Redlining Analysis: REMA vs CRA Assessment Area, and What Examiners Compare
Published 18 August 2026 · Quoted from the FFIEC Interagency Fair Lending Examination Procedures (August 2009) and the OCC Comptroller’s Handbook, Fair Lending, Version 1.0 (January 2023). Both linked at the foot of the page.
The short answer. A redlining analysis is not run on your CRA assessment area. It is run on your reasonably expected market area — where you actually lent, plus where you could reasonably have been expected to lend. The examination procedures are explicit that the assessment area “may be too limited,” and one step directs examiners to look at minority areas just outside it, on the theory that the institution may be avoiding them.
Which means the boundary you drew for yourself is not the frame of the examination. It is one of the things being examined.
Why the assessment area is the wrong unit
This is the passage that decides the whole subject, and it is worth reading in full because it explains the reasoning rather than just stating the rule:
The logic is that a CRA assessment area is a boundary the institution defined. If the question under examination is whether the institution avoided a geography, then a boundary it drew itself cannot be the only frame for answering it — the drawing of the boundary is part of the conduct at issue. Hence the reasonably expected market area, usually shortened to REMA.
Every step of the comparative analysis for redlining in the Interagency Procedures refers to areas within the institution’s CRA assessment area and reasonably expected market area. And one step goes outside both:
The OCC Comptroller’s Handbook makes the same point about the boundary itself even more directly, instructing examiners to review whether a prior CRA examination found the assessment area to have been influenced by prohibited factors, and then: “If there are minority group areas that the bank excluded from the assessment area(s), include them in the redlining analysis.”
Four geographies that get confused
| Term | What it is | Who defines it |
|---|---|---|
| CRA assessment area | The geography an institution delineates for CRA purposes, generally around its branches and deposit-taking facilities. | The institution — and reviewable, per both documents above |
| Reasonably expected market area (REMA) | Where the institution actually marketed and provided credit, plus where it could reasonably have been expected to. Can be wider than, narrower than, or simply different from the assessment area. | Examiners, for the purpose of the fair lending analysis |
| Market area | Used in the procedures alongside REMA, for the area actually served for a given product. The procedures distinguish market area by product — the phrase used is “market area for residential products.” | Follows the institution’s actual lending, per product |
| MSA / MD | Metropolitan Statistical Area and Metropolitan Division: Census-defined statistical geographies, reported as fields on the HMDA LAR. A unit of reporting, not a definition of anyone’s market. | The federal statistical agencies |
The practical consequence: an institution can be entirely comfortable with its CRA performance inside its assessment area and still have a redlining exposure, because the analysis is not confined to that area and may include geographies it deliberately left out.
What counts as a majority-minority area
The OCC handbook gives a working threshold, and a more granular alternative:
The handbook defines the minority population of an area as the population that is non-White or Hispanic, and adds that analysing concentrations in quartiles — zero to 25 percent, more than 25 to less than 50, more than 50 to less than 75, and more than 75 — may be helpful. If you are running a self-assessment, the quartile view is worth doing alongside the majority split, because it shows a gradient rather than a binary and gradients are what “how does treatment change as you cross this line” questions turn on.
The four categories of evidence
The OCC handbook enumerates what may be relevant to supporting or contradicting a finding of redlining. Paraphrased faithfully, with the distinctive points quoted:
1. Branching and delivery of services. The record of opening and closing branches, the history of providing services in the area, loan officer staffing of branches or regions, and expansion patterns. Where substantial lending goes through brokers or third parties, their locations and the areas they serve count too. And specifically: “Map the location of the bank’s branches and those of any competitors identified as appropriate lending peers.”
2. Marketing. This is stronger than most institutions expect. “Marketing decisions are affirmative acts to include or exclude areas,” and the handbook states that if sufficiently clear and supported by other evidence, “a difference in marketing to racially different areas could itself be treated as a violation of the FH Act.” The scope is broad: geographic distribution of pre-approved solicitations, advertising in local media and directories, business development calls to real estate brokers, meetings with community groups, telemarketing, direct mail and social media strategy. Examiners are also directed to consider whether opportunities to advertise in outlets targeted to high minority areas existed and whether the bank took them.
3. Lending performance compared to peers. Geographic analysis of lending activity, market share analysis and other peer comparisons. The handbook is direct about how much weight this carries: a bank’s lack of lending activity in a minority area as a proportion of its overall lending, compared with peers’ higher proportion of lending in that same area, “provides strong support for a finding of” redlining.
4. The comparative file review. Whether applicants of the same racial or national origin group as the suspected area were treated less favourably — and whether any identified victims applied for transactions in the suspected area. This is the same procedure covered in our separate piece on comparative file review versus matched-pair testing.
Why peer selection is the load-bearing choice
Two of the four categories above turn on the peer set: the branch mapping compares against “competitors identified as appropriate lending peers,” and the lending performance comparison is entirely relative to what peers achieved in the same geography. A redlining finding built on a peer set nobody can defend is a finding built on sand — and equally, a self-assessment that clears you against a conveniently chosen peer set has told you very little.
So the question worth asking internally before an exam is not “how do we look?” but “against whom, and could we justify that choice to someone who picked differently?” Peer analysis and mapping are where that work happens in practice; the geography is easier to argue about when you can see it.
When a redlining analysis is not appropriate
Worth knowing, because it saves work. The Interagency Procedures state that if there are no areas identifiable for their racial or national origin minority character within the institution’s CRA assessment area or reasonably expected market area for residential products, a redlining analysis is not appropriate. They add the alternative: “If there is a substantial but dispersed minority population, potential disparate treatment can be evaluated by a routine comparative file review of applicants.”
Redlining is a geographic argument. It needs a geography to argue about. Where the demographics are dispersed rather than concentrated, the analysis that fits is the file-based one.
If you are running this yourself
- Define the REMA before you look at the numbers, and write down why. Doing it afterwards invites the conclusion that the boundary was chosen to suit the result.
- Include the areas you left out of the assessment area. Both documents direct examiners to them specifically. If there is a defensible reason for the exclusion, document that reason now rather than reconstructing it under examination.
- Look just outside your boundary, per Step 4. A minority area adjacent to but excluded from where you lend is the pattern the step exists to catch.
- Do the marketing review, not just the lending review. It is the category most often skipped in self-assessments and the one the handbook says could constitute a violation on its own.
- Justify your peer set in writing. Two of the four evidence categories depend on it entirely.
Comply’s fair lending analysis scores redlining as one of the twelve FFIEC risk factors and works from the same database as mapping and peer analysis, so the geographic and peer halves of the review draw on the data you already filed. To be straight about the limits: RATA does not sell a dedicated redlining module, and the analysis above is assembled from those components rather than delivered as one packaged workflow. Definitions for REMA, assessment area and the other terms used here are in our compliance glossary.
Sources
- Interagency Fair Lending Examination Procedures, Federal Financial Institutions Examination Council, August 2009 (current; replaced the March 1994 procedures). Quoted from the comparative analysis for redlining, Steps 1 and 4. Copy hosted by NCUA (PDF).
- Comptroller’s Handbook, Consumer Compliance: Fair Lending, Version 1.0, Office of the Comptroller of the Currency, January 2023, as amended 14 July 2025. Quoted from the redlining procedures. Note that an earlier edition is also online and is stamped RESCINDED on every page; read the current booklet. Current booklet at the OCC.
Quotations were read from the primary documents on 18 August 2026. Examination procedures change; verify against the current edition before relying on any quotation here, including ours.
Frequently Asked Questions
What is a reasonably expected market area (REMA)?
It is the geography a redlining analysis is actually run on: the areas where an institution marketed and provided credit, plus the areas where it could reasonably have been expected to market and provide credit. The FFIEC Interagency Fair Lending Examination Procedures introduce it because the CRA assessment area may not be the right unit. Their note says the assessment area "can be a convenient unit for redlining analysis because information about it typically already is in hand. However, the CRA assessment area may be too limited," and that the areas best compared are those where the institution actually lent and where it could reasonably be expected to have lent, some of which "might be beyond or otherwise different from the CRA assessment area."
Is a redlining analysis run on the CRA assessment area?
Not on the assessment area alone. Every step of the comparative analysis for redlining in the Interagency Fair Lending Examination Procedures refers to areas within the institution's CRA assessment area AND reasonably expected market area. One step goes further and looks outside both: examiners are directed to identify minority areas located just outside the assessment area and market area, such that the institution may be purposely avoiding them. The OCC Comptroller's Handbook adds that if there are minority group areas the bank excluded from its assessment area, examiners should include them in the redlining analysis. The boundary you drew is itself part of what is examined.
What counts as a majority-minority census tract?
The OCC Comptroller's Handbook states that in general examiners should consider majority-minority geographies to be those with a minority population of greater than 50 percent, and low minority geographies to be those with a minority population of 50 percent or lower, where the minority population of an area means the population that is non-White or Hispanic. The handbook also notes that analysing concentrations in quartiles, such as zero to 25 percent, more than 25 to less than 50 percent, more than 50 to less than 75 percent, and more than 75 percent, may be helpful.
What evidence do examiners actually weigh in a redlining review?
The OCC Comptroller's Handbook names four categories. Branching and delivery of services: the record of opening and closing branches, loan officer staffing, expansion patterns, third-party and broker locations, mapped against the branches of appropriate lending peers. Marketing: whether the suspected area was omitted from marketing, since marketing decisions are affirmative acts to include or exclude areas. Lending performance compared to peers: geographic distribution of lending activity and market share against peers, where a lower proportion of lending in a minority area than peers achieve provides strong support for a finding. And the file review itself, checking whether applicants of the same group as the suspected area were treated less favourably.
When is a redlining analysis not appropriate?
The Interagency Fair Lending Examination Procedures state that if there are no areas identifiable for their racial or national origin minority character within the institution's CRA assessment area or reasonably expected market area for residential products, a redlining analysis is not appropriate. They add that where there is a substantial but dispersed minority population, potential disparate treatment can instead be evaluated by a routine comparative file review of applicants. Redlining is a geographic argument, so it requires a geography to argue about.
Schedule a Fair Lending Demo Comparative file review vs matched-pair testing
This page may be republished with attribution to RATA Associates and a link to rataassociates.com. If you believe a quotation or reading here is wrong, tell us — we would rather fix it than be cited incorrectly.
