Monitoring results of the Somalia cash and voucher transfer programme: Phase I

This article summarises the findings of a recent monitoring report on an emergency cash-based intervention in South Central Somalia. In what is thought to be the largest cash programme to be implemented by non-governmental organisations (NGOs), 14 NGOs (six international and eight local partners) distributed $50.6 million-worth of cash and commodity vouchers to 136,673 households affected by the famine of 2011. Approximately half the beneficiaries were located in parts of the country controlled by the Transitional Federal Government (TFG), and half were in areas controlled by the Islamist group Al Shabaab (AS). The monitoring exercise was undertaken by the Somalia Cash and Voucher Monitoring Group (CVMG) from September 2011 to March 2012. The CVMG (Phase I) is coordinated by UNICEF and made up of the six INGOs, their local NGO partners, three independent consultants contracted by the Humanitarian Policy Group/Overseas Development Institute and two organisations contracted to provide Independent Field Monitors (IFMs).

Why a cash-based intervention?

Although cash-based interventions are increasingly used in humanitarian response, they still make up only a small fraction of overall humanitarian assistance and are rarely used at scale except by governments responding to natural disasters and chronic poverty. The reasons for implementing cash and commodity vouchers on such a large scale in South Central Somalia were partly related to the challenges involved in distributing food aid, and the opportunities afforded by Somalia’s robust market system, large network of money transfer agents (hawala) and considerable experience of cash-based programming among NGOs in the country. In addition, the World Food Programme (WFP) had been forced to withdraw from South and Central Somalia in 2010 after repeated attacks on its offices and conveys and a subsequent ban imposed by Al Shabaab. With the withdrawal of WFP, deteriorating food security and rising malnutrition rates, cash-based programming was one of the only viable methods for providing assistance.

There were four main risks associated with large-scale cash-based interventions at the time: whether NGOs would be able to gain access to populations in need, without safety risks to their staff, particularly since many of these areas were controlled by Al Shabaab; whether the cash itself would be diverted by Al Shabaab and other authorities and local militias through taxation, intimidation and extortion (agencies were particularly concerned about the risk of diversion following a series of allegations of diversion against WFP-Somalia in 2009 and 2010); whether the market would be able to supply the quantities of food required; and whether the cash and voucher distribution would lead to inflation. In view of the risks involved, a joint CVMG monitoring exercise was established, managed by an independent organisation, the Overseas Development Institute (ODI), and undertaken together with the international and local NGO partners.

The Cash and Voucher Monitoring Group

The Somalia Cash and Voucher Monitoring Group (CVMG) was formed in September 2011. Monitoring activities involved the collection of both qualitative and quantitative data relating to the implementation process, and market and household impacts. Quantitative data were collected by the implementing NGOs, and qualitative data were collected by 18 IFMs hired specifically for the CVMG exercise.

For the cash projects, $43.9 million was distributed to 94,699 households through six hawala companies. The actual amount paid to beneficiaries varied among the NGOs and between regions according to the specific aim of the project and local market prices. The size of the cash transfer ranged from $75 to $125 a month, depending on the actual cost of the Minimum Expenditure Basket (MEB) in the local area. The MEB was developed by the Somalia Food Security and Nutrition Analysis Unit (FSNAU). It represents a household’s minimum daily food and non-food requirements.  For the voucher projects, over 7,000 tonnes and 655,000 litres of commodities were distributed to 41,974 beneficiary households. The commodities exchanged for the voucher were three or six litres of oil, 10kg of sugar, 25kg of wheat flour and 25kg of rice. The value of the commodities received through the vouchers was considerably less than the value of the cash, ranging from $51 to $65 depending on local conditions. The voucher projects aimed to meet 70% of an average family’s nutritional requirements. Vouchers were distributed to the beneficiaries by the NGO, and the beneficiaries then exchanged their vouchers for the specified food items with 45 shopkeepers selected to participate in the project. The shopkeepers were subsequently reimbursed for the value of the food distributed, based on the prices agreed in the shopkeeper contracts.

The biggest implementation challenges were access and security, particularly in Mogadishu and areas controlled by AS. Negotiations with the local authorities (AS, TFG and other local leaders) led to start-up delays in some areas, and two INGO projects were suspended when permission to operate was withdrawn by AS. Security deteriorated in AS areas during the course of the implementation period due to military incursions by the Kenyan and Ethiopian armies into Somalia. In cases where a planned distribution was not possible, beneficiaries were subsequently given two monthly transfers in one instalment, or the interval between transfers was shortened. The overall proportion of planned beneficiary transfers that was actually delivered was 64%. In areas controlled by AS, the proportion was 42%, and in TFG areas it was 85%.

The overall performance of hawala agents and voucher shopkeepers was generally good. The monitoring system identified projects and places where improvements could be made to the implementation process (e.g. the need for shorter waiting times and travelling times to distribution sites; the need to increase the amount of the transfer) and changes were subsequently made. All participating CVMG partners had a feedback mechanism in place to collect and respond to complaints or feedback from the communities with whom they worked. This system was effective in capturing simple operational issues relating to individual beneficiaries (e.g. corrections needed on ID cards), but there was a low level of awareness of the feedback mechanism among the surveyed beneficiaries (55%), and not all complaints were recorded on the feedback forms. Fundamental issues such as inclusion and exclusion errors were not captured through the feedback mechanism, which was not designed to gather feedback from non-beneficiaries and other stakeholders.

There were attempts on the part of local authorities (both TFG and AS) to influence the targeting and registration process, and to tax implementing NGOs and beneficiary households. Most of these attempts were successfully resisted or resolved by the NGOs concerned, but in one case a local NGO implementing partner had its operations suspended by AS because it refused to pay a 30% tax. Survey results indicated that, in total (across all post-distribution monitoring surveys), 2% of sample cash beneficiaries (5% in Mogadishu) reported having paid someone to access their cash, though in some cases these payments were given as gifts to relatives and it is clear that the respondent misunderstood the question. In Mogadishu, many of these payments are thought to have been made to gatekeepers (self-appointed ‘leaders’ of urban IDP camps who provide services to IDPs, such as access to land, security and access to aid). The total value of these payments is estimated to be up to $66,000, or 0.2% of the total cash amount transferred.

Regular market monitoring revealed that key food items were generally available in the markets, and the cash distribution allowed people in most areas to buy more, and more diverse, food. No inflationary effect was found as prices followed their normal seasonal pattern, declining considerably due to the good harvest season. There was, however, an appreciation of the Somali Shilling by 20% over the same period that counteracted some of the decline in price. This fluctuation in the currency rate was due to the massive influx of dollars into the market through relief operations, remittances from overseas, foreign investment and income from overseas livestock sales. Given that the cash and voucher transfers made up less than 3% of the total dollar inflows into Somalia, they therefore did not cause the inflation.

There were significant changes in household food consumption patterns over the course of the project. Not all of these changes are attributable to the intervention; the good harvest and the fall in prices also contributed substantially to the improvement in household food security. Before the projects began, households reported eating slightly more than one meal per day, largely consisting of cereals and oil (an average Household Dietary Diversity Score (HDDS) of 1.7). The World Bank estimates that the Somali diaspora transfers approximately $2 billion annually into Somalia through the hawala system (Somalia: From Resilience Towards Recovery and Development: A Country Economic Memorandum for Somalia, World Bank Report No. 34356-SO, 2005). By comparison, it has been estimated that only $1 billion in international aid is provided to Somalia annually (Associated Press, 6 May 2011).  By the end of the first three months of distributions, adults in the household were consuming two meals a day, and children three. In addition, dietary diversity had increased to at least four food groups, with cash-receiving households consuming a more varied diet (HDDS = 6) than households in receipt of commodity vouchers (HDDS = 4). There was also a rapid decrease in households facing severe food insecurity. At baseline more than 75% of households reported going to bed hungry, going a full 24 hours without food or having no food in the house. After six months of distributions, no household in the rural areas reported these conditions, while less than 10% of urban households were still facing these problems. In addition, household debts decreased substantially, opening up critical credit lines. The ways in which the cash transfer was spent became more diverse over the course of the project, with the proportion spent on food and debt repayment decreasing over time. Food expenditure was gradually replaced with spending on non-food items, including agricultural inputs, livestock, water, education, medicine, business investment and savings. Some of these patterns reflect normal seasonal changes, but they also suggest some level of recovery.

A number of secondary impacts were noted relating to household and community harmony. Community leaders interviewed by the IFMs reported that the cash intervention improved the status and dignity of the beneficiary families, allowing them to meet their priority needs. This feeling was reiterated by some of the case study households, who felt that their status in the community had been improved now that they had a little money or food to help others, rather than needing community support themselves. Jealousy between households was reported to be common, though this ill-feeling does not appear to have resulted in conflict and non-beneficiaries reported sharing meals with beneficiary households. This was confirmed by the household interviews, with 62% of cash beneficiaries and 67% of voucher beneficiaries reporting that they shared food with guests.

Conclusion

Overall, the experience of the CVMG partners was positive and confirms that effective implementation and monitoring of large-scale cash-based programming projects in a complex, conflict-affected environment like Somalia is possible provided that appropriate checks and balances are put in place to ensure transparency and accountability in targeting and cash distribution. Overall, the CVMG projects allowed for a significant level of humanitarian aid to reach those in need. The CVMG monitoring exercise has shown that large-scale, collaborative monitoring can be done in a complex, conflict-affected environment, and provides a basic model for how this can be achieved, including the improvements that are being made under Phase II of the CVMG programme.

Catherine Longley, Sophia Dunn and Mike Brewin are independent consultants.