Changing Demographics and the Impact to CRA (and Fair Lending)

We all know that the Community Reinvestment Act of 1977 (CRA) requires banks to meet the credit needs of the communities in which it operates. Other than the substantive changes in 1995 and 2005 and the public release of interagency CRA Questions & Answers, CRA has largely remained stable. But there is something brewing under the surface….

For decades, the demographics of individual counties and the census tracts that comprise them have largely remained static between decennial census periods (e.g., 1990 to 2000). That is no longer the case with the implementation of the American Community Survey (ACS) results that are released by the U.S. Census Bureau. From 2016 to 2017, the ACS resulted in significant changes to the income designations of census tracts across the U.S. and its territories. The income designations of census tracts (low-, moderate-, middle-, upper-, and at times unknown-income levels) align with CRA examinations and are pertinent to the Geographic Distribution criterion. The Geographic Distribution criterion is applicable to the Lending Test applied to banks of all sizes, with only a handful of exceptions (certain Strategic Plan and Wholesale/Limited Purpose banks). The Geographic Distribution criterion measures the dispersion of a bank’s loans within census tracts with the varying income designations, with emphasis placed on performance in low- and moderate-income (LMI) census tracts.

How much changed?

For 2017, there were over 75,000 census tracts across the U.S. and its territories. Of these census tracts, the ACS resulted in changes to the income designations of over 21,000 census tracts, alarmingly accounting for 28 percent of all census tracts.

What changed?

The census tracts with income designation changes resulted in a combination of both upward and downward trends. For example, some census tracts moved from low- up to moderate-income while others moved from middle- down to moderate-income. In some cases, the income designations changed by more than one level (i.e., from low- up to middle-income or from upper- down to moderate-income).

A Case Study

Let’s take one bank as an example. This bank has total assets of about $1 billion and it operates in 10 delineated assessment areas that cover portions of three states. These 10 assessment areas are comprised of 22 individual counties that contain a cumulative 518 census tracts. The ACS changes between 2016 and 2017 resulted in income designation changes to 174 of the 518 census tracts (34 percent). At the individual county level, 16 of the 22 counties (73 percent) had census tract income designation changes of 25 percent or greater, and 19 of the 22 counties (85 percent) had census tract income designation changes of 20 percent or greater. That bank’s Geographic Distribution baselines changed dramatically between 2016 and 2017. Has your bank been similarly affected?

Why does it matter?

There is a high probability that at least some of the census tract income designations of your delineated assessment area(s) have changed. Geographic Distribution is a major component of your CRA performance under the Lending Test, and you should be aware of the underlying changes to the demographics of your assessment area(s). For some banks, you have received credit for the loans that you have been making in LMI census tracts. Are they still categorized as LMI census tracts today? For other banks, your assessment area(s) may not have contained any LMI census tracts? Do they contain LMI census tracts now? You must be prepared for upcoming CRA examinations.

What about Fair Lending?

The ACS data described above also changed the minority concentration of census tracts across the U.S. and its territories. The minority concentration of census tracts is a critical component of redlining analyses conducted by the federal regulators. While the changes to the minority concentrations were not as significant as they were for CRA’s income level designations, each bank should be aware of the impact that ACS changes will have on the its Fair Lending risk profile.

How to react.

Analyze the demographics of each assessment area to determine the changes to the CRA income level designations and the minority concentrations of each census tract. The delineation of assessment areas should also be evaluated annually, based on your lending patterns, for both CRA and Fair Lending purposes. These measures, in concert with a system of self-evaluation, will help to ensure there are no surprises on future compliance and CRA examinations.

The use of “counties” also accounts for parishes, boroughs, the District of Columbia, and independent cities in the states of Maryland, Missouri, Nevada, and Virginia.
Particularly the 2012-2016 ACS 5-Year Public Use Microdata Sample files that were released on January 18, 2017.
The U.S. territories include American Samoa, Guam, Commonwealth of the Northern Mariana Islands, Puerto Rico, and the U.S. Virgin Islands.
CRA also includes an income designation for non-metropolitan, middle-income census tracts that are distressed and/or underserved.
The 2017 demographic changes are expected to remain in effect going forward until the release of 2020 Census data.

Michael Wallace

About Michael Wallace

Michael P. Wallace is a former regulator and is now President of Wallace Consulting Co., LLC. Wallace Consulting Co., LLC is dedicated to mitigating the CRA, Fair Lending, and HMDA risks at banks of all sizes. Contact Michael at (501) 508-5049 or mwallace@wallaceconsultingllc.com.

2 Responses to Changing Demographics and the Impact to CRA (and Fair Lending)

  1. Hi – do you know why some census tracts are noted as being “unknown” as it relates to income levels?

    • Hi Susan, sorry for the late reply! I didn’t see that I had a comment pending…

      The FFIEC page states, in essence, that “unknown income” is when the median income is 0 across a tract. So, my guess would be that a) nobody lives in the area, b) they did not fill out the census information, or c) there was not enough information in the tract to calculate a median income. Here’s the website. https://www.ffiec.gov/geocode/help3.aspx

Comments