Bryan Farris

An Interview with David Roodman (Part 2): Technology and Microfinance

This is part two of a three-part series; you can read the first part here.

Recently, I had the opportunity to have a discussion with David Roodman, who works at the Center for Global Development where he spends most of his time trying to understand the ins and outs of microfinance. He is trying to take a rigorous look at the space to understand what works, what doesn’t and how microfinance can be improved. David is best known for his blog posts about Kiva and about a potential repayment problem at Grameen. David Roodman is writing a book on microfinance which is likely to be published in the fall.

In part one of the series, David and I discussed the importance of credit bureaus and the phenomenon of multiple borrowing. David expressed concern that without proper tracking of credit, lenders may be unable to accurately assess the risk of their loan portfolio.

Bryan Farris, Do you think the introduction of credit bureaus and more structure around the lending process would remove a component of risk and enable banks to lower interest rates and be more competitive relative to high interest alternatives?

David Roodman: Portfolios of the Poor came out last year, and it is a really fine book that helps us better understand how poor people manage their money on a day to day basis. The authors found that a lot of poor people will borrow money from what we call money lenders or informal local creditors. The interesting thing is what they found happens, in general, when people pay those loans back late. Someone may borrow at 10% a month for four months, which you know, adding that up over a year can translate into a pretty high interest rate, upwards of 150% or so. When people pay those loans back late, they often don’t pay any extra interest. So, instead of paying 40% interest in four months, they pay 40% in six months or twelve months. What that means is that the actual interest rate that people are paying on informal credit as an alternative to micro-credit is often not as high as it seems.

Micro-credit still tends to be expensive; interest rates around 30% are not unusual and are higher than what most people have to pay in rich countries. Basic economics drive interest rates up; it does not cost a tenth as much to make a hundred dollar loan as it does to make a thousand dollar loan. Regardless of loan size, lenders need to spend time processing in the loan as well as finding borrowers and selecting them. As a result, the cost per dollar lent is higher and in order to be sustainable, MFIs have to charge more interest.

To answer your question, it’s hard to get around high interest rates unless there is a significant technological breakthrough which fundamentally changes how we can do micro-credit. It’s conceivable that a credit bureau system with a high-tech ID piece at the heart of it could be that breakthrough. If that occurs, lenders could summon a borrower’s credit history to identify reliable borrowers. Eventually, you could imagine it being the semi-automatic process.

Economists consider technology as a way of transforming inputs into outputs. Technological advance occurs whenever somebody figures out how to make more or higher quality output with the same input. Group credit is a technological advance that was discovered (or rediscovered) in the 1970s. With group credit, people are responsible for their peer’s loans.

If you and I were in a group together and you don’t pay back then I’ll have to cover for you and that turns me into a sort of unpaid employee of the bank. As a result, I’m going to be careful about who I enter into joint liability with, because I want to make sure that you’re a reliable borrower before I join with you. Once you take out a loan that I’m responsible for, I’m going to keep tabs on you to make sure that you’re still paying and you’re using your money responsibly. What happens with group credit is that the members of the group start to do the banker’s job: selecting each other, following up and trying to collect payments when things are difficult. That means that the banker has to work less, which helps cut down what would otherwise be even higher expenses. Isn’t it true though that the banker still meets with the group on a very regular basis, typically once a week? The point that you mentioned before about automating the credit and loan process is interesting; in theory a borrower could just go on the computer and look up their credit score and then get automatically approved for a loan. Do you think that it’s possible, with technologies such as M-Pesa, to eliminate the need for bankers and groups to meet once a week? If so, could that reduce costs significantly?

David Roodman: Conceivably yes, though I want to be very careful because it’s easy for me to sit here in Washington and say that all the poor need is this new high technology solution. I can imagine there can be a hundred reasons why it wouldn’t work.

Typically in classic group credit, a loan officer, usually a man, will meet with forty members of groups, and they are usually women, over the course of say an hour. During the meeting, they do all the week’s business, including dispersing new loans, taking repayments and maybe taking savings deposits too. The same officer might cover three, four, five villages in a day and thereby serve several hundred people per day. So it’s a kind of mass production.

You mentioned M-Pesa’s new mobile phone payment system for easy money transfers. If you want to send money to your relatives in the countryside from where you live in the city you can do that now by phone. Or you could even go to a store and pay for things, phone to phone. People are just starting to discover that they can use that system to save too. They can put money into the system-in a sense buy minutes-then not take it out and it’s turned into a savings device. When a user wants to convert electronic money to paper money or vice versa, they have to go to an agent of M-Pesa, who probably also runs a corner store -he or she doesn’t just do M-Pesa business. That agent in a way takes the role of or replaces the credit officer that we just described.

Like you suggested, that agent may be able to provide the essential service more economically because he or she is selling a lot of other things and is positioned to take advantage of economies of scale and diversification.

In fact, there is also another thing going on. The shopkeeper may lose money on their M-Pesa business but they still benefit when the M-Pesa customer withdraws cash and turns around to buy stuff in the store. In other words, the shopkeeper benefits from cross-selling. What is actually going on with the economics of these mobile phone money transfer systems is that shopkeepers are replacing dedicated credit officers.

– – –

This concludes the second part of the series. In the next part of the series, David and I discuss forms of microfinance beyond microcredit including microsavings and microinsurance. David shares his opinion on methods for encouraging savings at the base of the pyramid.