Delivering cash transfers in Zimbabwe’s liquidity crisis: persisting with a market-based approach

As part of the response to severe drought in Zimbabwe in 2015–17, the UK Department for International Development (DFID) provided funding to CARE (partnering with World Vision and two mobile network operators) to carry out a humanitarian cash transfer project to meet basic food needs for 400,000 people across 15 districts. Alongside the cash programme, the largest-ever in Zimbabwe, DFID designed and commissioned a market support mechanism to pay for commercial imports of maize. The response was adapted and expanded to cope with decreased liquidity, the risk of food scarcity and a deepening food crisis.

The programme broke new ground in the shared willingness of DFID and CARE to learn by doing, and to invest in a monitoring and evaluation system that would generate evidence for more appropriate programming. Humanitarian partners had been implementing commodity-based drought response in Zimbabwe for years. Only small-scale, direct cash pilots had so far been delivered, but the limitations of in-kind assistance were clear. First, there was a strong risk of political interference and difficulties in ensuring the impartiality of aid, with instances of partisan government distributions of food documented by the Zimbabwe Human Rights Commission. DFID’s view was that repeated food distributions had also led markets to function less well: markets were poorly supplied during the distribution period, but better supplied at other times. The move towards a market-based response to drought was also in line with CARE’s concern to maximise effectiveness, recognising the multiplier effects of cash and adhering to the ‘do no harm’ principle. An external evaluation identified the programme as a ‘game-changer in a country where food aid has dominated previous relief responses because it solidifies cash transfers as a viable, large-scale alternative to delivering food and offers a potential model for future cash responses’. Sarah Bailey et al., Zimbabwe ‘Cash First’ Humanitarian Response 2015–2017 Evaluation Report, Oxford Policy Management, 2017,  http://www.opml.co.uk/sites/default/files/Zimbabwe-Cash-First-Humanitarian-Response-2015%E2%80%932017.pdf.

Three apparent limitations emerged during implementation, calling into question whether an enabling environment for cash-based programming still existed:

• Capacity to deliver mobile money. CARE undertook mapping and due diligence of mobile network operators. Their capacity to take on a large-scale emergency programme with a high volume of transactions and strict deadlines, to partner with an international NGO and to ensure strong network coverage was uncertain. CARE emphasised the need to build and maintain strong relationships with network operators, to ensure we understood each other’s ways of working and how we needed to adapt to work together effectively.
• Liquidity. Zimbabwe has been in a tenuous economic situation since 2000. Hyper-inflation peaked in 2008 which led to the Zimbabwean dollar being taken out of circulation in 2009. The US dollar is the main currency in use, and the country cannot use monetary policy to address rising debt and the depletion of its foreign exchange reserves. In May 2016, the Reserve Bank began limiting cash withdrawals to $50 a day and restricting imports to stem the flow of cash out of the country. Agents began struggling to provide cash requested by programme beneficiaries. In November 2016, the long-anticipated ‘Bond Notes’ (a form of legal tender, pegged to the US dollar, designed to ease the liquidity shortage) were introduced. People were initially reluctant to accept them and feared they would become devalued, as they were not recognised outside Zimbabwe. We developed contingency options to physically import and distribute cash or food. Anxiety and rumours were compounded by the recent experience of hyper-inflation.
• Food supplies on local markets. Zimbabwe’s economic difficulties were reaching a crunch moment in a context of regional drought. Zambia was the only neighbouring country with a grain surplus, and commercial traders had made arrangements to import Zambian maize, but no export permits were being granted. Overseas shipments would be needed to meet Zimbabwe’s grain deficit. Restrictions on accessing foreign exchange to pay for imports meant that food scarcity and price hikes were possible, particularly in remote rural markets. FEWSNET’s early warning system was forecasting significant price increases at the peak of the hunger period.

Due to the nature of cash transfer programming, time and energy that would otherwise be absorbed in the logistics of delivering food aid could be devoted to ensuring that a robust and adaptive monitoring and feedback system was in place, including post-distribution monitoring at the household level and market monitoring of commodity prices and availability. We revised data collection tools to gather and analyse new information as the context shifted, and increased accountability measures such as extra household verifications. Over time, we collected additional data such as merchant and mobile agent mapping, cash-out rates, coping strategies under a worsening liquidity crisis and whether conditional cash-outs were being offered.

Equipped with this data, the programme could adapt to the changing environment. Changes brought in during implementation, based on updated response analysis, included revising the transfer value for very small households; raising the transfer value; extending the programme in light of a second consecutive year of crop failure; including a seasonal livelihoods transfer (‘multipurpose grant’) ahead of the cropping season; and supporting households severely affected by flooding. Specific adaptation measures, which were key to our ability to maintain a market-based response, are outlined below.

1. Shifting our messaging from ‘cash-outs’ to ‘e-payments’

• Despite challenges accessing cash, very few beneficiaries were struggling to obtain essential items. We found that, increasingly, businesses and individuals were accepting mobile money as payment for goods and services. As the liquidity crisis deepened, several government-controlled grain suppliers followed suit.
• We had been focusing too much on people’s ability to cash out as an indicator of the programme’s relevance. The fundamental question was whether people were able to access critical basic supplies, such as staple food items, without surcharges. We encouraged direct e-payments, at least for food items, and by March 2017 70% of beneficiaries were making e-purchases, compared to 17% at the start of the programme.
• Encouraging e-payments led to another set of risks linked to the acceptability of e-payments and the potential for unethical practices such as charging extra fees, conditional cash-outs or differential pricing. We again adapted data collection tools to determine whether shopkeepers and agents were accepting e-payments, and accepting them on a par with cash (in the majority of cases, they were).

2. A market support intervention to support commercial food imports

• To reduce the risk of a breakdown in food supply on rural markets, DFID developed an innovative macro-level intervention to support the private sector. The humanitarian grain market facility, managed by Crown Agents, leveraged $21 million of development and humanitarian funds to pay for maize imports ahead of the hunger gap. Commercial importers were able to apply for Crown Agents to pay for specific grain contracts on their behalf, using DFID funds. Importers had to deposit equivalent funds into a Zimbabwe account managed by Crown Agents, and agree to conditions on the pricing and distribution of the food in drought-affected districts. Some of DFID’s largest partners in country also agreed to receive project funds in country from Crown Agents.
• DFID acted on the specific bottleneck affecting the commercial sector (access to foreign exchange to pay for grain contracts) and matched it with a specific opportunity (partners were due to receive large volumes of DFID funds at this period of heightened need in Zimbabwe). The facility aimed to prevent the collapse of the private sector and maintain supply of staple food in areas where cash transfers were being delivered. Markets continued to be supplied, and the significant price hikes projected by FEWSNET did not materialise. The fund was managed at relatively minor cost (1% of funds leveraged) with no grant or subsidy to the private sector.

3. Open partnership, enhanced coordination and learning

• The willingness of DFID and CARE to work together in a genuine partnership was a key success factor from the design stage of the programme and through its many iterations of implementation. DFID and CARE had a shared focus on managing emerging risks; potential risks were closely analysed and proactively mitigated through regular meetings and report sharing. Flexibility on the part of the donor allowed for space to reflect, innovate and adapt when and where needed.
• CARE co-founded and co-led the Zimbabwe Cash Working Group with the World Food Programme (WFP) in order to coordinate the growing number of agencies (NGOs, government, UN) carrying out cash transfer programming in response to the drought. This became a crucial forum for sharing information and working through operational challenges linked to the cash crisis. A cash transfer value was harmonised across implementing agencies, and efforts were made to consolidate data collection based on what was being produced under the DFID/CARE programme, in order to provide a comprehensive picture of the market realities across the country.
• Working closely with network operators proved critical to the success of the programme. Some peer organisations were unable to launch or sustain similar mobile cash programmes, citing their inability to work effectively with operators.
• Two in-depth evaluations were conducted, a mid-year and a final evaluation, giving assurance that key lessons and good practice would be identified, generating evidence to underpin future programmes and confirming outcomes.

The final evaluation of the programme emphasised the importance of market monitoring and a comprehensive accountability and feedback system. It also underlined the potential for increasing or adapting transfers in periods of higher financial need – ‘once a cash transfer program is in place it is easy to provide an additional transfer’ – and confirmed that ‘mobile money is viable even when recipients are unable to fully cash out, if they can access goods/services via mobile money’. It identified a number of ways in which future responses could build on this experience and go further, for instance in delivering cash through a single pipeline, varying transfer values and further developing coordination, assessment and monitoring systems.

In terms of results, 87.5% of people said the project met their basic needs; hunger was reduced and there was an increase in dietary diversity. Overall there was a 21.7% reduction in negative coping strategies. Markets continued to function, and the final evaluation found evidence of increased market activity in programme areas. Finally, the programme provided increased exposure to and education on the use of mobile money in what has become one of the fastest-growing regions for electronic payments.

In the wake of the Grand Bargain by international donors, UN agencies and NGOs to improve the efficiency and effectiveness of humanitarian action, DFID and CARE felt conditions were right to persevere and innovate, in order to make best use of local capacities and deliver a more cost-effective, efficient and empowering programme.

Cristy McLennan was CARE’s Assistant Country Director in Zimbabwe, and led on this response. Mira Gratier was DFID’s Humanitarian Adviser in Zimbabwe during the drought response.