Case Study 25.1 — Redlining as a Live Theory: The Combating Redlining Initiative

A real, public enforcement program, examined for its analytical method rather than its numbers.

A note on figures before you begin. This case study deliberately contains no settlement amounts, no penalty figures, and no lending statistics. Those numbers exist, they are specific, and they are public — in the complaints, consent orders, and press releases published by the Department of Justice, the Consumer Financial Protection Bureau, and the Office of the Comptroller of the Currency. Read them there. A number recited from a textbook, a trade article, or a compliance seminar is a number you cannot defend if someone asks you where it came from, and this chapter has already told you why that matters. What is portable — what you can carry into your own practice — is the method. That is what this study is about.

Background

In October 2021, the United States Department of Justice announced what it called the Combating Redlining Initiative, describing it as the most aggressive coordinated effort to address lending discrimination in the Department's history. The announcement was made jointly with the Consumer Financial Protection Bureau and the Office of the Comptroller of the Currency, and it was accompanied by the announcement of a resolution involving Trustmark National Bank.

The Initiative's structure matters as much as any individual case. It formalized a referral pipeline that already existed in statute: a federal financial regulator that has reason to believe a lender has engaged in a pattern or practice of discrimination is obligated to refer the matter to the Department of Justice. What the Initiative added was coordination — a standing partnership among the Department, the CFPB, the prudential banking agencies, and state attorneys general, with a shared analytical approach and a public commitment to bring cases.

In the years since, the Department has announced a series of redlining resolutions. In January 2023 it announced a resolution with City National Bank that the Department described as its largest redlining settlement to that date. Subsequent matters have involved banks of varying size and geography, and — importantly for readers of this book — independent mortgage companies as well as depository institutions. The theory is not limited to banks with branch networks. A non-bank lender with loan production offices, a referral-based origination model, and a marketing budget can be, and has been, the subject of the same analysis.

Every one of those matters is documented in a public complaint and a public consent order. Go read two of them end to end. It is the single most educational afternoon available to a working loan officer, and it costs nothing.

The issue

The word "redlining" carries the weight of the history in Chapter 2 — the residential security maps, the underwriting standards of the period, the restrictive covenants. That history is real, documented, and not what these cases are about.

The modern claim is different in structure and, in a specific way, harder to see coming.

The allegation is avoidance. The government alleges that a lender, over a defined period, failed to provide mortgage lending services to majority-Black and majority-Hispanic neighborhoods within the market it actually served — and that the failure was not explained by legitimate business factors. The evidence is the lender's own record: where its offices were, where its loan officers worked, where its advertising ran, whose referrals it took, and where its applications came from.

Notice three things about that formulation.

First, no individual applicant needs to have been denied. A lender can have an underwriting record that is defensible file by file and still face this claim. The applications that would have revealed a problem were never taken, so there is nothing to compare.

Second, no one needs to have said anything. Internal communications appear in these complaints when they exist, and they are damaging when they do. But the core of the case is built from structured data, not from statements.

Third, the market is defined by the lender's own behavior. This is the piece practitioners most often misunderstand. The government's concept of the reasonably expected market area — the REMA — is derived from where the lender actually marketed, took applications, and lent. It is not the service area the lender declares, and it is not necessarily the assessment area a bank designates for Community Reinvestment Act purposes. A lender cannot narrow the market it will be measured against by declining to serve part of it. That would make the violation into its own defense.

What it shows: the analytical method, step by step

This is the part to learn. Strip away the specific institutions and the same sequence appears in matter after matter.

Step 1 — Define the market

Analysts establish the metropolitan area or areas in which the lender operated, then determine the REMA from the lender's own activity: office and branch locations, the geographic distribution of its applications and originations, its marketing footprint, and the territories assigned to its originators.

Step 2 — Map the record against demographics

Every application and every origination is plotted by census tract. Tract-level demographic data from the Census Bureau is overlaid. Tracts are typically grouped by the share of residents who are Black and Hispanic, with "majority-minority" tracts identified according to a stated threshold.

The lending data used in this step is HMDA data — supplied by the lender itself. This is the point at which §25.8 of the chapter stops being an administrative topic and becomes the evidentiary foundation of the entire case.

Step 3 — Compare against peers

A single lender's tract distribution means little in isolation. The analysis therefore constructs a peer group: lenders of comparable size and product mix, operating in the same market over the same period.

The question is then simple and difficult to answer away: what share of the peer group's applications came from majority-minority tracts, and what share of the subject lender's did? A gap that persists across multiple years, across multiple peer definitions, and across the whole market rather than one corner of it is the analytical core of a redlining case.

Step 4 — Look for the explanation

If a gap exists, the analysis turns to the operational record that would explain it — or that would show the lender knew and did nothing:

• Branch and loan production office locations, and their placement relative to majority-minority tracts.
• Loan officer siting and assignment — where originators were based, what territories they covered, and in some matters where they lived.
• Marketing: total spend, media placement, direct-mail geography, digital audience targeting and exclusions, and whether materials existed in languages other than English.
• Referral-source concentration — which real estate agents, builders, and partners generated the book of business, and what geography those partners served.
• Internal communications, including how prior fair-lending findings and internal analyses were discussed and acted on.
• Prior examination history and remediation.

Step 5 — Test the business justification

A lender is entitled to explain the pattern. Business models differ, product mixes differ, and some lenders genuinely do not operate in the retail purchase market at all. The questions asked are whether the offered justification is legitimate, documented, and substantial, and whether a less discriminatory alternative would have served the same business purpose.

Outcome: what the remedies tell you about the theory

Public redlining resolutions have consistently included a recognizable set of remedial commitments. Without any figures attached, the categories are:

Read that list as a description of the violation rather than a description of the punishment. Every remedy is an affirmative act of market participation. The remedy for having not been somewhere is being there — with an office, with people, with advertising, with capital, and with partnerships.

That is the clearest available statement of what modern redlining is understood to be: a claim about presence and absence, made from a lender's own data, and remedied by presence.

The lesson for a loan officer

You will never build a REMA. You will never run a peer analysis. And yet almost every input in Step 4 of that method is a description of an individual originator's working habits, aggregated.

Where you prospect. Which agents you cultivate and which calls you do not return. Whether your marketing reaches the whole market or the slice of it you find comfortable. Whether you have ever taken an application on a house priced well below your average. Whether the caller who spoke to you in Spanish got the same forty minutes as the caller who did not.

None of those feels like a lending decision. All of them appear on the map in Step 2, with your NMLS identifier attached to every row.

Three concrete practices follow, and none of them requires permission from anyone:

1. Know where your applications come from. Pull your own last fifty and map them by tract. If you cannot do that, ask your operations team; the data is in your loan origination system.
2. Widen the top of the funnel deliberately. Referral concentration is a business risk and a fair-lending risk, and the fix for both is the same: more sources, more varied, chosen on purpose. Chapter 38 makes the business case independently.
3. Never let a small loan be beneath you. The most common quiet mechanism by which an originator's book comes to mirror a redlining pattern is a personal minimum loan size that exists nowhere in writing.

Discussion questions

1. The Initiative's cases are built primarily from data the lenders themselves reported under HMDA. Argue both sides: does that make the enforcement theory more legitimate, or does it make compliance a matter of managing a dataset? What would you need to know to decide?

2. A lender argues that it is a wholesale-focused shop with no retail branches and therefore should not be measured against retail peers. Where in the five-step method does that argument belong, and what evidence would make it persuasive rather than convenient?

3. The REMA is defined by the lender's actual activity rather than its stated service area. Construct the strongest argument against that definition, then answer it.

4. Every remedy category in the outcome table is an affirmative obligation. Is that the right response to a finding of avoidance? What alternative remedy structure would you propose, and what would be lost?

5. You are a branch manager. Your top producer's book is heavily concentrated with three agents in one part of the metro, and their production is excellent. What do you do — specifically, this quarter — that addresses the fair-lending exposure without destroying the producer's business?

6. This case study contains no numbers. Was that the right editorial decision for a textbook? What would you have gained by including settlement figures, and what would you have risked?

Sources: the public complaints, consent orders, and announcements of the U.S. Department of Justice, the Consumer Financial Protection Bureau, and the Office of the Comptroller of the Currency; the Interagency Fair Lending Examination Procedures; HMDA data published by the CFPB. All characterizations of specific matters above are drawn from those public announcements. Verify current status and all figures at the source; enforcement matters develop and terms are sometimes modified.