Impact of milling vouchers on household food security in South Sudan

By Ruco Van Der Merwe

Ruco Van der Merwe is currently based in Nairobi, Kenya as a Food Assistance Adviser with Samaritan’s Purse International Relief (SP). He has been on numerous assignments with SP’s Disaster Assistance Response Team and is also a participant in the Global Food Security Cluster.

The author would like to acknowledge the contributions from Samaritan’s Purse International Relief and the country office in South Sudan. WFP South Sudan were especially helpful in providing information and reviewing the final product.

Location: South Sudan

What we know: Exchange of general food distribution commodities for non-food items or services is common in food assistance and may contribute to food insecurity.

What this article adds: A milling voucher scheme in a food assistance dependent refugee population reduced ration exchange for milling services and increased how long household rations lasted. Local markets and milling services successfully adjusted to the increase in programme demands. An unanticipated consequence was a fall in miller profits due to over-supply of millers. A routine monitoring tool that evaluates programme impact at the miller level and responds to this is recommended.

Yida refugee camp is located in Pariang County, Unity State, South Sudan. The population—primarily comprised of agro-pastoral Nubans—started fleeing to the location in 2011 when fighting resumed in the neighbouring South Kordofan State between the Sudanese Armed Forces (SAF) and the Sudanese People’s Liberation Movement-North (SPLM-N). The initial year and a half of the camp’s existence saw the population rapidly rise to over 65,000; these have largely stabilised since March 2013. Thousands of people arrived to Yida camp with few household assets and in poor general health. Samaritan’s Purse (SP) was the first agency on the ground and rapidly mobilised an emergency response. As the camp population grew, SP partnered with WFP and UNHCR to provide emergency food assistance and nutritional support to the refugees, as well as expanding into water, sanitation and hygiene (WASH) and protection activities.Women in Yida camp using a traditional milling method for sorghum

Mid-2012 saw fluctuations in global acute malnutrition (GAM) rates, which led to several assessments seeking to establish the root causes of continued food insecurity, despite the provision of 100% general food distribution (GFD) rations (SPHERE standard), 100% food assistance coverage of the population in need, and a full array of nutritional services¹. Investigation found that households were significantly reducing their total caloric intake due to costs associated with milling - commercial mills were charging up to 7 SSP (South Sudanese Pound) to mill a single malwa (USAID vegetable oil tin equal to about 3.5kg of cereal) or in-kind trade ratios between ¼ to one and 1:1². Follow-up studies³ revealed that in July 2012, 39% of households noted grinding costs as a food-security constraint and by December 2012, 63.6% of mothers with children under five years old in the household were paying (cash or in-kind) to have their GFD cereal ration milled⁴. Although grinding costs were never confirmed to be the primary reason affecting nutrition and food-security status of households, it certainly was a contributing factor.

In response to these concerns, WFP and UNCHR launched the interagency milling assessment in January 2013 to “determine the cost and availability of milling for beneficiaries in Yida.”⁵ The assessment went further to analyse the milling capacity within the camp to ensure that any potential programmes would not run into supply side (milling-service provision) challenges or adversely affect the market. In May 2013, SP, in partnership with WFP and UNHCR, rolled out a milling-voucher programme targeting the entire GFD caseload (70,292 as of October 2013)⁶ of Yida camp with milling services for 70% of their ration. The primary objective of the programme was to ensure that beneficiaries maintained full access to their Sphere-size ration entitlement and to protect infringement of costs associated with the milling of the rations.

Study objectives

Once the programme had been operational for six months, SP undertook a further study with the following objectives:

• Contribute to the larger body of case studies focusing on market-integrated relief. The project itself had the following unique elements, which would hopefully add additional insights: voucher distribution integrated into GFD, blanket voucher targeting total refugee population, and utilisation of a rural refugee marketplace.
• Evaluate the impact that milling vouchers have had on household food security for refugees living in Yida Refugee Camp, South Sudan.
• Evaluate the impact the milling-voucher programme has had on millers.

As milling vouchers were only introduced in May 2013 after the food assistance and nutritional programmes had been in implementation for almost two years, this allowed SP to analyse the pre- and post-voucher household food security using routine data sources throughout the time period.

Methodology

Two weeks were spent in South Sudan meeting with relevant actors, field staff, and beneficiaries. Secondary data covering 12 months were analysed, which allowed for comparison of indicators before and after the voucher programme commenced. The case study used both primary and secondary data from the following six sources:

1. A Rapid Miller Survey: Fifteen out of the 59 certified millers were interviewed. Yida camp is currently divided into 20 blocks; 1 miller was selected from each of 15 blocks.
2. Market Price Monitoring: SP routinely tracks key commodity and service prices in Yida market. For this specific study, diesel and petrol prices were analysed.
3. Post-Distribution Monitoring (PDM) data: As part of the GFD, SP routinely carries out PDMs two weeks following the food distributions. Data were collected and analysed from January to September 2013.
4. Nutrition data: This study examined routine admission data from an SP supported targeted supplementary feeding programme (TSFP) in Yida camp.
5. Food-security and nutrition assessments: Analysis of data that had been collected since the establishment of Yida camp (ACF’s Integrated Nutrition and Retrospective Mortality Survey; Samaritan’s Purse and Liverpool Associates in Tropical Health’s Yida Joint Household Assessment: LQAS Survey of Key Household Indicators Related to Current Malnutrition Trends; and Yida Interagency Milling Assessment).
6. Focus Group Discussions (FGD): Five FGDs were carried out on October 17th in Yida camp with groups disaggregated as follows: three groups of women, one group of youth, and one group of community leaders. Each group had 8-10 representatives.

Limitations

The data presented in this article should be reviewed keeping in mind several limitations and factors influencing the results.

• Seasonality plays a significant role in Yida camp where the rainy season has traditionally impacted access and mobility. Population mobility is generally limited during the rainy-season months (June–September) meaning that most new arrivals show up during the dry season (October–May). Road access has recently improved following improvements to the Yida-Bentiu route, which has resulted in better availability of goods on the local market and stabilisation of prices. Household-level food production is not currently encouraged in Yida due to its proximity to the border and security concerns.
• Data presented from the PDM and rapid miller assessment is not statistically representative of the entire population and is only intended to give general insights.
• Nutrition is influenced by a host of complex factors so that the findings in no way assume direct causality between the voucher programme and subsequent nutritional outcomes.

Implementation

Targeting

A total of 59 millers were selected to be certified service providers for the camp. Beneficiaries were entitled to redeem their vouchers at any one of the participating certified-milling locations.

Voucher Value

Vouchers were restricted to milling services and entitled each person to mill 3 malwas of cereal (approximately 10.5 kg or 70% of a beneficiary’s ration). The average price of milling 1 malwa is approximately 3 SSP, meaning that the entire voucher value is equivalent to 9 SSP (3 USD) per person per month (Figure 1). Vouchers were distributed according to household size.

Figure 1: Voucher value

Partners

The implementation of the project was divided between five actors. WFP and UNHCR were jointly responsible for printing and providing the vouchers to SP and for monitoring market prices and analysing PDM data. SP was responsible for distributing the vouchers to the beneficiaries and carrying out PDM. Certified millers were responsible for receiving vouchers in exchange for milling services. Equity bank was responsible for paying all the millers in accordance to the number of vouchers they presented. Payments should occur on a bimonthly basis to ensure that millers have enough cash flow for continuous milling operation. In November 2013, SP took over miller payments from Equity Bank.

Process⁷

1. Receiving Vouchers:
   Every month, SP, in partnership with the local WFP office, submitted a VRN (voucher-release note) to WFP Juba. The VRN is based on the anticipated beneficiary caseload for the upcoming month, as determined by UNHCR’s registration figures. WFP was responsible for delivering the vouchers to the field location before the distribution cycle begins. Upon delivery, vouchers are counted, verified, and signed over to SP.
2. Distributing Vouchers:
   Voucher distribution was integrated into the existing GFD system to improve cost-efficiency and accountability. The distribution process in Yida used UNCHR’s innovative rapid food-distribution system, which integrates bar-code scanners during the distribution process. In order to receive their ration, beneficiaries had to present their ration card, which had a bar-coded strip on it. This strip was scanned upon entrance into the GFD area to pull up a photo of the individual for visual verification. Once a beneficiary was visually verified, the bar-coded vouchers were scanned and the vouchers were linked to that specific household. If it was a household size of four, then the software required four vouchers to be scanned (one per person); if household size was seven, then seven vouchers; etc. At the end of the GFD, distribution records were compared with voucher-inventory records for accountability.
3. Payment:
   Once all of the distributions were completed, a final voucher distribution figure was presented to WFP, along with empty voucher books and any remaining vouchers. Equity bank was then notified of the final amount to be transferred and coordinated with WFP and the millers to arrange for an appropriate time to carry out the cash transfer.
4. Reporting and Monitoring:
   Project output with regard to vouchers received, distributed and returned, was monitored in the same way that food-commodity stocks were tracked and inventoried. Voucher impact at the household level was also monitored. This was accomplished via monthly post-voucher-distribution monitoring. Monitoring typically occurred two weeks after distribution.

Market conditions

A woman with her vouchers following distribution

The area surrounding Yida camp is extremely prone to flooding during the rainy season, which led to significant access constraints. The 2011-2012 rainy seasons saw camp access reduced almost exclusively to air support. Significant road work was done, which has since allowed for semi-reliable access to the nearest regional town and market of Bentiu, approximately three hours away by road in the rainy season. The camp itself has one primary market and three secondary markets that function on a daily basis. There are generally a wide variety of goods including primary commodities (cereals, pulses, and oil), livestock (cows, goats, and sheep), and fuel (diesel and petrol), although seasonal fluctuations do exist. The markets primarily serve the refugee population, although there is a very small host community that also access them. Most goods that reach Yida are sourced from Bentiu. There are no banks or money-transfer services available in the camp. Cell phone network provision by Zain has significantly improved, although no mobile money-transfer system exists in South Sudan. The map in Figure 2 gives a general idea of community assets, including grinding mill locations, in the camp.

Figure 2: Community assets, including mill locations, in Yida camp

Impact

Millers

Business owner next to his milling machine

One of the primary concerns facing the voucher programme was whether there would be market impact as a result of accessibility during the rainy season. In order for the project to function effectively, millers would require access to key inputs, such as diesel, petrol, machine parts, and maintenance services. If markets were not well integrated, or were inaccessible due to the rain, market disruptions would be likely. Millers were asked based on previous experiences, what were the major factors contributing to price fluctuations for their milling services. Both direct inputs of fuel and maintenance were stated to be the primary contributing factors; only 27% of millers felt that overall prices were indirectly affected by access (see Figure 3).

Using routine data collected from market price monitoring at Yida’s primary market, the survey was able to evaluate the impact that the voucher programme had on petrol and diesel prices. Figure 4 displays the commodity-price trends over time, where the red dots indicate the month in which the voucher programme commenced. There was inflation for both commodities (increase of 66% for petrol and 35% for diesel) between the months of May and July. This was likely due to initial increase in demand related to the programme beneficiaries, but also additional demand in relation to prepositioning for the rainy season. After July, markets stabilised and adjusted to the new levels of demand.

An interesting finding from the study was that major improvements in access to Yida (compared to previous years) resulted in a stable market environment. Most millers noted that they could source fuel directly from Yida market (93%) without having to seek out regional markets (7%). Also, 77% of all millers could access maintenance for their equipment from within the camp, either by servicing the equipment themselves (12%) or finding someone locally to do it (65%).

Millers further noted that access was not limited to the dry season only; 100% reported being able to access fuel at least occasionally during the rainy season. Most of the millers (93%) prepositioned fuel stocks in anticipation of the rainy season. The improved accessibility addressed a lot of the initial concerns regarding an increase in milling service demand within an insulated market.

The initial assumption was that the increase in demand for milling service would have an overwhelmingly positive impact on the millers. In reality, the miller survey showed that 60% of millers believed that business actually got worse. Millers reported an average decrease in profits of 41%, or 620 SSP (see Figure 5), and an average reduction of two hours of operation per day (from 10 hours to 8 hours daily). There is the potential for some recall bias, as the question regarding average weekly profits was conducted after the programme was already operational. The assessment itself did not examine the reasons behind the decrease in business, however discussions with project staff suggested that this was primarily due to an increase of milling-service providers during the project start up, as well as pre-voucher price-gouging⁸ that inflated profits. This would explain the decrease in profits as business was spread out across an oversupply of milling-service providers and prices for milling services became fixed.

All of the millers surveyed felt that the current payment system was not functioning well and asserted that this was due to delayed payment. Several millers reported running out of prepositioned fuel stocks resulting in a temporary closure of business until they were paid and able to restock. Additionally, millers noted a lack of adequate information being communicated by the partners in regard to schedule of payment or reasons for delays. It appears that there was little differentiation between the roles of SP, WFP, UNHCR, and Equity Bank; the perception of poorly managed payment schedule negatively affected all partners. All of the millers stated that they preferred the current system of vouchers for cash as the transfer modality, over banks or alternative systems. Overall, the millers were unwilling to consider alternative cash-transfer modalities, although this was probably influenced by the unreliability of the existing transfer process.

Households

Data for the evaluation of voucher impact on household food security primarily came from FGDs, PDMs, and nutritional data. All FGDs noted that the voucher programme resulted in their food rations lasting longer. Groups noted that rations are lasting between 4-7 days longer every month, since they no longer had to sell or exchange rations for milling services. One FGD noted it has improved feeding for young children, since they can now easily serve them porridge. The PDM data corroborates information from the FGDs (Figure 6). Cereals were reported to last on average four days longer (24 rather than 20 days) following the start of the voucher programme.

Prior to the voucher programme, households were exchanging approximately 27% (average across FGDs) or 8.1 days of rations for milling services. Since the voucher programme started, that figure had dropped to 2% (average across FDGs). One FGD noted that 10% of beneficiaries sell or exchange their vouchers when they desire access to alternative goods or foods (such as meat and onions). The “youth” FGD added that vouchers could be exchanged for food with millers at a rate of one malwa of ground sorghum for one voucher.

Nutritional data from the SP TSFP programme was collected and analysed to see whether any patterns arose in connection with the start of the voucher programme. Average TSFP new admissions (Sept 2012-Oct 2013) declined by 165 per month following the start of the voucher programme although attribution cannot be assumed (see Figure 7).

Recommendations

Pre-voucher market assessments should carefully consider whether supply meets programming demands, but also ensure that the project does not encourage the oversupply of services. The Yida case highlighted that a rapid and unchecked increase on supply side can actually have a net-negative impact on profits for service providers.

Millers voiced the need for routine information on the status of payments and project updates. This would be an easy component to integrate into all future voucher programmes and will likely result in higher confidence in the partnership.

Apply ‘do-no-harm’ principles to all actors involved in the project cycle. Ensure payments are done on time, and include a penalty in contractual agreement for delayed payments. The Yida case exposed the fact that some millers were suffering business losses due to delayed payments and an inability to find capital to restock milling inputs. NGOs and business partners should be held accountable to ensure that their performance does not negatively impact the markets and individuals they are intending to assist.

It would be beneficial to integrate a routine-monitoring tool that evaluates programme impact at the miller level and allows for process adjustments.

Conclusion

The findings from this assessment indicate that the voucher programme has improved household food access. Households are significantly less likely to exchange their ration for milling services and rations now stretch an extra 4-7 days per month. Despite initial concerns regarding market integration and access due to Yida’s remote location, the majority of millers indicated routine access to goods on local and regional markets. The markets have been able to adjust to the new levels of demands for programme inputs (fuel and maintenance services). Future milling programmes should carefully consider the millers as a key beneficiary group and ensure monitoring tools are in place to routinely evaluate project impact on their businesses.

For more information, contact: Ruco Van Der Merwe, email: RVanDerMerwe@samaritan.org

¹Samaritan’s Purse and Liverpool Associates in Tropical Health, Yida Joint Household Assessment: LQAS Survey of Key Household Indicators Related to Current Malnutrition Trends, 11.

²WFP, Mission Report: Escalating Malnutrition-Related Mortality at Yida Camp, Unity State; WFP, Meeting Minutes: Discussion on Maize.

³ACF, Integrated Nutrition and Retrospective Mortality Survey, 22.

⁴See footnote 1.

⁵WFP, Yida Interagency Milling Assessment.

⁶UNHCR. (2013, October). Regional Overview. UNHCR.org. Retrieved October 2013, from http://data.unhcr.org/SouthSudan

⁷Riddell, Voucher Distribution Manual, Yida, South Sudan

⁸This term describes where a vendor prices commodities at a level much higher than is considered reasonable or fair. This often happens after a demand or supply shock and tends to be relatively short term and affect essentials.