The Use of Social Data Raises Issues for Consumer Lending

Jonathan Zim – With over 1.54 billion monthly active users, Facebook continues to be the leader among social media platforms, but what do they do with the wealth of personal data gathered? The knowledge of users’ “likes,” preferences, political views, friends, acquaintances, and colleagues is used to create a digital identity—a collection of data in an identifying profile encapsulating the unique details of ones life.  In August 2015, Facebook secured a patent which may benefit their users by filtering out SPAM emails, offensive content, or even improve the accuracy of searches—but it can also be used to reject your next loan application. This new patent will permit lenders to determine credit worthiness based on the credit scores of those within a social network In other words, credit underwriters may be able to garner a borrower’s creditworthiness based on their digital identity. This has piqued the interest of regulators in Washington D.C. who must ensure this practice is accurate and free of discrimination.

According to the patent, when an individual applies for a loan, the lender will examine the credit ratings of authorized members from a user’s social network, or in other terms, your “friends.” If the average credit rating of those on your “friends list” is at least the minimum credit score, the lender will continue to process the loan application. If not, the loan application will be denied. This has the potential for great bias. Many have argued that a friend on Facebook does not establish a financial relationship, nor is it necessarily determinative of a user’s creditworthiness. At a recent press conference, an advisor to the Treasury Secretary, Antonio Weiss stated, “The increasing amount of nontraditional data used in credit underwriting decisions may raise consumer protection questions, specifically those questions regarding compliance with fair lending obligations. Just because a credit decision is made by an algorithm, does not mean its fair.” Regulators must carefully examine this undisclosed algorithm to eliminate the potential for discrimination.

In the 20th century, banks denied loans on the basis that an individual lived in a minority neighborhood.  This is analogous to this patent, which similarly has the potential for discrimination based on those within a social network. Federal laws such as the Equal Credit Opportunity Act, the Electronic Funds Transfer Act, and the Fair Credit Reporting Act, were enacted to curtail discriminatory lending practices, and to govern a creditor’s determination of a loan approval. The National Consumer Law Center has raised its concerns of possible racial discrimination in Facebook’s undisclosed algorithm, and stated that black individuals tend to have lower incomes and lower credit scores, and determining their credit worthiness based on their social network will clearly lead to discrimination against minorities, effectively returning to the biases of the 20th century. If Facebook along with credit underwriters utilize this patent, they must ensure that the algorithms are disclosed to ensure compliance with the federal laws in place and to prevent discrimination similar to that of the 20th century.

Despite these criticisms, proponents for the patent argue that this patent may benefit new small businesses. Utilizing the data gathered from a business’s digital identity can assist in the comparison to businesses of the same category to more accurately determine the risk of default. Eric Haller, Executive Vice President of Esperian DataLabs, has said that this patent can remedy the difficulty of new startups seeking credit. DataLabs’ Research suggests that high Yelp reviews or positive social media comments from customers will benefit their financing endeavors and improve the predictability of commercial loan defaults by 40 percent.

Proponents for the patent also claim that access to an individual or small business’s social network will actually give the online lender a more comprehensive view of the applicant and permit lenders to make credit decisions more quickly. In its response to a request for information from the U.S. Department of Treasury, Kabbage, Inc., a company which provides an automated platform for loan applications, found that these methods will improve accuracy of loan decisions due to the wealth of information related to the volume, character, and stability of the business. It further elaborated that these methods focus on the quality of service provided to customers rather than other socioeconomic variables utilized in the past. Although businesses normally do need to disclose a substantial amount of information in the patent application, this patent does not disclose the algorithm used to determine creditworthiness, it only discloses the method upon which the algorithm is intended to operate.

On the whole, this patent has the potential to affect the economy on a massive scale as it relates to individuals and businesses.  The government should intervene to ensure its compliance with federal discrimination and lending law protections afforded to individuals and businesses. If first meticulously evaluated under these grounds, this patent can serve a useful purpose to the future of credit underwriting.  Ultimately, the federal government should compel Facebook to disclose the algorithm to the U.S. Department of Treasury in an effort to protect consumers from potential discrimination, and to ensure a non-biased, accurate online loan-approval system for businesses and consumers alike.

Leave a Reply

Your email address will not be published. Required fields are marked *