• The Kakuma refugee camps have become popularly associated with entrepreneurship. From the International Finance Corporation’s Kakuma as a Marketplace study1 to the bustling retail markets and restaurants in Kakuma One, the proposed Huduma-Biashara Business Centre, and the new market based Kalobeyei settlement, the Kakuma camps have become an exemplar of the shift towards private sector-led development in refugee contexts.2
• In 2016, the Kalobeyei settlement was opened 3.5 kilometres away from the Kakuma camps, with the intention of promoting the self-reliance of refugees and the host population, and delivering integrated services to both. Its development is guided by the Kalobeyei Integrated Social and Economic Development Programme (KISEDP), which offers a range of innovative, market-based approaches to refugee protection that diverge from the conventional aid model implemented in Kakuma. There so far have been few studies that examine the emergence of refugee-led markets at the business level, whether in the Kakuma camps, in the Kalobeyei settlement, or elsewhere.
In order to address this gap, our research aimed to study one particular sector: the food market. This sector is of particular interest because it is such a significant part of economic life in refugee camps, and because it is heavily shaped by the modalities of food assistance provided by the international community. Kakuma is currently undergoing a gradual transition from in-kind food assistance to cashbased assistance, and as an interim step, it has introduced a food provision model called Bamba Chakula.
• The Bamba Chakula (‘get your food’ in Swahili) programme is a cash-based intervention designed by World Food Programme (WFP) as an alternative to in-kind food assistance. It represents a transitional arrangement between in-kind and full-cash assistance. It began in 2015. By providing refugees with mobile currency supplied through Safaricom, it allows recipients to choose the food items that suit their preferences, with some restrictions relating to commodities like alcohol and tobacco, while supporting the growth of local markets. The currency is only redeemable from contracted traders, who have been able to apply for a Bamba Chakula (BC) contract during a series of competitive application processes. In Kalobeyei, refugees receive nearly all food assistance through BC, while in Kakuma, about 70% of food assistance is in-kind and the rest is through BC. In this report, we focus particularly on the role that BC has played in shaping the food market.
• Our study is based mainly on a business survey of three groups of food retailers: successful BC applicants, unsuccessful BC applicants, and food retailers who have not applied to be BC traders. The survey targeted all traders in WFP’s registry of applicants to BC and a random sample Executive summary of non-applicant food retailers, sampled from a Norwegian Refugee Council (NRC) census. In total, we interviewed 730 entrepreneurs (of whom 629 currently have a business). We complemented the survey with qualitative data collection based on semi-structured interviews and focus groups.
Our aim was to examine what role BC status, among other factors, has played in influencing business performance and market structure.
• We find that Somalis and Ethiopians are the most likely nationality groups to be engaged in the food retail business, whereas South Sudanese refugees tend to be the least represented. Refugee men are more likely to own shops than refugee women; however, the opposite pattern is found for Kenyans. The group of shop owners who never applied for a BC contract are inherently different from those who applied (whether or not they were successful) in a range of ways, such as the relative composition of nationalities, gender, education level, and previous experience and training. Those who did not apply were sometimes hindered by information, language, or literacy constraints.
• We find that having a BC contract gives a huge advantage.
This is unsurprising given that it gives selected traders exclusive access to the purchasing power provided to refugees through all cash-based assistance. Having a BC contract is correlated with differences in operational competence and better business outcomes. BC shops generally have better outcomes in terms of profits, sales, stock levels, the variety of goods they offer, and the estimated current value of the business and its assets. It is, however, important to highlight that these differences cannot solely be attributed to the BC contract itself, as the groups are inherently different in some aspects such as nationality, gender, family background, education, training and prior (shop) experience. BC traders also invested more start-up capital in their businesses to begin with.
• Kalobeyei-based BC traders do better than BC traders in Kakuma, in terms of profits and sales. This is unsurprising given that there are fewer BC retailers operating in Kalobeyei even though the volume of aid distributed in the form of BC is the same (about 500,000 USD per month per site) in both Kakuma and Kalobeyei.3 Within Kakuma, 19% of household survey respondents admit selling part of their in-kind food aid in order to acquire cash, and this is likely to be under-reported. However, the selling price is relatively low and the additional purchasing power does not appear to create a major opportunity for retailers in Kakuma.
• There is a high concentration of market power. A relatively small number of mainly Kenyan wholesalers supply the many small retailers. 69% of traders source exclusively from the five largest wholesalers. Refugee traders often organise in ‘buying groups’ in order to balance wholesalers’ market dominance. In addition, WFP provides price guidelines to wholesalers and retailers, and retailers regularly meet to agree on common prices. These dynamics are among the factors explaining why prices do not vary much across traders.
• Nationality plays a role in exchange. A preference for doing business with one’s own nationality seems apparent in hiring employees and when customers choose which shop to buy from, but nationality has little importance when retailers decide from which wholesaler to source goods.
Refugee-host interaction is limited between retailers and customers. Refugees source from Kenyans, but Kenyans rarely source from refugees.
• Credit-based purchases are common and credit flows exist between both wholesalers and retailers, and between retailers and consumers. Having access to credit through wholesalers (due to the predictability of demand) is part of what lets BC traders operate and compete differently than others. Many BC shopkeepers retain the refugees’ BC SIM card as collateral, and in return distribute products on credit if BC transfers are delayed or if customers run short of food and money at the end of month. Trust and loyalty play a key role in shaping market-based interactions.
• Alongside the introduction of BC, WFP has introduced the Kenya Retail Engagement Initiative (KREI), which is a holistic approach to support and develop the local economic systems involved in the broader transition from in-kind aid to cash-based assistance. In addition to activities for supply chain development, KREI includes business training. WFP is one of several actors who offer business training to food retailers in Kakuma and Kalobeyei.
• Our data shows that access to business training in general is correlated with improved business performance. It is associated with a 20% higher level of sales and profit, although this association is not necessarily reflecting a causal relationship: it may be driven by businesses with better performances being more likely to opt into business training. Nevertheless, insofar as the relationship is causal, the mechanism through which it operates appears to be by creating better business practices. The practices that matter include: giving special offers and bulk discounts (correlated with higher sales), asking customers whether there are products they would like (correlated with higher profits), asking suppliers for preferential terms (correlated with higher sales and higher stock variety), and book-keeping (correlated with higher stock variety).
• Although refugees have been more likely to purchase items from other refugees, initiatives by WFP to provide additional opportunities to Turkana traders have helped to increase the number of refugee frequenting Turkanaowned shops. In particular, WFP paying Turkana shops to distribute corn soya blend (CSB) has helped increase the number of refugees going to Turkana shops.
• Our findings suggest that introducing full cash-based assistance would help to ‘level the playing field’ by removing some of the advantages afforded to BC traders. This might lead to possible tensions, particularly among some Turkana traders who would stand to lose the most. However, this would partly be mitigated by the advantages that BC shops have already enjoyed, which have led to greater investment and stock accumulation. A range of steps are needed to carefully manage the transition to cash assistance, both in Kakuma and elsewhere.
• Overall, understanding the BC experience in Kakuma offers insights of wider relevance into how markets emerge and develop in refugee camps and settlements, the process of transition from in-kind to cash-based assistance, and what determines entrepreneurial success in a refugee camp. It also poses challenging questions about the appropriate role for humanitarian organisations in regulating markets. We offer a series of practical recommendations relating to the transition to cash assistance, the future role of BC as an interim measure, and the role of the Retail Engagement Initiative.