Lower-income families make up 39% of the country but received only 25% of 2024 mortgage originations. That 14-point national gap hides enormous variation, and the county you operate in should set your benchmark, not the headline number.
Nationally, 39% of American families are lower-income, but only about 25% of 2024 mortgage originations went to a lower-income borrower, representing a 14-point gap. That’s the number that tends to get quoted. It’s also the wrong number for almost every individual CRA exam, because the gap varies enormously by county.
- Median county gap: 11.8 points
- National gap: 14 points
- Widest jurisdiction (Puerto Rico): ~33 points
FairPlay mapped the borrower-distribution test, which is the gap between the share of mortgage originations to lower-income borrowers and the share of families that are lower-income. The study looked at every U.S. county using 2024 HMDA originations and 2025 FFIEC census data. The pattern that emerges isn’t random. The gaps are widest in expensive urban markets, along the southern border, and in farm valleys. They’re narrowest across much of the Midwest.
Housing and banking economics at play
The report suggests the gap reflects housing economics as much as banking. In high-cost markets, fewer lower-income families can qualify for today’s home prices regardless of which bank they approach. Several factors likely compound in lower-gap regions:
- Lower home prices relative to income
- Smaller required down payments
- Greater availability of entry-level housing stock
- Smaller appraisal gaps
- Higher baseline homeownership rates among lower-income households
None of that is something a bank’s lending policy can change on its own. A bank in a $1.2 million median-home-price metro is working with a fundamentally different applicant pool than a bank in a market where starter homes run $180,000. This holds true even if both banks are underwriting with identical rigor and identical intent to serve lower-income borrowers.
Why the map itself matters more than its color key suggests
One nuance worth noting: the map centers on the median county’s gap of 11.8 points, not zero. A county shaded as ‘narrower than median’ can still have a real, meaningful lending gap. That just means it’s simply better than most other counties, and not necessarily good in absolute terms. Readers shouldn’t interpret a favorable color the same as a closed gap.
Puerto Rico is a stark outlier, with a median gap of roughly 33 points across its municipalities, which is more than double the median. It is excluded from the map’s projection. Any bank with Puerto Rico exposure should treat it as its own analytical category rather than folding it into a national or regional peer comparison.
What this means for benchmarking your CRA exam
A few considerations when benchmarking your own CRA exam:
- Stop benchmarking against the national 14-point gap. Pull the gap for your specific assessment area counties and compare against that instead.
- Build peer comparisons around banks operating in markets with similar affordability profiles, not simply similar asset size. A bank in Des Moines, for example, faces structurally different lending-test ceilings than a bank in Los Angeles.
- When presenting borrower-distribution results to examiners or your board, frame performance against the local benchmark explicitly, since examiners themselves assess each bank against the demographics of the communities it serves.
- Flag high-cost or low-affordability assessment areas early in exam prep, which points to a gap that would be alarming in Ohio, but may be close to the local norm in coastal California.
The underlying point connects back to the first three findings in the study: CRA outcomes are shaped by what a bank does, but also by where it operates. Geography sets the ceiling and floor for what’s achievable on the lending test. This is exactly why the report also finds that community development, not lending volume, tends to be where differentiation actually happens as banks scale.
This closes our four-part series on the CRAi Insights Report. Read the full report, which includes more than 35 years of Outstanding-rate history, at fairplay.ai/crai-insights. Find our previous posts here:
Finding 1: Why Large Banks Achieve Outstanding CRA Ratings 10x More Often Than Small Banks
Finding 2: The Secret to an Outstanding CRA Rating
Finding 3: Banks with Outstanding CRA Ratings Deploy Capital More Intentionally
Sources: 2024 HMDA mortgage originations; FFIEC 2025 census data. Family shares interpolated from FFIEC income brackets, accurate to approximately one percentage point. Counties with fewer than 25 originations are excluded. From FairPlay’s CRAi Insights Report, Volume One (fairplay.ai/crai-insights).
