All sorts of stuff factors into your credit score: How many credit cards you have; whether or not you’ve paid bills late in the past; the age of your oldest account; what types of loans you have; whether you’ve maxed out your credit cards; and so on.
But a handful of high-tech startups now see an opportunity to sift through mountains of other data about you, including some drawn from your social media profiles, to determine whether or not you’re the type of person who’s going to pay back a lender or not. And you’d be surprised what kind of information can have an effect. Are you the kind of person who “shouts” in all caps or never bothers to capitalize at all? How many friends or connections do you have? And even: Are you the type to fire off negative or racist commentary?
Wait — being racist could cost you a loan? What does prejudice have to do with whether or not you’ll pay back a debt?
“On a public forum like Facebook, it’s like talking in public,” says Navin Bathija, founder and CEO of Neo, a company that evaluates car-loan applicants (with their permission) based not only on traditional lending metrics but on the strength of prospective borrowers’ professional networks on LinkedIn.
Bathija says there’s a strong correlation between a robust network and the likelihood of repayment. A person with a bigger network and more recommendations, along with a track record of steady employment and advancement without job-hopping means the borrower probably has the means to pay back the loan; and more connections mean they’ll be in a better position to land a new job if they do get laid off.
The link to racism is something the team at Neo is still testing. “Mr Bathija reckons that within a year there will be enough evidence to determine if making racist comments on Facebook is correlated with a lack of creditworthiness,” The Economist says. The theory goes like this: People who express racists beliefs online probably vocalize them in the real world, as well. In the workplace, this could lead to conflicts with co-workers, clashes with suppliers or customers, and even legal trouble.
The link between social attitudes and job stability isn’t as far-fetched as it might sound, says Heidi Golledge, CEO and co-founder of CareerBliss.com. “We can correlate almost exactly productivity, their success at work, to positivity on social media pages,” she says. People who gripe about their lives — even if they don’t complain about their jobs — and vent about religion, politics, or race don’t do as well on the job. “With creditworthiness, it’s the same thing,” Golledge says. People who exercise the self-discipline and control to keep it dignified on their social media profiles are exhibiting the same qualities lenders look for.
Neo isn’t the only such company dipping its toes into big-data waters. A startup called LendUp is experimenting with the idea that people with stronger social networks, as exhibited by having more friends and more frequent interactions with them, are more dependable borrowers. “If you have a very strong, close geographic network, that’s helpful to you,” CEO Sasha Orloff told TIME last year.
ZestFinance is a lending site that uses a mashup of “Google-style machine learning” and “Capital One-style credit scoring,” according to its website, to yield default rates 40% lower than comparable services that target underbanked Americans. The company’s founder and CEO Douglas Merrill, former CIO at Google, tells The Economist that people who either ignore or abuse the caps-lock key are less likely to pay back their debts, when other factors are the same.
The biggest argument for slicing and dicing these seemingly unrelated snippets of online activity is that people with sparse credit histories — often young people or recent immigrants — would no longer be penalized by a lending system that takes a “guilty until proven innocent” approach. Currently, you pretty much have to earn the right to get a good rate on a loan, which means plenty of people who will turn out to be diligent borrowers wind up paying inflated rates until they prove themselves by establishing a standard credit history.
For lenders, the other advantage is that social media could give them a crystal ball to predict which borrowers look good on paper but will turn out to be duds.
Bathija points out that there’s a right way and a wrong way to go about this. Certain behavioral patterns could wind up being proxies for factors lenders aren’t allowed to discriminate against, like race.
Although social media experts frequently warn consumers against letting it all hang out online, many of us ignore that advice. At its worst, the idea of mining our online selves for slivers of data could seem reminiscent of the Miranda warning recited to criminals: Anything you say can and will be used against you.
“It’s scary to some extent,” Bathija says, “But this conversation needs to happen with consumers because, whether they know it or not, [social media] information is being used.”