Community Reinvestment Act (CRA) compliance and fair lending risk management are closely linked, yet the requirements can pull in different directions. Nowhere is this tension clearer than in the delineation of a bank’s CRA assessment area (AA). For CRA purposes, the AA must include the bank’s facilities and areas of substantial lending activity while avoiding the arbitrary exclusion of low- or moderate-income (LMI) census tracts. At the same time, fair lending analysis—particularly redlining reviews—scrutinizes whether the bank’s geographic footprint, marketing, and lending patterns exclude majority-minority census tracts (MMCTs). These two sets of tracts do not always overlap, and both frameworks require the bank to define areas it can reasonably and legitimately serve.
Getting the delineation right is therefore more than a technical mapping exercise. It is a critical control that can either reduce residual risk or create examination exposure under both regimes.
The Dual Requirements and the Core Tension
Under CRA rules, a bank must delineate one or more assessment areas consisting of whole geographies - typically whole census tracts or, for larger institutions under updated standards, whole counties. The AA must encompass the main office, branches, and deposit-taking facilities, plus surrounding areas where the bank has originated or purchased a substantial portion of its loans.
Critically, the AA may not arbitrarily exclude LMI tracts; however, they may take into account the bank’s size, financial condition, and capacity. Examiners review the delineation for compliance with these limitations but do not score the boundary choice itself as a performance criterion.
Fair lending analysis, guided by the Interagency Fair Lending Examination Procedures, treats assessment area delineation as a potential redlining risk indicator. Redlining occurs when a lender provides unequal access to credit—or unequal terms—because of the race, color, national origin, or other prohibited characteristics of the residents of an area. Detection methods rely heavily on HMDA data, peer comparisons of application and origination rates in MMCTs (generally tracts that are more than 50 percent minority), multi-year trend analysis, and geographic mapping that looks for “doughnut” or “horseshoe” patterns of activity surrounding but avoiding minority areas. An AA that appears drawn to exclude high-minority geographies can support a finding of potential redlining when combined with lending disparities, limited branching, or marketing gaps.
The practical difficulty is that LMI tracts and MMCTs are not perfectly aligned. A bank that carefully includes all LMI tracts to satisfy CRA may still leave out nearby MMCTs that are not LMI, creating a fair lending vulnerability. Conversely, expanding the AA to capture every MMCT may produce an area larger than the bank can reasonably serve given its size, product mix, delivery channels, and competitive position—raising both CRA capacity questions and the risk of underperformance metrics that themselves become redlining indicators.
Key Challenges in Practice
Several recurring pitfalls arise at this intersection:
- Misalignment of demographic layers. Mapping only LMI tracts can leave majority-minority areas outside the AA even when those areas are contiguous to the bank’s facilities or historical lending. Peer comparisons then show shortfalls relative to institutions that do serve those tracts.
- Capacity versus appearance. Expanding the AA to eliminate any exclusion risk can dilute the bank’s ability to demonstrate strong performance inside the area, while a tighter footprint invites questions about intentional avoidance.
- Inconsistent treatment of the “reasonably expected market area.” Fair lending analysis examines the broader REMA—the area where the bank could reasonably be expected to market and lend—alongside the formal CRA AA. A material mismatch between the two can undermine explanations for lending patterns.
- Multi-year and visual evidence. Regulators look for patterns over multiple years. A single year’s data may be explainable; persistent underperformance in MMCTs relative to peers, especially when maps show clear avoidance, is harder to defend.
- Documentation gaps. Without contemporaneous analysis of capacity, demand, competition, and product suitability, examiners may view boundary choices as post-hoc rationalizations.
These challenges are heightened for banks with hybrid delivery models, limited branch networks, or rapid growth into new markets.
Keys to Avoiding the Pitfalls
A disciplined, documented, data-driven approach can reconcile the two frameworks and reduce residual risk. The following practices have proven effective:
1. Start with dual-layer geographic analysis. Map both LMI tracts and MMCTs (and, where relevant, high-minority tracts) against the bank’s facilities, historical loan applications and originations, and peer activity. If there is one particular minority group that is prominent in the area, other breakdowns should be considered as well. Use current Census and MSA boundaries. Identify overlaps, gaps, and contiguous areas that the bank can reasonably reach given its size, channels, and product set. This analysis should inform—not merely justify after the fact—the proposed AA.
2. Document capacity and legitimate business constraints rigorously. CRA explicitly allows consideration of the bank’s size and financial condition when assessing whether LMI exclusion is arbitrary. Fair lending explanations similarly require credible, non-prohibited-basis reasons for differences in treatment. Maintain contemporaneous records of delivery-channel limitations, product suitability, competitive intensity, housing stock characteristics, and applicant flow. Avoid relying solely on “we do not have branches there” without supporting market analysis.
3. Align the CRA AA with the reasonably expected market area used for fair lending. Treat the formal AA and the REMA as related but not identical constructs. Where the bank’s marketing, broker relationships, or digital channels create a broader expected market, ensure the AA does not create unexplained exclusions of MMCTs within that market. Periodic reconciliation of the two geographies prevents later inconsistencies during examinations.
4. Incorporate peer and trend monitoring from the outset. Once the AA is set, monitor application and origination percentages in MMCTs (and LMI tracts) against appropriately defined peers—typically institutions of similar volume operating in the same MSA or market. Track results over multiple years. Early detection of persistent shortfalls allows course correction through targeted outreach, product adjustments, or refined marketing before patterns harden into examination findings.
5. Use mapping and visual analytics as a control, not merely a reporting tool. Regularly plot applications, originations, and physical presence against demographic layers. Look specifically for doughnut or horseshoe patterns. Visual evidence that activity surrounds but avoids MMCTs is difficult to overcome even when statistical disparities are modest.
6. Review and update the AA as the bank evolves. Growth, acquisitions, new delivery channels, or shifts in lending volume require reassessment. Document the rationale for any boundary changes. Institutions should confirm that updated delineations continue to satisfy both CRA technical rules and fair lending risk considerations.
7. Embed the process in a broader Compliance Management System. AA delineation should not sit in isolation. Link it to ongoing fair lending monitoring, CRA performance tracking, board reporting, and training. Residual risk after controls should be quantified and communicated in plain language to senior management and the board.
Summary
CRA assessment area delineation and redlining risk management share a common foundation: the expectation that banks serve the communities in which they operate without excluding areas on prohibited bases or arbitrarily overlooking LMI neighborhoods.
The demographic misalignment between LMI tracts and majority-minority tracts, combined with real capacity constraints, creates genuine operational challenges. Banks that treat delineation as a static mapping exercise expose themselves to findings under both frameworks. Those that approach it as a dynamic, dual-purpose risk-management process—supported by rigorous data analysis, contemporaneous documentation, peer monitoring, and clear alignment between the formal AA and the broader market—position themselves to satisfy CRA obligations while minimizing fair lending exposure.
Proactive institutions do not wait for an examination to test the coherence of their geographic footprint. They continuously evaluate whether the areas they have committed to serve are both inclusive of the communities they are expected to reach and realistic given their ability to deliver credit safely and soundly.
That disciplined approach remains the most effective way to navigate the intersection of these two critical regulatory regimes.
