Racism in Peer-to-Peer Lending

Not only did the subprime meltdown almost send our financial system into cardiac arrest, but it also happened to show that racism is alive and doing just fine.

A number of investigations across the country have found that non-whites were disproportionately steered into subprime loans.

You might wonder that perhaps this was caused by lenders' due diligence. The combination of lower incomes and relatively worse credit histories -- which are more common among blacks and latinos than whites -- means that lenders can justifiably charge higher rates. (This is called "statistical" discrimination.) But when economists and sociologists controlled for these factors, non-whites still made up an oddly high portion of subprime loans, hinting that some race-based discrimination was taking place.

I know. You're probably thinking: "Institutional racism, so what else is new? What's your point!!?"

Well, now let's look at the world of peer-to-peer lending where ordinary people can play the role of banker. Does racial discrimination exist here too?

Yes. And no. And the no is the way more interesting part.

For the uninitiated, a number of websites have popped up over the last few years that allow regular folks to lend to one another. (Go here for a summary of these sites.)

This year, the volume of peer-to-peer loans is expected to break $1 billion and by 2010 should hit $5.8 billion. The main attraction of these sites is that users can borrow at rates that are lower than they can get from a bank.

The largest lending facilitator, Prosper.com, is a cross between eBay and Facebook. Potential borrowers set up a profile listing how much money they want to borrow and the maximum interest rate they're willing to pay. They also have the option to add a picture and a reason for why they need the loan -- typical ones include paying down credit-card debt or funding a movie or business venture. Users' profiles list the potential borrowers' credit grades, which Prosper assigns based on credit checks. Incomes are also listed, but the information is unverified by Prosper.

Next, lenders bid the lowest rate they're willing to offer based on each borrower's profile. A typical loan is funded by a group of lenders that have bid the lowest rates. It's rare for a loan to be funded by only one person, and not all loans find funding.

Now back to our discrimination story.

Looking at data from Prosper, Devin Pope of Wharton and Justin Sydnor of Case Western Reserve University found that a listing with a picture of a black person was 25 to 35 percent less likely to get funded than a listing with a white person with similar a profile. And when a black person did get funding, the interest rate charged was about 0.6 to 0.8 percent higher.

(More likely to find funding were women and those with photos that showed a link to the military. The funding success for latinos and asians wasn't statistically different from the average. Another recent study of Prosper by Enrichetta Ravina of NYU found that attractive people are more successful in landing loans and lower rates than their homelier peers.)

So this shows that blacks are being discriminated against, right? Not quite.

Just like the subprime market, the lenders would be offering higher rates to poor credit risks in order to protect against not getting their money back. What this means is that when you adjust for risk, the returns on high interest rates loans should be equal to the returns on the average loan.

Most studies don't have access to loan performance data, but luckily Prosper allows researchers to tap in. And it turns out that the risk-adjusted return on loans to blacks was lower than the average return.

I found the implications here quite astonishing. Lenders, in some ways, aren't being "racist" enough. Given the higher propensity for loans to blacks to default on Prosper, lenders should be charging higher rates.

Why don't they?

"Clearly skin color is not a causal factor in loan default," Pope and Sydnor write. "Higher default rates for blacks must stem from some difference in the background and financial characteristics of these borrowers that is not fully reflected in the standard" methods used by lenders to evaluate applicants. (emphasis mine)

They think the results show that lenders on Prosper -- although they understand the correlation between ethnicity and default rates -- don't "appreciate the strength of these correlations."

And so it seems peer lenders might be more equitable than professional ones -- though this could change with better loan performance data.

UPDATE
Commenter Tom left a good link about someone's experience with Prosper and discrimination, but it got cut off, so here it is again.

Add Your Comment

Report Abuse

Archive

Categories

Links