Kenya’s Central Bank Governor Njuguna Ndung’u recently urged the country’s mobile money transfer (MMT) operators to reduce their transaction fees. According to the Governor, “There is no way one can send 50 Shilling at 35 Shillings”. This translates into a seemingly exorbitant 70 percent fee for a small transaction equivalent to less than $1. Safaricom, the telecom operator that offers M-Pesa service (a highly successful Kenyan venture with more than 13 million clients), and other Kenyan MMT operators, however, maintained that the services they provide represent value for money. So are mobile money transfer costs too high?
Before we answer this question, it’s worth pointing out that even if mobile money transfer costs are fixed, average costs expressed as a share of the amount sent can rise if the average size of transactions falls. That is what happened with M-Pesa. M-Pesa charges a fixed fee per transaction within pre-specified fee brackets (see tariff poster). As the use of M-Pesa spread, Kenyans started using it for smaller and smaller transactions. The average amount sent through M-Pesa declined from the equivalent of about $50 in March 2007 to less than $30 by March 2009. The fees charged by M-Pesa, including withdrawal charges, expressed as a share of the average amount, rose correspondingly until mid-2008 (see chart). Because the average transaction size fell to the lowest fixed fee bracket in mid-2008, there was a downward jump in the fee. Then average costs rose again up until March 2009.
Chart 1: Mobile money transfer fees rise with a decline in transaction size
Note: Transfer fee includes charges to send money and the fee for a registered receiver to withdraw cash. Sources: Pulver, Jack and Suri 2009 and Safaricom’s tariff poster.
With the expansion of mobile money transfer services to remote, rural areas in Kenya (a country where the average monthly income is $63), it is possible that the poorest are using M-Pesa to send and receive even smaller amounts. For example, for sending 800 Kenyan shilling (equivalent to about $10) through M-Pesa, the average transfer fee including withdrawal charges rises to 7 percent for transfers to recipients who have an M-Pesa account (and to 9.4 percent for unregistered recipients). Mobile money transfer costs relative to the incomes of the poorest could be much higher than what the average cost implies.
Whether mobile money transfer costs are reasonable, transparent and affordable is important in the quest for increasing financial inclusion of the poor. Increased competition and technological advancements will no doubt put downward pressure on mobile money transfer costs in a dynamic industry. A relevant question is the break-even point for providing these services for the operators. An earlier analysis by Ratha and Riedberg of the cost structure of the cross-border remittance industry concluded that the overall international remittance costs (including staff, infrastructure, marketing, and compliance) could be as little as $3 per transaction. A similar analysis of the cost structure of the mobile money transfer industry could provide policymakers guidance on the feasibility of reducing costs further.
The Afrique en ligne (Africa Online) newspaper features an interesting article by Isabelle Gross which suggests substantial room for improvement in the pricing of mobile money transfer (MMT) systems in Africa and the need for interoperability across mobile money providers.
It says mobile money transfer service providers in Kenya, Uganda, and Cote d'Ivoire use a variety of pricing schemes to make the transfer fees appear low: "For registered users, mobile operators split the cost of transfer and withdrawal between the sender and the receiver. By spreading the cost, the service also appears cheaper."
Mobile money transfer service providers also charge more for unregistered recipients: "The "walled garden" commercial approach of mobile operators for payment services gets tougher when it comes to the fees charged for [mobile money] transfers. The common denominator among the four operators [in Uganda] is: If you transfer money to another mobile user registered to the service, the fee is rather small but if you transfer money to a person that is not registered to the service then the fee can be five to 20 times more."
The article argues that since "The processes behind the transfer and the withdrawal are the same for the two categories of users and therefore for the mobile operators, the cost should be the same too."
Moreover, the lack of interoperability across mobile money transfer providers often implies that it expensive and sometimes impossible to transfer money to a mobile subscriber with another network. It asks:
"Is it interoperability between competing mobile payment services that will finally get rid off this “walled garden” commercial approach?"
While the article is well-reasoned and asks the right questions, there is need for more analysis and research into the pricing and industry structure of mobile money transfer services, so that regulations and policies can be tailored to the welfare of the ultimate beneficiaries.
The full article is available at: http://www.afriquejet.com/news/africa-news/africa's-mobile-money-pricin… and at http://mobilemoneyafrica.com/?p=3247
What kind of significant policy changes should be recommended (on the various levels) to improve how remittances are transmitted.
Do you have constructive and practical recommendations pertaining to the infrastructure for transmission of remittances?