Consumer credit reporting agency TransUnion launched this week an updated version of its proprietary resident scoring algorithm, called ResidentScore 3.0.
The revamped tool uses artificial intelligence-powered machine learning to predict the likelihood a tenant will be evicted or miss rental payments within 12 months.
“Property management companies can face significant losses when evictions are high,” Maitri Johnson, vice president of multifamily at TransUnion said, in a statement. “With screening as the gateway to all things that occur on a property, the best way to identify which applicants may pose a risk is to have a more efficient and effective process in place for vetting potential residents.”
“Having the proper checks and balances in place to reduce costly involuntary turnover can lead to huge savings,” Johnson added.
Approximately 4 percent of rental properties result in an eviction at an average cost of $5,000 per unit, according to TransUnion.
The tool has been in the market since 2013, but TransUnion says the new model is 4 percent more predictive than the previous iteration and outperforms standard credit models by 16 percent. TransUnion did not disclose the variables it considers, but did say it leverages multiple variables and new data elements.
The ResidentScore ranges between 350 and 850, with a higher score indicating that a tenant is less likely to be evicted. With a score under 520, there’s a 28.79 percent the tenant will be evicted or skip out on a rent payment, but as that score climbs about 600, the chance falls to less than 3 percent.
Traditional credit reporting has been linked to racial disparities in housing, although it’s not clear what factors TransUnion considers and if it would have any positive impact on helping communities of color obtain housing more quickly.
A 2019 study from ShelterForce, an independent publication focusing on community development and affordable housing, found including on-time rent and utility payments would increase credit scores and reduce the number of borrowers considered to be subprime by half.
“Because a majority of African Americans and Latinos are renters, this change could reduce racial disparities in credit scores as well,” the study reads. “VantageScore estimates that mortgage lending to African Americans and Latinos could increase by 16 to 32 percent over 2013 levels if all credit scores included information on rent and utility payments.”