IN THE SUPREME COURT OF TEXAS
════════════
No. 10-0245
════════════
Patrick O. Ojo, On Behalf of Himself and
All Others Similarly Situated, Appellant
v.
Farmers Group, Inc., Fire Underwriters Association,
Fire Insurance Exchange, Farmers Underwriters Association,
and Farmers Insurance Exchange, Appellees
════════════════════════════════════════════════════
On Certified Question from the United States
Court of Appeals for the Ninth Circuit
════════════════════════════════════════════════════
Argued October 14, 2010
Justice Green delivered the opinion of the Court, in which Chief Justice Jefferson, Justice Wainwright, Justice Medina, Justice Johnson, Justice Guzman, and Justice Lehrmann joined, and in which Justice Willett joined as to Parts I, II, III.A–B, IV, and V.
Chief Justice Jefferson filed a concurring opinion.
Justice Willett filed an opinion concurring in part.
Justice Hecht did not participate in the decision.
The United States Court of Appeals for the Ninth Circuit certified to this Court the following question:
Does Texas law permit an insurance company to price insurance by using a credit-score factor that has a racially disparate impact that, were it not for the [McCarran-Ferguson Act],1 would violate the federal Fair Housing Act, 42 U.S.C. §§ 3601–19, absent a legally sufficient nondiscriminatory reason, or would using such a credit-score factor violate Texas Insurance Code sections 544.002(a), 559.051, 559.052, or some other provision of Texas law?
Ojo v. Farmers Group, Inc., 600 F.3d 1201, 1204–05 (9th Cir. 2010) (en banc) (per curiam). Pursuant to Article 5, section 3-c of the Texas Constitution and Texas Rule of Appellate Procedure 58.1, we answer that Texas law prohibits the use of race-based credit scoring, but permits race-neutral credit scoring even if it has a racially disparate impact.
I. Introduction
Patrick Ojo, an African-American resident of Texas, carries a homeowner’s property-and-casualty insurance policy issued by Farmers Group, Inc. Id. at 1202. Although Ojo has never made a claim on his homeowner’s policy, Farmers raised Ojo’s insurance premium by nine percent. Id. Ojo alleges that Farmers increased the premium as a result of unfavorable credit information acquired though its automated credit-scoring system. Id.
On behalf of himself and other racial minorities whose premiums increased as a result of Farmers’ use of a credit-scoring system, Ojo sued Farmers and its affiliates, subsidiaries, and reinsurers in federal court. Id. Ojo alleges that the defendants’ credit-scoring systems employ several “undisclosed factors” which result in disparate impacts for minorities and violate the federal Fair Housing Act (FHA), 42 U.S.C. §§ 3601–3619. Ojo, 600 F.3d at 1202. Ojo does not assert that he or any other member of the putative plaintiff class has suffered intentional discrimination at the hands of the defendants. Id.
Citing Federal Rule of Civil Procedure 12(b)(1) and 12(b)(6), the defendants moved to dismiss all of Ojo’s claims. Id. Applying the McCarran-Ferguson Act’s (MFA) reverse-preemption standard, 15 U.S.C. § 1012(b), the district court concluded that the Texas Insurance Code preempted Ojo’s FHA claims. Id. at 1203. Accordingly, the district court declined to answer whether Ojo’s disparate-impact discrimination claim sufficiently complied with Federal Rule of Civil Procedure 12(b)(6). Id. at 1202. On appeal to the United States Court of Appeals for the Ninth Circuit, a divided three-judge panel held that Texas law did not reverse-preempt Ojo’s FHA claim, initially reversing the district court. Ojo v. Farmers Group, Inc., 565 F.3d 1175, 1178 (9th Cir. 2009). Subsequently, the Ninth Circuit ordered the case reheard en banc. Ojo v. Farmers Group, Inc., 586 F.3d 1108, 1108 (9th Cir. 2009). The Ninth Circuit’s rehearing en banc resulted in the certified question now before us. See Ojo, 600 F.3d at 1204–05.
II. Background
Ojo sued in federal court based on the FHA, under which it is unlawful “[t]o discriminate against any person in the terms, conditions, or privileges of sale or rental of a dwelling, or in the provision of services or facilities in connection therewith, because of race.” 42 U.S.C. § 3604(b). Federal courts of appeals have interpreted this FHA provision to prohibit not just intentional acts of discrimination, but also race-neutral actions that have discriminatory effects on racial minorities (disparate-impact discrimination).2 Several courts of appeals have also held that the FHA applies in the underwriting of homeowner’s property insurance, given the FHA’s prohibition of discrimination “in the provision of services . . . in connection” with the “sale or rental of a dwelling.” 42 U.S.C. § 3604(b); see, e.g., Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351, 1360 (6th Cir. 1995); NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 301 (7th Cir. 1992). Ojo’s cause of action asserts this type of disparate impact liability in Farmers’ pricing of homeowner’s insurance based on credit scoring.
Ojo’s disparate impact claim, however, may be “reverse-preempted” by Texas law under the MFA, which provides that “[n]o Act of Congress shall be construed to invalidate, impair, or supersede any law enacted by any State for the purpose of regulating the business of insurance, . . . unless such Act specifically relates to the business of insurance.” 15 U.S.C. § 1012(b). Under the MFA, state law reverse-preempts a federal statute if: “(1) the federal law does not specifically relate to insurance; (2) the state law is enacted for the purpose of regulating insurance; and (3) the application of federal law to the case might invalidate, impair, or supersede the state law.” Ojo, 600 F.3d at 1208–09 (citing Humana Inc. v. Forsyth, 525 U.S. 299, 307 (1999)). The Ninth Circuit, hearing this case en banc, held that “it is undisputed that the FHA does not specifically relate to insurance,” thus satisfying the first prong of MFA reverse-preemption.3 Id. at 1203. It is also undisputed that “the relevant provisions of Texas law . . . are enacted for the purpose of insurance regulation,” thus satisfying the second prong. Id. The certified question before us specifically deals with the third prong, and asks whether allowing Ojo’s claim under the FHA might invalidate, impair, or supersede Texas law. See id. at 1204–05. In light of the fact that Texas only prohibits the use of credit score factors or rates based on race, or rates that differ because of race, we answer that application of the FHA to permit a cause of action for disparate impact resulting from the use of credit scoring in the field of insurance certainly might invalidate, impair, or supersede Texas law.
III. The Texas Insurance Code Does Not Provide for a Cause of Action
Based on a Racially Disparate Impact
The Texas Insurance Code expressly prohibits “unfair discrimination” and specifically states that “[a] person may not charge . . . an individual a rate that is different from the rate charged to other individuals for the same coverage because of the individual’s race, color, religion, or national origin.” Tex. Ins. Code § 544.002(a)(2). An exception to this provision provides that “[a] person does not violate Section 544.002 if the refusal, limitation, or charge is required or authorized by law or a regulatory mandate.” Id. § 544.003(c). Farmers points out that § 559.051 authorizes the use of race-neutral credit score factors, and that this authorization is the exception to § 544.002,which is recognized in § 544.003. Section 559.051 permits an insurer to “use credit scoring, except for factors that constitute unfair discrimination, to develop rates, rating classifications, or underwriting criteria.” Id. § 559.051; see also id. § 559.052(a)(1) (“An insurer may not use a credit score that is computed using factors that constitute unfair discrimination . . . .”). The factors that “constitute unfair discrimination” are not defined in the Texas Insurance Code. However, the Code does define an “unfairly discriminatory” rate as one that “is based wholly or partly on the race, creed, color, ethnicity, or national origin of the policyholder or an insured.” Id. § 560.002(c)(3)(C) (emphasis added).
Under Texas Insurance Code § 559.201, the use of credit score factors defined by § 559.052(a)(1) that constitute “unfair discrimination” is deemed an “unfair practice in violation of Chapter 541.” Id. § 559.201 (making violations of Chapter 559 an unfair practice under Chapter 541). Unfair practices under Chapter 541 are subject to private civil suits, including class actions. Id. §§ 541.151 (Private Action for Damages Authorized), 541.251(a) (Class Action Authorized); see Farmers Group, Inc. v. Lubin, 222 S.W.3d 417, 421–22 (Tex. 2007).
No Texas courts have interpreted whether these Insurance Code provisions prohibit only intentional discrimination or also discrimination based on disparate impact. We derive from these provisions that insurance rates may not be “based wholly or partly on” race, and that an individual may not be charged a rate that is “different from the rate charged to other individuals for the same coverage because of the individual’s race.” Tex. Ins. Code §§ 544.002(a)(1) (emphasis added), 560.002(c)(3)(C). Additionally, while credit scoring is authorized, it may not be based on “factors that constitute unfair discrimination.” Id. §§ 559.051, 559.052(a)(1). We can only assume that a credit score factor constitutes unfair discrimination if it is “based wholly or partly on” race, or if it is used to arrive at an insurance rate that is “different from the rate charged to other individuals for the same coverage because of the individual’s race.” See id. §§ 544.002(a)(1), 560.002(c)(3)(C). Ojo alleges these provisions not only prohibit intentional discrimination—the use of race-based classifications to price insurance differently—but also prohibit disparate impact discrimination—the use of race-neutral pricing schemes that effectuate disparate results (in this case, racial minorities alleging they have suffered higher premium rates as a direct consequence of race-neutral credit scoring). However, nothing in the Insurance Code prohibits the use of race-neutral credit scoring. In fact, the Code requires that the factors used in credit scoring to price insurance be race-neutral, or not based on race. See id. §§ 544.002(a)(1), 560.002(c)(3)(C). The nature of Ojo’s disparate impact claim presupposes that these factors are race neutral, which is exactly what the Code requires. Nevertheless, to support his argument that the “based on” and “because of” race language in the Texas Insurance Code implies the availability of a cause of action for disparate impact discrimination, Ojo draws our attention to the same language used in the FHA, an act which has been interpreted to provide for disparate impact protection. See 42 U.S.C. § 3604; see, e.g., City of Black Jack, 508 F.2d at 1184. Ojo also relies on the United States Supreme Court’s interpretation of Title VII of the Civil Rights Act as providing for a disparate impact cause of action, an act that also prohibits discrimination “because of” race. See 42 U.S.C. § 2000e-2; Smith v. City of Jackson, 544 U.S. 228, 240 (2005); Griggs v. Duke Power Co., 401 U.S. 424, 436 (1971). However, given the numerous other considerations, addressed below, that have led federal courts to broadly interpret the FHA and Title VII, we find this argument unavailing. We are also guided by the fact that the use of the “because of” race and “based on” race language in Texas case law and the Texas Labor Code has been more in association with intentional discrimination claims than claims for disparate impacts. We first address the use of this language within Texas statutes and case law.
A. The Language of the Insurance Code Is Inconsistent
with a Disparate Impact Theory of Liability
The Texas Insurance Code prohibits “unfairly discriminatory” insurance rates as those that charge differently “because of” or “based wholly or partly on” race. See Tex. Ins. Code §§ 544.002(a)(1), 560.002(c)(3). Texas courts considering this language in the employment context have used the “because of” and “based . . . on” race language in the disparate treatment context, but not in the area of disparate impacts. In University of Texas v. Poindexter, 306 S.W.3d 798 (Tex. App.—Austin 2009, no pet.), the court of appeals held that “[d]isparate-treatment discrimination addresses employment actions that treat an employee worse than others based on the employee’s race, color, religion, sex, or national origin. In such disparate-treatment cases, proof and finding of discriminatory motive is required.” Id. at 804 n.1 (emphasis added); accord Massarsky v. Gen. Motors Corp., 706 F.2d 111, 117 (3d Cir. 1983) (noting that a plaintiff could establish intentional discrimination when his “employer applied an expressly race-based or sex-based standard in its treatment of the plaintiff” (emphasis added)). In Wal-Mart Stores, Inc. v. Davis, 979 S.W.2d 30 (Tex. App.—Austin 1998, pet. denied), the court of appeals described disparate impact claims as ethose that “involve facially neutral practices . . . that operate to exclude a disproportionate percentage of persons in a protected group and cannot be justified by business necessity. . . . Disparate treatment [exists where] the defendant . . . treats some people less favorably than others because of their race, color, religion, sex, or national origin.” Id. at 44 (emphasis added). The United States Supreme Court has similarly distinguished between disparate treatment discrimination and disparate impact discrimination, noting that the former is discrimination against others “because of their race,” while the latter encompasses “practices that are facially neutral . . . but that in fact fall more harshly on one group than another.” Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 335–36 n.15 (1976); see also Smith, 544 U.S. at 239 (plurality opinion).
Significantly, the phrase “because of race” is also used in the Texas Labor Code, which makes an employer liable for taking action adverse to an employee “because of race.” See Tex. Lab. Code § 21.051. Under the Labor Code, a plaintiff must show causation by demonstrating that race was a motivating factor in the employer’s decision, the standard for proving intentional discrimination, or disparate treatment. See id. § 21.125(a) (“Except as otherwise provided by this chapter, an unlawful employment practice is established when the complainant demonstrates that race . . . was a motivating factor for an employment practice, even if other factors also motivated the practice . . . .”); cf. Quantum Chem. Corp. v. Toennies, 47 S.W.3d 473, 480 (Tex. 2001) (holding that in a claim for age discrimination, an employee must show that age was a motivating factor in the employer’s decision to terminate the employee); Herbert v. City of Forest Hill, 189 S.W.3d 369, 375 (Tex. App.—Fort Worth 2006, no pet.) (holding that to prove causation in a race discrimination case, a plaintiff “must establish that race was a motivating factor for [the] employment practice” (internal quotation marks omitted)). In addition, the Texas Legislature expressly provided for disparate impact protection in Texas Labor Code § 21.122(a)(1),4 where it defined the burden of proof for disparate impact cases in the employment context:
An unlawful employment practice based on disparate impact is established under this chapter only if a complainant demonstrates that a respondent uses a particular employment practice that causes a disparate impact on the basis of race . . . and the respondent fails to demonstrate that the challenged practice is job-related for the position in question and consistent with business necessity . . . .
Id. No such section appears in the Texas Insurance Code. The Texas Legislature, well aware of how to create a cause of action for disparate impact discrimination, chose not to do so in the field of insurance, specifically with regards to the use of credit scoring.5 Because the Legislature chose not to include a section expressly providing for or defining a disparate impact claim in the Texas Insurance Code, but did do so in the Texas Labor Code, we conclude that the Legislature did not intend to provide for disparate impact liability for the use of credit scoring in pricing insurance.
B. The Use of “Because of” Race in the Federal FHA and Title VII
Did Not Alone Prompt Federal Courts to Hold That These Acts
Create Causes of Action Based on Racially Disparate Impacts
Ojo relies on federal case law interpreting the FHA and Title VII to provide for disparate impact protection, arguing that the Texas Insurance Code should also be interpreted to provide for disparate impact protection because it uses the same “because of race” language as those federal acts. See 42 U.S.C. § 2000e-2 (prohibiting discrimination by employers of individuals “because of such individual’s race” ); 42 U.S.C. § 3604(b) (prohibiting discrimination in the provision of services in connection with housing “because of race”); Tex. Ins. Code § 544.002(a) (defining unfair discrimination as providing insurance coverage differently “because of the individual’s . . . race”). Although Ojo has pointed us to a wealth of federal authority holding that the FHA and Title VII provide for disparate impact protection, the reasons supporting those holdings extend far beyond the use of the phrase “because of race.” See, e.g., Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights, 558 F.2d 1283, 1289 (7th Cir. 1977) (focusing on the policy goals of the FHA in deciding on a broad interpretation of its provisions); see also Peter E. Mahoney, The End(s) of Disparate Impact: Doctrinal Reconstruction, Fair Housing and Lending Law, and the Antidiscrimination Principle, 47 Emory L.J. 409, 425 (1998) (describing the origins of the disparate impact standard under the FHA as partially “borrowed” from the case law on Title VII, and also noting the standard’s diverse and inconsistent application by federal courts). In determining whether a statute provides for disparate impact protection, federal and state courts have looked first to the language of the statute to assess whether “the thrust of the Act [is] to the consequences of . . . practices, not simply the motivation.” Griggs, 401 U.S. at 432; see also Tex. Parks & Wildlife Dep’t. v. Dearing, 240 S.W.3d 330, 352 (Tex. App—Austin 2007, pet. denied). The United States Court of Appeals for the Seventh Circuit actually regarded the phrase “because of race” as a potential obstacle to disparate impact protection before considering the policy goals behind the FHA:
The major obstacle to concluding that action taken without discriminatory intent can violate section 3604(a) is the phrase “because of race” contained in the statutory provision. The narrow view of the phrase is that a party cannot commit an act “because of race” unless he intends to discriminate between races. . . . The broad view is that a party commits an act “because of race” whenever the natural and foreseeable consequence of that act is to discriminate between races, regardless of his intent.
Vill. of Arlington Heights, 558 F.2d at 1288; accord Resident Advisory Bd. v. Rizzo, 564 F.2d 126, 146 (3d Cir. 1977) (“[W]e note that the ‘because of race’ language might seem to suggest that a plaintiff must show some measure of discriminatory intent.”). The Seventh Circuit declined to take a narrow view of the “because of race” language because of the congressional mandate within the FHA “to provide, within constitutional limitations, for fair housing throughout the United States.” Vill. of Arlington Heights, 558 F.2d at 1289 (quoting 42 U.S.C. § 3601). The Seventh Circuit also relied on previous interpretations of the FHA, and its goal to “promote ‘open, integrated residential housing patterns and to prevent the increase of segregation, in ghettos, of racial groups whose lack of opportunities the Act was designed to combat.’” Id. (quoting Otero v. N.Y. City Hous. Auth., 484 F.2d 1122, 1134 (2d Cir. 1973)). Other federal circuit courts applying disparate impact protections under the FHA have also relied upon this congressional mandate.6 Numerous courts have also noted that the need for disparate impact protection under the FHA arose from the lack of disparate impact liability under the Fourteenth Amendment after the United States Supreme Court’s decision in Washington v. Davis, 426 U.S. 229 (1976), and the difficulty of proving intentional discrimination.7
In determining whether discriminatory impact liability exists within the FHA, Title VII, and the Age Discrimination in Employment Act (ADEA), state and federal courts have also focused on the breadth and reach of prohibitory language, and have refused to find disparate impact liability when a statute focuses only on the nature of an action, and not on its effects. See, e.g., Monson v. Rochester Athletic Club, 759 N.W.2d 60, 67 (Minn. Ct. App. 2009) (holding that there was no disparate impact liability where “the [state law] does not include such effects-based language”); see also Smith, 544 U.S. at 235–36 (holding that the ADEA provides for disparate impact liability because it not only prohibits employers’ actions that “limit, segregate, or classify” persons, but rather, also prohibits actions that “deprive any individual of employment opportunities or otherwise adversely affect his status as an employee” (citing 29 U.S.C. § 623(a))); Dearing, 240 S.W.3d at 339 (quoting Smith, 544 U.S. at 235). Both Title VII and the ADEA have been interpreted by the United States Supreme Court to provide for disparate impact liability because they go so far as to prohibit practices that “tend to deprive employees of opportunities.” See Smith, 544 U.S. at 235–36 (ADEA); Griggs, 401 U.S. at 430–32 (Title VII); see also Huntington, 488 U.S. at 18 (declining to determine whether the FHA provides for disparate impact protection, stating: “Since appellants conceded the applicability of the disparate-impact test for evaluating the zoning ordinance under Title VIII, we do not reach the question whether that test is the appropriate one.”).
Sections 544.002(a) and 560.002(c)(3) of the Texas Insurance Code do not include the type of broad prohibitory language that gives rise to disparate impact claims. Rather, both sections focus exclusively on the manner in which insureds are classified; that is, they prohibit classifications because of or based on race. Neither statute broadens its application so as to prohibit practices that may “otherwise adversely affect” or “tend to deprive” an insured of an opportunity, or any other similarly expansive language, as was the case in the federal acts at issue in Griggs and Smith. See Smith, 544 U.S. at 235–36 (ADEA); Griggs, 401 U.S. at 430–32 (Title VII). Rather, the Texas Insurance Code authorizes actions that classify individuals based on credit score in order to affect insurance pricing, as long as such classifications are not based on race or because of race. See Tex. Ins. Code §§ 544.002(a), 559.051, 560.002(c)(3). As long as insurers use race-neutral factors in credit scoring to set insurance rates, they do not run afoul of the Texas Insurance Code in the way an employer would run afoul of Title VII for using race-neutral testing that adversely affects employees of a certain race. See Griggs, 401 U.S. at 430 (“Under [Title VII], practices, procedures or tests neutral on their face, and even neutral in terms of intent, cannot be maintained if they operate to ‘freeze’ the status quo of prior discriminatory employment practices.”). Because the Texas Insurance Code expressly authorizes credit scoring, it cannot be subject to the same breadth of interpretation applied to Title VII or the FHA simply because it uses the phrases “because of” or “based . . . on” race. Ojo’s argument that federal interpretations of these acts should control our interpretation of the Texas Insurance Code is unavailing in light of the additional considerations, other than some similar language, present in the federal case law.
C. The Legislative History of the Insurance Code Is Inconsistent
with a Disparate Impact Theory of Liability
In addition to the express language of the statute, courts have looked to a statute’s legislative history when determining whether the statute gives rise to a disparate impact theory of liability. See, e.g., Smith, 544 U.S. at 238 (“[W]e think the history of the enactment of the ADEA . . . supports the . . . consensus concerning disparate-impact liability.”); Gen. Bldg. Contractors Ass’n, Inc. v. Pennsylvania, 458 U.S. 375, 389 (1982) (examining the legislative history of 42 U.S.C. § 1981 and holding the statute did not give rise to a disparate impact claim); Griggs, 401 U.S. at 436 (“From the sum of the legislative history relevant in this case, the conclusion is inescapable that the [agency’s] construction . . . comports with congressional intent.”); Dearing, 240 S.W.3d at 351 (“When ascertaining legislative intent, we may also consider . . . the law[’s] . . . history . . . .”); see also Tex. Gov’t Code § 311.023(3) (allowing courts to consider legislative history when construing statutes). We also look to legislative history in this instance because the declared policy of the MFA is to ensure that state legislatures are able to regulate the business of insurance without unintended federal interference.8
The legislative history of the credit scoring bill and the arguments of its opponents indicates that the Texas Legislature was aware of the possibility of a disparate impact on racial minorities, yet did not expressly provide for a disparate impact claim as it did in the Texas Labor Code. Despite its longstanding prohibition of unfair discrimination, the Legislature first expressly authorized the use of credit scoring in setting insurance rates in 2003. Act of June 2, 2003, 78th Leg., R.S., ch. 206, § 3.01, 2003 Tex. Gen. Laws 916, 916–21, repealed by Act of May 24, 2005, 79th Leg., R.S., ch. 728, § 11.020(b), 2005 Tex. Gen. Laws 2188, 2217 (recodifying the relevant credit scoring sections of the Insurance Code into Tex. Ins. Code chapter 559) (originally codified at Tex. Ins. Code Ann. art. 21.49-2U, § 7(a) (West Supp. 2003)). Opponents of the credit scoring bill admonished:
The state should ban the practice of credit scoring altogether. Tornadoes do not strike homeowners on the basis of their credit scores, and no independent studies have proven any statistical relationship between a consumer’s credit history and his or her ability to drive or maintain an automobile. . . . Credit scoring is discriminatory, especially against women, minorities, low-income consumers, and consumers who conduct all of their personal business on a cash basis.
House Research Org., Bill Analysis, Tex. S.B. 14, 78th Leg., R.S., 20 (May 21, 2003). Despite those concerns, the Legislature decided to authorize credit scoring in pricing insurance, but addressed some of the concerns with certain statutory restrictions. In addition to prohibiting the use of “factors that constitute unfair discrimination,” Tex. Ins. Code Ann. art. 21.49-2U, § 7(a) (West Supp. 2003) (current version at Tex. Ins. Code § 559.051), the Legislature prohibited insurers from denying, cancelling, or refusing to renew a policy “solely on the basis of credit information,” as well as from denying coverage solely because the consumer does not have a credit card account. Id. § 3(a) (current version at Tex. Ins. Code § 559.052). Also, certain information could not be used as a negative factor in an insurer’s scoring methodology, such as a collection account with a medical industry code. Id. § 4(a)(3) (current version at Tex. Ins. Code § 559.101). However, even with these restrictions, the Legislature included no language expressly providing for a cause of action based on disparate impact.
The Legislature also directed the Commissioner of the Texas Department of Insurance (TDI) to conduct a study and submit a report to state officials and the 79th Legislature before January 1, 2005, containing, among other things:
• a summary statement regarding the use of credit information, credit reports, and credit scores by insurers . . . ;
• any disproportionate impact on any class of individuals, including classes based on income, race, or ethnicity . . . ; and
• recommendations from the department to the [L]egislature regarding the use of credit information by insurers.
Act of June 2, 2003, 78th Leg., R.S., ch. 201, § 3.01, sec. 15(a), (b)(1), (b)(5)–(6), 2003 Tex. Gen. Laws 916, 920–21 (expired Mar. 1, 2005) (emphasis added) (previously located at Tex. Ins. Code Ann. art. 21.49-2U, § 15 (West Supp. 2003)). Insurance Commissioner Jose Montemayor completed this credit scoring study and submitted his findings in December 2004, stating in part:
Similar to other published studies,9 there appears to be a strong relationship between credit score and insurance risk (or loss). . . . [With regards to auto insurance,] as credit scores improve, the frequency decreases, i.e. people have fewer accidents or claims.
Tex. Dep’t of Ins., Report to the 79th Legislature: Use of Credit Information by Insurers in Texas 18–20 (Dec. 2004), http://www.tdi.state.tx.us/reports/documents/creditrpt04.pdf. These findings were supplemented with a report to the Legislature, which explained that under a multivariate analysis:
For both personal auto liability and homeowners, credit score was related to claim experience even after considering other commonly used rating variables. . . . For both personal auto liability and homeowners, the difference in claims experience by credit score was substantial. Typically, the claim experience for the 10 percent of policyholders with the worst credit scores was 1.5 to 2 times greater than that of the 10 percent of policyholders with the best credit scores. The magnitude of the variation noted in the earlier report remains unchanged even after considering other commonly used rating variables.
Tex. Dep’t of Ins., Supplemental Report to the 79th Legislature: Use of Credit Information by Insurers in Texas: The Multivariate Analysis 6 (Jan. 2005), http://www.tdi.state.tx.us/reports/documents/credit05sup.pdf.
In a letter accompanying the report, Commissioner Montemayor explained that while disparate impacts result from the use of credit scoring, he was without authority to ban or regulate the use of credit scoring that produces disparate impacts as long as it is actuarially sound and not intentionally discriminatory.10 Commissioner Montemayor stated that “credit scoring, if continued, is not unfairly discriminatory as defined in current law because credit scoring is not based on race, nor is it a precise indicator of one’s race.” Letter from Jose Montemayor to the 79th Texas Legislature (Jan. 31, 2005) (accompanying Tex. Dep’t of Ins., Supplemental Report to the 79th Legislature: Use of Credit Information by Insurers in Texas: The Multivariate Analysis (Jan. 2005), http://www.tdi.state.tx.us/reports/documents/credit05sup.pdf). In addition, Commissioner Montemayor stated that the use of credit scoring in pricing insurance inevitably carried the risk of disproportionate impacts just as any risk-based assessment would, and that to discontinue insurers’ assessment of risk factors would effectively homogenize the risk and essentially charge everyone the same insurance rate, something that would “be a set-back to all Texans, of all races, especially those of moderate to lower income whose risk remains low.” Id.
“Even when a statute is not ambiguous on its face, we can consider other factors to determine the Legislature’s intent, including . . . administrative construction of the statute . . . .” Helena Chem. Co. v. Wilkins, 47 S.W.3d 486, 493 (Tex. 2001) (citing Tex. Gov’t Code § 311.023). We cite the Commissioner Montemayor’s letter and report here, however, more for evidence of the Texas Legislature’s awareness of potential disparate impacts, and to show that the Legislature, knowing this, still chose not to expressly provide for disparate impact protection as it did in the Labor Code. The Texas Legislature expressly directed the Commissioner to analyze the effects of credit scoring in insurance pricing and report back during the next legislative session. It was during this subsequent session that the Legislature re-codified various portions of the Insurance Code, including the sections on credit scoring now codified at Texas Insurance Code chapter 559, and made no relevant changes to the Code, despite the Commissioner’s warnings of the potential for disparate impacts. In fact, two bills banning credit scoring (H.B. 23 and S.B. 167), which were introduced by members of the 79th Legislature before the submission of Commissioner Montemayor’s January 2005 report, died in committee after the report was submitted. See Tex. H.B. 23, 79th Leg., R.S. (2005); Tex. S.B. 167, 79th Leg., R.S. (2005).
Given the Legislature’s and the Insurance Commissioner’s awareness of the potential for disparate impacts, and the Legislature’s decision to not enact any express prohibition of disparate impact discrimination in the Insurance Code, we can only conclude that the Legislature did not intend to create a cause of action for disparate impact discrimination in insurance pricing based on credit scoring.
IV. The Texas Fair Housing Act (TFHA) Does Not Change Our Conclusion
Regarding the Lack of Disparate Impact Liability in the Texas Insurance Code
The certified question also asks us to consider other provisions of Texas law that may provide for disparate impact protection. Ojo argues that the Texas Fair Housing Act (TFHA) should provide such protection because the FHA, which the TFHA was intended to mirror, provides for disparate impact liability in the provision of housing. See Tex. Prop. Code § 301.002(3) (“The purposes of this chapter are to provide rights and remedies substantially equivalent to those granted under federal law.”). Federal courts have interpreted the FHA to apply to the provision of homeowner’s insurance. See, e.g., Am. Family Mut. Ins. Co., 978 F.2d at 299. Ojo is correct that Texas courts will generally construe Texas statutes implementing federal rights consistently with federal case law. See, e.g., Quantum Chem. Corp., 47 S.W.3d at 476 (holding that the Texas Commission on Human Rights Act (TCHRA) was enacted to implement the policies of Title VII, and thus federal case law interpreting Title VII guides this Court’s reading of the TCHRA). However, the TFHA contains a “carve-out” provision, which provides that provisions of the TFHA “do[] not affect a requirement of nondiscrimination in any other state or federal law.” Tex. Prop. Code § 301.044(b). In addition, Ojo cannot direct us to, and indeed there is very little, federal authority confirming the existence of disparate impact liability even under the FHA in the field of insurance. See, e.g., Saunders v. Farmers Ins. Exch., 537 F.3d 961, 964 (8th Cir. 2008) (“Applying [HUD] standards, we have recognized a disparate impact [FHA] claim against private actors in another context. But at least with respect to insurers, the question is not free from doubt. However, the Insurers have not raised the issue and therefore we assume, without deciding, that private insurers may be liable under the [FHA] on a disparate impact theory.” (internal citations omitted)); Dehoyos v. Allstate Corp., 345 F.3d 290, 299 n.7 (5th Cir. 2003) (“We . . . decline to differentiate claims of disparate impact and claims of intentional discrimination at this preliminary stage of litigation”); Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351, 1362 (6th Cir. 1995) (stating that “HUD has never applied a disparate impact analysis to insurers”); Allstate Fair Hous. Opportunities of Nw. Ohio v. Am. Family Mut. Ins. Co., 684 F. Supp. 2d 964, 967–70 (N.D. Ohio 2010) (holding that plaintiffs failed to make a prima facie case of disparate impact discrimination regarding the use of a specific valuation method as an underwriting criterion). Because the relevant provisions of the Texas Insurance Code are more recent and specific regarding discriminatory liability in the field of insurance, and because we have determined that the Insurance Code does not provide for disparate impact liability, we conclude that Ojo’s argument that Texas provides for disparate impact liability in the field of insurance under the TFHA lacks merit.
V. Conclusion
The Texas Insurance Code is void of any language creating a cause of action for a racially disparate impact. The Texas Legislature has demonstrated that it is well aware of how to create a cause of action for disparate impact in other contexts, but it has chosen not to do so in the field of insurance. When dealing with issues of policy, this Court has consistently deferred to the judgment of the Legislature, and has not created causes of action where the Legislature did not clearly express a desire to do so. See Edgewood Indep. Sch. Dist. v. Meno, 917 S.W.2d 717, 726 (Tex. 1995) (recognizing that it is not our responsibility “to judge the wisdom of the policy choices of the Legislature, or to impose a different policy of our own choosing.”). The decision to either allow or prohibit the use of credit scoring in pricing insurance that creates disparate impacts properly rests with the Legislature, and we leave to the Legislature’s judgment the question of whether to expressly create a cause of action for disparate impact in the field of insurance, as it expressly created within the Texas Labor Code. See Tex. Labor Code § 21.122(a)(1) (defining the burden of proof in asserting a cause of action for discrimination based on disparate impact). Allowing a claim against Texas insurers for using completely race-neutral factors in credit scoring would frustrate the regulatory policy of Texas that the MFA is meant to protect, which is the continued regulation of the field of insurance by the states without unintentional congressional intrusion. See 15 U.S.C. §§ 1011 (“[T]he continued regulation and taxation by the several States of the business of insurance is in the public interest . . . .”), 1012(b) (“No Act of Congress shall be construed to invalidate, impair, or supersede any law enacted by any State for the purpose of regulating the business of insurance, . . . unless such Act specifically relates to the business of insurance . . . .”). Therefore, we answer the certified question by holding that Texas law does not prohibit an insurer from using race-neutral factors in credit-scoring to price insurance, even if doing so creates a racially disparate impact.
______________________________
Paul W. Green
Justice
OPINION DELIVERED: May 27, 2011
1 The McCarran-Ferguson Act (MFA) allows state insurance law to “reverse-preempt” federal law that does not directly relate to insurance. See 15 U.S.C. § 1012(b); Ojo v. Farmers Group, Inc., 600 F.3d 1201, 1203 (9th Cir. 2010) (en banc) (per curiam).
2 See Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights, 558 F.2d 1283, 1293–94 (7th Cir. 1977) (recognizing that a village’s refusal to rezone plaintiffs’ property to accommodate federally financed low-cost housing had the potential to effect a strong discriminatory impact capable of violating the federal FHA); United States v. City of Black Jack, 508 F.2d 1179, 1184 (8th Cir. 1974) (“Title VIII [the FHA] is designed to prohibit all forms of discrimination, sophisticated as well as simple-minded.” (internal quotation marks omitted)); cf. Pfaff v. U.S. Dep’t of Hous. and Urban Dev., 88 F.3d 739, 747–50 (9th Cir. 1996) (holding that the Department of Housing and Urban Development (HUD) failed to establish a prima facie case against a private landlord that a facially neutral, numerical occupancy restriction illegally discriminated against families with children, and admonishing HUD for alleging such restrictions were discriminatory); Simms v. First Gibraltar Bank, 83 F.3d 1546, 1555–56 (5th Cir. 1996) (holding that a jury verdict awarding damages for disparate impact discrimination under the federal FHA was not supported by sufficient evidence when the plaintiff identified only a bank’s rejection of his loan application, rather than a specific bank policy or practice, as having adverse, discriminatory effects on minorities). But see Town of Huntington v. Huntington Branch, NAACP, 488 U.S. 15, 16 (1988) (per curiam) (refusing to address whether a town’s refusal to rezone violated the federal FHA and provided a cause of action based on disparate impact).
3 We note the conundrum this creates: without reverse-preemption of the federal FHA by Texas law, Ojo would have a disparate impact cause of action for insurance pricing under the federal FHA, and yet, reverse-preemption is only at issue if the federal FHA does not “specifically relate[] to the business of insurance.” 15 U.S.C. § 1012(b). However, this is not an issue the certified question requires us to resolve. We instead focus our attention on whether Texas law provides for a disparate impact cause of action for insurance pricing based on credit scoring.
4 The Texas Government Code also prohibits fire departments from administering tests that disparately impact “any group defined by race.” Tex. Gov’t Code § 419.103. These tests must also comply with Chapter 21 of the Labor Code, which occurs when “the disparate impact on a group is the result of a bona fide occupational qualification.” Id. § 419.103(b).
5 See Harris County Hosp. Dist. v. Tomball Reg’l Hosp., 283 S.W.3d 838, 847 (Tex. 2009) (“The judiciary’s task is not to refine legislative choices . . . . The judiciary’s task is to interpret legislation as it is written.”); Cameron v. Terrell & Garrett, Inc., 618 S.W.2d 535, 540 (Tex. 1981) (“It is a rule of statutory construction that every word of a statute must be presumed to have been used for a purpose . . . [and] we believe every word excluded from a statute must also be presumed to have been excluded for a purpose.”); cf. Tex. Natural Res. Conservation Comm’n v. IT-DAVY, 74 S.W.3d 849, 854 (Tex. 2002) (similarly holding that in the realm of statutory waiver of sovereign immunity, it is the Texas Legislature’s task to “weigh the conflicting public policies” in enacting statutes providing for such waiver).
6 See, e.g., Rizzo, 564 F.2d at 147 (noting that “[a]lthough the legislative history of Title VIII [the FHA] is somewhat sketchy, the stated congressional purpose demands a generous construction of Title VIII”); City of Black Jack, 508 F.2d at 1184 (recognizing that the FHA was “passed pursuant to the congressional power under the Thirteenth Amendment to eliminate the badges and incidents of slavery,” and noting that the United States Supreme Court has treated the entire Civil Rights Act of 1866, another act passed under the power of the Thirteenth Amendment, broadly (citing Jones v. Mayer Co., 392 U.S. 409, 442–43 (1968) (“[W]hen racial discrimination herds men into ghettos and makes their ability to buy property turn on the color of their skin, then it too is a relic of slavery.”))).
7 See, e.g., Rizzo, 564 F.2d at 146 (“Given the increased burden of proof which Washington v. Davis and Arlington Heights now place upon equal protection claimants, we suspect that Title VIII will undoubtedly appear as a more attractive route to nondiscriminatory housing, as litigants become increasingly aware that Title VIII rights may be enforced even without direct evidence of discriminatory intent.”); Vill. of Arlington Heights, 558 F.2d at 1290 (“[A] requirement that the plaintiff prove discriminatory intent . . . is often a burden that is impossible to satisfy.”); City of Black Jack, 508 F.2d at 1185 (“Effect, and not motivation, is the touchstone, in part because clever men may easily conceal their motivations . . . .”); see also Peter E. Mahoney, The End(s) of Disparate Impact: Doctrinal Reconstruction, Fair Housing and Lending Law, and the Antidiscrimination Principle, 47 Emory L.J. 409, 425–26 (1998) (noting that many federal courts in the years after enactment of the FHA “aggressively expand[ed] the scope and application of equal protection analysis to a host of local governmental activities” because of the “difficulty of proving an overt intent to discriminate,” which led to a similar expansion of the protection afforded by the FHA).
8 See 15 U.S.C. § 1011 (“Congress hereby declares that the continued regulation and taxation by the several States of the business of insurance is in the public interest, and that silence on the part of the Congress shall not be construed to impose any barrier to the regulation or taxation of such business by the several States.”); W. & S. Life Ins. Co. v. State Bd. of Equalization of Cal., 451 U.S. 648, 654 (1981) (stating that Congress passed the MFA “believing that the business of insurance is ‘a local matter, to be subject to and regulated by the laws of the several States’” (citing H.R. Rep. No. 143, at 2 (1945))); In re Title Ins. Antitrust Cases, 702 F. Supp. 2d 840, 871 (N.D. Ohio 2010) (“Indeed, it would seem that this type of activity [setting insurance rates] is precisely the kind that the McCarran-Ferguson Act meant to leave to the state legislatures to regulate.”).
9 In 2003, EPIC Actuaries, LLC published a study reviewing more than 2.7 million auto insurance policies and found that an insured’s credit-based insurance score is directly connected to the insured’s likelihood of filing a claim, and that credit scoring measures risk not previously measured by other rating factors and is among the top predictors of risk. See Michael J. Miller & Richard A. Smith, The Relationship of Credit-Based Insurance Scores to Private Passenger Automobile Insurance Loss Propensity (June 2003), http://www.ask-epic.com/Publications/Relationship%20of%20Credit%20Scores_062003.pdf.
10 The letter specifically stated:
Disproportionate impact is a lack of symmetry, or unequal percentages. In other words, disproportionate impact is an uneven distribution of each racial group with a given risk factor, although the uneven distribution is not caused by one’s race. . . . By the nature of risk-based pricing and underwriting, all factors used in insurance have a disproportionate impact to some extent. One could make a convincing argument to ban the use of all risk-related factors based solely on disproportionate impact. Effectively, we would ban risk-based pricing and underwriting and revert to a pricing system where we homogenize the risk and essentially charge everyone the same price—regardless of risk. That would be a set-back to all Texans, of all races, especially those of moderate to lower income whose risk remains low.
As Commissioner, I have the authority to end a practice that is either unfairly or intentionally discriminatory. However, I do not have a legal basis to ban a practice that has a disproportionate impact if it produces an actuarially supported result and is not unfairly or intentionally discriminatory. Prior to the study, my initial suspicions were that while there may be a correlation to risk, credit scoring’s value in pricing and underwriting risk was superficial, supported by the strength of other risk variables. Hence, there would be evidence that credit scoring was a coincidental variable that served as a surrogate for an unlawful factor in rating and underwriting. If this were proven to have been the case, I would have had a legal basis to make the connection between disproportionate impact and intentional discrimination, and . . . ban credit scoring outright . . . .
The study, however, did not support those initial suspicions. Credit scoring, if continued, is not unfairly discriminatory as defined in current law because credit scoring is not based on race, nor is it a precise indicator of one’s race. . . .
. . . .
Allowing credit scoring to be used . . . will ensure its link to risk under some of the strongest consumer protections in the nation, especially for people that suffer hardship. However, if the presence of credit scoring in insurance will only feed suspicion and divide us as Texans, its continued use to any degree may simply not be worth it. If the Legislature determines that credit scoring should be eliminated, then I recommend that it be phased out over time.
Letter from Jose Montemayor to the 79th Texas Legislature (Jan. 31, 2005) (accompanying Tex. Dep’t of Ins., Supplemental Report to the 79th Legislature: Use of Credit Information by Insurers in Texas: The Multivariate Analysis (Jan. 2005), http://www.tdi.state.tx.us/reports/documents/credit05sup.pdf) (emphasis added).