UNDERSTANDING DISPARATE IMPACT IN LENDING

Most lenders are well-versed on what discrimination looks like and how to prevent blatant forms of it from occurring within their portfolios. Yet to properly comply with fair lending regulations and legislation, it’s important to also know how to prevent less direct discrimination like disparate impact. In lending, when a policy or practice that’s otherwise neutral to a protected minority but then disproportionately burdens or excludes that minority, it’s known as disparate impact. Lending policies or practices that inadvertently lead to such discrimination may be justified due to “business necessity,” so just because there is disparate impact in how a lender does business doesn’t mean it’s prohibited.

Yet this justification can’t just be abstract or theoretical but must rather directly affect aspects of how a lender does business, such as by increasing expenses or decreasing profits. Even if a practice or policy can be justified by business necessity, however, if different practices or policies can instead be applied to deal with the issue to make it less discriminatory to protected minorities, this could prohibit a particular case of disparate impact. In lending, businesses can still be held accountable for using practices or policies without intending to be discriminatory and, in the process, inadvertently violate fair lending legislation. This is why it’s so important to understand what disparate impact in lending is and how to prevent it.

Discrimination & Disparate Impact in Lending

There are two primary pieces of legislation that help establish fairness in the lending industry and which deal with the problems inherent in disparate impact. These are the Equal Credit Opportunity Act (ECOA) and Fair Housing Act (FHAct). While the latter applies only to transactions related to residential real estate, the ECOA prohibits discriminatory practices with any transactions involving credit.

US courts have recognized three types of lending discrimination under this legislation: 

Type	Definition
Obvious indications of unequal treatment	These are clear and overt signs that individuals or groups are being treated differently based on their race, gender, religion, or other protected characteristic. An example of an obvious indication of unequal treatment in auto lending might be a loan officer at a car dealership who refuses to offer a loan to a customer based on their race, gender, or other protected characteristics.
Proportional evidence of unequal treatment	This type of evidence refers to statistical disparities that suggest unequal treatment. An example of proportional evidence of unequal treatment in auto lending might be if data analysis shows that a particular group of people is consistently offered higher interest rates on auto loans than other groups of people with similar creditworthiness.
Proof of unequal effects	This type of evidence involves demonstrating that the effects of a particular policy or practice disproportionately affect a particular group. For instance, if a lender requires a down payment of 20% for all auto loans, but this requirement disproportionately affects low-income individuals, it may be considered proof of unequal effects.

It’s this last one that equates to a disparate impact in lending towards protected minorities, as it goes beyond looking at the treatment of a potential borrower. In other words, a lender needs to also look at the impact of their policies and procedures and how they affect protected minorities, not just their treatment of them.

The CFPB & Disparate Impact In Lending

The Consumer Financial Protection Bureau (CFPB) is one of the primary enforcement agencies when it comes to the enforcement of fair lending practices within the consumer finance industry. The agency’s Regulation B describes what lending practices are specifically permitted and prohibited, along with requirements for lenders. Additionally, the 2010 Dodd–Frank Act expanded upon what the ECOA covers regarding consumer protection.

Specifically, the ECOA prohibits discrimination in lending based on:

• Age
• Exercising any Consumer Credit Protection Act rights in good faith
• Income from public assistance programs
• Marital status
• National origin
• Race or skin color
• Religion
• Sex

Manual processes conducted by a lender’s employees aren’t always consistently applied, sometimes leading to inadvertently discriminatory practices. Whether a decision regards funding, servicing, or underwriting, manually applying decision rules can lead to inconsistency in a lender’s decisioning. This alone may result in a disparate impact on lending applicants, which can result in non-compliance. 

How Fintech and Automation Can Help

To prevent discriminatory practices, many lenders turn to financial technology (fintech) that automates their lending processes. While using algorithms that are driven by data has definite benefits that will speed assessments and decrease expenses, if used without considering possible disparate impacts, it can lead to a violation of fair lending practices. Lenders who turn towards technology to keep their operations compliant should thus also ensure that there’s nothing within these automated underwriting decisions that could cause disparate impacts to either applicants or customers. It’s for this reason that lenders utilizing fintech in their business model should partner with a seasoned provider that understands the industry. 

Using Technology to Limit Disparate Impact in Lending

Adopting new technology can limit the possibility of any disparate impact in lending when applied correctly. By using algorithms to support the loan decisioning process rather than human discretion, creditworthy applicants that otherwise wouldn’t get approved can obtain credit without increasing the risk to the lender. With new modeling techniques and properly designed algorithms, human bias in the decisioning process can be significantly reduced.

By automating decision rules, lenders can then evaluate all loan applicants using the same criteria while also recording how, when, and why decisions were made concerning a specific customer. This provides solid evidence to regulators that a lender is complying with regulations. Automated lending systems also create an auditable history that shows who implemented a decision rule. This is especially helpful in cases where modifications are made to how a regulation is interpreted and applied.

Modern lending software offers advantages over manual processes by:

• Alerting lenders automatically when a decision looks to be non-compliant.
• Establishing a consistent decisioning process.
• Instituting processes that are easy to audit, evaluate, and track.
• Preventing errors due to manual processes.

The most advanced lending software allows lenders to take advantage of reporting capabilities that utilize data analytics, which then provides accurate statistics on a lender’s processes and how they impact their customers. Such software platforms also include many inbuilt rules that allow lenders to automatically comply with applicable regulations, including by automatically sending out messages concerning any adverse actions on applicants. This creates a record of communications that can later be used as proof during an audit as to why an applicant wasn’t approved for a loan.

Since the burden of proof is on lenders regarding compliance, the best way to achieve it involves utilizing technology that helps lenders maintain it. Though many current lending platforms allow lenders to configure decision rules and workflow around regulatory requirements, it’s important to ensure that these are flexible enough to deal with any regulatory changes. Not only does implementing such software allow lenders to ensure they’re maintaining regulatory compliance, but these tools also enable lenders to more easily modify their decision rules to deal with any changes to how rules and regulations are interpreted. It’s through the latest technology that lenders can then attain, support and validate compliance with existing rules and regulations.