Kenya is the first country in the world to use mobile phones for cash transfers. The service, called M-PESA, was developed by Safaricom Limited. Concern Worldwide pioneered the use of M-PESA for bulk cash transfers during the post-election emergency in early 2008 in the Kerio Valley, one of the remotest parts of Kenya. During the violence, cattle rustlers attacked communities in the Kerio Valley, looting their livestock and displacing them. Concerns initial response was to provide food aid, but carrying and distributing food proved very costly and insecure. Cash transfers were seen as a way of overcoming the challenges posed by the terrain and the security situation. In all, 570 households were targeted with cash transfers. A total of 36,000 (about $53,000 or 3,600,000 Kenya Shillings) was disbursed in two instalments.
An overview of M-PESA
The M-PESA service does not require users to have bank accounts. All they need to do is register at an authorised M-PESA agent by providing their Safaricom mobile number and their identification card. Once registered, the user can buy digital funds at any M-PESA agent and send that electronic cash to any other mobile phone user in Kenya by SMS. Recipients can either redeem this for conventional cash at M-PESA agent outlets or buy Safaricom airtime for themselves and other subscribers. An M-PESA-enabled mobile phone can also function as an electronic wallet, holding up to 500 (50,000 Kenyan shillings).
The first step involved developing a customised system allowing for bulk transfers to hundreds of beneficiaries simultaneously. Safaricom set up Concern as a corporate user of M-PESA. Concerns computers were enabled to access M-PESAs administrative website, from which the disbursement was done. Money for disbursement was deposited into Concerns M-PESA account, including service charges. Next, the list of beneficiaries, amount due and their mobile phone numbers were entered into a database that could be uploaded into the M-PESA system. Safaricom staff trained Concern on using the M-PESA system for bulk payment and to generate relevant reports.
The emergency response team of Concern and its partner, the Catholic Diocese of Eldoret, conducted a number of sensitisation meetings to establish women as beneficiaries of the project. The team also discussed the nature of the intervention, targeting, timeframe and resource allocation. Sensitisation meetings were conducted with the help of the local administration and religious leaders. In association with community volunteers, the team also conducted household assessments to verify the number of people in each household and how the household was affected by the crisis; possession of national identification cards and numbers; and household access to mobile phones and numbers. Further verification and targeting exercises were conducted in public community meetings to ensure that the most vulnerable people and households were included.
Registration and cash transfers through M-PESA
Registration was a painstaking process, presenting the team with a number of challenges. Although women were targeted as recipients of the cash transfer, most lacked identification documents, which are a key requirement for receiving cash at M-PESA centres. Second, M-PESA runs on an SMS platform, meaning that users need to be literate, but a large number of beneficiaries could not read or write. Lastly, about 60% of the targeted beneficiaries did not have access to mobile phones.
To overcome these problems, targeted households were clustered into groups of ten or less, and one literate person was nominated as cluster leader. Beneficiaries without identification documents nominated a trustworthy adult member of the household to receive cash on their behalf, with the close monitoring of the cluster leader. Where no cluster member had access to a mobile phone, cluster leaders were provided with one, along with a solar charger, and trained on how to use the equipment. It was made clear to cluster leaders that they would not individually own the phones, and could not sell them or charge cluster members a fee to use them. Although the equipment was shared by all cluster members, each beneficiary received his or her own SIM card to register for M-PESA, to reduce the risk of cash transfers falling into the wrong hands.
The standard procedure is for people to go to an M-PESA agent with their identification documents to register and withdraw money. In Kerio Valley, the nearest M-PESA agent was 80km away from the target area. Travel costs were prohibitive, and security risks high. A remote M-PESA service was accordingly established at the Kinyach Police Station in Kerio Valley. Kinyach is a secure, central market for all the targeted communities. The Safaricom agent was dispatched there during market days, giving beneficiaries easy access to the cash and allowing them to use it immediately to buy their food. It was estimated that households had a food deficit of about 50%, so the size of the cash transfer was calculated to cover this. Based on local market conditions, it was calculated that each person needed 3.20 every two weeks to enable households to meet 100% of their food needs half with the cash transfer, the other half with food from other sources. The transfer, equivalent to two weeks food ration for every household, was calculated based on the number of members in each household multiplied by 3.20.
Costbenefit analysis of the M-PESA-based cash transfer
Analysis shows that the M-PESA-based cash transfers were secure, enabled a quick emergency response, were cost-effective, respected peoples choices and empowered communities.
1. Direct costsThe cost of transporting and distributing emergency food from Eldoret to the Kerio Valley (relief food for 100 families for two weeks) was 350 or 18.5% of the cost of the food (1,900). A cash transfer of a similar amount to 100 households cost 55, mainly in Safaricom transaction charges. In addition, six out of every ten household clusters received a mobile phone and a solar charger, i.e. six phones and chargers per 100 beneficiaries, which cost 250. The transaction and equipment costs came to 305, or 16% of the 1,900 assistance package. While this is lower than the food distribution costs it should be noted that providing the equipment was a one-off cost, as opposed to the food distribution costs which recur with each distribution.
2. The recipients and the local market A market assessment in Kerio Valley revealed that sufficient food was available on the market, and that prices of basic items like maize and beans were half those in Eldoret town. This meant that the cash transfer gave targeted households access to a wider range of commodities at a lower cost. The injection of cash also strengthened the local economy.
3. The role of information technology in empowering the poorEvery beneficiary received a SIM card and a shared mobile handset with solar charger. The ability of mobile phones to send SMS expands the range of cheap communication available to the poor. This isolated community now knows how to receive remittances by using M-PESA. Communities also see advantages in using their mobile phones to develop early warning and security alarm systems to prevent cattle-rustling. Women also see income-generating opportunities by offering informal payphone services to other members of the community.
4. NGOprivate sector partnershipThe private sector has a unique ability to enhance the effectiveness of emergency response. Although Safaricom waived none of the M-PESA transaction charges for cash transfers, it brought its talent and technology to bear in the partnership by developing a customised version of M-PESA for bulk transfers. Safaricom also provided support in training the community on the use of M-PESA and mobilising the M-PESA agent in the target area.
5. Might alternative delivery mechanisms have been simpler? Concern had three options: continuing direct food distributions, implementing a food voucher scheme and cash transfers. We have already seen that food distribution was not the best option for Concern. A voucher scheme was not a viable option in the absence of any local partner organisation to set up and run it at the local level. In addition, a large number of beneficiaries were displaced and living in temporary camps, making it difficult to establish a voucher scheme because people were expected to move back to their original settlements as the security situation improved. For a voucher scheme to run effectively it requires a formal agreement between a local organisation and established businesses, mainly shops. This was not possible since the main food items purchased were bought from open markets, which are normally too fluid to allow such arrangements. Finally, direct cash distribution was not viable: security risks for staff and beneficiaries were too high, and Concern would have incurred high personnel costs during every round, just as in a food distribution. With M-PESA, neither Concern nor Safaricom staff members needed to maintain a long-term presence in the area after beneficiaries had been registered.
Lessons learned and common concerns
Some of the key concerns over cash transfers include: cash is difficult to target because everyone wants money; a cash injection may cause inflation and may increase security risks; women may not have control over the income; and it may be spent on the wrong things. However, field experience shows that many of these perceived risks and fears are not borne out in practice.
Taking into account peoples natural attraction to cash, a deliberate decision was made to conceal the nature of the assistance until beneficiary identification was complete. Targeting women for the assistance generated some resistance from the men, but the sensitisation meetings helped people to understand why women were being targeted and to motivate the community. Once the cash transfer was completed, Concerns emergency team conducted a number of focused group discussions and interviews with beneficiaries and the wider community to assess the impact of the cash transfer. Concern also engaged an external evaluator. Neither the emergency response team nor the evaluator registered any gender-related incidents either at the household or community level. There was no evidence that men had taken any money from the women or used it inappropriately.
The direct observation and evaluation findings show that more than 95% of families spent most of the money received on basic foodstuffs. Some beneficiaries bought meat or chicken in addition to basic food. A few families used a large portion of the money to pay school fees and buy school uniforms for their children. The families who saved some livestock from the raid spent a large chunk of the cash on medicine to treat their animals. A few families also tried to revitalise their businesses by restocking goats, starting mobile phone airtime businesses or setting up shops.
A number of households mentioned that the cash was not enough to buy food for two weeks. Some households that had moved away from the target area to escape the violence were not included. Beneficiaries explained that they shared their food or cash not only with these returnees but also with neighbours, friends and relatives, using up the cash within the first week. Culturally, the community does not believe in food rationing, but rather eats large meals when food is available.
A number of beneficiaries complained about an increase in the price of goods after the first round of cash transfers. A market survey and interviews with community members indicated that food commodities such as maize and beans had increased in price by an average of 25%. Community members and some traders explained that the rises were due to the impact of the post-election violence. Price inflation was common all over Kenya during this period. Given the size of the cash injection it is unlikely that it could have caused the levels of inflation witnessed, though it may have been a contributory factor.
The overall success of this pilot project offers opportunities to scale-up. However, the following issues need to be considered:
- The beneficiary database needs to be improved to detect errors relating to mobile phone numbers and national identification numbers. The capacity of M-PESA to cope with bulk transfers in the scale-up programme needs to be strengthened.
- Mobile phones are not robust enough to cope with frequent changes of SIM card. The ratio of mobiles to families will have to be increased. The best option for a long-term cash transfer project is to ensure that the targeted households have mobile phones and M-PESA-compliant SIM cards. However, literacy levels need to be taken into account.
- The second transfer round was expected to be easy. Unfortunately, 20 of the 570 beneficiaries lost their SIM cards. Concern had to recall the money and then do the transfer again once those who had lost their cards received replacements. A quick mechanism to deal with lost SIM cards needs to be developed.
Having successfully implemented this project in one of the remotest parts of Kenya, it can be concluded that it is possible to do similar work anywhere in the country as long as the area has a mobile phone network. The challenges of implementing this project in urban environments or in developed rural areas will be less significant, as more people in these kinds of areas are educated and have access to mobile phones. Although Concern used M-PESA in an emergency, this model could be replicated in development contexts for a variety of social protection activities. The project was also a valuable learning experience for Safaricom. The company was made familiar with the needs of communities in Kerio Valley, and given an opportunity to test new ways in which M-PESA could be used. This experience remains an asset in the future development of the M-PESA cash transfer system, and has shown the products potential.
Dipankar Datta is Assistant Country Director for Concern Worldwide Kenya. Anne Ejakait is Concerns Programme Officer and Monica Odak is Assistant Programme Manager. The authors acknowledge the assistance of Edgar M. Andagalu, Key Accounts, M-PESA Department, Safaricom Limited. Christine Bukania, Learning Documentation and Advocacy Officer of Concern, helped with editing. The views expressed in this article are the authors, and do not necessarily reflect the views of Concern or any of its partners, including Safaricom Limited. Correspondence to: firstname.lastname@example.org and email@example.com.