Under the Fair Credit Reporting Act ("FCRA"), insurance companies are required to send adverse action notices to consumers whenever they increase the rates for insurance on the basis of information contained in consumer credit reports. 15 U.S.C. §§ 1681a(k)(1)(B)(i), 1681m(a).1 The principal question before us is straightforward: Does FCRA's adverse action notice requirement apply to the rate first charged in an initial policy of insurance? We hold that the answer is yes: The Act requires that an insurance company send the consumer an adverse action notice whenever a higher rate is charged because of credit information it obtains, regardless of whether the rate is contained in an initial policy or an extension or renewal of a policy and regardless of whether the company has previously charged the consumer a lower rate.

We also resolve five ancillary questions. First, we hold that FCRA's adverse action notice requirement applies whenever a consumer would have received a lower rate for insurance had his credit information been more favorable, regardless of whether his credit rating is above or below average. Specifically, the requirement covers those whose credit information is disregarded and replaced for purposes of a rate computation by an average or neutral credit figure, so long as the insurance rates would have been lower had the credit information been more favorable. Second, we hold that charging more for insurance on the basis of a transmission stating that no credit information or insufficient credit information is available constitutes an adverse action based on information in a consumer report and therefore requires the giving of notice under FCRA. Third, we hold that, to comply with FCRA's notice requirement, a company must, inter alia, communicate to the consumer that an adverse action based on a consumer report was taken, describe the action, specify the effect of the action upon the consumer, and identify the party or parties taking the action. Fourth, we hold that when a consumer applies for insurance with a family of companies and is charged a higher rate for insurance because of his credit report, two or more companies within that family may be jointly and severally liable. The notice requirement applies to any company that makes a decision that a higher rate shall be imposed, issues a policy at a higher rate, or refuses to provide a policy at a lower rate, if the company's action is based in whole or in part on the consumer's credit information.2 Finally, we adopt the Third Circuit's definition of "willfully," as that term is employed in FCRA, and hold that a company is liable for a willful violation of FCRA if it "knowingly and intentionally committed an act in conscious disregard for the rights of others." Cushman v. Trans Union Corp., 115 F.3d 220, 226 (3d Cir.1997) (quoting Philbin v. Trans Union Corp., 101 F.3d 957, 970 (3d Cir.1996) (as amended)). Like the Third Circuit, we hold that conscious disregard means "either knowing that policy to be in contravention of the rights possessed by consumers pursuant to the FCRA or in reckless disregard of whether the policy contravened those rights." Id. at 227.

The Fair Credit Reporting Act seeks to ensure the "[a]ccuracy and fairness of credit reporting" through a variety of means. 15 U.S.C. § 1681. Central to this goal, FCRA limits the persons who may obtain consumer credit reports and requires users of such reports to notify consumers when, in reliance on a consumer report, "adverse action" has been taken. 15 U.S.C. §§ 1681a, 1681b, 1681m. Specifically, § 1681m(a) provides: "If a person 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," the person shall provide "notice of the adverse action to the consumer." Adverse action notices advise consumers that an adverse action has been taken against them, and the nature of that action, and alerts them that they may view a copy of the consumer report that triggered the adverse action free of charge and correct any errors affecting their economic well-being. Even if reports are free from error, adverse action notices give consumers important information about how improved credit information may benefit them and how they can avoid receiving unfavorable credit ratings in the future.

To resolve the various issues that have arisen regarding FCRA's notice of adverse action requirement in a set of related cases, we have consolidated two appeals for purposes of this opinion: Reynolds v. Hartford Financial Services Group, Inc., 2005 WL 2714503 and Edo v. GEICO Casualty co., 2004 WL 2731226. Reynolds presents the principal issue: May a rate first charged in an initial policy of insurance constitute an increased rate for purposes of the FCRA adverse action notice requirement? Hartford Fire asserts that a rate cannot qualify as increased unless a lower rate has previously been charged to the customer. Reynolds also presents the issues whether a communication stating that no credit information or insufficient credit information is available constitutes a "consumer report" under the statute and whether an adverse action notice that does not tell the consumer that an adverse action has been taken against him, describe that action and its effect upon the consumer, and identify the parties taking the action is sufficient under FCRA. Edo presents the issue whether an adverse action occurs whenever a consumer would have received a lower rate if his credit information had been more favorable; or whether an insurance company's practice of providing an adverse action notice only if the consumer's credit information is below average (or "neutral") and that factor results in the imposition of a higher rate than if his credit rating had been average, is consistent with FCRA. Both Edo and Reynolds also require us to decide which companies are liable under FCRA for the failure to give notice of an increased rate when several affiliated companies are involved in the process of rate-setting and policy issuance. Finally, defendants in both cases seek summary judgment on the alternative ground that their actions were not willful as a matter of law. To address this last contention, we must define the meaning of the term "willfully" as it applies in FCRA.

A. Reynolds v. Hartford Financial Services Group, Inc.

Jason Reynolds is the sole remaining named-plaintiff in this class action against Hartford Fire Insurance Company ("Hartford Fire").3 He seeks statutory and punitive damages, as well as reasonable attorneys fees for the company's violation of FCRA's adverse action notice requirement. Reynolds' claims relate to two insurance policies he obtained, one for automobile and the other for homeowners insurance. On the record before us, Hartford Fire set the rates to be charged for both policies. Hartford Property and Casualty Insurance Company of Hartford ("PCIC Hartford") issued Reynolds the homeowners insurance policy and Hartford Insurance Company of the Midwest ("Hartford Midwest") issued him the automobile insurance policy. We refer to Hartford Fire, Hartford PCIC, and Hartford Midwest as the "Hartford Companies."

Reynolds originally sued Hartford Fire and later sought to amend his complaint to add PCIC Hartford and Hartford Midwest.4 Hartford Fire sought summary judgment, which the district court granted on two grounds. First, it held that "the entity contracting with the policyholder is the only possible statutory taker of adverse action because only the contracting entity is capable of increasing the premium for or changing the terms of the insurance contract with the insured." Second, and in the alternative, it held that an insurance company that issues a policy to a new policy-holder "cannot [be held to] `increase' a charge for insurance unless the insurer makes an initial demand for payment of the insured and subsequently increases the amount of that demand based on information in the insured's credit report." The second and alternative holding relies on a previous decision by the same court, Mark v. Valley Insurance Co., 275 F. Supp. 2d 1307, 1317 (D.Or.2003). On the basis of that earlier decision, the district court also denied Reynolds leave to amend, reasoning that he could not "state viable FCRA claims against the proposed defendants," PCIC Hartford and Hartford Midwest. In other words, leave was denied on the ground that the policies were initial issues and no previous charge had been made to the customer at a lower rate.

During the relevant time period, Hartford Fire and the American Association of Retired Persons ("AARP") had an agreement under which Hartford Fire or one of its subsidiaries would issue automobile and homeowners insurance to AARP members at a premium rate if those individuals enjoyed favorable credit ratings. While the procedures used for issuing the two kinds of insurance varied slightly, they were the same in most relevant respects. In both cases, employees of Hartford Fire would make all of the decisions concerning AARP members' insurance policies for all of its subsidiaries, including Hartford Midwest, which issued automobile insurance, and PCIC Hartford, which issued homeowners insurance. In doing so, Hartford Fire's employees would obtain credit information from Trans Union, a consumer information bureau, through a contract to which Hartford Fire and Trans Union were signatories. This information would be conveyed to Hartford Fire through the risk assessment and data supply firm, ChoicePoint, in the form of an "insurance score." High insurance scores correlated with more favorable credit reports. With regard to automobile insurance, if an AARP member had a high enough insurance score, he would qualify for a ten percent discount. With regard to homeowners insurance, only if the member obtained a top insurance score could he be assigned to the top tier of insurance with the best rate.

If, when Hartford Fire sent a request for an insurance score, no credit information matched the name and address of the consumer or if the information that did match was insufficient to generate an insurance score, this information would be transmitted to the company, and the consumer would be labeled a "no hit" or a "no score" and would not be assigned an insurance score. Without an insurance score, the consumer could not qualify for the ten percent discount with Hartford Midwest, nor could he be placed in the top insurance tier with PCIC Hartford. As a result, a "no hit" or "no score" consumer would in numerous instances pay more for insurance than if he had received a high insurance score.5

Hartford Fire is the only one of the Hartford Companies to have developed or sent adverse action notices. The parties dispute whether Hartford Fire actually sent an adverse action notice to Reynolds, but that is a question of fact for the fact-finder. Whether the notice Hartford Fire contends it sent was adequate under FCRA is a question of law that we discuss below.

Reynolds applied for both automobile and homeowners insurance by contacting the Hartford Companies. He had no existing policy with that group. An employee of Hartford Fire collected personal information and attempted to obtain Reynold's insurance score twice, once for each insurance application. The credit bureau reported both times that Reynolds was a "no hit." See n. 5, supra. Although Hartford Midwest issued him an automobile insurance policy and Hartford PCIC issued him a homeowners insurance policy, as a result of his "no hit" status Reynolds did not receive either of the AARP premium rates.

B. Edo v. GEICO Casualty Co.

The second of the consolidated cases relates to an automobile insurance policy obtained by Ajene Edo. Like Reynolds, Edo seeks statutory and punitive damages, as well as reasonable attorney fees, on behalf of a class of consumers for violation of FCRA's adverse action notice requirement. He, too, is the sole remaining named-plaintiff. Edo appeals the district court's grant of summary judgment to defendants Government Employees Insurance Company ("Government Employees"), GEICO General Insurance Company ("GEICO General"), and GEICO Indemnity Corporation ("GEICO Indemnity").6 These are affiliated companies, all of which are subsidiaries of the GEICO Corporation and are referred to collectively by the parties as the "GEICO Companies." We sometimes refer to that group of companies by that designation and sometimes simply as GEICO.

Unlike Hartford Fire, the GEICO Companies concede that adverse actions can occur with respect to the first rates charged in an initial policy of insurance. They do not assert that in order for an adverse action to occur there must be an increase to a rate that the consumer has previously been charged. Nevertheless, the district court granted summary judgment with respect to the various GEICO entities on a number of different grounds. First, the court held that Edo did not have standing to bring a FCRA claim against Government Employees because he "was not eligible for insurance coverage from [that company] regardless of his consumer credit score because Government Employees offers insurance coverage only to government employees or military personnel." Next, it granted summary judgment in favor of GEICO General because that company "did not contract with Plaintiff to issue or to underwrite an insurance policy." This ruling was in accord with the district court's previous holdings in other related cases that only the company that issues the insurance policy can be held to have taken an adverse action under FCRA. See Ashby v. Farmers Group, Inc., 261 F. Supp. 2d 1213, 1222 (D.Or.2003); Razilov v. Nationwide Mut. Ins. Co., 242 F. Supp. 2d 977, 989-90 (D.Or.2003). Finally, it granted summary judgment to GEICO Indemnity because "the premium charged to [Edo] by GEICO Indemnity would have been the same even if GEICO Indemnity did not consider information in Plaintiff's consumer credit history."

The GEICO Companies are organized by risk. GEICO General provides preferred policies with low rates for those who are lesser insurance risks. Government Employees also provides preferred policies, but only to government employees. GEICO Indemnity issues standard policies with mid-level rates for moderate risk consumers. Finally, GEICO Casualty issues non-standard policies with high rates for those who are greater risks. The GEICO Companies began using consumer credit reports in early 1999.

In order to purchase insurance, consumers call a toll-free number and talk to a GEICO sales counselor. The sales counselor is employed by Government Employees but works on behalf of all of the GEICO Companies. Indeed, all of the work of the GEICO Companies is performed by Government Employees workers, as the other companies do not have any employees. Upon learning that a customer wishes to purchase automobile insurance, the sales counselor elicits basic information and asks whether he may use the customer's credit information when arranging for his policy. If the customer acquiesces, the sales counselor obtains the credit information in the form of an insurance score calculated by the data analysis firm Fair Isaacs from information supplied by Trans Union. Government Employees is the only GEICO company that has a contract with Trans Union and Fair Isaacs to provide this information. Using a Government Employees computer system, the sales counselor converts the insurance score to a credit weight and combines it with other weights assigned to other insurance factors, such as age and number of accidents to arrive at a final total insurance weight. Based on that final weight, the sales counselor assigns the customer to one of the GEICO Companies and determines the appropriate insurance tier. This determination serves to establish the rate the consumer will be charged. After the information the customer has provided has been verified, he is issued an insurance policy at that rate.

The GEICO Companies' original FCRA policy, adopted in 1999, was to send adverse action notices to all consumers whose credit reports were used in making insurance decisions. Later that same year, GEICO changed its policy, at least in part to reduce costs. Instead of sending adverse action notices to everyone, GEICO developed a system for determining which actions it deemed adverse by comparing the rate charged to the rate that it would have charged had the credit information been "neutral."

GEICO's new system, however, did not comply with FCRA's requirements. The GEICO Companies' policy during the period relevant to this case was to compare the consumer's actual company and tier placement (which, as described above, was based in part on his credit rating) with the company and tier to which he would have been assigned had a "neutral" credit weight been substituted for his actual credit weight when calculating the final total insurance weight. The GEICO Companies' "neutral" credit weight was defined, generally speaking, as the weight that reflected the average credit rating of all consumers. The GEICO Companies would calculate two final total insurance weights, using only one variable—the actual credit weight in one case, and the "neutral" credit weight in the other. Only if the final total insurance weight using the "neutral" credit weight would have resulted in the consumer's placement with a different company or in a different tier than that to which the consumer was actually assigned, and only if such different placement would have resulted in the consumer's being charged a lower rate, would GEICO Companies issue an adverse action notice. In other words, the GEICO Companies' policy was to refrain from sending a statutory notice if use of the consumer's actual credit information caused the applicant to be placed with an entity and in a tier that resulted in the charging of the same or a lower rate than the rate that he would have been charged had the calculation and the ensuing assignment been based on a "neutral" or average credit rating. Under this policy, even if the rate ultimately charged was higher than the rate to which the consumer would have been entitled had he had a more favorable credit rating, the statutory notice was not sent if use of the "neutral" and the actual credit data would have led to an assignment to the same entity and tier. Thus, it was not GEICO's policy to send adverse action notices to all consumers who would have been charged lower rates had they enjoyed a more favorable credit rating.

Following Edo's call to GEICO's toll-free number, the sales counselor used the credit information he obtained to place its new customer with GEICO Indemnity. The GEICO Companies then applied its policy for determining whether an adverse action had occurred. GEICO calculated that, had the neutral credit weight been used instead of Edo's actual credit weight, the resulting final total weight would still have resulted in Edo's being placed with GEICO Indemnity. That Edo's placement, and the rate charged for his insurance, did not improve when the "neutral" weight was used is not surprising, as Edo's actual credit weight was better than average. Under its policy, however, the GEICO Companies did not issue him an adverse action notice.

It is uncontested that if the GEICO Companies had used the highest credit weight that a consumer could receive rather than the neutral credit rate to determine Edo's alternate placement, GEICO would have placed Edo with GEICO General, a preferred company, and offered him a lower insurance rate. In short, if Edo's credit information had been more favorable (even though it was already above average), he would have been charged less for his insurance. In 2002, the GEICO Companies changed its policy and began to issue adverse action notices whenever a report with more favorable credit information would have resulted in a lower insurance rate. Under the new policy, Edo would have received the statutory notice.

The principal question in this and a number of related cases7 constitutes a matter of first impression: Does FCRA's adverse action notice requirement apply to the rates first charged in an initial policy of insurance or is it limited to an increase in a rate that the consumer has previously been charged? As with all statutory interpretation, we begin with the text of the statute. See, e.g., Hartford Underwriters Ins. Co. v. Union Planters Bank, N.A., 530 U.S. 1, 6, 120 S. Ct. 1942, 147 L. Ed. 2d 1 (2000). An adverse action with respect to insurance is defined by 15 U.S.C. § 1681a(k)(1)(B)(i) as "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, in connection with the underwriting of insurance."

Specifically, we must decide whether charging a higher price for initial insurance than the insured would otherwise have been charged because of information in a consumer credit report constitutes an "increase in any charge" within the meaning of FCRA. First, we examine the definitions of "increase" and "charge." Hartford Fire contends that, limited to their ordinary definitions, these words apply only when a consumer has previously been charged for insurance and that charge has thereafter been increased by the insurer. The phrase, "has previously been charged," as used by Hartford, refers not only to a rate that the consumer has previously paid for insurance but also to a rate that the consumer has previously been quoted, even if that rate was increased before the consumer made any payment. Reynolds disagrees, asserting that, under the ordinary definition of the term, an increase in a charge also occurs whenever an insurer charges a higher rate than it would otherwise have charged because of any factor—such as adverse credit information, age, or driving record8 — regardless of whether the customer was previously charged some other rate. According to Reynolds, he was charged an increased rate because of his credit rating when he was compelled to pay a rate higher than the premium rate because he failed to obtain a high insurance score. Thus, he argues, the definitions of "increase" and "charge" encompass the insurance companies' practice. Reynolds is correct.

"Increase" means to make something greater. See, e.g., OXFORD ENGLISH DICTIONARY (2d ed. 1989) ("The action, process, or fact of becoming or making greater; augmentation, growth, enlargement, extension."); WEBSTER'S NEW WORLD DICTIONARY OF AMERICAN ENGLISH (3d college ed.1988) (defining "increase" as "growth, enlargement, etc[.]"). "Charge" means the price demanded for goods or services. See, e.g., OXFORD ENGLISH DICTIONARY (2d ed. 1989) ("The price required or demanded for service rendered, or (less usually) for goods supplied."); WEBSTER'S NEW WORLD DICTIONARY OF AMERICAN ENGLISH (3d college ed. 1988) ("[T]he cost or price of an article, service, etc."). Nothing in the definition of these words implies that the term "increase in any charge for" should be limited to cases in which a company raises the rate that an individual has previously been charged.

While no court has considered whether an increase requires a previous charge within the meaning of FCRA, the Sixth Circuit has employed the term "increase" in an analogous circumstance, stating, "An increase in the base price of an automobile that is not charged to a cash customer, but is charged to a credit customer, solely because he is a credit customer, triggers [the Truth in Lending Act's] disclosure requirements." Cornist v. B.J.T. Auto Sales, Inc., 272 F.3d 322, 327 (6th Cir.2001). Defined in this manner, an increased charge is a charge that is higher than it would otherwise have been but for the existence of some factor that causes the insurer to charge a higher price.

Second, the statutory definition of "adverse action," as it is made applicable to insurance, explicitly encompasses "any insurance, existing or applied for." 15 U.S.C. § 1681a(k)(1)(B)(i) (emphasis added). Congress' use of the latter phrase demonstrates its intent that "adverse actions" apply to all insurance transactions—from an initial policy of insurance to a renewal of a long-held policy. The text of the statute does not permit the imposition of any temporal limitation. Hartford has suggested no sensible alternative reading of "existing or applied for." Thus, reading the terms "increase" and "charge" in the context of the provision as a whole, particularly the "existing or applied for" phrase, supports affording them their ordinary meaning.

Third, our interpretation of the terms at issue best comports with the stated purpose of FCRA: to ensure the "[a]ccuracy and fairness of credit reporting." 15 U.S.C. § 1681. FCRA's adverse action notice requirement is an important tool that Congress created, using broad, encompassing language. Through this requirement, Congress sought to promote the rights of consumers by giving them essential information about how their credit report is used, information that they could obtain in no other way. The information Congress mandated serves two important ends. First and foremost, once consumers possess this information they can check and correct any errors in their credit reports. This increases the chances that a consumer's financial stability will not be hampered by faulty credit information. It also improves the overall accuracy of credit reports, which facilitates the operation of our markets. Second, even when credit reports are accurate, informing consumers when their credit rating is hurting them in the marketplace gives them important information about the benefits of improving their credit rating in the future and may even assist them in learning how to do so.

Hartford Fire's contention that FCRA does not apply to the rate charged in initial insurance policies would seriously undermine Congress's clear purpose. The use of credit reports to help determine the rates to be charged for initial insurance policies is common. Moreover it is these policies that the economically unsophisticated are most likely to purchase. Congress did not create such strong protections for consumers only to render them inapplicable in so critical a circumstance. Furthermore, as FCRA is a consumer protection statute, we must construe it so as to further its objectives. Guimond v. Trans Union Credit Info. Co., 45 F.3d 1329, 1333 (9th Cir.1995). While our interpretation is the plain one, this canon supports our result.

We hold that whenever because of his credit information a company charges a consumer a higher initial rate than it would otherwise have charged, it has increased the charge within the meaning of FCRA. Therefore, the fact that Reynolds' policy was an initial one, and his rate was the initial rate charged, is of no consequence. Reynolds' rate was increased above that which it would have otherwise been because of his credit report. As the statute's text is clear, we need not resort to either the agency's interpretations9 or the statute's legislative history. The district court erred in granting summary judgment to Hartford Fire on the ground that FCRA does not apply to the rate first charged in an initial policy.10

The GEICO Companies contend that their method of determining which consumers were entitled to receive adverse action notices comported with FCRA, while Edo asserts that under GEICO's procedure numerous consumers who were charged increased rates because of their credit rating failed to receive the statutorily required notice. At the time Edo sought an initial insurance policy, it was GEICO's practice to send an adverse action notice to a consumer only if the use of his actual credit information resulted in his placement with an entity and tier that provided a higher insurance rate than the entity and tier to which he would have been assigned if "neutral" or average credit information had been used instead. In short, it was GEICO's policy to send adverse action notices only to some of the consumers who would have received more favorable rates had they enjoyed a better credit rating. Specifically, notices were sent only to those with below average credit who would have been charged a lower rate for insurance had they received an average credit rating. GEICO contends that only in such circumstance has an adverse action occurred. GEICO is incorrect.

FCRA does not limit its adverse action notice requirement to actions that result in the customer paying a higher rate than he would otherwise be charged because his credit rating is worse than the average consumer's. Instead, it requires such notices whenever a consumer pays a higher rate because his credit rating is less than the top potential score. In other words, if the consumer would have received a lower rate for his insurance had the information in his consumer report been more favorable, an adverse action has been taken against him.11 Such is the case with Edo. Because Edo would have been placed with GEICO General instead of GEICO Indemnity and thus would have been charged a lower rate if his credit rating had been higher, an adverse action occurred and an adverse action notice was required under FCRA.12 Under the GEICO formula, the fact that the credit rating Edo actually received was higher than the average rating did not mean that Edo would not be charged a higher rate than he would have been charged had he had an even better credit report, but it ensured that he would not receive an adverse action notice when he was charged that increased rate. The district court erred in granting GEICO Indemnity summary judgment on the ground that Edo's rate was not increased on the basis of his credit report.

Hartford Fire makes a separate argument as to why in Reynolds' case no adverse action was taken. Specifically, the company argues that no adverse action was taken against Reynolds "based in whole or in part on any information contained in a consumer report" within the meaning of 15 U.S.C. § 1681m(a). When Hartford Fire requested credit information about Reynolds, Trans Union did not possess the necessary information to generate an insurance score and transmitted this finding to the insurer. Reynolds was therefore considered a "no hit." See n. 5, supra. Because he was so designated, Reynolds was rendered ineligible for the premium rates available to AARP members with qualifying credit ratings, and, as a result, was charged a higher rate in his initial policies. Hartford Fire argues, however, that these were not adverse actions because, it contends, an "adverse action" occurs only if it is based on "information contained in a consumer report" and no such report was received with respect to Reynolds. We reject Hartford Fire's argument.

FCRA's definition of "consumer report" is broad. It unquestionably encompasses a credit reporting agency's communication to an insurance company that a consumer does not have enough information on file for an insurance score to be calculated. Specifically, 15 U.S.C. § 1681a(d)(1) explains that "[t]he term `consumer report' means 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 ...." (emphasis added). Reporting that an agency cannot obtain any information regarding a consumer or that a consumer has insufficient credit information on file conveys a message regarding the consumer's creditworthiness, standing, and capacity that makes his obtaining of credit far more difficult. Such a report suggests that the consumer cannot show that he pays debts in a timely manner. That information may be false: The credit agency may have used the wrong name or searched the wrong records, missing data that would have shown that the applicant is indeed creditworthy. Providing notice therefore serves the statutory purpose of allowing the consumer to correct errors in credit reports. Accordingly, we hold that a communication that a consumer has no information available or an insufficient credit history to permit the calculation of a credit rating qualifies as "a consumer report" within the meaning of FCRA. Because it is uncontested that Reynolds would have been charged lower insurance rates had he received qualifying credit ratings and because the application of the "no hit" rule precluded him from receiving such ratings, we hold that an adverse action was taken against him on the basis of information contained in a credit report. Accordingly, the district court's order of summary judgment may not be affirmed on the ground that its actions with respect to Reynolds were not based on such information.

Hartford Fire also urges us to affirm the district court's grant of summary judgment on the alternative ground that, although (in its view) it was not required under FCRA to send adverse action notices, the notices that the Hartford Companies did send were sufficient to meet its FCRA responsibilities. We reject this argument because the notices were inadequate as a matter of law. Under 15 U.S.C. § 1681m(a)(1), a company that takes adverse action on the basis of a consumer report must "provide oral, written, or electronic notice of the adverse action to the consumer" as well as meet a number of other specific requirements.13 While the term "notice of an adverse action" is not defined in the statute, we hold that, at a minimum, such a notice must communicate to the consumer that an adverse action based on a consumer report was taken, describe the action, specify the effect of the action upon the consumer, and identify the party or parties taking the action.14 See Fischl v. General Motors Acceptance Corp., 708 F.2d 143, 150 (5th Cir.1983) (requiring disclosure of "reliance on data contained in [a consumer's] credit report" when providing notice of an adverse action).

The notices Reynolds received did not comply with any of the above requirements. They did not tell him that any adverse action had been taken against him. They simply stated that "[t]he Hartford's eligibility and pricing decisions are based in part on consumer report(s) from a consumer reporting agency" and allowed him to make a written request in order to find out more. Reynolds was entitled to be informed that his rate for insurance was increased because of information in his credit report. He was also entitled to be told that Hartford Fire made the pricing decision and that Hartford PCIC and Hartford Midwest issued him policies at those higher rates. FCRA recognizes the difference between telling a consumer that his credit information could affect his insurance rate and that it did adversely affect his rate, and requires notice of the latter. We therefore reject Hartford Fire's alternative argument for upholding the district court's order.

The defendants all contend that only one company can be liable when an insurance policy contains an increase in rates—the issuing company. The plain text of the statute, as well as its purposes, are to the contrary. Here, we hold that all of the defendants are potentially liable under the statute.

FCRA requires that "any person" who takes an adverse action is liable. 15 U.S.C. § 1681m(a). The definition of "any" includes the plural. See, e.g., WEBSTER'S NEW WORLD DICTIONARY OF AMERICAN ENGLISH (3d college ed.1988) ("one, a, an, or some; one or more without specification or identification"). With regard to insurance transactions, liability attaches whenever an adverse action is taken "in connection with the underwriting of insurance." 15 U.S.C. § 1681a(k)(1)(B)(i). This broad "in connection with" language confirms that a variety of entities may be liable. No provision in the statute nor comment in the legislative history suggests that Congress intended that only a single company be responsible under FCRA when a consumer is charged an increased rate for insurance. Therefore, the defendants find themselves in the difficult position of persuading us that Congress intended something different from what it wrote. We analyze their three arguments separately.

First, GEICO argues that the words "applied for" in the definition of adverse action, § 1681a(k)(1)(B)(i), demonstrate that only the issuing company is liable under the statute. On that basis, GEICO asks us to hold that Edo "applied for" insurance only with GEICO Indemnity because that was the company that issued him a policy. The argument is frivolous, both factually and legally. As a matter of fact, Edo did not apply to one company but instead requested insurance from the GEICO family of companies. He did not specifically ask to be placed with GEICO Indemnity, and the GEICO Companies did not interpret his telephone call as requesting a policy with that company in particular, as evidenced by the evaluation by Government Employees of his eligibility for a policy from several of the GEICO affiliates. That he was placed with GEICO Indemnity and not another GEICO entity was the result of a decision made by Government Employees personnel, not the result of a limited application by Edo. Thus, GEICO's argument has no basis in fact. Furthermore, as a matter of law, we refuse to turn the words "applied for" into a legal term of art that refers only to the issuing company. The clearest indication that Congress did not intend the words "applied for" to be used in such an unusual manner is that this interpretation would eliminate all potential FCRA liability for denials of insurance. No one disputes that FCRA defines the term adverse action to include denials. 15 U.S.C. § 1681a(k)(1)(B)(i). However, under GEICO's interpretation, because a consumer has not "applied for" insurance unless a policy has been issued and because a company that denies insurance has not, by definition, issued a policy, adverse action notices would never be required for denials of insurance. This is manifestly contrary to the statute, as it would eliminate from its coverage an important set of actions that Congress clearly intended to subject to FCRA's requirements.

Second, all the defendants argue that "takes any adverse action" limits FCRA's adverse action notice requirement to companies that actually issue an insurance policy. We find no such limitation in the statute by virtue of Congress's use of the word "takes" or otherwise. 15 U.S.C. § 1681m(a). To the contrary, adverse action is defined far more broadly than just "issuance." 15 U.S.C. § 1681a(k). The statutory definition specifically includes denials and cancellations as well as increases in rates, and other unfavorable changes, whenever and by whomever made. The word "takes" neither adds to nor detracts from that definition. It describes the act of engaging in the conduct that gives rise to the notice requirement. As discussed below, all of the companies at issue here took "adverse actions," as that term is defined in the statute.

Third and finally, all the defendants argue that we should hold liable only the issuing company because holding several companies liable for FCRA violations arising out of the issuance or denial of a single application will result in multiple, confusing adverse action notices, which would thwart rather than further FCRA's purpose. Such is not the case. Joint and several liability simply imposes the obligation on all of the affiliated companies responsible for taking an adverse action to ensure that the affected consumer receives a statutory notice describing the adverse affect of his credit report within that family of companies. Multiple notices are not required; a single notice from the companies involved identifying those companies and their respective roles will suffice.

Holding all the companies that take adverse action against a consumer jointly responsible for issuing a notice furthers FCRA's objectives. For example, joint responsibility substantially increases the prospect that an adverse action notice will be sent and that a customer who seeks to obtain insurance from a group of affiliated companies will be informed as to the manner in which his credit information adversely affected him. By imposing joint and several liability, Congress also improved the quality of information consumers receive, because each of the companies that takes an adverse action against the consumer must say so in the notice. We doubt that many consumers understand how a group of affiliated insurance companies operates or how consumers are assigned to specific entities within their overall structure. By having the organizations explain the actions each affiliated company took, Congress made it more likely that consumers would comprehend what transpired with respect to the increased cost of their policy.

On the basis of the record before us all three GEICO Companies and Hartford Fire may be held liable under FCRA, as may the two other Hartford entities as to which leave to amend was denied.15 Two of the GEICO Companies, working together, are responsible for increasing Edo's charge for insurance: Government Employees, which made the decision as to which of the GEICO family of companies would issue the insurance to Edo and, in so doing, determined that he would be charged at an increased rate, and GEICO Indemnity, which then issued the insurance policy at that increased rate. GEICO General is responsible because it denied Edo insurance for the reason that his credit rating was not sufficiently high. Hartford Fire, like Government Employees, made the critical rate-to-be-charged decision. It determined that, on the basis of Reynolds' credit report, he was not eligible for the lower rates afforded by its affiliates to the qualifying AARP members and that he would be charged for his insurance at a higher rate. Hartford Fire may therefore be held liable for increasing Reynolds' charges for insurance on the basis of his credit rating. Hartford PCIC and Hartford Midwest issued the policies to Reynolds at the increased rates determined by Hartford Fire, and may, accordingly, be held liable as well.

In sum, Government Employees, GEICO General, and GEICO Indemnity may be held jointly and severally liable for failing to issue an adverse action notice to Edo. Likewise, Hartford Fire may be held liable for failing to issue a notice to Reynolds, and Reynolds may also properly state claims against Hartford PCIC and Hartford Midwest. Thus, Reynolds should be permitted to amend his claims on remand.

Each of the defendants asks that we affirm the district court's grant of summary judgment on the alternative ground that, as a matter of law, its conduct was not willful. We must first define "willfully" as it appears in FCRA.16 Interestingly, there is no legislative history to explain what Congress intended by the use of that term.

We begin by following all five of the other circuits that have addressed the issue of the mens rea that is required with regard to the act that allegedly violates FCRA and hold that the act must have been performed "knowingly and intentionally." See Phillips v. Grendahl, 312 F.3d 357, 370 (8th Cir.2002); Dalton v. Capital Associated Indus., Inc., 257 F.3d 409, 418 (4th Cir.2001); Cousin v. Trans Union Corp., 246 F.3d 359, 372 (5th Cir.2001); Duncan v. Handmaker, 149 F.3d 424, 429 (6th Cir.1998); Cushman v. Trans Union Corp., 115 F.3d 220, 226 (3d Cir.1997). An act that is merely negligent is not willful. See McLaughlin v. Richland Shoe Co., 486 U.S. 128, 133, 108 S. Ct. 1677, 100 L. Ed. 2d 115 (1988) ("The word `willful' is widely used in the law, and, although it has not by any means been given a perfect consistent interpretation, it is generally understood to refer to conduct that is not merely negligent."). Additionally, we adopt the position of four of the five other circuits and hold that, although the act must be intentional, it need not be the product of "malice or evil motive." See, e.g., Dalton, 257 F.3d at 418 (holding that a plaintiff need not show malice or evil motive); Cousin, 246 F.3d at 372 (same); Bakker v. McKinnon, 152 F.3d 1007, 1013 (8th Cir.1998); Cushman, 115 F.3d at 226 (same). But see Duncan, 149 F.3d at 429 (requiring "`a motivation to injure'"). In this respect, for purposes of willfulness we distinguish civil from criminal liability. See Bryan v. United States, 524 U.S. 184, 191, 118 S. Ct. 1939, 141 L. Ed. 2d 197 (1998) ("Most obviously [willfulness] differentiates between deliberate and unwitting conduct, but in the criminal law it also typically refers to a culpable state of mind." (emphasis added)).

Next, we address the more difficult question: What is the nature of the mens rea that is required with respect to the law? Here, we follow the Third Circuit. Specifically, we hold that as used in FCRA "willfully" entails a "conscious disregard" of the law, which means "either knowing that policy [or action] to be in contravention of the rights possessed by consumers pursuant to the FCRA or in reckless disregard of whether the policy [or action] contravened those rights." Cushman, 115 F.3d at 227. We adopt this holding for two principal reasons.

First, we believe that the Third Circuit's definition best comports with Supreme Court precedent. The Court has consistently stated that willfulness for civil liability requires either knowledge or reckless disregard with respect to whether an action is unlawful. See Trans World Airlines, Inc. v. Thurston, 469 U.S. 111, 128, 105 S. Ct. 613, 83 L. Ed. 2d 523 (1985); see also Hazen Paper Co. v. Biggins, 507 U.S. 604, 614, 113 S. Ct. 1701, 123 L. Ed. 2d 338 (1993) (quoting Thurston and holding that, for an alleged civil violation of the Age Discrimination in Employment Act (ADEA), "willful" requires only a "`reckless disregard for the matter of whether its conduct was prohibited by the ADEA'"); McLaughlin, 486 U.S. at 134 n. 13, 108 S. Ct. 1677 (using the Thurston definition of "willful" in interpreting the Fair Labor Standards Act); United States v. Illinois Cent. R.R. Co., 303 U.S. 239, 242-43, 58 S. Ct. 533, 82 L. Ed. 773 (1938) (holding civil defendant's failure to unload a cattle car was "willful" because it showed a disregard for governing statute and an indifference to its requirements). The Court's rule with respect to civil cases differs from its rule in criminal proceedings. In criminal cases, actual knowledge of illegality is required for a willful violation of a criminal statute. See Bryan, 524 U.S. at 196, 118 S. Ct. 1939 (requiring "knowledge that the conduct is unlawful" in a criminal case); Ratzlaf v. United States, 510 U.S. 135, 149, 114 S. Ct. 655, 126 L. Ed. 2d 615 (1994) (requiring proof that the criminal defendant "knew the structuring [of financial transactions] in which he engaged was unlawful"); Cheek v. United States, 498 U.S. 192, 201, 111 S. Ct. 604, 112 L. Ed. 2d 617 (1991) (requiring proof that a criminal "defendant knew of the duty purportedly imposed by the provision of the statute or regulation he is accused of violating").17

Second, the Third Circuit's approach best furthers the purposes and objectives of the Act. It is fair and balanced; it is practical as well. It avoids the two extremes of excusing non-compliance even though the answer to a previously undecided question is objectively apparent and imposing liability not withstanding a truly excusable inability to predict future developments in the evolving construction of a statute by the courts. It encourages companies that use consumer credit reports to make the necessary effort to inform themselves fully and fairly as to their statutory obligations and, as a result, to carry out the statutory mandate of ensuring that consumers are notified when their credit information has been used against them. Unlike the defendants' preferred definition, the Third Circuit's standard does not create perverse incentives for companies covered by FCRA to avoid learning the law's dictates by employing counsel with the deliberate purpose of obtaining opinions that provide creative but unlikely answers to "issues of first impression." Because a reckless failure to comply with FCRA's requirements can result in punitive damages, insurance and other companies will more likely seek objective answers from their counsel as to the true meaning of the statute.

In sum, if a company knowingly and intentionally performs an act that violates FCRA, either knowing that the action violates the rights of consumers or in reckless disregard of those rights, the company will be liable under 15 U.S.C. § 1681n for willfully violating consumers' rights. A company will not have acted in reckless disregard of a consumers' rights if it has diligently and in good faith attempted to fulfill its statutory obligations and to determine the correct legal meaning of the statute and has thereby come to a tenable, albeit erroneous, interpretation of the statute. In contrast, neither a deliberate failure to determine the extent of its obligations nor reliance on creative lawyering that provides indefensible answers to issues of first impression is sufficient to avoid a conclusion that a company acted with willful disregard of FCRA's requirement.18 We hold that reliance on such implausible interpretations constitutes reckless disregard for the law and therefore amounts to a willful violation of the law.

Here, Hartford Fire, GEICO General, GEICO Indemnity, and Government Employees' interpretation of the statute was untenable. Indeed, the companies' exceedingly narrow interpretations of their obligations under FCRA were counter to the statute's plain text as well as its purpose. Relying on reasoning that was plainly unmeritorious, these companies sought to benefit from the privilege of using consumers' private credit information and yet be free of the attendant obligations. Specifically, Hartford Fire's contention that FCRA's adverse action notice requirement does not apply when a company charges a higher price for an initial insurance policy, but only when it has previously charged the consumer a lower rate, finds no basis in the statute. Likewise, its argument that adverse action notices are not required when a credit transmission reports that a consumer does not have adequate credit information to generate a score is clearly contrary to FCRA's language and to the purpose of the statute. Furthermore, Hartford Fire's argument that an adverse action notice sufficiently complies with FCRA's requirement even if it does not disclose that an adverse action was taken or the nature of that action is nonsensical. The GEICO Companies' argument that they are free to send adverse action notices only to those consumers whose credit rating is below average and not to those with a higher rating also finds absolutely no support in the statute. Likewise, GEICO General does not make even a colorable argument that it did not take adverse action against Edo when, acting in conjunction with the other affiliated companies, it failed to provide him with a policy at a preferred rate; denials are clearly covered by the adverse action definition. Finally, Hartford Fire and Government Employees' argument that they are not liable because, even though they may have made the decision to increase the consumers' rates, they did not issue the policies is not colorable as the statute on its face is not limited to issuers. We therefore hold that on the record before us the defendants all acted in reckless disregard of the consumers' statutory rights.19 Accordingly, defendants' argument for summary judgment on the basis that they did not act willfully also fails.

In conclusion, we hold that FCRA applies, inter alia, to the first rates charged in initial insurance policies. We also hold that FCRA requires insurance companies to send adverse action notices whenever they charge a higher rate for insurance, in initial policies or otherwise, because of the consumer's credit information, not simply when the consumer's credit rating is below average. Furthermore, we hold that a communication that there is a lack of sufficient credit information regarding a consumer is a credit report within the meaning of FCRA. In addition, we hold that adverse action notices must communicate to the consumer that an adverse action based on a consumer report was taken, describe the action, specify the effect of the action upon the consumer, and identify the party or parties taking the action. With respect to which companies in a group may be liable under FCRA, we hold that a company that makes the rate-setting decision, a company that issues the insurance policy, and any company that denies insurance at a more favorable rate may be held jointly and severally liable, and that such companies may provide a single adverse action notice to consumers containing all of the requisite information. Finally, we adopt the Third Circuit's definition of "willfully": Reckless disregard is sufficient.

As a consequence of these rulings, we hold that the district court erred in granting summary judgment to Hartford Fire on the basis that increased charges for insurance in an initial policy do not constitute adverse actions, and in denying Reynold's request for leave to amend his complaint to add Hartford PCIC and Hartford Midwest for that same reason. Likewise, we hold that the district court erred in granting summary judgment to GEICO Indemnity on the basis that the actions it took were not adverse and granting summary judgment to Hartford Fire, Government Employees, and GEICO General on the basis that only the issuer of insurance can be liable under FCRA. Finally, we hold that summary judgment may not be granted on the alternative grounds that a transmission that a consumer has insufficient credit information to generate a score is not a credit report, that Hartford Fire's adverse action notices were sufficient, or that the defendants' actions were not willful. In sum, we reverse the district court's grant of summary judgment with respect to all defendants in both Edo and Reynolds, reverse its denial of Reynolds' request to amend his complaint to add Hartford PCIC and Hartford Midwest, and remand to the district court for further proceedings consistent with this opinion.

REVERSED and REMANDED.

I join the majority opinion except so much of Section II.F that concludes that GEICO and Hartford insurance companies willfully violated FCRA. The district court awarded judgment to the insurance companies on summary judgment without reaching the question whether the companies "willfully fail[ed] to comply" with the requirements of FCRA. 15 U.S.C. § 1681n(a). The insurance companies urged the district court, as an alternative ground, to hold that the companies did not act willfully as a matter of law. They urged the same position to us on appeal as an alternative ground for affirming the judgment. The appellants did not brief the question of willfulness until the reply briefs, and in each case the appellants requested that we decline to rule for the insurance companies and remand the case to the district court for further proceedings.

I would not decide as a matter of fact that the insurance companies behaved willfully on the basis of their lawyers' arguments on appeal. I cannot conclude on the basis of the record before us that the companies' actions here were so "indefensible," "implausible," "untenable," "plainly unmeritorious," "clearly contrary to FCRA's language," "nonsensical," and without "colorable argument" that we can conclude that the companies' "reliance on such implausible interpretations constitutes reckless disregard for the law" and decide their willfulness ourselves without the benefit of findings of fact or full briefing.

If I thought our record was complete, I would not find that the companies willfully failed to comply with FCRA as a matter of law. The majority holds that we can reach the question because the district "necessarily found their actions not to be willful as a result of their reliance on an untenable interpretation of the statute." The reason the district court "necessarily found their actions not to be willful" is because the district court agreed with their legal position, a position the majority now declares to be "nonsensical" and "untenable." While I agree with the majority that the district court's understanding of FCRA was wrong, I cannot go so far as to conclude that the district court's conclusion, like the companies' position, was also untenable.

Accordingly, I would remand the question of whether the companies willfully failed to comply with FCRA to the district court for further proceedings. I respectfully dissent as to that portion of the opinion.