Non-English speakers represent a growing and important segment of the U.S. population, yet they remain an oft-ignored group in housing and mortgage policy discourse. According to the 2013 American Community Survey, nearly 62 million people—approximately 1 in 5 households—do not speak English at home, representing nearly a 100 percent increase since 1990. While Spanish speakers constitute the largest group, with more than 38 million speakers, millions more speak other languages. Language barriers and immigration status pose overlapping and distinct challenges in housing and mortgage access, and though non-English speakers are not exclusively immigrants, the challenges faced by non-English speakers are likely to be exacerbated among immigrants. This article examines the barriers non-English speakers face in mortgage access, and provides concordant policy recommendations.
Limited Information in the Primary Language
Non-English speakers have unequal mortgage access due to limited information available in their primary language. The process of obtaining a mortgage is often lengthy and requires the ability to communicate with credit grantors and credit counselors, complete required documentation, and understand credit disclosures. The National Council of La Raza cites a lack of consumer-oriented help as the primary barrier to homeownership among Spanish-speaking communities, and several prominent lawsuits highlight these challenges. For example, in 2002 the Federal Trade Commission brought suit against a debt collection company for failing to make clear and conspicuous disclosures to their Spanish-speaking customers. Some cases, such as Brooklyn Union Gas Co. v. Jimeniz (1975) and Sitarz v. Drexel Burnham Lambert Inc., et al. (1991) resulted in void loan contracts because the borrower did not understand the language of the loan obligation.
Though a growing number of mortgage lenders provide services in Spanish to increase market share among Hispanics, lenders vary widely in their willingness and ability to solicit, underwrite, and grant loans in Spanish. Services provided to non-Spanish speaking language minorities are even more limited. To date, only Arizona, California, the District of Columbia, Oregon, and Texas have adopted any legislation requiring the translation of mortgage loan documents into the borrower’s primary language.
Non-English speakers also may experience discrimination at multiple stages throughout the mortgage lending process, including the initial advertising and outreach stage, the pre-application and inquiry stage, the loan approval/denial stage, setting of the terms and conditions, and throughout loan administration. Discrimination may take the form of differential treatment discrimination—that is, treating borrowers differently based on their language background—or disparate impact discrimination—that is, engaging in lending practices that disparately impact non-English speakers but do not serve a legitimate business purpose. Lending practices that disparately impact minorities are considered discriminatory under federal law if they do not serve a legitimate business purpose, or if they cannot be replaced by alternative policies that result in less disparate impact.
Significant evidence documents widespread mortgage discrimination on the basis of race and national origin—two factors that are highly correlated with language minority status—though direct evidence of discrimination on the basis of language background is relatively limited because federal policies do not currently regulate discrimination on the basis of language status. To date, only a single legal case study documents credit discrimination against non-English speakers specifically. In 2001, the Justice Department settled a lawsuit against Associates National Bank for holding Spanish-language credit card applicants to stricter underwriting standards and excluding them from promotional credit services with more favorable terms and conditions.
Sizable socioeconomic and credit score gaps between English and non-English speakers likely explain a significant portion of the mortgage access gap. English fluency is one of the predominant factors explaining income gaps between the white and Hispanic populations, and English fluency is also a strong predictor of homeownership and associated home equity wealth. In addition, non-English speakers are less likely to establish banking relationships with formal financial institutions, which would enable them to establish credit histories and obtain mortgages. Though no studies currently examine credit score differences on the basis of primary language status, a recent study found that average credit scores for Hispanic homeowners are 15.5 points lower than for white homeowners.
Further, the majority of language minorities are immigrants who face unique hurdles in accessing credit. In addition to being 13 percent less likely than native-born populations to have checking accounts, recent immigrants do not have established credit histories in the U.S., and will encounter difficulties obtaining credit. Recent research from the Federal Reserve reports that credit scores produce a substantial negative disparate impact on recent immigrants, whose shorter credit histories reflect their recent arrival in the country. Disparate impact also may result from high wealth transfer rates among immigrants. Nearly 40 percent of immigrant households send money to family and friends in their country of origin; in 2005, transfers from the U.S. to Latin America and the Caribbean amounted to more than $40 billion. These transfers and the costs of sending money can substantially reduce immigrant wealth levels and exacerbate the difficulties faced by immigrants in accessing mortgage credit. Another hurdle is that credit scores are linked and located using Social Security Numbers; immigrants without an SSN often cannot obtain credit because they cannot be located in credit bureau files.
It is difficult to determine whether business necessity justifies lending practices that disparately impact non-English speakers. Though the Home Mortgage Disclosure Act (HMDA) requires mortgage lenders to report information on certain borrower demographic characteristics, HMDA does not require lenders to report borrower language status or several other loan underwriting variables, such as credit score. As a result, it is impossible to examine disparate impact on non-English speakers and/or to distinguish whether mortgage disparities result from discriminatory practices or from systematic differences in other characteristics used in loan underwriting.
A number of consumer advocacy organizations promote fair lending initiatives and consumer credit education, including HUD, the Consumer Financial Protection Bureau, the National Fair Housing Alliance (NFHA), the Center for Responsible Lending (CRL), and the Community Reinvestment Coalition (CRC). While many of these organizations offer services in Spanish, their efforts could be expanded to serve members of other language minority communities and to advocate on behalf of expanding federal and state policies requiring that mortgage lenders provide disclosures in the borrower’s primary language. Though HUD and the NFHA frequently conduct audits to identify racial discrimination in housing and mortgage lending, these efforts should be expanded to document and address discrimination against non-English speakers. To specifically protect non-English speakers against discrimination, the Equal Credit Opportunity Act could be amended to include language status as a protected class. Expanding HMDA data reporting requirements to include additional data on mortgage applicant demographic, financial, and credit characteristics will enable policymakers and researchers to disentangle whether disparities in mortgage approval and pricing among non-English speakers and/or immigrants may be attributed to discrimination or to other factors underlying the loan underwriting process. Though the Consumer Financial Protection Bureau proposed substantial expansions to HMDA data reporting requirements in 2014, the proposal currently does not include plans to collect information on mortgage applicants’ primary language or immigration status.
Finally, mortgage lenders should examine their underwriting criteria to determine whether lending policies accurately reflect credit risk. It may be possible to relax certain underwriting criteria if they disparately impact non-English speakers but do not appear related to loan performance. Increasingly, large national banks are accepting the Income Tax Identification Number (ITIN) from immigrants and the Matricula Consular identification card from Mexican immigrants in lieu of an SSN, and are employing ITIN data from federal tax returns to underwrite mortgage loans. In addition, mortgage lenders should investigate alternative instruments for evaluating credit risk for borrowers with “thin” or nonexistent credit histories. These alternative instruments include the Fair Isaac Expansion Score, the First American Corporation Anthem Score, the eFund DebitBureau tool, and the Payment Reporting Builds Credit (PBRC) system. The Expansion and Anthem scores use information such as credit bureau data from thin credit history files, supplemented with data on borrower rental history, utility and insurance payments, child care expenses, and information from credit-granting agencies that do not report to national credit data repositories. The eFund DebitBureau tool provides lenders with information about checking account behavior, and includes data on bounced checks, overdrafts, and check reorders. The PBRC system allows households to improve their credit scores by providing proof of timely bill and rent payments. These alternative measures of credit risk may expand credit in underserved markets serving language minority and immigrant households.