Abstract

Excerpted From: Paul Arrowood, HMDA Demographics: Racial and Ethnic Discrimination in the Mortgage Application and Approval Process, 27 Transactions: Tenn. J. Bus. L. 91 (Spring 2026) (Full Document)

Buying a home is a confusing and uncertain process for nearly all buyers. That concern is often heightened by race- or ethnicity-based concerns of discrimination during the application process. Congress enacted the Home Mortgage Disclosure Act (HMDA) to address general issues of redlining and housing inequality, and it quickly grew to include the collection and retention of racial and ethnic data on home loan applicants. HMDA data collection is now essential to conveying transparency and public confidence in the mortgage lending process, but its implementation is flawed in some of its most crucial aspects. Visual or surname observation, disclosure of “free form” racial or ethnic category, inconsistent denial labeling in loan data reporting, and other intangibles of attrition early in the process undermine the usefulness of the data. Ultimately, HMDA data is best used for its ability to monitor, calm consumer doubts, and disprove facial discrimination in lending, while serving as a poor legal remedy due to the erosion of disparate impact liability. Congress can make practical updates to HMDA to preserve its monitoring and good faith assurances and minimize the inconsistencies and problems identified throughout this paper.

As demonstrated throughout this paper, HMDA (and the federal regulation enacting it - Regulation C) has heightened awareness of racial disparities for decades in the mortgage industry, particularly the imbalances present before the rigidification of the credit underwriting process in the late 90s, and again before the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (Dodd-Frank Act). Although both racial and ethnic *92 discrimination have decreased and HMDA has improved transparency, the Act still has several shortcomings in its dealings with demographic data. It is both overinclusive and underinclusive of those who are truly denied loans and who may be racially discriminated against in less direct ways, and forces loan originators to make uncomfortable and inexact judgment calls when reporting. All of this occurs while some demographics are told to “fill in the blank” with their information, thereby falling through the cracks of the data grouping system.

This paper examines the practical implications of a cause of action brought under the Fair Housing Act (FHA) or the Equal Credit Opportunity Act (ECOA) in an antidiscrimination lawsuit, while forecasting the bleak chances of success on such a claim in the current political climate. It will also focus on the benefits and drawbacks of the data that is collected under the demographic section of HMDA and Regulation C, and attempt to recommend actionable changes to improve the literal effectiveness of the HMDA, while minimizing the frictions that come with any additional burden in processes or regulation, in an industry that is already extremely highly regulated.

The HMDA has come a long way from its humble beginnings as a method of curbing overt redlining in the 1970s. In its current form, the demographic portion of HMDA is an essential, *120 robust system, tracked through a standardized application process with a large variety of datapoints. However, the demographic collection framework is outdated and often relies on inexact and rudimentary methods to track those specifics.

Instances where disparate treatment (facial discrimination) exists in mortgage lending have largely gone away since the advent of modern underwriting standards and financial legislation. The data tracked by HMDA, combined with case law and political trends, also indicate that disparate impact liability is going by the wayside very soon in the mortgage industry. Due to the weak ““teeth” in disparate impact claims and causal limits from the data that is tracked, the true power of HMDA demographics lies in monitoring, not in litigation. Just as the IRS takes on criticism for not having evidence to rebut discrimination claims, HMDA should be used by the lending industry to increase confidence in the tolerance and goodwill of lenders.

To do this, Congress should use the means described in the proposals of this paper to improve data integrity: promote online applications, remove the visual or surname observation requirement, narrow and clarify when an application may be marked “declined,” and broaden the race and ethnicity categories while removing the “other” category, to improve aggregability. Additionally, Congress, Fannie Mae and Freddie Mac, and individual lenders alike should seek to promote the monitoring aspect of the HMDA demographics as a ward against lender prejudices while using the data as a safeguard and assurance of the open-mindedness and tolerance of the industry to minorities.

There are tensions defined and explained throughout this paper, with the most glaring of these being the tension between a desire for colorblindness and a desire for trustworthy data. Although *121 there are compelling arguments on both sides as well as from the middle, today it is enough to conclude that HMDA demographics are necessary despite their faults. The fear of the unknown can be the worst fear of them all. How to draw the line between legitimate disparate impact and a person who has simply been raised in a good home environment is a very difficult line to draw and seems to be a distinction our legal system is unlikely to parcel out any time soon. Inaccurate and overbroad records are misleading, and if HMDA data is to be collected, it should be done the right way. HMDA should not be abandoned or diluted. Instead, it should be reshaped to capture accurate and modern demographics while acting as a key tool in the CFPB’s toolbox for promoting fairness and confidence in the mortgage industry.