Philanthropy Action

News & Commentary


See A Mostly Comprehensive Guide to the Kiva and Donor Illusions Debate for links to the debate on this topic from around the web.

Update: See the bottom of the post

David Roodman of the Center for Global Development recently wrote a brilliant post on how Kiva, the “person-to-person” microlending site, is not exactly what it seems. The quick version: while it appears that Kiva lenders are choosing who they loan to, in reality the loans to the borrowers on Kiva’s site have already been given. Roodman rightly points out that Kiva operates the way it should it if wants to be maximally helpful to both borrowers and MFIs. The problem is the subtle misleading of Kiva’s users.

In Kiva’s defense, this subtle misleading is not unique to Kiva; most NGO’s operate this way especially in disaster relief, child sponsorship, and alternative gifts (like giving a cow or a goat). The reason it’s so prevalent is that the donors demand it—and they vote with their dollars if the NGO is unwilling to provide the illusion of a person-to-person connection. Kudos to Roodman for exposing the illusion in a comprehensive and thoughtful way. While I think trafficking in such illusions is wrong, I understand why they are perpetrated in the name of the “greater good.“ I wish Kiva and others would abandon this practice, but I also acknowledge that they can’t until donors stop requiring NGOs to mislead them.

But today I saw a Kiva document that, for me, points to a far bigger problem. Roodman shared a document Kiva uses to explain its operations to MFIs. Two points in the document floored me. First, all losses from Kiva-securitized loans are borne by the Kiva user. Second, Kiva’s monthly repayment reports are not based on actual repayment data.

The first is a problem because of fungibility. MFIs have no liability for delinquent loans underwritten by Kiva users, while they often are liable for capital from other sources. I’m sure I’m not the only one who immediately saw that this encourages MFIs to book delinquent loans to Kiva rather than to other capital providers. It seems to me we have pretty recent experience with what can happen when you separate liability from lenders.

The second, in my mind, isn’t subtle misleading of users—it’s just plain misleading. Essentially, Kiva is not receiving reports from the MFIs on whether specific loans are being repaid. Kiva is reporting repayment to its users based on the assumption that the loan is being repaid—and Kiva will assume that the loan is repaid until the MFI decides to notify Kiva otherwise. For anyone who has been inside the books of an MFI this should be really disturbing. No two MFI’s define default the same way and often default is up to the discretion of a particular loan officer. Combined with the zero-liability noted above, these are accounting loopholes big enough to drive an entire fleet of trucks through.

In the big picture, this is just another illustration of the dangers of the illusion model of person-to-person connection in charitable giving. Like little white lies there is often a perfectly plausible, even good, reason for creating the illusion. And like little white lies, the illusion has to keep growing to continue to be credible. And as the illusion becomes all encompassing you open the door to deception, corruption and fraud. That’s what happened with child sponsorship back in the ‘90’s. Is that what will happen to person-to-person microfinance in the ‘10’s?

Update: Thanks to David Roodman for his comment below, helping clarify two things: one, I’m not suggesting that lenders are currently booking loses to Kiva; two, there’s already some evidence that MFIs are shifting losses around to benefit Kiva lenders, so this is not just a theoretical discussion.

But that led me to thinking more about how Kiva is handling repayments. The linked document says that MFIs don’t necessarily repay Kiva. Kiva deducts repayments from the next loan to the MFI. What that means is that some new lenders money is not just not going to a specific borrower, it’s never leaving Kiva’s bank accounts. It’s being shifted from one Kiva user’s account to another. Now, because of the fungibility of money there’s technically nothing wrong with this. In fact, it’s a good thing because it minimizes overhead costs (such as wire fees and currency conversion). But it is different from the story and it seems pretty clear that many Kiva users would not be happy if they knew the reality of the situation.


Tim, I’d gotten the impression from Kiva that MFIs tend to go out of their way *not* to pass losses to Kiva, because they want to keep their official Kiva repayment rates high. currently lists a 98.42% repayment rate, so it doesn’t seem as if a lot of MFIs are dumping delinquent loans on Kiva users.

As for your second point, I think Kiva would respond that the MFIs are supposed to manually update the data to reflect any discrepencies between the repayments data shown on their site and actual repayments. Of course a lot of times they don’t—-see above.

What I found interesting about this document is that it shows that Kiva has been capable for at least half its history of explaining how it actually works in a very clear, accessible way. That Kiva shows a different picture on its web site therefore reflects a strategic choice, not an accident caused by scrambling to keep up with its growth.

October 12, 2009


thanks for the comment. I should clarify that I’m not suggesting that there is currently a situation where lenders are booking their losses to Kiva but that the incentive is clearly there (as commenter Tim Rann on your blog made clear). As you note though there is already some suggestion that MFIs are shifting around their losses and whether they are doing that to Kiva’s benefit or Kiva’s detriment doesn’t matter all that much. It’s still disturbing given Kiva’s marketing.

Thinking about this more it also occurred to me that the way Kiva handles repayments means that new lenders money is often not going to the borrower or even the MFI. It’s being used to repay prior lenders.

Again this is the right thing to do to minimize transaction costs. But I wonder how many of Kiva’s users would be comfortable and happy if they knew their new loans were going into other lenders accounts?

October 13, 2009

I love and as long as they continue to promote and capitalize microfinance better than the next-best alternative, they’ll be #1 in my book.

October 13, 2009

Well, I have been telling my former students to be cautious of their unconditional love to any NGO business model for the past 5 years. 

There is nothing wrong with raising questions, if Kiva is telling you the truth they will come out the crisis stronger.  However, any real professional with 10+ years of real world experience in Economic Development can attest you that, albeit appealing, this model once brought up to scale is misleading to say the best.

Curiosity is a sign of intelligence, innocence, but intolerance and stubbornness tends to show precisely the opposite… not only for people but also for companies and charities!

All the best,

November 13, 2009

It is a tough issue for sure - microfinance can only really be microfinance at small scale, and growing it up to larger sizes tends to remove or at least reduce many of the advantages it enjoyed earlier.

The model for my own nonprofit venture,, is similar in that it allows microdonations to nonprofits; but by spreading the risk around, it should help givers avoid putting all their eggs in Kiva’s basket, or any other.


November 13, 2009

Remember my personal information

Notify me of follow-up comments?

Comments may be edited for length. Inappropriate comments will not be published.