fails to comply with FCRA if it acts in reckless disregard of a consumer’s FCRA rights, remanding for further proceedings on the reckless dis regard issue. Relying on its decision in GEICO’s case, the Ninth Cir cuit rejected the District Court’s position in the Safeco case and re manded for further proceedings. Held: 1. Willful failure covers a violation committed in reckless disregard of the notice obligation. Where willfulness is a statutory condition of civil liability, it is generally taken to cover not only knowing viola tions of a standard, but reckless ones as well. See, e.g., McLaughlin v. Richland Shoe Co., 486 U. S. 128, 133. This construction reflects common law usage. The standard civil usage thus counsels reading §1681n(a)’s phrase “willfully fails to comply” as reaching reckless FCRA violations, both on the interpretive assumption that Congress knows how this Court construes statutes and expects it to run true to form, see Commissioner v. Keystone Consol. Industries, Inc., 508 U. S. 152, 159, and under the rule that a common law term in a statute comes with a common law meaning, absent anything pointing an other way, Beck v. Prupis, 529 U. S. 494, 500–501. Petitioners claim that §1681n(a)’s drafting history points to a reading that liability at taches only to knowing violations, but the text as finally adopted points to the traditional understanding of willfulness in the civil sphere. Their other textual and structural arguments are also un persuasive. Pp. 6–10. 2. Initial rates charged for new insurance policies may be adverse actions. Pp. 10–17. (a) Reading the phrase “increase in any charge for . . . any insur ance, existing or applied for,” §1681a(k)(1)(B)(i), to include a disad vantageous rate even with no prior dealing fits with the ambitious objective of FCRA’s statement of purpose, which uses expansive Cite as: 551 U. S. ____ (2007) 3 Syllabus terms to describe the adverse effects of unfair and inaccurate credit reporting and the responsibilities of consumer reporting agencies. See §1681(a). These descriptions do nothing to suggest that remedies for consumers disadvantaged by unsound credit ratings should be denied to first-time victims, and the legislative histories of both FCRA’s original enactment and a 1996 amendment reveal no reason to confine attention to customers and businesses with prior dealings. Finally, nothing about insurance contracts suggests that Congress meant to differentiate applicants from existing customers when it set the notice requirement; the newly insured who gets charged more ow ing to an erroneous report is in the same boat with the renewal appli cant. Pp. 10–13. (b) An increased rate is not “based in whole or in part on” a credit report under §1681m(a) unless the report was a necessary condition of the increase. In common talk, “based on” indicates a but-for causal relationship and thus a necessary logical condition. Though some textual arguments point another way, it makes more sense to suspect that Congress meant to require notice and prompt a consumer chal lenge only when the consumer would gain something if the challenge succeeded. Pp. 13–14. (c) In determining whether a first-time rate is a disadvantageous increase, the baseline is the rate that the applicant would have re ceived had the company not taken his credit score into account (the “neutral score” rate GEICO used in Edo’s case). That baseline com ports with the understanding that §1681m(a) notice is required only when the credit report’s effect on the initial rate is necessary to put the consumer in a worse position than other relevant facts would have decreed anyway. Congress was more likely concerned with the practical question whether the consumer’s rate actually suffered when his credit report was taken into account than the theoretical question whether the consumer would have gotten a better rate with the best possible credit score, the baseline suggested by the Govern ment and respondent-plaintiffs. The Government’s objection to this reading is rejected. Although the rate initially offered for new insur ance is an “increase” calling for notice if it exceeds the neutral rate, once a consumer has learned that his credit report led the insurer to charge more, he need not be told with each renewal if his rate has not changed. After initial dealing between the consumer and the insurer, the baseline for “increase” is the previous rate or charge, not the “neutral” baseline that applies at the start. Pp. 15–17. 3. GEICO did not violate the statute, and while Safeco might have, it did not act recklessly. Pp. 18–21. (a) Because the initial rate GEICO offered Edo was what he would have received had his credit score not been taken into account, 4 SAFECO INS. CO. OF AMERICA v. BURR Syllabus GEICO owed him no adverse action notice under §1681m(a). P. 18. (b) Even if Safeco violated FCRA when it failed to give Burr and Massey notice on the mistaken belief that §1681m(a) did not apply to initial applications, the company was not reckless. The common law has generally understood “recklessness” in the civil liability sphere as conduct violating an objective standard: action entailing “an unjusti fiably high risk of harm that is either known or so obvious that it should be known.” Farmer v. Brennan, 511 U. S. 825, 836. There be ing no indication that Congress had something different in mind, there is no reason to deviate from the common law understanding in applying the statute. See Beck v. Prupis, 529 U. S., at 500–501. Thus, a company does not act in reckless disregard of FCRA unless the action is not only a violation under a reasonable reading of the statute, but shows that the company ran a risk of violating the law substantially greater than the risk associated with a reading that was merely careless. The negligence/recklessness line need not be pinpointed here, for Safeco’s reading of the statute, albeit erroneous, was not objectively unreasonable. Section 1681a(k)(1)(B)(i) is silent on the point from which to measure “increase,” and Safeco’s reading has a foundation in the statutory text and a sufficiently convincing justification to have persuaded the District Court to adopt it and rule in Safeco’s favor. Before these cases, no court of appeals had spoken on the issue, and no authoritative guidance has yet come from the Federal Trade Commission. Given this dearth of guidance and the less-than-pellucid statutory text, Safeco’s reading was not objectively unreasonable, and so falls well short of raising the “unjustifiably high risk” of violating the statute necessary for reckless liability. Pp. 18– 21. No. 06–84, 140 Fed. Appx. 746; No. 06–100, 435 F. 3d 1081, reversed and remanded. SOUTER, J., delivered the opinion of the Court, in which ROBERTS, C. J., and KENNEDY and BREYER, JJ., joined, in which SCALIA, J., joined as to all but footnotes 11 and 15, in which THOMAS and ALITO, JJ., joined as to all but Part III–A, and in which STEVENS and GINSBURG, JJ., joined as to Parts I, II, III–A, and IV–B. STEVENS, J., filed an opinion concurring in part and concurring in the judgment, in which GINSBURG, J., joined. THOMAS, J., filed an opinion concurring in part, in which ALITO, J., joined. Cite as: 551 U. S. ____ (2007) 1 Opinion of the Court NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Wash ington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press. SUPREME COURT OF THE UNITED STATES _________________ Nos. 06–84 and 06–100 _________________ SAFECO INSURANCE COMPANY OF AMERICA, ET AL., PETITIONERS 06–84 v. CHARLES BURR ET AL. GEICO GENERAL INSURANCE COMPANY, ET AL., PETITIONERS 06–100 v. AJENE EDO ON WRITS OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT [June 4, 2007] JUSTICE SOUTER delivered the opinion of the Court.* The Fair Credit Reporting Act (FCRA or Act) requires notice to any consumer subjected to “adverse action . . . based in whole or in part on any information contained in a consumer [credit] report.” 15 U. S. C. §1681m(a). Any one who “willfully fails” to provide notice is civilly liable to the consumer. §1681n(a). The questions in these consoli dated cases are whether willful failure covers a violation committed in reckless disregard of the notice obligation, and, if so, whether petitioners Safeco and GEICO commit ted reckless violations. We hold that reckless action is covered, that GEICO did not violate the statute, and that —————— * JUSTICE SCALIA joins all but footnotes 11 and 15 of this opinion. 2 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court while Safeco might have, it did not act recklessly. I A Congress enacted FCRA in 1970 to ensure fair and accurate credit reporting, promote efficiency in the bank ing system, and protect consumer privacy. See 84 Stat. 1128, 15 U. S. C. §1681; TRW Inc. v. Andrews, 534 U. S. 19, 23 (2001). The Act requires, among other things, that “any person [who] takes any adverse action with respect to any consumer that is based in whole or in part on any information contained in a consumer report” must notify the affected consumer.1 15 U. S. C. §1681m(a). The notice must point out the adverse action, explain how to reach the agency that reported on the consumer’s credit, and tell the consumer that he can get a free copy of the report and dispute its accuracy with the agency. Ibid. As it applies to an insurance company, “adverse action” is “a denial or cancellation of, an increase in any charge for, or a reduc tion or other adverse or unfavorable change in the terms of coverage or amount of, any insurance, existing or applied for.” §1681a(k)(1)(B)(i). FCRA provides a private right of action against busi nesses that use consumer reports but fail to comply. If a violation is negligent, the affected consumer is entitled to actual damages. §1681o(a) (2000 ed., Supp. IV). If willful, however, the consumer may have actual damages, or statutory damages ranging from $100 to $1,000, and even —————— 1 So far as it matters here, the Act defines “consumer report” as “any written, oral, or other communication of any information by a consumer reporting agency bearing on a consumer’s credit worthiness, credit standing, [or] credit capacity . . . which is used or expected to be used or collected in whole or in part for the purpose of serving as a factor in establishing the consumer’s eligibility for . . . credit or insurance to be used primarily for personal, family, or household purposes.” 15 U. S. C. §1681a(d)(1) (footnote omitted). The scope of this definition is not at issue. Cite as: 551 U. S. ____ (2007) 3 Opinion of the Court punitive damages. §1681n(a) (2000 ed.). B Petitioner GEICO2writes auto insurance through four subsidiaries: GEICO General, which sells “preferred” policies at low rates to low-risk customers; Government Employees, which also sells “preferred” policies, but only to government employees; GEICO Indemnity, which sells standard policies to moderate-risk customers; and GEICO Casualty, which sells nonstandard policies at higher rates to high-risk customers. Potential customers call a toll-free number answered by an agent of the four affiliates, who takes information and, with permission, gets the appli cant’s credit score.3 This information goes into GEICO’s computer system, which selects any appropriate company and the particular rate at which a policy may be issued. For some time after FCRA went into effect, GEICO sent adverse action notices to all applicants who were not offered “preferred” policies from GEICO General or Gov ernment Employees. GEICO changed its practice, how ever, after a method to “neutralize” an applicant’s credit score was devised: the applicant’s company and tier placement is compared with the company and tier place ment he would have been assigned with a “neutral” credit score, that is, one calculated without reliance on credit history.4 Under this new scheme, it is only if using a —————— 2 The specific petitioners are subsidiary companies of the GEICO Cor poration; for the sake of convenience, we call them “GEICO” collectively. 3 The Act defines a “credit score” as “a numerical value or a categori zation derived from a statistical tool or modeling system used by a person who makes or arranges a loan to predict the likelihood of certain credit behaviors, including default.” 15 U. S. C. §1681g(f)(2)(A) (2000 ed., Supp. IV). Under its contract with its credit information providers, GEICO learned credit scores and facts in the credit reports that signifi cantly influenced the scores, but did not have access to the credit reports themselves. 4 A number of States permit the use of such “neutral” credit scores to 4 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court neutral credit score would have put the applicant in a lower priced tier or company that GEICO sends an ad verse action notice; the applicant is not otherwise told if he would have gotten better terms with a better credit score. Respondent Ajene Edo applied for auto insurance with GEICO. After obtaining Edo’s credit score, GEICO offered him a standard policy with GEICO Indemnity (at rates higher than the most favorable), which he accepted. Be cause Edo’s company and tier placement would have been the same with a neutral score, GEICO did not give Edo an adverse action notice. Edo later filed this proposed class action against GEICO, alleging willful failure to give notice in violation of §1681m(a); he claimed no actual harm, but sought statutory and punitive damages under §1681n(a). The District Court granted summary judgment for GEICO, finding there was no adverse action when “the premium charged to [Edo] . . . would have been the same even if GEICO Indemnity did not consider information in [his] consumer credit history.” Edo v. GEICO Casualty Co., CV 02–678–BR, 2004 U. S. Dist. LEXIS 28522, *12 (D. Ore., Feb. 23, 2004), App. to Pet. for Cert. in No. 06– 100, p. 46a. Like GEICO, petitioner Safeco5 relies on credit reports to set initial insurance premiums,6 as it did for respon dents Charles Burr and Shannon Massey, who were of fered higher rates than the best rates possible. Safeco —————— ensure that consumers with thin or unidentifiable credit histories are not treated disadvantageously. See, e.g., N. Y. Ins. Law Ann. §§2802(e), (e)(1) (West 2006) (generally prohibiting an insurer from “consider[ing] an absence of credit information,” but allowing it to do so if it “treats the consumer as if the applicant or insured had neutral credit informa tion, as defined by the insurer”). 5 Again, the actual petitioners are subsidiary companies, of Safeco Cor poration in this case; for convenience, we call them “Safeco” collectively. 6 The parties do not dispute that the credit scores and credit reports relied on by GEICO and Safeco are “consumer reports” under 15 U. S. C. §1681a(d)(1). Cite as: 551 U. S. ____ (2007) 5 Opinion of the Court sent them no adverse action notices, and they later joined a proposed class action against the company, alleging willful violation of §1681m(a) and seeking statutory and punitive damages under §1681n(a). The District Court ordered summary judgment for Safeco, on the understand ing that offering a single, initial rate for insurance cannot be “adverse action.” The Court of Appeals for the Ninth Circuit reversed both judgments. In GEICO’s case, it held that whenever a consumer “would have received a lower rate for his insur ance had the information in his consumer report been more favorable, an adverse action has been taken against him.” Reynolds v. Hartford Financial Servs. Group, Inc., 435 F. 3d 1081, 1093 (2006). Since a better credit score would have placed Edo with GEICO General, not GEICO Indemnity, the appeals court held that GEICO’s failure to give notice was an adverse action. The Ninth Circuit also held that an insurer “willfully” fails to comply with FCRA if it acts with “reckless disre gard” of a consumer’s rights under the Act. Id., at 1099. It explained that a company would not be acting recklessly if it “diligently and in good faith attempted to fulfill its statutory obligations” and came to a “tenable, albeit erro neous, interpretation of the statute.” Ibid. The court went on to say that “a deliberate failure to determine the extent of its obligations” would not ordinarily escape liability under §1681n, any more than “reliance on creative lawyer ing that provides indefensible answers.” Ibid. Because the court believed that the enquiry into GEICO’s reckless disregard might turn on undisclosed circumstances sur rounding GEICO’s revision of its notification policy, the Court of Appeals remanded the company’s case for further proceedings.7 —————— 7 Prior to issuing its final opinion in this case, the Court of Appeals had issued, then withdrawn, two opinions in which it held that GEICO 6 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court In the action against Safeco, the Court of Appeals re jected the District Court’s position, relying on its reasoning in GEICO’s case (where it had held that the notice re quirement applies to a single statement of an initial charge for a new policy). Spano v. Safeco Corp., 140 Fed. Appx. 746 (2005). The Court of Appeals also rejected Safeco’s argument that its conduct was not willful, again citing the GEICO case, and remanded for further proceedings. We consolidated the two matters and granted certiorari to resolve a conflict in the Circuits as to whether §1681n(a) reaches reckless disregard of FCRA’s obliga tions,8 and to clarify the notice requirement in §1681m(a). 548 U. S. ___ (2006). We now reverse in both cases. II GEICO and Safeco argue that liability under §1681n(a) for “willfully fail[ing] to comply” with FCRA goes only to acts known to violate the Act, not to reckless disregard of statutory duty, but we think they are wrong. We have said before that “willfully” is a “word of many meanings whose construction is often dependent on the context in which it appears,” Bryan v. United States, 524 U. S. 184, 191 (1998) (internal quotation marks omitted); and where willfulness is a statutory condition of civil liability, we have generally taken it to cover not only knowing viola tions of a standard, but reckless ones as well, see McLaughlin v. Richland Shoe Co., 486 U. S. 128, 132–133 (1988) (“willful,” as used in a limitation provision for —————— had “willfully” violated FCRA as a matter of law. Reynolds v. Hartford Financial Servs. Group, Inc., 416 F. 3d 1097 (CA9 2005); Reynolds v. Hartford Financial Servs. Group, Inc., 426 F. 3d 1020 (CA9 2005). 8 Compare, e.g., Cushman v. Trans Union Corp., 115 F. 3d 220, 227 (CA3 1997) (adopting the “reckless disregard” standard), with Wantz v. Experian Information Solutions, 386 F. 3d 829, 834 (CA7 2004) (con struing “willfully” to require that a user “knowingly and intentionally violate the Act”); Phillips v. Grendahl, 312 F. 3d 357, 368 (CA8 2002) (same). Cite as: 551 U. S. ____ (2007) 7 Opinion of the Court actions under the Fair Labor Standards Act, covers claims of reckless violation); Trans World Airlines, Inc. v. Thurston, 469 U. S. 111, 125–126 (1985) (same, as to a liquidated damages provision of the Age Discrimination in Employment Act of 1967); cf. United States v. Illinois Central R. Co., 303 U. S. 239, 242–243 (1938) (“willfully,” as used in a civil penalty provision, includes “ ‘conduct marked by careless disregard whether or not one has the right so to act’ ” (quoting United States v. Murdock, 290 U. S. 389, 395 (1933))). This construction reflects common law usage, which treated actions in “reckless disregard” of the law as “willful” violations. See W. Keeton, D. Dobbs, R. Keeton, & D. Owen, Prosser and Keeton on Law of Torts §34, p. 212 (5th ed. 1984) (hereinafter Prosser and Keeton) (“Although efforts have been made to distinguish” the terms “willful,” “wanton,” and “reckless,” “such distinc tions have consistently been ignored, and the three terms have been treated as meaning the same thing, or at least as coming out at the same legal exit”). The standard civil usage thus counsels reading the phrase “willfully fails to comply” in §1681n(a) as reaching reckless FCRA viola tions,9 and this is so both on the interpretive assumption —————— 9 It is different in the criminal law. When the term “willful” or “will fully” has been used in a criminal statute, we have regularly read the modifier as limiting liability to knowing violations. See Ratzlaf v. United States, 510 U. S. 135, 137 (1994); Bryan v. United States, 524 U. S. 184, 191–192 (1998); Cheek v. United States, 498 U. S. 192, 200– 201 (1991). This reading of the term, however, is tailored to the crimi nal law, where it is characteristically used to require a criminal intent beyond the purpose otherwise required for guilt, Ratzlaf, supra, at 136– 137; or an additional “ ‘bad purpose,’ ” Bryan, supra, at 191; or specific intent to violate a known legal duty created by highly technical stat utes, Cheek, supra, at 200–201. Thus we have consistently held that a defendant cannot harbor such criminal intent unless he “acted with knowledge that his conduct was unlawful.” Bryan, supra, at 193. Civil use of the term, however, typically presents neither the textual nor the substantive reasons for pegging the threshold of liability at knowledge of wrongdoing. Cf. Farmer v. Brennan, 511 U. S. 825, 836–837 (1994) 8 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court that Congress knows how we construe statutes and ex pects us to run true to form, see Commissioner v. Keystone Consol. Industries, Inc., 508 U. S. 152, 159 (1993), and under the general rule that a common law term in a stat ute comes with a common law meaning, absent anything pointing another way, Beck v. Prupis, 529 U. S. 494, 500– 501 (2000). GEICO and Safeco argue that Congress did point to something different in FCRA, by a drafting history of §1681n(a) said to show that liability was supposed to attach only to knowing violations. The original version of the Senate bill that turned out as FCRA had two stan dards of liability to victims: grossly negligent violation (supporting actual damages) and willful violation (sup porting actual, statutory, and punitive damages). S. 823, 91st Cong., 1st Sess., §1 (1969). GEICO and Safeco argue that since a “gross negligence” standard is effectively the same as a “reckless disregard” standard, the original bill’s “willfulness” standard must have meant a level of culpa bility higher than “reckless disregard,” or there would have been no requirement to show a different state of mind as a condition of the potentially much greater liabil ity; thus, “willfully fails to comply” must have referred to a knowing violation. Although the gross negligence stan dard was reduced later in the legislative process to simple negligence (as it now appears in §1681o), the provision for willful liability remains unchanged and so must require knowing action, just as it did originally in the draft of §1681n. Perhaps. But Congress may have scaled the standard for actual damages down to simple negligence because it thought gross negligence, being like reckless action, was covered by willfulness. Because this alternative reading is —————— (contrasting the different uses of the term “recklessness” in civil and criminal contexts). Cite as: 551 U. S. ____ (2007) 9 Opinion of the Court possible, any inference from the drafting sequence is shaky, and certainly no match for the following clue in the text as finally adopted, which points to the traditional understanding of willfulness in the civil sphere. The phrase in question appears in the preamble sen tence of §1681n(a): “Any person who willfully fails to comply with any requirement imposed under this sub chapter with respect to any consumer is liable to that consumer . . . .” Then come the details, in paragraphs (1)(A) and (1)(B), spelling out two distinct measures of damages chargeable against the willful violator. As a general matter, the consumer may get either actual dam ages or “damages of not less than $100 and not more than $1,000.” §1681n(a)(1)(A). But where the offender is liable “for obtaining a consumer report under false pretenses or knowingly without a permissible purpose,” the statute sets liability higher: “actual damages . . . or $1,000, whichever is greater.” §1681n(a)(1)(B). If the companies were right that “willfully” limits liabil ity under §1681n(a) to knowing violations, the modifier “knowingly” in §1681n(a)(1)(B) would be superfluous and incongruous; it would have made no sense for Congress to condition the higher damages under §1681n(a) on know ingly obtaining a report without a permissible purpose if the general threshold of any liability under the section were knowing misconduct. If, on the other hand, “will fully” covers both knowing and reckless disregard of the law, knowing violations are sensibly understood as a more serious subcategory of willful ones, and both the preamble and the subsection have distinct jobs to do. See United States v. Menasche, 348 U. S. 528, 538–539 (1955) (“ ‘[G]ive effect, if possible, to every clause and word of a statute’ ” (quoting Montclair v. Ramsdell, 107 U. S. 147, 152 (1883))). The companies make other textual and structural ar guments for their view, but none is persuasive. Safeco 10 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court thinks our reading would lead to the absurd result that one could, with reckless disregard, knowingly obtain a consumer report without a permissible purpose. But this is not so; action falling within the knowing subcategory does not simultaneously fall within the reckless alterna tive. Then both GEICO and Safeco argue that the refer ence to acting “knowingly and willfully” in FCRA’s crimi nal enforcement provisions, §1681q and §1681r, indicates that “willfully” cannot include recklessness. But we are now on the criminal side of the law, where the paired modifiers are often found, see, e.g., 18 U. S. C. §1001 (2000 ed. and Supp. IV) (false statements to federal investiga tors); 20 U. S. C. §1097(a) (embezzlement of student loan funds); 18 U. S. C. §1542 (2000 ed. and Supp. IV) (false statements in a passport application). As we said before, in the criminal law “willfully” typically narrows the oth erwise sufficient intent, making the government prove something extra, in contrast to its civil-law usage, giving a plaintiff a choice of mental states to show in making a case for liability, see n. 9, supra. The vocabulary of the crimi nal side of FCRA is consequently beside the point in con struing the civil side. III A Before getting to the claims that the companies acted recklessly, we have the antecedent question whether either company violated the adverse action notice re quirement at all. In both cases, respondent-plaintiffs’ claims are premised on initial rates charged for new in surance policies, which are not “adverse” actions unless quoting or charging a first-time premium is “an increase in any charge for . . . any insurance, existing or applied for.” 15 U. S. C. §1681a(k)(1)(B)(i). In Safeco’s case, the District Court held that the initial rate for a new insurance policy cannot be an “increase” Cite as: 551 U. S. ____ (2007) 11 Opinion of the Court because there is no prior dealing. The phrase “increase in any charge for . . . insurance” is readily understood to mean a change in treatment for an insured, which as sumes a previous charge for comparison. See Webster’s New International Dictionary 1260 (2d ed. 1957) (defining “increase” as “[a]ddition or enlargement in size, extent, quantity, number, intensity, value, substance, etc.; aug mentation; growth; multiplication”). Since the District Court understood “increase” to speak of change just as much as of comparative size or quantity, it reasoned that the statute’s “increase” never touches the initial rate offer, where there is no change. The Government takes the part of the Court of Appeals in construing “increase” to reach a first-time rate. It says that regular usage of the term is not as narrow as the District Court thought: the point from which to measure difference can just as easily be understood without refer ring to prior individual dealing. The Government gives the example of a gas station owner who charges more than the posted price for gas to customers he doesn’t like; it makes sense to say that the owner increases the price and that the driver pays an increased price, even if he never pulled in there for gas before. See Brief for United States as Amicus Curiae 26.10 The Government implies, then, that reading “increase” requires a choice, and the chosen reading should be the broad one in order to conform to what Congress had in mind. We think the Government’s reading has the better fit with the ambitious objective set out in the Act’s statement —————— 10 Since the posted price seems to be addressed to the world in gen eral, one could argue that the increased gas price is not the initial quote. But the same usage point can be made with the example of the clothing model who gets a call from a ritzy store after posing for a discount retailer. If she quotes a higher fee, it would be natural to say that the uptown store will have to pay the “increase” to have her in its ad. 12 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court of purpose, which uses expansive terms to describe the adverse effects of unfair and inaccurate credit reporting and the responsibilities of consumer reporting agencies. See §1681(a) (inaccurate reports “directly impair the efficiency of the banking system”; unfair reporting meth ods undermine public confidence “essential to the contin ued functioning of the banking system”; need to “insure” that reporting agencies “exercise their grave responsibili ties” fairly, impartially, and with respect for privacy). The descriptions of systemic problem and systemic need as Congress saw them do nothing to suggest that remedies for consumers placed at a disadvantage by unsound credit ratings should be denied to first-time victims, and the legislative histories of FCRA’s original enactment and of the 1996 amendment reveal no reason to confine attention to customers and businesses with prior dealings. Quite the contrary.11 Finally, there is nothing about insurance contracts to suggest that Congress might have meant to differentiate applicants from existing customers when it set the notice requirement; the newly insured who gets charged more owing to an erroneous report is in the same boat with the renewal applicant.12 We therefore hold that —————— 11 See S. Rep. No. 91–517, p. 7 (1969) (“Those who . . . charge a higher rate for credit or insurance wholly or partly because of a consumer report must, upon written request, so advise the consumer . . .”); S. Rep. No. 103–209, p. 4 (1993) (adverse action notice is required “any time the permissible use of a report results in an outcome adverse to the interests of the consumer”); H. R. Rep. No. 103–486, p. 26 (1994) (“[W]henever a consumer report is obtained for a permissible purpose . . . , any action taken based on that report that is adverse to the interests of the consumer triggers the adverse action notice requirements”). 12 In fact, notice in the context of an initially offered rate may be of greater significance than notice in the context of a renewal rate; if, for instance, insurance is offered on the basis of a single, long-term guar anteed rate, a consumer who is not given notice during the initial application process may never have an opportunity to learn of any adverse treatment. Cite as: 551 U. S. ____ (2007) 13 Opinion of the Court the “increase” required for “adverse action,” 15 U. S. C. §1681a(k)(1)(B)(i), speaks to a disadvantageous rate even with no prior dealing; the term reaches initial rates for new applicants. B Although offering the initial rate for new insurance can be an “adverse action,” respondent-plaintiffs have another hurdle to clear, for §1681m(a) calls for notice only when the adverse action is “based in whole or in part on” a credit report. GEICO argues that in order to have adverse ac tion “based on” a credit report, consideration of the report must be a necessary condition for the increased rate. The Government and respondent-plaintiffs do not explicitly take a position on this point. To the extent there is any disagreement on the issue, we accept GEICO’s reading. In common talk, the phrase “based on” indicates a but-for causal relationship and thus a necessary logical condition. Under this most natural reading of §1681m(a), then, an increased rate is not “based in whole or in part on” the credit report unless the report was a necessary condition of the increase. As before, there are textual arguments pointing another way. The statute speaks in terms of basing the action “in part” as well as wholly on the credit report, and this phrasing could mean that adverse action is “based on” a credit report whenever the report was considered in the rate-setting process, even without being a necessary condi tion for the rate increase. But there are good reasons to think Congress preferred GEICO’s necessary-condition reading. If the statute has any claim to lucidity, not all “adverse actions” require notice, only those “based . . . on” informa tion in a credit report. Since the statute does not explicitly call for notice when a business acts adversely merely after consulting a report, conditioning the requirement on 14 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court action “based . . . on” a report suggests that the duty to report arises from some practical consequence of reading the report, not merely some subsequent adverse occur rence that would have happened anyway. If the credit report has no identifiable effect on the rate, the consumer has no immediately practical reason to worry about it (unless he has the power to change every other fact that stands between himself and the best possible deal); both the company and the consumer are just where they would have been if the company had never seen the report.13 And if examining reports that make no difference was supposed to trigger a reporting requirement, it would be hard to find any practical point in imposing the “based . . . on” restriction. So it makes more sense to suspect that Congress meant to require notice and prompt a challenge by the consumer only when the consumer would gain something if the challenge succeeded.14 —————— 13 For instance, if a consumer’s driving record is so poor that no in surer would give him anything but the highest possible rate regardless of his credit report, whether or not an insurer happened to look at his credit report should have no bearing on whether the consumer must receive notice, since he has not been treated differently as a result of it. 14 The history of the Act provides further support for this reading. The originally enacted version of the notice requirement stated: “When ever . . . the charge for . . . insurance is increased either wholly or partly because of information contained in a consumer report . . . , the user of the consumer report shall so advise the consumer . . . .” 15 U. S. C. §1681m(a) (1976 ed.). The “because of” language in the original statute emphasized that the consumer report must actually have caused the adverse action for the notice requirement to apply. When Congress amended FCRA in 1996, it sought to define “adverse action” with greater particularity, and thus split the notice provision into two separate subsections. See 110 Stat. 3009–426 to 3009–427, 3009–443 to 3009–444. In the revised version of §1681m(a), the original “because of” phrasing changed to “based on,” but there was no indication that this change was meant to be a substantive alteration of the statute’s scope. Cite as: 551 U. S. ____ (2007) 15 Opinion of the Court C To sum up, the difference required for an increase can be understood without reference to prior dealing (allowing a first-time applicant to sue), and considering the credit report must be a necessary condition for the difference. The remaining step in determining a duty to notify in cases like these is identifying the benchmark for determin ing whether a first-time rate is a disadvantageous in crease. And in dealing with this issue, the pragmatic reading of “based . . . on” as a condition necessary to make a practical difference carries a helpful suggestion. The Government and respondent-plaintiffs argue that the baseline should be the rate that the applicant would have received with the best possible credit score, while GEICO contends it is what the applicant would have had if the company had not taken his credit score into account (the “neutral score” rate GEICO used in Edo’s case). We think GEICO has the better position, primarily because its “increase” baseline is more comfortable with the under standing of causation just discussed, which requires notice under §1681m(a) only when the effect of the credit report on the initial rate offered is necessary to put the consumer in a worse position than other relevant facts would have decreed anyway. If Congress was this concerned with practical consequences when it adopted a “based . . . on” causation standard, it presumably thought in equally practical terms when it spoke of an “increase” that must be defined by a baseline to measure from. Congress was therefore more likely concerned with the practical ques tion whether the consumer’s rate actually suffered when the company took his credit report into account than the theoretical question whether the consumer would have gotten a better rate with perfect credit.15 —————— 15 While it might seem odd, under the current statutory structure, to interpret the definition of “adverse action” (in §1681a(k)(1)(B)(i)) in 16 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court The Government objects that this reading leaves a loophole, since it keeps first-time applicants who actually deserve better-than-neutral credit scores from getting notice, even when errors in credit reports saddle them with unfair rates. This is true; the neutral-score baseline will leave some consumers without a notice that might lead to discovering errors. But we do not know how often these cases will occur, whereas we see a more demonstra ble and serious disadvantage inhering in the Govern ment’s position. Since the best rates (the Government’s preferred base line) presumably go only to a minority of consumers, adopting the Government’s view would require insurers to send slews of adverse action notices; every young appli cant who had yet to establish a gilt-edged credit report, for example, would get a notice that his charge had been “increased” based on his credit report. We think that the consequence of sending out notices on this scale would undercut the obvious policy behind the notice require ment, for notices as common as these would take on the character of formalities, and formalities tend to be ig nored. It would get around that new insurance usually —————— conjunction with §1681m(a), which simply applies the notice require ment to a particular subset of “adverse actions,” there are strong indica tions that Congress intended these provisions to be construed in tan dem. When FCRA was initially enacted, the link between the definition of “adverse action” and the notice requirement was clear, since “adverse action” was defined within §1681m(a). See 15 U. S. C. §1681m(a) (1976 ed.). Though Congress eventually split the provision into two parts (with the definition of “adverse action” now located at §1681a(k)(1)(B)(i)), the legislative history suggests that this change was not meant to alter Congress’s intent to define “adverse action” in light of the notice re quirement. See S. Rep. No. 103–209, at 4 (“The Committee bill . . . defines an ‘adverse action’ as any action that is adverse to the interests of the consumer and is based in whole or in part on a consumer report”); H. R. Rep. No. 103–486, at 26 (“[A]ny action based on [a consumer] report that is adverse to the interests of the consumer triggers the adverse action notice requirements”). Cite as: 551 U. S. ____ (2007) 17 Opinion of the Court comes with an adverse action notice, owing to some legal quirk, and instead of piquing an applicant’s interest about the accuracy of his credit record, the commonplace notices would mean just about nothing and go the way of junk mail. Assuming that Congress meant a notice of adverse action to get some attention, we think the cost of closing the loophole would be too high. While on the subject of hypernotification, we should add a word on another point of practical significance. Al though the rate initially offered for new insurance is an “increase” calling for notice if it exceeds the neutral rate, did Congress intend the same baseline to apply if the quoted rate remains the same over a course of dealing, being repeated at each renewal date? We cannot believe so. Once a consumer has learned that his credit report led the insurer to charge more, he has no need to be told over again with each renewal if his rate has not changed. For that matter, any other con struction would probably stretch the word “increase” more than it could bear. Once the gas station owner had charged the customer the above-market price, it would be strange to speak of the same price as an increase every time the customer pulled in. Once buyer and seller have begun a course of dealing, customary usage does demand a change for “increase” to make sense.16 Thus, after initial dealing between the consumer and the insurer, the base line for “increase” is the previous rate or charge, not the “neutral” baseline that applies at the start. —————— 16 Consider, too, a consumer who, at the initial application stage, had a perfect credit score and thus obtained the best insurance rate, but, at the renewal stage, was charged at a higher rate (but still lower than the rate he would have received had his credit report not been taken into account) solely because his credit score fell during the interim. Although the consumer clearly suffered an “increase” in his insurance rate that was “based on” his credit score, he would not be entitled to an adverse action notice under the baseline used for initial applications. 18 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court IV A In GEICO’s case, the initial rate offered to Edo was the one he would have received if his credit score had not been taken into account, and GEICO owed him no adverse action notice under §1681m(a).17 B Safeco did not give Burr and Massey any notice because it thought §1681m(a) did not apply to initial applications, a mistake that left the company in violation of the statute if Burr and Massey received higher rates “based in whole or in part” on their credit reports; if they did, Safeco would be liable to them on a showing of reckless conduct (or worse). The first issue we can forget, however, for al though the record does not reliably indicate what rates they would have obtained if their credit reports had not been considered, it is clear enough that if Safeco did vio late the statute, the company was not reckless in falling down in its duty. While “the term recklessness is not self-defining,” the common law has generally understood it in the sphere of civil liability as conduct violating an objective standard: —————— 17 We reject Edo’s alternative argument that GEICO’s offer of a stan dard insurance policy with GEICO Indemnity was an “adverse action” requiring notice because it amounted to a “denial” of insurance through a lower cost, “preferred” policy with GEICO General. See §1681a(k)(1)(B)(i) (defining “adverse action” to include a “denial . . . of . . . insurance”). An applicant calling GEICO for insurance talks with a sales representative who acts for all the GEICO companies. The record has no indication that GEICO tells applicants about its corporate structure, or that applicants request insurance from one of the several companies or even know of their separate existence. The salesperson takes information from the applicant and obtains his credit score, then either denies any insurance or assigns him to one of the companies willing to provide it; the other companies receive no application and take no separate action. This way of accepting new business is clearly outside the natural meaning of “denial” of insurance. Cite as: 551 U. S. ____ (2007) 19 Opinion of the Court action entailing “an unjustifiably high risk of harm that is either known or so obvious that it should be known.”18 Farmer v. Brennan, 511 U. S. 825, 836 (1994); see Prosser and Keeton §34, at 213–214. The Restatement, for exam ple, defines reckless disregard of a person’s physical safety this way: “The actor’s conduct is in reckless disregard of the safety of another if he does an act or intentionally fails to do an act which it is his duty to the other to do, knowing or having reason to know of facts which would lead a reasonable man to realize, not only that his conduct creates an unreasonable risk of physical harm to another, but also that such risk is substan tially greater than that which is necessary to make his conduct negligent.” Restatement (Second) of Torts §500, p. 587 (1963–1964). It is this high risk of harm, objectively assessed, that is the essence of recklessness at common law. See Prosser and Keeton §34, at 213 (recklessness requires “a known or obvious risk that was so great as to make it highly prob able that harm would follow”). There being no indication that Congress had something different in mind, we have no reason to deviate from the common law understanding in applying the statute. See Prupis, 529 U. S., at 500–501. Thus, a company subject to FCRA does not act in reckless disregard of it unless the action is not only a violation under a reasonable reading of the statute’s terms, but shows that the company ran a risk of violating the law substantially greater than the risk associated with a reading that was merely careless. Here, there is no need to pinpoint the negli gence/recklessness line, for Safeco’s reading of the statute, —————— 18 Unlike civil recklessness, criminal recklessness also requires sub jective knowledge on the part of the offender. Brennan, 511 U. S., at 836–837; ALI, Model Penal Code §2.02(2)(c) (1985). 20 SAFECO INS. CO. OF AMERICA v. BURR Opinion of the Court albeit erroneous, was not objectively unreasonable. As we said, §1681a(k)(1)(B)(i) is silent on the point from which to measure “increase.” On the rationale that “increase” presupposes prior dealing, Safeco took the definition as excluding initial rate offers for new insurance, and so sent no adverse action notices to Burr and Massey. While we disagree with Safeco’s analysis, we recognize that its reading has a foundation in the statutory text, see supra, at 11, and a sufficiently convincing justification to have persuaded the District Court to adopt it and rule in Safeco’s favor. This is not a case in which the business subject to the Act had the benefit of guidance from the courts of appeals or the Federal Trade Commission (FTC) that might have warned it away from the view it took. Before these cases, no court of appeals had spoken on the issue, and no au thoritative guidance has yet come from the FTC19 (which in any case has only enforcement responsibility, not sub stantive rulemaking authority, for the provisions in ques tion, see 15 U. S. C. §§1681s(a)(1), (e)). Cf. Saucier v. Katz, 533 U. S. 194, 202 (2001) (assessing, for qualified immunity purposes, whether an action was reasonable in light of legal rules that were “clearly established” at the time). Given this dearth of guidance and the less-than pellucid statutory text, Safeco’s reading was not objec tively unreasonable, and so falls well short of raising the “unjustifiably high risk” of violating the statute necessary —————— 19 Respondent-plaintiffs point to a letter, written by an FTC staff member to an insurance company lawyer, that suggests that an “ad verse action” occurs when “the applicant will have to pay more for insurance at the inception of the policy than he or she would have been charged if the consumer report had been more favorable.” Letter from Hannah A. Stires to James M. Ball (Mar. 1, 2000), http://www.ftc.gov/os/statutes/fcra/ball.htm (as visited May 17, 2007, and available in Clerk of Court’s case file). But the letter did not canvas the issue, and it explicitly indicated that it was merely “an informal staff opinion . . . not binding on the Commission.” Ibid. Cite as: 551 U. S. ____ (2007) 21 Opinion of the Court for reckless liability.20 * * * The Court of Appeals correctly held that reckless disre gard of a requirement of FCRA would qualify as a willful violation within the meaning of §1681n(a). But there was no need for that court to remand the cases for factual development. GEICO’s decision to issue no adverse action notice to Edo was not a violation of §1681m(a), and Safeco’s misreading of the statute was not reckless. The judgments of the Court of Appeals are therefore reversed in both cases, which are remanded for further proceedings consistent with this opinion. It is so ordered. —————— 20 Respondent-plaintiffs argue that evidence of subjective bad faith must be taken into account in determining whether a company acted knowingly or recklessly for purposes of §1681n(a). To the extent that they argue that evidence of subjective bad faith can support a willful ness finding even when the company’s reading of the statute is objec tively reasonable, their argument is unsound. Where, as here, the statutory text and relevant court and agency guidance allow for more than one reasonable interpretation, it would defy history and current thinking to treat a defendant who merely adopts one such interpreta tion as a knowing or reckless violator. Congress could not have in tended such a result for those who followed an interpretation that could reasonably have found support in the courts, whatever their subjective intent may have been. Both Safeco and GEICO argue that good-faith reliance on legal advice should render companies immune to claims raised under §1681n(a). While we do not foreclose this possibility, we need not address the issue here in light of our present holdings. Cite as: 551 U. S. ____ (2007) 1 Opinion of STEVENS, J. SUPREME COURT OF THE UNITED STATES _________________ Nos. 06–84 and 06–100 _________________ SAFECO INSURANCE COMPANY OF AMERICA, ET AL., PETITIONERS 06–84 v. CHARLES BURR ET AL. GEICO GENERAL INSURANCE COMPANY, ET AL., PETITIONERS 06–100 v. AJENE EDO ON WRITS OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT [June 4, 2007] JUSTICE STEVENS, with whom JUSTICE GINSBURG joins, concurring in part and concurring in the judgment. While I join the Court’s judgment and Parts I, II, III–A, and IV–B of the Court’s opinion, I disagree with the rea soning in Parts III–B and III–C, as well as with Part IV– A, which relies on that reasoning. An adverse action taken after reviewing a credit report “is based in whole or in part on” that report within the meaning of 15 U. S. C. §1681m(a). That is true even if the company would have made the same decision without looking at the report, because what the company actually did is more relevant than what it might have done. I find nothing in the statute making the examination of a credit report a “necessary condition” of any resulting increase. Ante, at 13. The more natural reading is that reviewing a report is only a sufficient condition. The Court’s contrary position leads to a serious anom 2 SAFECO INS. CO. OF AMERICA v. BURR Opinion of STEVENS, J. aly. As a matter of federal law, companies are free to adopt whatever “neutral” credit scores they want. That score need not (and probably will not) reflect the median consumer credit score. More likely, it will reflect a com pany’s assessment of the creditworthiness of a run-of-the mill applicant who lacks a credit report. Because those who have yet to develop a credit history are unlikely to be good credit risks, “neutral” credit scores will in many cases be quite low. Yet under the Court’s reasoning, only those consumers with credit scores even lower than what may already be a very low “neutral” score will ever receive adverse action notices.1 While the Court acknowledges that “the neutral-score baseline will leave some consumers without a notice that might lead to discovering errors,” ante, at 16, it finds this unobjectionable because Congress was likely uninterested in “the theoretical question of whether the consumer would have gotten a better rate with perfect credit.” Ante, at 16.2 The Court’s decision, however, disserves not only those consumers with “gilt-edged credit report[s],” ante, at 16, but also the much larger category of consumers with better-than-“neutral” scores. I find it difficult to believe —————— 1 Stranger still, companies that automatically disqualify consumers who lack credit reports will never need to send any adverse action notices. After all, the Court’s baseline is “what the applicant would have had if the company had not taken his credit score into account,” ante, at 15, but from such companies, what the applicant “would have had” is no insurance at all. An offer of insurance at any price, however inflated by a poor and perhaps incorrect credit score, will therefore never constitute an adverse action. 2 The Court also justifies its deviation from the statute’s text by rea soning that frequent adverse action notices would be ignored. See ante, at 16–17. To borrow a sentence from the Court’s opinion: “Perhaps.” Ante, at 8. But rather than speculate about the likely effect of “hyper notification,” ante, at 17, I would defer to the Solicitor General’s posi tion, informed by the Federal Trade Commission’s expert judgment, that consumers by and large benefit from adverse action notices, however common. See Brief for United States as Amicus Curiae 27–29. Cite as: 551 U. S. ____ (2007) 3 Opinion of STEVENS, J. that Congress could have intended for a company’s unre strained adoption of a “neutral” score to keep many (if not most) consumers from ever hearing that their credit re ports are costing them money. In my view, the statute’s text is amenable to a more sensible interpretation. Cite as: 551 U. S. ____ (2007) 1 THOMAS, J., concurring in part SUPREME COURT OF THE UNITED STATES _________________ Nos. 06–84 and 06–100 _________________ SAFECO INSURANCE COMPANY OF AMERICA, ET AL., PETITIONERS 06–84 v. CHARLES BURR ET AL. GEICO GENERAL INSURANCE COMPANY, ET AL., PETITIONERS 06–100 v. AJENE EDO ON WRITS OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT [June 4, 2007] JUSTICE THOMAS, with whom JUSTICE ALITO joins, concurring in part. I agree with the Court’s disposition and most of its reasoning. Safeco did not send notices to new customers because it took the position that the initial insurance rate it offered a customer could not be an “increase in any charge for . . . insurance” under 15 U. S. C. §1681a(k)(1)(B)(i). The Court properly holds that regard less of the merits of this interpretation, it is not an unrea sonable one, and Safeco therefore did not act willfully. Ante, at 18–21. I do not join Part III–A of the Court’s opinion, however, because it resolves the merits of Safeco’s interpretation of §1681a(k)(1)(B)(i)—an issue not neces sary to the Court’s conclusion and not briefed or argued by the parties.