Ideas to Action:

Independent research for global prosperity


David Roodman's Microfinance Open Book Blog


This post continues my previous report on my trip to Kenya to see M-PESA, the mobile phone--based money transfer service. Here's the full set.

On our second day, last Tuesday, we flew to Kisumu, a town on Lake Victoria, that vast body of water straddling the borders between Kenya, Tanzania, and Uganda. There we met Frederik Eijkman, co-founder of a local company called PEP Intermedius. Like M-PESA itself, PEP began with microfinance, in 2004. But when M-PESA came along in 2007, Frederik and the other founder, Paul Otieno, seized the business opportunity. To launch M-PESA, Safaricom needed to quickly build a national network of M-PESA points analogous to Western Union outlets, where people could swap paper and e-money. Today PEP manages just over 100 M-PESA outlets; it owns eight and supports the rest as franchises. PEP outlets compete side-by-side with other M-PESA points.

I first learned about phone-to-phone e-money in 2006 at a conference in the basement of the International Finance Corporation. The impressions made upon me then naturally centered on the phones. In a photograph from the Philippines, a customer and a merchant stood facing on opposite sides of a display case in a food shop, each holding up a phone in order to do their bit of electronic business. Brian Richardson, creator of the phone-based bank WIZZIT in South Africa, pulled a phone out of his pocket and before our eyes, in seconds, transferred money to his wife back home. So in my mind before I went to Kenya, phone banking was electronic banking. Poor countries were leapfrogging the rich, leaving cash behind.

There is some of that going in Kenya, but in Kisumu, and in a paper by Frederik and the Gates Foundation's Ignacio Mas and Jake Kendall, I saw a far richer story.

Consider the challenge facing this woman, who runs a busy PEP-backed shop:

All day, people come to her expecting to swap paper money for electronic (put cash in) or vice versa. Her business depends on her dependability. If she runs out of cash, she will disappoint customers wanting to withdraw, and they make take their business elsewhere, permanently. So she starts her day with several thousand dollars on hand---her own capital. And yet she could run out during the day and have to go to the PEP office for more. Or she could accumulate more cash than she feels safe holding. If you click the picture to enlarge it, you'll see that she does not operate alone, and that the two women have two posted M-PESA agent numbers (04514 and 32698). By sharing the storefront, one can cover when the other has to go to PEP or do other errands, virtually guaranteeing availability to customers at all times.

Now, this shop is just across the street from PEP. But other PEP outlets we saw were farther, up to half an hour away by car. For many M-PESA shop owners, a visit to PEP---or to the nearest branch of a bank where PEP also has an account---involves time, money, and risk of robbery. (Ten PEP stores experience robberies last year.) Look at this fine portrait of another PEP agent:

The day-to-day work of PEP is analogous to that of the M-PESA shop owners, just a step up in the hierarchy. Over the Web, PEP monitors the transactions of the M-PESA points it supports in order to anticipate who will show up to get or deliver cash. PEP needs to have enough cash on hand to meet their needs promptly, and may have to visit the bank during the day to manage its cash stock. The banks in turn do the same thing on a national scale (and in fact some bank branches have registered as M-PESA agents like PEP). Thus M-PESA can be see as a highly filigreed extension of Kenya's banking system, a new system of capillaries to extend the established veins and arteries. Banks do many things; one is to manage the movement of cash around the country. M-PESA is bringing this important function to millions of poor people who before could not afford formal bank accounts and had to rely on other, more costly and risky, ways of sending money cross-country.

As I see it, M-PESA is a high-quality service: instant, on-demand cash-to-electronic-money conversion, almost anywhere in the country. The work needed to provide this service consists primarily in maintaining presence (having 16,000 people in the shops, on the ready) and moving cash behind the scenes in anticipation of demand. From this point of view, the heart of M-PESA is not, as I originally supposed, in the phones. The phones, rather, are a wrapper. But they do hold the system together, for:

  • The phones make transactions credible. The instant you turn cash into an electronic balance on your phone, Safaricom verifies your new balance by text message. That makes the new-fangled e-money seem real and trustworthy. More generally, the phones let customers hold Safaricom and its agents accountable for promises of service to a degree impossible in the informal methods of sending money. If you try to send money by van and it disappears along the way, what do you do?
  • Trust in turn allows the professionalization of cash transport. Before Gaudencia (above), her dozens of customers may have had to trek individually to Kisimu to pick up cash sent home by husbands in Nairobi. Now Gaudencia can effectively do it for all of them, all at once, saving many person-hours of time. I suppose that is an economy of scale: one person can carrying $1,000 is far more efficient (if a bit more risky) than ten carrying $100 each. That is the real technological breakthrough, for it cuts the work needed to transfer money long-distance, point to point.
  • The phones allow real-time aggregation of transaction information so that the likes of PEP can anticipate cash demand more efficiently.

Late in the day, Frederik took us to the small market town, or cross-roads, of Holo. I think it is on the same road, winding northwest from Kisumu to Uganda, that passes by the Bumala market, where microsavings was first put to the randomized test. Frederik told us that the arrival of M-PESA had transformed the Holo market. Before, people trekked to Kisumu to get their cash, and spent much of it there. Now the cash comes to them, and they spend it locally; the market thrives.

It struck me then that, ironically, the principal impact of the new form of mobile money may have been to increase access to the ancient form. As I related in my first M-PESA post, Safaricom currently charges a steep price for small electronic transfers---30 shillings to move 100 (about $1.20) from one person's phone account to another's. Only a fool would pay for his yams in the market that way. (Using M-PESA to pay bigger bills is, however, being promoted.) For now at least, Kenya is not leapfrogging to an all-electronic, cashless economy. As a result, for the poor who can access M-PESA, the new system may be doing more to expand access to cash than to supplant it. I never would have guessed.

The funny interplay between currencies reminded me of a rhyme I learned around age six:

Make new friends

But keep the old.

One is silver

And the other gold.


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