In recent years, the Department of Justice (DOJ) and various federal, state, local and non-profit entities have generated many headlines about identifying and pursuing cases of redlining against financial institutions. Recent cases include:
- a $33 million settlement between Hudson City Savings Bank and the DOJ;
- an $825,000 settlement between Evans Bank and the Attorney General of New York; and
- a $200 million settlement between Associated Bank and the Department of Housing and Urban Development (HUD).
In addition to these cases, municipalities – such as Los Angeles and Miami – and consumer or community advocacy groups have brought lawsuits against mortgage lenders alleging discriminatory and predatory lending practices against minority applicants and high-minority geographies. And all of this recent activity seems to have captured the public’s attention, as highlighted by Figure 1, which shows Google’s index of search volume for terms related to the topic of redlining in the U.S. since 2010. The search data show a general increase in interest since the fourth quarter of 2014 that continues through June 2016.
Figure 1. Google U.S. Search Index for the Topic of Redlining, January 2010 – June 2016
Redlining is a Key Focus Area of Fair Lending Compliance
In Fair Lending compliance, redlining is a form of discrimination in the provision and terms of credit on the basis of race, national origin, or other prohibited basis characteristics of the geographic areas in which the applicants will reside after receiving credit. Redlining risk exists when a financial institution’s lending patterns result in disparate outcomes of marketing, underwriting, or pricing decisions between geographic areas with high and low concentrations of minorities.
In its 2016 Fair Lending Report, the Consumer Financial Protection Bureau (CFPB) identified redlining as a key compliance area of focus, stating that:
In 2015 the Bureau had a number of ongoing investigations and authorized lawsuits against institutions that are focused on fair lending. In particular, as mortgage lending is among the Bureau’s top priorities, the Bureau focused its efforts on addressing the unlawful practice of redlining…At the end of 2015, the Bureau had a number of authorized enforcement actions in settlement negotiations and pending investigations.
The DOJ also identified redlining as an area “coming back into stark focus.” The DOJ’s prioritization and the CFPB’s continued, multi-year focus on redlining and the accumulation of complaints, lawsuits and related settlements in recent years, requires that financial institutions understand and address their potential redlining risk exposure.
Evaluating Redlining Risk in Mortgage Lending
To assess for redlining risk among mortgage lenders, ADI developed the Redlining Sweep℠ to simulate how the CFPB and other agencies may use publicly available data reported under the Home Mortgage Disclosure Act (HMDA) to screen for potential cases of redlining. Figures 2 and 3 show the top 20 metropolitan statistical areas for Hudson City Savings Bank and Associated Bank using the Redlining Sweep℠.
This level of data analysis represents a first step in a multi-step process of evaluating redlining risk. With this tool, we can quickly identify the geographic markets that present the greatest exposure to redlining risk by comparing a lender to similarly-situated peers within each market.
Figure 2. Redlining Sweep℠ – Hudson City Savings Bank Top 20 MSAs by Application Volume, 2009 – 2013
The results shown in Figures 2 and 3 highlight how federal regulators, as well as state, local and non-profit agencies and entities can use HMDA data to identify mortgage lenders with potential instances of redlining. In both cases, the geographic markets highlighted in their respective settlements are identified as presenting high exposure to redlining risk.
For Hudson City Savings Bank, the report identifies 16 of the 20 markets shown as presenting exposure to redlining risk at greater than a 90% confidence level, when comparing its activities to similarly-situated peers. These results are consistent with the consent order, which identified high-risk areas throughout New York, New Jersey, Connecticut and Pennsylvania.
The Associated Bank report also identified those markets named in its consent order: Chicago, Milwaukee, Minneapolis, Racine and Kenosha.
Figure 3. Redlining Sweep℠ – Associated Bank Top 20 MSAs by Application Volume, 2008 – 2010
After identifying the geographic markets that present redlining risk, a deeper analysis should be conducted to determine what factors contributed to the observed lending patterns. Additional activities include:
- Analyzing market penetration and branch distribution maps to pinpoint geographic gaps in lending and service distribution;
- Evaluating the assessment area to determine whether a person could reasonably conclude that it arbitrarily excludes high-minority areas;
- Reviewing marketing, advertising and sales activities that may exclude residents in high-minority areas;
- Examining branches, originators, brokers, etc. for patterns of risk within certain organizational channels or units;
- Developing regression models to identify where risk is concentrated, while controlling for various explanatory variables; and
- Assessing the strength of compliance monitoring policies, procedures and systems.
The deeper analysis will lead to conclusions about whether the observed patterns can be explained by mitigating factors or require corrective actions.
Evaluating Redlining Risk in Consumer Lending
Redlining risk is not limited to mortgage lending. Consumer credit products – such as auto loans, credit cards and personal unsecured loans – can present redlining risk for financial institutions. While Fair Lending examiners may not have public data with which to screen institutions for potential redlining, they can blend loan origination data with geodemographic data during an examination to analyze several outcomes for redlining-related discrimination.
Recent settlements involving lending divisions of Honda and Toyota have shown that the CFPB has clearly demonstrated its willingness to use geodemographic data against consumer lenders in Fair Lending cases, the agency’s minority BISG proxy methodology relies on such data and can be easily applied to evaluate lenders for redlining also.
Even community advocacy groups have found ways to identify redlining in the absence of institutional data. The National Fair Housing Alliance used information collected during mystery shops to evaluate realtors. Mystery shops, or “matched-pair testing”, has been used for decades in Fair Lending compliance to gather data, and an enterprising agency or group may utilize it or other methods (e.g., surveys, consumer complaints, etc.) within their markets to collect data that can be used to file complaints or bring lawsuits against consumer lenders.
Consumer lenders should stay ahead of such efforts by conducting their own evaluation of their exposure to redlining risk. The activities to undertake would be similar to those outlined above for mortgage lending.
Loan origination data should be analyzed based on the geodemographic characteristics of the applicants for each market. This analysis compares the distribution of applications and loans as well as underwriting and pricing outcomes between high- and low-minority geographic areas. The results of this effort will identify and highlight potential exposure to risk and help prioritize the markets, neighborhoods, products and decisions that require a deeper analysis, as previously described (e.g., market penetration maps, regression modeling, etc.). Lenders can also use testers to gain qualitative data and insight through a controlled study. The value of this dual-perspective is a deeper understanding of the risk, as well as information that can be used in the strategy to mitigate that risk.
Redlining Needs to be a Focal Point in Fair Lending Compliance Programs
All available evidence indicates a continued focus on redlining as part of the Fair Lending compliance front. The CFPB has identified redlining as a key focus area in recent years, and its willingness to apply geodemographic data in evaluating consumer lenders shows that the issue will not be limited to mortgage lending.
For mortgage lenders, the new HMDA rule may open up more avenues for scrutiny. The additional data fields that lenders will be required to report will allow regulators to develop more sophisticated screening methodologies that can help them build a stronger case for potential redlining than under current reporting requirements.
The growing collection of recent settlements, complaints, agency updates and articles focused on redlining gives lenders insight into how serious the issue has become among regulators, the DOJ and the public. ADI helps lenders by screening for risk exposure using our Redlining Sweep℠ and conducting a deeper analysis of the markets that drive the potential risk. The conclusions drawn from the overall analysis will identify where redlining risk, if any, is concentrated and inform next steps and corrective actions.