Ideas to Action:

Independent research for global prosperity


Just over three weeks have passed since I first blogged on Kiva. The journey since then can be measured in other ways, and at extents I never imagined: several hundred tweets, 50 comments on this blog, nearly a score of blog posts elsewhere, 10,000 hits to the original post, a reply from Matt Flannery, and a Kiva web site revision. I have written little more till now on Kiva; I have been busy, I already had a long turn at the microphone, and I preferred to watch the reaction blossom.

[Update: I just noticed Kiva changed the tag line on its home page. Was: "Kiva lets you lend to a specific entrepreneur, empowering them to lift themselves out of poverty." Is: "Kiva connects people through lending to alleviate poverty." Nice.]

What to make of it all? For one, the experience shows the attractions of specificity and controversy. A blog entitled "Nonprofits Sometimes Simplify Explanations to Donors of What They Do" would have been a bigger story but would not have garnered much attention. And it helps to blog topics of natural interest to those who tweet. More generally it demonstrate the remarkable power of electronic social networks to propagate ideas.

All that is fine. I am thrilled with the interest. But I also know that the greatest occupational hazard in my business is letting (modest) success like this go to one's head, distorting one's judgment and one's judgment about the quality of one's judgment. What matters more is whether I have done good. Will my writing make any poor people better off?

The case can be made. Tim Ogden, who follows matters of philanthropy full time, commented that "the longer we gently mislead the public about both microfinance and philanthropy, the more painful the hangover will ultimately be." By prompting to Kiva to become more transparent, I have saved Kiva and other peer-to-peer non-profits it inspires from an even bigger backlash down the road. Also, I have helped people who I would not have otherwise reached think more critically about charity and microfinance, in particular about the gaps between the simple, confident pitches for particular programs and the complexities of carrying those programs out and determining their impacts.

On the other hand, I take seriously the arguments of Sasha Dichter of the Acumen Fund and Ashley at SocialEarth that there is only so much to gain from bemoaning human nature. We like stories, and if serving them up makes us feel good about giving, is that so bad? How much charity is really selfless anyway?

The attention has forced me to ponder whether the message of the post—that Kiva should be more transparent about the person-to-person relationships it seems to construct—is worthy of this attention, not to mention the time I put into writing it; and if not, whether a larger point that has been missed.

Let me tell you about my visit to Kiva 9 days ago. By luck I was in San Francisco for personal reasons. I took advantage of the opportunity to visit Kiva's office, spending about an hour each with Premal Shah, the president, and Ben Elberger, who manages relations with microfinance institutions (MFIs) in Anglophone Africa and used to work here at CGD. Unfortunately, Matt Flannery was not around. We actually spent few words on the blog controversy. And because everyone was wonderfully friendly and open and we set no ground rules about what was "on the record," I hesitate to report too specifically on some matters, lest I violate confidences. Some thoughts:

  • The people behind the sexy web site could hardly be more informal. Imagine T-shirts, jeans, dull carpet, cubicles, sub-Ikea-grade furniture, minimal decoration. Entering the building and office suite is about as ceremonious as visiting your friend's appointment. Except worse, because the phone system you use to dial the office to be buzzed in doesn't work.
  • I know this is stating the obvious, but the web site is an elaborate façade. There are the people who run Kiva, the people who use it, and in between the web machine. Users mentally equate the site with the organization, as I did before I visited. But the site is like the glass that separates customer and teller. Kiva employees too think of the site as distinct from themselves, as an electronic object that they manage and through which they see their customers. (Except of course that it's not quite so transparent...) Largely hidden behind the façade is an organization with personalities, history, divisions, uncertainties, and struggles (not least to grow fast). However unitary its site, Kiva is not a monolith, and its actions do not emanate from a single, all-knowing intentionality. I got the impression, for instance, that many people within Kiva had never seen the MFI-facing diagram of how it works. If so, then I overreached somewhat in commenting that the existence of this diagram "reflected a choice, not just an accident caused by scrambling to keep up with its growth."
  • That said, I stand by the spirit of the comment. Even when an organization acts in ways that are not entirely premeditated, those acts reflect organizational values. That the "How It Works page" languished unchanged for perhaps three years despite the problems I pointed to says that the organization had higher priorities. In particular, the Kiva organization has two main divisions: the people like Ben who work with MFIs and the tech guys who do the web site. I got the impression that the tech side drives the organization. Matt and Premal both come out of the Silicon Valley/Bay Area Internet start-up milieu. The site is what makes Kiva special. By training, people coming out TiVo, eBay, and Yahoo! don't think as much about the product they're selling as they do about how to sell it—how to maximize the "conversion rate" (% of visitors who buy). One way to do that, as an example, is not to mention on the loan page that the borrower will pay 30% or 50% interest. I know people at Kiva care deeply about microfinance and are learning more all the time, but I think the dominance of the Internet commerce culture is one root of the current controversy.
  • Ben persuaded me that I probably overestimated the costs for MFIs of working with Kiva. Yes, they must collect, translate, and post stories, and pictures too, but this costs something closer to 1% of the loan amount than the roughly 10% I feared. I had figured that the labor was comparable to that of making a loan. But a) the MFI only has to do it once for each loan, whereas it must typically meet with the borrower every week to collect loan installments; and b) for group loans, it must only perform the act once for the group, not separately for each borrower. (I'm not certain where I got the erroneous figure that Kiva charges MFIs 2% interest. I know I didn’t completely make it up, but I apologize for the mistake.) On the other hand, the cost of operating the web site may be substantial. Kiva's expenses (registration required) for the first months of 2009 were $1.9 million, against a throughput I calculate at $20.7 million during the same time, for a non-trivial 9% expense ratio. (This snapshot might not be indicative of the long term trend.) In theory, if Kiva users stopped asking for stories, Kiva could drop its P2P (person-to-person) fiction, shrink its operations, and pass substantial savings on to MFIs and their borrowers.
  • I learned that when users complain to Kiva about feeling misled, it is less often about what I wrote about, that loans are disbursed before posting, than about Kiva borrowers having to pay what seem like high interest rates. (Colleague John Simon, a friend of Kiva, first pointed this issue out to me.) On the page for the loan to Phong Mut in Cambodia, who I wrote about before, the interest rate she pays appears nowhere. Since Kiva users earn 0%, and since nothing on the loan pages implies otherwise, some (many?) Kiva users assume borrowers pay 0%. But if you dig around, you can learn that Phong Mut's lender, MAXIMA, typically charges 29%/year. It keeps all that to cover the costs of providing microcredit. How much interest Phong Mut will pay is unclear. The repayment history at the bottom of her loan page shows principle payments only. Nor is interest in the new How Kiva Works diagram. While it is appropriate for MFIs to charge enough interest to cover their costs, which are typically substantial relative to the small loans being made, the higher the interest rate, the more frequently will credit leave borrowers worse off. Rates do matter and Kiva users ought to understand this aspect of the transactions to which they are party.

Kiva fellow videotaping borrower in Sierra LeoneOne clever piece of the Kiva organization, and a special strength, is the Kiva Fellows, volunteers who, at their own expense, visit MFI offices and borrowers to collect and check stories. They are the eyes and ears of Kiva. Among other things, they sniff for fraud. (They are also a hidden cost, in the sense that Fellow's time and expenses are an economic cost of charity in the Kiva model.) The Fellows program page contains a picture (right) of a Fellow videotaping an interview with a borrower in Sierra Leone. The young American is viewing the man through the lens of a camera that probably cost more than the Sierra Leone's GDP/capita, having arrived their on airfare priced even higher. He is having a good experience. The borrower may not have heard of Kiva until shortly before this interview. Chiefly on his mind must be how to protect his access to capital, which means saying whatever it is he thinks the young American wants to hear. Probably he understands that he must talk about his enterprise even if he used the loan for to pay for his father's funeral. Probably he understands that he should accentuate the positive to come off to the extent he can as the "entrepreneur empowered to lift himself out of poverty," to paraphrase Kiva's old home page. He is to play a role. Though probably older and wiser, he is less lucky in life, so he is the object. [Note: Some Kiva Fellows have strongly objected to my projections here.]

The picture crystallizes my feelings about Kiva's story-constructing ethos, which I emphasize is emblematic of microfinance groups and charities generally. It makes me uncomfortable. Perhaps that means I am a self-righteous critic, a haughty intellectual enjoying his pretended superiority over the majority of the human race that craves stories, while not having to get his hands dirty actually helping people. Maybe so. Regardless, I would say this: If you are comfortable enough with this picture, and understand and like how Kiva operates, interest rates and all, then lend through Kiva. If not, then both you and I need to learn how Kiva compares to other microfinance conduits on overhead rates. Do groups such as the Grameen Foundation and Acción International and MicroPlace channel funds to MFIs more efficiently? Or has Kiva, with its high-tech shoestring culture found a way to harness modern technology to bring the cost of the P2P experience close to zero?

Taking a step back, I see that the important question here is not whether Kiva is adequately transparent. It is whether and when microcredit helps people. Kiva, like all microfinance outfits, faces powerful incentives against being too inquisitive about the latter question. This is why the groups I most admire are the ones that are inquisitive enough to perform serious research, such as Freedom from Hunger and Bangladesh-based BRAC. BRAC, in addition to being a Kiva partner in Africa, is an extraordinary learning organization, as I am learning now from Ian Smillie's book. I believe that BRAC founder Fazle Abed ought someday to stand next to Muhammad Yunus as a Nobel laureate. In this light, Kiva's hospitality and responsiveness to me in the last few weeks bode well. What I want to see more of in this business is what Abed infused in BRAC: a culture of public, critical thinking about what is being done.


CGD blog posts reflect the views of the authors drawing on prior research and experience in their areas of expertise. CGD does not take institutional positions.


David Roodman's Microfinance Open Book Blog