Fed Strategies for Managing Fair Lending Risks of Digital Redlining
Any time a regulator shares its observations of where violations are occurring and what you can do to prevent them, it’s worth taking note.
This includes the Fed’s December 2019 Consumer Compliance Supervision Bulletin. The publication’s goal is to increase transparency into what the Fed is seeing in the consumer compliance arena and offer practical advice for reducing risk—and this month fair lending risk related to internet-based marketing takes a front seat.
Internet-Based Marketing Is A Fair Lending Risk
It’s no secret that data analytics have made it increasingly easy to target specific market segments online. But it also creates consumer protection and fair lending risks, as the Fed notes.
We already know that redlining is prohibited by the Equal Credit Opportunity Act (ECOA) and the Fair Housing Act (FHA). Consumers are entitled to equal access to credit. But the kind of data collected and analyzed is often incomplete. It relies on information like browsing histories, locations, habits, and predictive analytics. This can lead to inadvertently making higher-priced offers to consumers than they would normally be entitled to, the Fed points out.
Examples cited by the Fed include Mac users being steered towards more expensive hotels than Windows users, a college prep course that varied price by zip code even within the same city, or Amazon not providing same-day delivery service in majority black ZIP codes.
In the housing arena, the Department of Housing and Urban Development hit Facebook with a formal complaint in late 2018 for “digital redlining” by letting landlords and home sellers use its advertising platform to participate in housing discrimination by making it possible to post an ad that excluded minorities or discriminate based on gender. HUD followed up with a housing discrimination charge in March 2019 and Facebook ended up settling with a variety of civil rights organizations.
Fair Lending Risk Management Tips
As the entity with supervisory authority under ECOA for state member banks with assets of $10 billion or less and FHA for all state member banks, the Fed is reminding financial institutions to consider the fair lending implications of any advertising technology that relies on algorithms and filters.
While internet-based marketing algorithms and analytics are constantly evolving, the Fed notes that existing best practices for fair lending risk management can serve as effective controls. Steps for managing this risk include:
▪ Reviewing for and eliminating any geographic filters used in internet-based advertising, such as filters by ZIP code or branch radius, that would exclude predominantly minority neighborhoods;
▪ Reviewing other elective filters in use to ensure that the they do not prevent (even unintentionally) viewers with characteristics protected by the ECOA or the FHA from seeing an online ad;
▪ Reviewing any information available from the internet-based advertising service regarding the audience that viewed each online ad;
▪ Establishing policies and procedures to evaluate fair lending risk for all marketing and outreach initiatives, including those that are internet-based and those that are not;
▪ Monitoring marketing and outreach activities to ensure that those activities are reaching the entire assessment or credit market area, including any predominantly minority neighborhoods;
▪ Ensuring that vendor management policies are in place; and
▪ Understanding what marketing practices are being employed by the bank or on its behalf, and having appropriate controls to assess and manage any risk associated with those practices.
The Fed’s recommended strategies aren’t new, but they are an important reminder that analyzing redlining and fair lending risk is essential for effective risk management. You don’t want a consumer complaint to the regulators to uncover a problem.
Proactively uncovering and addressing a problem is the best way to manage risk.