How to root out discrimination in the mortgage industry

“What’s interesting about intersecting it with our data is we have the boundaries of every property in the country,” Zach Wade (pictured), vice president, Data Science, at LightBox, told MPA in a recent telephone interview. “So, by intersecting the publicly available data set with our structural level data set – which also includes things like home prices, what the building is, how big it is – we’ve been able to go in and look at a time period of transactions to see if there’s a persistent difference in areas that were historically redlined either by that loan corporation.”

While illegal redlining is now not overt, Wade noted, data comparisons can help detect remnants of unsavory lending practices of the past. “What we did is look at, given historical redlining that was made illegal, if there were still artifacts of it,” he explained. “That particular home loan corporation [Home Owners’ Loan Corp.] was pretty explicit of what was desirable versus undesirable, and race played an active role in their definition of it.”

The aim of overlaying new data with old maps is to detect any lingering underinvestment or economic stagnation in certain areas once redlined, he explained. “We didn’t actually do too much of a detailed analysis or take an opinion as much as create this data set others could look at to draw their own conclusions from.”

The precision of such granular data also can detect areas most prone to flooding, or parts of the country where there is a lack of broadband – other potential signs of inequity.