Will Kiva Zip’s trustee model work?

Several weeks ago, I wrote some quick thoughts on building reputations and trust networks within the service I am creating.  My thoughts echo Kiva Zip’s pursuit to find a substitute for credit scores. The Fast Company article I mentioned in my last post writes,

“Shah’s hoping ‘trustees’ will emerge as stamps of approval… [and] argues that [they] can be a substitute for a high credit score. He suggests that the administrator of a beauty school could, say, vouch for the top three graduates each year. So does that mean they will be safe loans? ‘I don’t know, is the short answer,’ Shah concedes. ‘This experiment could dramatically increase risk to lenders.’”

Premal Shah’s ideas about the role of trustees are interesting within this context, but I think part of the reason he is unable to provide a more definite answer lies in the kind of relationships that exist between a trustee, lender and borrower in the Kiva Zip model.

Part of the risk to the lender is the fact that they have no pre-existing relationship with the borrower and don’t have a definite measure by which to evaluate the borrower’s ability to pay them back. A trustee steps in and mitigates this risk, but there is nothing that connects the trustee with the lender. A trustee can be validated by a 4th party as a reliable trustee, but there is no assurance that the potential lender trusts these people either. Because there is no relationship between lender and borrower, no relationship between trustee and lender, and also no way to validate the opinions of “recommenders” there are some huge social leaps that participants of this system have to make… Leaps that are potentially very risky.

By working with existing relationships and communities, my concept uses Mark S. Granovetter’s idea of the strength of weak ties to mitigate some of that risk.

I’ve included a rough sketch of how these relationships work in my concept, but the essential idea is that there will be a strong tie between a borrower and a “trustee” (for consistency, I’m using Kiva’s language, but in the diagram this person is labeled as “friend”). This strong tie will allow the trustee to vouch for the borrower in ways that are meaningful and authentic. However, this strong tie prevents the “trustee” from wanting to lend to the borrower because they fear that their deep relationship will be compromised if the borrower defaults. Instead, the trustee uses their endorsement to attract potential lenders that might be outside of the borrowers immediate network. The tie between trustee and lender can be strong or weak depending, but most likely there will be some relationship between the two. This way, if the borrower defaults, they are not only hurting their own reputation but also that of their friend. If the trustee doesn’t make a good recommendation, it publicly devalues their worth in the system and also affects their relationship with their friends and connections. This kind of feedback loop seems to be entirely missing within the Kiva Zip model.