CLOSING CASE
Send Money Home: M-PESA and the Kenya Experience
The slang term unbanked refers to people who do not use banks or other
financial institutions. Instead of using checks and cards, they conduct most
of their financial transactions in cash. While there are many reasons for
people to be unbanked, most of them are poor and either lack the credit
standing to have a bank account or they are located in poorer regions where
there are no banking services. In many countries around the world, moving
from the ranks of the unbanked is almost a necessary step for getting out of
poverty. From 2011 to 2014, the number of “unbanked” adults declined an
astounding 20% to 2 billion (World Bank 2015). The drop was not the
result of declines in the number of unbanked living in the more advanced or
growing economies. Instead it was almost solely attributable to shifts in the
number of unbanked residing in sub-Saharan Africa and more specifically
in Kenya. It was due to a program called M-Pesa originally intended to
provide person-to-person international remittances conducted via cell
phones. This closing case describes in detail the M-Pesa program including
the problems it was designed to address, the structure and operation of the
program, and its long-term results on the unbanked poor in Kenya and other
parts of the world.
The Problem
In developing countries, immigration is “a way of life.” When done
voluntarily, people often migrate for the express purpose of finding work or
taking advantage of opportunities outside of the countries or areas where
they live. Worldwide, this migration results in a massive transfer of money
from workers to their families and friends back home. These transfers are
known as remittances. While any single remittance is usually small, the
aggregate amounts are substantial. For example, according to figures from
the World Bank (2016), “official” global remittances involving foreign
workers from developing countries totaled over $430 billion in 2015. To
put this in perspective, for many developing countries, the yearly total is
often more than the development assistance they receive from all sources
and larger the monies from direct foreign investments. These sums not only
represent big money from the developing countries’ perspective; they also
represent big money for the money transfer operators like Western Union
who handle these transfers. The MTOs charge fees for their services and are
making money from currency conversion. While most of these operators
follow strict guidelines and rules, even small charges can have a major
impact on the amounts received by the individual families. In addition to
these global remittances, developing countries also have sizeable “internal”
remittances generated by workers who have moved from the rural parts of
the country to take jobs in the cities. Because the workers and their families
are unbanked, most of these remittances take place the old fashion way.
Either the workers take the cash home themselves or they have someone do
it for them. While this certainly avoids the fees charged by the MTOs, it is
still a costly and dangerous undertaking—it takes time to transport the
money (usually by bus) and places both the person and the money at risk
given the high rates of robbery in many of these countries. Also, because
the transfers occur outside any formal financial system, there is no way to
measure the numbers of people, cash, and transactions involved. In the past
a number of countries have sought to address either directly or indirectly
the issues and inefficiencies in global and domestic remittances. Many of
these programs revolved around the idea of somehow turning the
“unbanked” into the “banked.” Theoretically, if both the sender and the
receiver had bank accounts, this could certainly simplify the domestic
transfers of funds and would open other possibilities for international
transfers. It might also address the larger issues of poverty, thus alleviating
the need for family members to separate in the first place. Ignoring the fact
that it costs money to have a bank account and the fact that banking systems
in many developing countries are suspect, this approach is a massive,
costly, and long-term undertaking. As the last few years have demonstrated,
the real answer might be in mobile money and the “un-banking” of the
banking system. At least, that is what the experience in Kenya over the last
few years have shown.
The Solution
The history of the M-Pesa is well documented in Money, Real Quick by
Omwansa and Sullivan (2012) and touched on more recently by Runde
(2016). The discussion that follows briefly covers some of the key events
that these discussions highlight. Kenya is an East African country of
approximately 40 million people with high unemployment and poverty.
Approximately 10 years ago, an in-depth survey of the Kenyan financial
sector found, much to the surprise of the Central Bank of Kenya, that only
20% of the adults in the country were “banked.” Basically, the bank was
servicing urban elites and slowly dying in the process. The same could be
said for the government-owned telecom system which historically had
provided landlines in urban areas. As a result, only 2% of the population
had phone service. In contrast, Kenya’s mobile operators, of which
Safaricom Network Company was by far the largest, had managed in a
relatively short period of time to get ten million mobile phones in the hands
of the Kenyans (for a 35% penetration rate). So, instead of thinking about
how to spread a dying landline business or branch banking business to the
rural populace or for that matter to the slums of the cities, maybe it would
be easier to figure out how to use mobile phones to help bring financial
services to the poor.
The original pilot for Kenya’s mobile money system was run under the
auspices of Britain’s governmental development agency (DFID) and
focused on reducing the costs of microfinance loan repayments and
lowering the associated interest rates. After the initial foray, control of the
program was shifted to Safaricom, and the focus morphed away from loan
repayments toward person-to-person money transfer. The new system was
called M-PESA—the “M” stands for mobile and “PESA” is the Swahili
word for money. The marketing slogan for the new system was simply
“Send money home,” although the system had broader financial
capabilities. The task of sending money was relatively straightforward.
First, the sender and the receiver had to have mobile phones that supported
texting, which virtually every mobile phone has regardless of the
underlying technology. Next, they had to get Safaricom SIM cards. Once
they had the SIM cards, they had to register with an M-PESA agent. All
that was required to register was an identity card, something that every
Kenyan had. Once registered, the network sends an updated menu to the
registered customer’s mobile phone. At this point, the system is ready to go.
To actually send money, a registered customer first deposits cash to his or
her account. This is done by giving cash to an M-PESA agent who
immediately credits it to the customer’s account. The network sends a text
message to verify the deposit. Once it’s in the customer’s account, he or she
can send money at any time by selecting “Send money” from the M-PESA
menu and then entering the recipient’s phone number. At this point, the
sender is prompted for his or her M-PESA PIN and then selects “Ok.” At
this point, the system sends a message to the sender confirming the transfer
and the recipient’s name. In turn the recipient receives a message with the
sender’s name and the amount transferred to his or her account. On the
other end, a recipient can now go to a local M-PESA agent to pick up the
money. Actually, the recipient is basically making a cash withdrawal from
his or her account. It’s done by showing the agent his or her identity card,
choosing “Withdraw” cash from menu, entering the agent’s ID number, and
then entering his or her M-PESA PIN. Once the transaction is confirmed,
the agent distributes the cash. Obviously, a major key to the success of the
system is the M-PESA agent network (Stahl 2015). It has been described as
a network of “human ATMs.” An agent might be a local grocer, or gas
station owner, or a post person, etc. They were recruited through a thorough
selection and vetting process. Many of them were also selling mobile
airtime for Safaricom. They receive regular training and are frequently
monitored. They are also restricted from doing business with other mobile
operators. Outside of the due diligence process, the major hurdle for an
agent is monetary. Agents have to pre-purchase mobile money so they can
sell it to customers for cash. Likewise, they have to sell cash for mobile
money so customers can withdraw funds. Both the cash and the mobile
money they manage are theirs, not M-PESA’s. Some agents do well, but for
the majority, it is a part-time job. Less obvious is the fact that M-PESA is
modeled after a “prepaid” mobile phone system where consumers pay
upfront for minutes rather than relying on credit and paying for minutes
after they are used. In M-PESA you don’t need credit approval upfront to
open and use the system. Essentially, you open an account and then deposit
money. There are no fees for making deposits nor are there fees for adding
airtime to a phone. There are fees for transferring and withdrawing funds.
They pale in comparison to fees charged by the MTOs and by ATMs. There
are also restrictions on the maximum amount of money you can store in
your account and the amount of money you can transfer at any given time.
There are a variety of reasons for these restrictions. First, a majority of
customers are poor so the system is focused on their needs. Second, they
don’t want the system to be used for illegal purposes (e.g., money
laundering). Third, and most importantly, M-PESA is not a bank. Monies
are held in a trust owned by Vodafone (the major shareholder) and
deposited in commercial banks.
