You are here

DOJ and CFPB Reach Settlement with BancorpSouth over Redlining

Adam Rust's picture

Posted June 30, 2016

News that BancorpSouth has reached a settlement with the Department of Justice and the CFPB over redlining should make people question how well the Community Reinvestment Act, as currently written, can prevent a bank from deciding to walk away from low-income and minority neighborhoods located within its service area. 

Redlining describes the practice of a bank to avoid making loans in certain neighborhoods consciously. The term derives from the red lines used by regulators to draw "residential security maps" which were meant to serve as guidelines for underwriting. It contributed to the disinvestment in urban areas and in particular to neighborhoods of color. In the 1970s, community groups from Chicago organized to protest the practice both locally and across the country. 

The settlement's details, which reveal how the bank made an institutional practice of redlining minority neighborhoods, are unusual in the concrete nature of the allegations. In an era where a claim of bias can often be explained away by pointing to a logarithm, redlining accusations are difficult to prove. 

The details of the DOJ's complaint make a strong case to support their argument:

96.5 percent of loan officers working in Memphis were white. 

When comparing loans made to African-Americans with those originated to whites, after controlling for credit characteristics, the African-American borrowers paid an additional 30.3 basis points in interest on first lien mortgages and 63.9 basis points on second-lien mortgage loans.

During DOJ testing, the same loan officer told a white tester that "620 is the minimum" credit score, but then told a black tester that "normally you want to have a 640 credit score."

The DOJ complemented the preceding points with excerpts from audio recordings made at staff meetings.

  • A manager told loan officers and processors that mortgage applications from minorities and other protected classes should be turned down in 21 days, and that "borderline" customers should be turned down quickly, but loans from white applicants could be excluded from the shortened review period.  
  • When asked by a person at the meeting as to why there would be a race-based policy in place, the manager said: "I think it's lawsuits, it's lawsuits, and we just dodged a really large bullet." The "bullet" described was a threat made by a minority applicant to sue the bank. The lawsuit was not filed. 
  • The use of the n-word when discussing the recent hire of an African-American.
  • When asked to explain the bank's 'race-based denial policy," a loan comments that "they need to get their credit up" and "stop paying their damn bills late." 

All of these details serve to underscore how BancorpSouth had a systemic problem with race.

Part of the Issue is the Construction and Enforcement of the CRA Assessment Area

The FDIC allowed BancorpSouth to pick the areas where its CRA performance would be assessed, and as a result, the regulator made it possible for the bank to implement its disinterest in serving the credit needs of minority neighborhoods without running afoul of the law. Bancorp South drew the assessment areas for its 2013 CRA exam in such a way as to exclude every high-minority neighborhood and 137 of 142 majority-minority neighborhoods in the Memphis MSA. To do that, the bank had to draw most of downtown Memphis outside of its assessment area while also including most of the region's wealthier suburbs inside its AAs. Fewer than half of the residents living inside the boundaries of the Memphis MSA are white, and in Shelby County, where eight percent of the metropolitan area's minority population lives, only 27.5 percent are white. 

But even though the bank was redlining, BancorpSouth received a "satisfactory" in its 2013 exam.  The FDIC's exam notes that "the bank displayed an adequate record regarding its borrower profile loan distribution...The institution established an adequate record regarding its geographic loan distribution....the bank displayed a good responsiveness to credit and community development needs."

BancorpSouth changed its assessment areas in January 2013 to include all of the census tracts where it had a branch, both within the city limits of Memphis as well as inside five other counties in Tennessee. But according to the DOJ, the bank still maintained its policy of "discouraging prospective applicants for credit in minority communities."  But by gerrymandering its assessment areas, the Bank was able to pick census tracts where it had branches but still avoid having any branches in a minority neighborhood.

The catchment area for CRA is considered to be in all areas within an MSA where a bank has a branch - not just in selected census tracts. So while the bank could claim that it was meeting credit needs in the immediate census tracts where it had branches that should not have been enough to pass the expectations of a CRA exam. Nonetheless, the bank got a "satisfactory." When the DOJ reviewed the bank's lending from the lens of a broader area (loans made within 5 miles of a branch), applications from majority-minority areas were the exception. Only 13 percent of loan applications came from majority-minority areas - and the majority of applications received in those areas were from white applicants. 

This mismatch between the demographics of an assessment area says less about the FDIC and more about a shortcoming in the legal construction of the CRA.  As implemented, banks are given the choice of picking their assessment area. Thus, the bank was able to pick almost all of Tipton and Desoto Counties for its assessment, even though they had only one majority-minority census tract, while selectively excluding most of the City of Memphis. 

The DOJ notes that BancorpSouth's General Loan Policy states that its CRA assessment area and its primary trade area are one and the same. The DOJ says that the bank said that loans made outside of the trade area are "undesirable." While the idea of a trade area in and of itself does not contradict the intent of the CRA, the idea that it could be so disaggregated is contrary to the spirit of the law. 

A negative CRA exam would have to build its case upon a lack of lending, investment, and service within low-income areas and to low-income borrowers. As written, the CRA is not about race. This is an important caveat. Thus, the CRA exam process does not test for violations of the Equal Credit Opportunity Act or the Fair Housing Act. Thus, the FDIC did not ignore those laws, as it was not in its capacity to hold the bank accountable for compliance in these areas. But if it is the case that minority neighborhoods were also low-income neighborhoods, then the CRA exam missed something very significant.