Opinion
MASTER FILE NO. 08-MD-1974, MDL No. 1974.
October 13, 2011
MEMORANDUM OPINION AND ORDER
Before the Court is a Motion for Class Certification filed by plaintiffs Gillian Miller, Lakisha Austin, Arthur and Luella Davis, Gabriel Garcia, Sonia Jenkins, Anton Griffin, German Pena, Bernice Smith, Guadalupe Marchan, and Rosendo Quinones ("Plaintiffs"). Defendant Countrywide Bank, N.A. ("Countrywide") opposes the motion, and both parties have filed supplemental briefs discussing the impact of Wal-Mart Stores Inc. v. Dukes, 131 S. Ct. 2541 (2011) on the question of class certification. After reviewing all of the submissions and for the reasons set forth in the opinion below, the Court will deny Plaintiff's Motion for Class Certification.
Also listed as defendants are Countrywide Home Loans, Inc., Countrywide Correspondent Lending, Countrywide Financial Corp., Full Spectrum Lending, Inc., Loans for Residential Homes Mortgage Corp., and Summit Mortgage, LLC. Bank of America acquired Countrywide's home lending and mortgage servicing businesses on July 1, 2008.
I.
Plaintiffs are African-American or Hispanic consumers who obtained one or more home mortgage loans from Countrywide between 2005 and 2007. Pls.' Class Cert. Report ¶ 72. They allege Countrywide discriminated against them by charging higher rates and other costs on home mortgage loans than loans provided to similarly-situated non-minority customers. According to Plaintiffs, a facially discriminatory policy did not produce this unfair treatment. Instead, their theory is that Countrywide's race-neutral policy of allowing its loan officers and third-party mortgage brokers to exercise circumscribed discretion to increase the cost of any given mortgage loan produced an unlawful disparate impact on minority borrowers. They seek to certify a class defined as "[a]ll African-American and Hispanic borrowers to whom Countrywide originated a residential-secured loan, including correspondent loans, between January 1, 2002 and the present." Pls.' Mot. Class Certification 18.
Countrywide was one of the nation's largest mortgage lenders during the class period. Countrywide originated loans through three primary channels: retail, wholesale, and correspondent. In the retail channel, Countrywide employees ("loan officers") originated mortgage loans through brick-and-mortar branch offices and call centers. In the wholesale channel, third-party brokers originated mortgage loans through Countrywide's Wholesale Lending Division. The broker would collect information from the borrower and submit it to Countrywide for underwriting and approval, then negotiate any additional broker fees with the borrower. Countrywide did not control how brokers charged such fees, except to place a cap on the total amount of fees it would allow on a Countrywide loan. Finally, Countrywide also purchased completed loans through its correspondent lending channel. Countrywide conditioned purchase of these loans on conformity with its underwriting guidelines.
Countrywide's loan pricing contained both objective and subjective components. As Plaintiffs describe it, the objective component used a number of risk-based underwriting criteria-for example, credit score, income, property value, and loan amount, among others-to establish a "par rate" interest rate. Loan officers and mortgage brokers then had discretion to adjust this par rate and add fees, subject to a cap. Increasing the interest rate and fees on a loan also increased the compensation to the broker or loan officer. Likewise, correspondent lenders had discretion to add charges and increase interest rates and Countrywide paid more for higher-rate loans. The total cost of the loan to the borrower, expressed as an Annual Percentage Rate ("APR"), was a combination of the par rate established through a standardized underwriting process and the discretionary fee and rate mark-ups negotiated between the borrower and loan officer or broker.
Plaintiffs rely upon the report of Dr. Ian Ayres, a legal scholar and economist well-versed in statistical analysis, in support of their disparate impact claim. Based on data obtained from Countrywide on nearly three million loans made between July 2005 and June 2007, Dr. Ayres concludes that the data "show[] that minorities paid more for Countrywide mortgage loans than whites with similar risk-characteristics." Pls.' Class Cert. Report ¶ 10. More specifically, his regression analysis finds that African-Americans and Hispanics paid 11.64 and 12.50 basis points more in APR, respectively, than similarly-situated whites. The analysis controlled for a host of risk-based factors. Dr. Ayres notes that Countrywide regularly performed similar tests to measure potential disparities in loan costs between minority and white borrowers, using similar regression techniques. These tests also showed African-Americans and Hispanics received higher-cost loans than white borrowers with similar risk characteristics. Pls.' Class Cert. Report ¶ 70.
A "basis point" is equal to one hundredth of a percentage point and is used to denote differences in interest rates.
Dr. Ayers' preferred model to test for disparate impact controlled for the same factors as other models-including month of interest rate lock, lien status (subordinate v. first), and credit score-as well as loan amount, loan type (government, conventional, HELOC, unknown), loan product category (including bond product and the interaction of loan term, initial fixed-rate period for ARMs, interest-only, pay option, and balloon payment presence), LTV bin (for first lien loans), CLTV bin (for subordinate lien loans), total debt-to-income ratio bin, cash-out refinance, property type interacted with occupancy type, self-employed borrower or co-borrower, lender-paid mortgage insurance, escrow waiver, presence of a co-applicant, prepayment penalty presence and length, documentation type, credit report items (bankruptcy, foreclosure, obligation, delinquency, and lawsuit), concurrent funding/close, state, and metropolitan area (MSA). Pls.' Class Cert. Report Table 5.
Plaintiffs challenge the subjective component of Countrywide's loan pricing, which they term the "Discretionary Pricing Policy." They do not object to the underwriting process that produces the par rates. Plaintiffs allege that giving loan officers and brokers pricing discretion, while tying their compensation to higher rates and fees, caused minority borrowers to pay higher APRs on mortgage loans than similarly-situated non-minority borrowers. Based on this alleged disparate impact, Plaintiffs claim that Countrywide has violated the anti-discrimination provisions of the Equal Credit Opportunity Act, 15 U.S.C. § 1691, the Fair Housing Act, 42 U.S.C. § 3601, and the Civil Rights Act, 42 U.S.C. §§ 1981- 1982.
II.
Before certifying a putative class, the Court must perform a "rigorous analysis" to ensure the class first meets the requirements of Rule 23(a) and then one of the Rule 23(b) subsections. Reeb v. Ohio Dep't of Rehab. Corr., 435 F.3d 639, 644 (6th Cir. 2006). In light of the Supreme Court's Wal-Mart decision, Plaintiffs seek only injunctive relief under 23(b)(2) and monetary relief under 23(b)(3). However, the Wal-Mart holding as to 23(a) requirements in the context of a disparate impact theory of discrimination obviates any need for discussion of 23(b), as Plaintiffs in this case do not clear the 23(a) threshold. See Wal-Mart, 131 S. Ct. 2541, 2550-57 (2011).
A.
Rule 23(a) imposes four prerequisites before a party may proceed as a representative of a class: numerosity, commonality, typicality, and adequate representation. Coleman v. Gen. Motors Acceptance Corp., 296 F.3d 443, 446 (6th Cir. 2002) (citing Fed.R.Civ.P. 23(a)). The class must be "so numerous that joinder of all members is impracticable . . . [with] questions of law or fact common to the class." Fed.R.Civ.P. 23(a). The representative parties' claims or defenses must be typical of those of the class and the representative parties must be able to "fairly and adequately protect the interests of the class." Id. These requirements work to "limit the class claims to those fairly encompassed by the named plaintiff's claims." Wal-Mart, 131 S. Ct. at 2550 (quoting Gen. Tel. Co. of Southwest v. Falcon, 457 U.S. 147, 156 (1982)).The proposed class easily meets the numerosity requirement. Therefore, the requirements of commonality and typicality become the focus of the Court's attention. Though these two factors tend to merge, Falcon, 457 U.S. at 157, the "crux of this case," as in Wal-Mart, "is commonality." Wal-Mart, 131 S. Ct. at 2550. "Commonality requires the plaintiff to demonstrate that the class members 'have suffered the same injury.'" Id. at 2551 (quoting Falcon, 457 U.S. at 157). It is not enough merely to claim that all class members suffered a disparate-impact injury is insufficient; "their claims must depend upon a common contention," the determination of which "will resolve an issue that is central to the validity of each one of the claims in one stroke." Id. A court's 23(a) analysis will frequently "overlap with the merits of the plaintiff's underlying claim." Id. (citing Falcon, 457 U.S. at 160).
B.
Here, the underlying claim is that Countrywide's "Discretionary Pricing Policy," though facially race-neutral, had a disparate impact on minority borrowers. Disparate impact suits, like disparate treatment suits, are claims for unlawful discrimination, which "are often by their very nature class suits, involving classwide wrongs." Falcon, 457 U.S. at 157. However, the mere fact that "racial discrimination is . . . class discrimination" does not define a class of persons that have suffered the same injury as a representative plaintiff. Id. The putative class representative must demonstrate that the class members suffered the same discrimination that he or she suffered. Class certification will not obtain where the only commonality among members is their race and their relation to the defendant.Plaintiffs' expert calculated a disparate impact by first comparing the average APR paid by African-American and Hispanic borrowers (8.464% and 8.123%, respectively) to the average APR paid by white borrowers (7.602%). Plaintiffs' expert then controlled for "business-justified" factors. See supra n. 3. The disparate impact that Dr. Ayres found is an average of the cost differences between minority borrowers and similarly-situated non-minority borrowers. After controlling for the "business-justified" factors, the difference between the average APR paid by white borrowers and that paid by non-white borrowers narrowed to 11.64 and 12.5 basis points, or 0.116 and 0.125 percentage points.
Although Dr. Ayres bases his model on classwide data, that fact alone does not support a conclusion that every member of the class suffered the same injury, or any injury at all. The analysis does not show that every minority borrower would have paid 11.64 or 12.50 basis points less in APR, but for their minority status. Some may have paid even less than what the average measured disparate impact predicts; some may not have paid less at all. Moreover, the average differences between groups may well be within the range expected from the exercise of non-discriminatory discretion among thousands of loan officers and brokers working from hundreds of separate offices. More would be required to prove any individual or collective claim in these circumstances.
The Wal-Mart decision makes clear that, absent a showing of a common direction or common method of exercising discretion, statistical evidence of average disparities will not suffice to meet Rule 23(a)'s commonality requirement. When plaintiffs are suing "about literally millions of . . . decisions at once," there must be "some glue holding the alleged reasons for all those decisions together," Wal-Mart, 131 S. Ct. at 2552, so that the claims of the representative party and the class are capable of resolution by a "common answer." Id. at 2551.
Here, even accepting Plaintiffs' statistical analysis, they have not "identified a common mode of exercising discretion" to support commonality. Id. at 2554. Too many variables prevent the conclusion that even an unconscious discriminatory motive or thought similarly animated thousands of mortgage rate decisions. However, the idea that thousands of loan officers in hundreds of separate locations around the country would exercise their discretion in a similar discriminatory fashion as to each purported class member defies belief. Whether an individual loan officer or a single office did so, might be a different question.
Although certification of this type of class was more debatable pre- Wal-Mart, the Supreme Court has explained that Rule 23(a) commonality requires more than what Plaintiffs have demonstrated here. Compare Ramirez v. Greenpoint Mortg Funding, Inc., 268 F.R.D. 627, 635 (N.D. Cal. 2010) (pre- Wal-Mart certification of class challenging discretionary mortgage pricing on disparate impact theory), and Barrett v. H R Block, No. 08-10157-RWZ, 2011 WL 1100105, at *5 (D. Mass. March 21, 2011) (same), with In re Wells Fargo Residential Mortg. Lending Discrimination Litigation, No. 08-MD-01930, 2011 WL 3903117 (N.D. Cal. Sept. 6, 2011) (post- Wal-Mart denial of class certification of plaintiffs alleging disparate impact discrimination based on a policy of discretionary mortgage loan pricing). Plaintiffs do not meet the commonality requirement and cannot be certified as a class.
Being otherwise sufficiently advised,
IT IS HEREBY ORDERED that Plaintiffs' Motion for Class Certification is DENIED.
The Court will set a conference in the near future.
October 11, 2011