Enabling Vulnerable Children to Learn. The Namibian School Feeding Programme: Cost Analysis September Republic of Namibia Ministry of Education

Transcription

1 Enabling Vulnerable Children to Learn The Namibian School Feeding Programme: Cost Analysis September Republic of Namibia Ministry of Education

2 Republic of Namibia Namibian School Feeding Programme Cost Analysis was conducted with technical assistance from World Food Programme Cover image by Justin Ellis: A primary school learner preparing to have his midmorning meal at one of the schools benefiting from the school feeding programme Enquiries: Directorate of Programmes and Quality Assurance Private Bag Windhoek, NAMIBIA Tel: ( ) Fax: ( )

3 Table of Contents Abbreviations... 4 Executive Summary Introduction Methodology Central Level Regional Level School Level Financial Implementation of NSFP Current Expenditures for NSFP Assessment of NSFP Costs Comparison with Other Countries Distributing CSB Rations Future and Better Recommendations Conclusion Annex I. Procurement Costs by Region and Term Annex II. Ration Differentiation Options

4 Abbreviations CSB MAWF MOA MOE MOF MOL MOU NAB NSFP OVC PQA WFP Corn Soy Blend Ministry of Agriculture, Water and Forestry Ministry of Agriculture Ministry of Education Ministry of Finance Ministry of Labour Memorandum of Understanding Namibian Agronomic Board Namibian School Feeding Programme Orphans and Vulnerable Children Programmes and Quality Assurance World Food Programme 4

5 Executive Summary The Namibia School Feeding Programme (NSFP) is a nationally-run and funded government programme providing meals to approximately school children across the country. The meal consists of 125 grams (475 kilocalories or 1,988 kilojoule) of fortified maize blend that is cooked at schools and served as a midmorning meal. The Ministry of Education (MOE) at head office is charged with the implementation of the NSFP. NSFP is a highly centralized programme. Through the national tendering process, funds are disbursed to three tenders. School feeding operations occur through funding flows at three levels: central, regional, and school-level. This analysis assesses the actual costs for school feeding based on 2011 expenditures. 1 The cost of school feeding borne by MOE at all levels is estimated at NAD 242 per child per year. The total cost of school feeding including community level contributions at the school level, is NAD 282 per child per year based on six school visits. Expenditures are presented for each level of implementation (central, regional, and school) and are broken down by cost category: commodities, logistics and storage, management and administration, staff costs, and capital costs. The program is very centralized: 79% of costs occur at central level. In addition, this analysis assesses the projected budget of implementing school feeding incorporating the minimal recommendations outlined in the operational review. The operational review provides an in-depth assessment of the strengths and challenges of the NSFP. It makes recommendations for improving program quality as the program considers future expansion. The minimal recommendations include supplying schools with cooking equipment and primary school learners with eating utensils, compensating cooks in cash, and increasing staffing. The total projected budget given the addition of these minimal recommendation ranges from approximately NAD 83.5 million to NAD 85.2 million. However, many of these costs may in fact yield cost-savings, as the provision of fuel efficient stoves, for example, may save on cooking-time and fuel costs. Finally, this analysis sets forth recommendations for further analysis into ways to contain costs. This objective should be considered in parallel with other objectives, such as supporting national production, ensuring food quality, and diversifying the food basket. In addition, it sets forth a recommendation for conducting a study of school level cooking arrangements. Such a study can assist the MOE with ascertaining where and how to focus its financial resources to the greatest benefit. 1 The calendar year and school year in Namibia coincide. 5

6 1. Introduction The Namibia School Feeding Programme (NSFP) is a nationally-run and funded government programme providing meals to approximately 270,000 primary school learners across the country. The World Food Programme (WFP) initiated school feeding in Namibia in Since 1997, school feeding has been completely taken over by the Government. Currently, the school feeding programme reaches approximately 270,000 beneficiaries at primary-level in more than 1,200 state and private schools. The programme provides a mid-morning meal to learners at participating schools. The meal consists of 125 grams (475 kilocalories or kilojoule) of fortified maize blend that is cooked at schools. WFP and the Ministry of Education (MOE) signed a two-year Memorandum of Understanding (MOU) in May, 2012 for WFP to provide technical assistance on the school feeding programme. As part of this MOU, the Government of Namibia requested that WFP conduct an operational review of the Namibian School Feeding Programme and a historical analysis of the transition to national ownership. This paper presents an analysis of the cost of the school feeding programme in Namibia that complements the findings of the school feeding review. This cost analysis provides the MOE with information it can use in its management decisions in relation to the planned expansion of the Namibian School Feeding Programme. Specifically, it provides information on the following aspects of the cost of school feeding: 1. An estimation of the actual costs for school feeding based on 2011 expenditures. 2 The cost of school feeding borne by MOE at all levels is estimated at NAD 242 per child per year. The total cost of school feeding including community level contributions at the school level, is NAD 282 per child per year. Expenditures are presented for each level of implementation (central, regional, and school) and are broken down by cost category. 2. An evaluation of the efficiency of Namibian school feeding as compared to other countries providing similar rations. 3. An estimation of how much it would cost to better implement NSFP in the future by incorporating the minimal recommendations identified in the operational review. 3 2 The calendar year and school year in Namibia coincide. 3 Justin Ellis, The Namibian School Feeding Programme: A Case Study, forthcoming

7 2. Methodology Costs were estimated at each level of NSFP implementation. Costs are based on 2011 expenditures, and divided into five cost categories: commodities, logistics and storage, management and administration, staff costs, and capital costs. Costs are also aggregated by cost category to show total expenditures on each category for all levels of implementation. For items that are one time expenses (i.e. are not recurrent but are paid once and last for a certain number of years), the cost of the item was divided by its lifespan to determine the annual expenditure. In cases when the cost of the item was unknown (such as the cost of the office building construction), the rental rate was used as a proxy. In instances when individuals did not know the cost of a given item (such as the costs incurred for electricity), the cost incurred by another comparable region or school was used as an estimate. As many of the cost categories include items that are not wholly used for NSFP, the share of the cost that is attributable to NSFP was extracted by multiplying the total cost by the NSFP share. For example, although there are four posts for NSFP at the central level, only two staff members work full-time on NSFP. The share of the time the other two staff members devote to NSFP was multiplied by their salary plus benefits, to identify the central-level costs for NSFP staff. Similarly, the share of offices used by NSFP staff, of total offices in the building, was multiplied by MOE s utilities bill to extract the NSFP utilities costs. 2.1 Central Level Costs at the central level were assessed through interviews with human resources personnel, staff implementing NSFP, and other staff that are responsible for relevant payments (e.g. utilities, vehicles, etc.). Data on the cost of the tenders were provided by MOE staff although the original source is the suppliers. At the central level, the following costs were assessed: Commodity costs costs for the tender to provide protein blend, maize meal, sugar and salt. Logistics and storage costs for transportation and storage to regional warehouses, and from warehouses to schools. Management and administration costs for training events and utilities (water, electricity, phones). Staff costs costs for salary plus benefits of all individuals implementing school feeding. 7

8 Capital costs costs for office equipment (computers, cell phones, laptops, and printers), vehicle lease, and office building construction (or rental rate). 2.2 Regional Level Costs at the regional level are partially based on estimates. Regional level staff costs were provided by the human resources personnel at central level. However data on the regional cost of capital, management and administration for NSFP are not maintained at the central level only data for total MOE budget for the regions are maintained. Therefore, capital, management and administration costs were identified through phone and in-person interviews with hostel officers and / or their clerical assistants. Four regions were able to provide information on the cost of NSFP in their regions: Khomas, Omaheke, Kavango, and Hardap. Regional level costs presented in this analysis are thus based on the experience of implementing NSFP in these four regions. As such, regional level costs should be considered an estimate of the cost of school feeding at the regional level. At the regional level, the following costs were assessed: Management and Administration costs for utilities and school visits (mileage plus fuel). Capital costs for office equipment, vehicle lease, and office building construction (or rental rate). Staff - costs for salary plus benefits of all individuals implementing school feeding. Estimated costs are calculated using weighted averages based on the number of beneficiaries in each region surveyed. 2.3 School Level To determine the expenditures on NSFP at the school level, six school visits were conducted in the Khomas, Omaheke, Otjozondjupa, and Hardap regions. Due to the small number of schools visited, cost estimates presented in this analysis are not representative of school feeding costs at the school level. One school, Five Rand Primary School in the Otjozondjupa region, was dropped from the analysis. The school is not considered to be representative of other school feeding schools in Namibia because it delivers the meals as take-home rations to learners whereas the meals generally are cooked at schools. Costs at school level include both costs incurred by MOE (specifically utilities but sometimes other cost items) and costs borne by the community. Costs at the school 8

9 level are distinguished by whether they are incurred by MOE or the community. School level costs include: Utilities costs for electricity (used by stoves) and water (used for cooking and cleaning). Cooking equipment costs for stoves, pots, ladles, fuel / wood for cooking, etc. Shelters costs for constructing cooking and storage shelters, if such costs were incurred. Cleaning supplies scourers, soap, wash buckets, etc. Transport costs of fuel etc. for transport of any necessary items (such as cleaning supplies) to schools. Note that the cost of transporting the blend is captured in central level costs. Cooks remuneration in most cases cooks are compensated with maize blend bags, and thus these costs are not included in the school level costs as the commodity costs are already captured in the central level costs. However, cooks remuneration is accounted for in cases where cooks are paid. Utensils costs for plates and spoons for learners to eat the meal. Food commodities - food items (e.g. oil that is added to the blend) Estimated costs are calculated using weighted averages based on the number of beneficiaries at each school surveyed. 3. Financial Implementation of NSFP School feeding operations occur through funding flows at three levels: central, regional, and school-level. The Ministry of Education at head office is charged with the implementation of the NSFP. Currently the NSFP is a sub-division, under the Division for Management Planning, Appraisal and Training, in the Directorate of Programmes and Quality Assurance (PQA), in the Department of Schools and Formal Education. The Ministry of Finance funds NSFP by distributing funds to the Ministry of Education. PQA authorizes the spending of funds for NSFP through a national tendering process. Although the PQA has authority to spend money, the funds are actually centrally controlled by the Ministry of Education. NSFP is a highly centralized program. Through the national tendering process, funds are disbursed to three tenders to (1) provide protein blend, sugar and salt, (2) mix the protein blend, sugar and salt with maize meal and fortification, and transport it to regional warehouses, and (3) transport 12.5 kg bags of the maize blend from the 9

10 regional warehouses to schools. Each tender, however, is awarded per region, and a tenderer can be awarded several regions. The tender is issued for three years. However, implementation of education programmes is decentralised to the regions, to the thirteen Regional Councils. MOE disburses funds to the regional education office for operational expenses. The budget includes hostel officers requirements. Although hostel officers are not dedicated to school feeding (and there is no specific school feeding budget for the regions), hostel officers spend some of their time coordinating the day-to-day operations of NSFP at the regional and school levels. The operational review found that hostel officers spend on average 25 percent of their time on NSFP. Figure 1: Financial Flows of NSFP Ministry of Education Ministry of Finance Protein Blend (Soya, Sugar, Salt) CENTRAL Directorate of PQA Maize Blend (Maize Meal & Protein Blend) REGIONAL Regional Council (13) Director of Education Transport & Storage Provider SCHOOL Schools MOE Funding Community Contribution Community Contribution 10

11 4. Current Expenditures for NSFP Table 1 below provides a summary of current expenditures on NSFP at the different levels of implementation in Column four shows total cost per child. The number of beneficiaries in 2011 was The next column shows the share of costs incurred at central, regional, and school levels of total MOE costs (i.e. total costs excluding community contributions). The final column shows the share of costs incurred at each level of total costs (including community contributions). Table 1. Summary of school feeding cost per child by level Level Funding Source Total Cost Cost per Child per Year Share of Total (I + II +III) Share of Total (I + II +III +IV) I. Central MOE NAD 59,800,000 NAD % 79% II. Regional (estimate) MOE NAD 3,210,000 NAD 11 4% 4% III. School (estimate) MOE NAD 10 4% 3% IV. School (estimate) Community NAD 40 14% MOE Total (I + II +III) NAD % TOTAL (I + II +III+IV) NAD % Costs were also aggregated by cost category for each level of implementation as shown in Table 2. Figure 2 shows the distribution of costs to MOE by cost category at central and regional levels. Figure 2. Distribution of MOE Costs at Central and Regional Level by Cost Category 11

12 Table 2. Summary of school feeding cost per child by level and category Cost Category Level Central and Regional Costs All Levels Central Regional (estimate) School- MOE (estimate) Total Cost Per Year Cost Per Child Per Year Cost Per Child Per Year Total Cost Per Year Cost Per Child Per Year Total Cost Per Year Cost Per Child Per Year Total Cost Per Year Cost Per Child Per Year School- Community (estimate) Total Cost Per Year Cost Per Child Per Year Capital N$ 729,000 N$ 3.24 N$ 33 N$ 159,000 N$ 0.59 N$ 570,000 N$ 2.65 NA N$ 2.52 NA N$ Logistics & Storage N$ 9,900,000 N$ N$ 37 N$ 9,900,000 N$ NA NA NA NA NA N$ 0.20 Management & N$ 869,000 N$ 1.55 N$ 9 N$ 129,000 N$ 0.48 N$ 740,000 N$ 1.07 NA NAD 7.84 NA NA Administration Staff N$ 2,460,000 N$ 9.09 N$ 12 N$ 580,000 N$ 2.13 N$ 1,880,000 N$ 6.96 NA NA NA N$ 2.81 Commodity N$ 49,000,000 N$ N$ 191 N$ 49,000,000 N$ NA NA NA NA NA N$ 9.50 TOTAL (NAD) N$ 63,000,000 N$ 232 N$ 282 N$ 59,800,000 N$ 222 N$ 3,200,000 N$ 11 NA N$ 10 NA N$ 40 TOTAL (USD) US$ 7,505,665 US$ 28 US$ 34 US$ 7,124,425 US$ 26 US$ 381,240 US$ 1 NA US$ 1 NA US$ 5 12

13 Central level costs comprise the majority of total NSFP costs 79%. Central level costs include expenditures for the tenders to supply the commodity, as well as its transportation and storage. Table 3 shows the costs at the central level of operating NSFP. Table 3. Summary of Central Level Costs of Implementing NSFP 2011 Cost Category Cost Total Beneficiaries Cost per Child per Year Share of Total Commodity N$ 49,000, ,585 N$ % Logistics and Storage N$ 9,900,000 N$ 37 17% Mgmt and Admin Costs N$ 129,000 N$ % Staff N$ 575,000 N$ % Capital N$ 159,000 N$ % TOTAL N$ 59,800,000 N$ % Expenditures on the maize blend constitute the bulk of the expenses: approximately 82 percent of central level costs. Logistics costs (primarily for transportation) are also high in comparison to other countries. Management and staff costs together represent less than 2% of the expenditure, which is very low. Table 4 shows the breakdown of the costs of the maize blend by commodity. The protein blend is the most expensive component of the maize blend. Table 4. Cost Distribution of Maize Blend Commodity Total Quantity (Metric ton/year) Total Cost Cost per MT Share of Total Cost MAIZE MEAL 4266 N$ N$ % PROTEIN BLEND 1693 N$ N$ % SUGAR 731 N$ N$ % SALT 81 N$ N$ % TOTAL 6772 N$ N$ % At regional level, thirteen regional officers implement NSFP. Regional level costs are presented in Table 5. Total regional level costs per child in 2011 were only NAD in the four regions surveyed. 4 The costs are primarily for staffing. The staff at the regional level includes thirteen hostel officers, their clerical assistants, and 66 regional inspectors. Based on the interviews conducted with all the regional level hostel officers as part of the operational review, it was estimated that hostel officers spend 25 percent of their time on 4 Total costs for capital and operations are estimated by calculating the weighted average capital and operations cost per region and multiplying this weighted average by 13. Total cost per child is a weighted average for the four regions. 13

14 NSFP, their clerical assistants also spend 25 percent, and regional inspectors only five percent of their time. The largest expenditures on capital are for office space and vehicle leasing. Table 5. Summary of Estimated Regional Level Costs Cost Category Total Cost Total Cost per Child per Year Share of Total Staff N$ 1,900,000 N$ % Capital (estimate) N$ 570,000 N$ % Mgmt and Admin (estimate) N$ 740,000 N$ % Total N$ 3,210,000 N$ % School level costs comprise a significant share of total NSFP costs. The total cost per child of implementing school feeding at the school level in 2011 was on average NAD 49. Of this total, average costs borne by community members were approximately 40 and the remainder was borne by MOE. 5 Table 6 shows the summary of school feeding costs at the school level. As is evident from the figures, there is significant variation in costs by school. The estimates presented below should accordingly be interpreted with caution when considering applications of this cost analysis to other schools. Table 6. Summary of Namibia School Feeding Costs at School Level Region School Number of Beneficiaries Total Cost Cost per Child per Year I. Costs Incurred by MOE Omaheke Gobabsis 546 N$ 860 N$ 2 Omaheke Nossob 525 N$ N$ 10 Hardap Origo 100 N$ N$ 12 Khomas Dr. Frans Aupa Indongo 400 N$ N$ 13 Khomas Theo Katjimune 40 N$ N$ 77 Weighted Average for MOE Costs N$ 10 II. Costs Incurred by Community Members Omaheke Gobabsis 546 N$ N$ 8 Omaheke Nossob 525 N$ N$ 43 Hardap Origo 100 N$ N$ 137 Khomas Dr. Frans Aupa Indongo 400 N$ N$ 16 Khomas Theo Katjimune 40 N$ N$ 425 Weighted Average for Community Member Costs N$ 40 Weighted Average for School Level Costs (I + II) N$ 49 5 Total cost for MOE and community contributions do not sum to 50 due to rounding. A weighted average cost was calculated for school level costs borne by MOE and community members due to the high variation in costs and number of learners benefiting from school feeding at each school. 14

15 Theo Katjimune primary school in the Khomas region incurred particularly high costs for implementing school feeding. This was due to a number of factors. First, the school uses a gas stove and the cost for gas is estimated to be approximately NAD per year. Second, the school is the only school to provide a cash salary to the cook. Third, the school only feeds on average 40 learners a day (whereas the other schools fed over 100 and two, over 500). The schools generally spend money from their own funds (the school development fund) or get in-kind donations to support school feeding. School development funds are contributed by learners parents. MOE indirectly supports school feeding at the school level in a number of ways. First, MOE pays for water and electricity that help school feeding operate. Second, maize blend bags, which are purchased by MOE at the central level, are the primary method of remuneration for cooks. Third, teachers spend part of their working hours helping to implement school feeding by supervising learner feeding, keeping records of stocks, and making purchases of necessary additional items (cooking and cleaning supplies). As teachers are provided a salary by MOE, their time working on school feeding is a hidden cost for MOE at the school level. This hidden cost has not been reflected in the cost analysis as it is very difficult to estimate and is covered by other funding sources. 5. Assessment of NSFP Costs In 2011, the MOE s allocation for school feeding was NAD 50 million. Actual expenditures, however, for only the tenders (commodity, storage, and transport) was NAD The cost of the commodity rations (excluding transport and storage) are currently approximately 80 percent of MOE s costs of implementing NSFP at the central and regional levels. There may be some opportunities for cost-cutting in supplying the rations. The cost of supplying the different commodities for the maize blend is very consistent across the different terms of the school year. The only exception is that there may be some small cost-saving opportunities for purchasing the protein blend during the first term of the year rather than in the second or third terms in the Ohangwena, Omusati, Oshana, and Oshikoto regions. The protein blend is cheaper in these regions in the first term of the year. 6 However, the costs vary widely by region. The costs in these four regions for the protein blend are generally 125 percent of the costs in the other regions. The cost difference per metric ton of sugar and salt in these regions, as compared to other regions, is also more 6 Future analysts may want to investigate the underlying causes behind the seasonal price fluctuations. 15

16 than ten percent higher. While the cost of the maize meal does not follow this pattern, there are still significant regional variations in costs that could be captured in order to minimize expenditures on NSFP. Annex I shows the average supplier price by region. As shown in the table, the regions with the higher costs are supplied by the same corporation. Further research is needed to assess the reason behind the price differentials and whether there may be opportunities for MOE to cut its commodity purchasing costs. The need for such an analysis is detailed in the Recommendations section (section 8) below. Although the costs at the regional level are quite minimal, this is generally due to a lack of staffing, monitoring, and evaluation that was noted in the operational review. At the regional level, there is a need for a staff dedicated to school feeding. Section 7 of this report provides an assessment of the costs of increasing staffing at the regional level. The operational review also made a number of recommendations for school feeding implementation whose costs are analyzed in Section 7 below. However, one opportunity for cost saving that particularly stands out is minimizing expenditures by switching to fuel efficient stoves. Schools with access to electricity and cooking shelters can switch from gas or wood to electric stoves which are far more efficient. In schools that lack this capacity, there are options for more fuel efficient stoves that will necessitate less wood consumption, engendering health and environmental benefits. 6. Comparison with Other Countries Distributing CSB Rations NSFP provides a maize blend similar to Corn-Soy Blends (CSB) used in some school feeding programmes, in particular by WFP. In 2011, WFP distributed CSB-based rations as meals in Ethiopia and Tanzania. 7 The standardised cost of CSB based school feeding rations per child per year in 2011 was USD 30 in Ethiopia and USD 27 in Tanzania for a standardized ration providing 700 kilocalories or 2,929 kilojoules per day. 8 The cost per child incurred by MOE for implementing NSFP in 2011, standardized to 700 kilocalories/200 days in order for the costs to be comparable to costs for WFP-designed rations, was approximately USD 42 per child. NSFP costs are therefore more expensive than WFP costs for implementing CSBbased school meals. 9 Countries implementing their own school feeding programs are 7 Although other countries use CSB, only these two countries have rations with CSB as the main commodity. The ration in Tanzania is 120 grams of CSB and 20 grams of vegetable oil 20. In Ethiopia the ration is 100 grams of Faffa and 20 grams of sugar. 8 WFP procurement office The WFP standardised cost of providing school meals in 2011 was 40 USD per child per year. CSB rations are usually less expensive than other modalities. CSB purchase prices are usually cheaper than other commodities used in school feeding and may require less preparation time. 16

17 expected to incur higher costs than WFP does for running comparable school meals programs as they are unlikely to have WFP s logistics capacity for tracking commodity prices. One reason that may be underlying the cost differential between WFP and NSFP costs is the cost of the maize blend. The cost of procuring maize blend 10 in Namibia is far higher than the cost to WFP of purchasing CSB elsewhere in southern Africa. The cost of the NSFP maize blend in 2011 was NAD 7,200 per metric ton, equivalent to approximately USD 858. Yet, WFP procured CSB in 2011 at an average price of USD 615 per metric ton in southern Africa. Figure 2 below shows the difference in purchase prices in the southern African countries in Figure 3. Purchase Rates for CSB in Southern African Countries (2011) 11, 12 Although maize blend purchase prices may be higher in Namibia, school feeding may provide a way for the Namibian Government to spur national production. The Namibian Government favors national procurement in its supply arrangements. The NSFP can therefore be used by the Government to provide a steady source of demand for Namibian food producers. According to the operational review analysis, less than half of the maize meal procured for the NSFP is of Namibian origin and even if it is provided by Namibian mills, the millers are obliged to use both Namibian and foreign suppliers to ensure year-round supply. 13 The average price for the maize meal for NSFP in 2011 was NAD 4025 per metric ton. Namib 10 Maize blend refers to the maize and protein blend served in the NSFP. Maize meal refers to only the maize meal component of this blend. 11 Namibia purchases a maize blend which is similar to the CSB blend used by WFP. 12 WFP procurement office Justin Ellis, The Namibian School Feeding Programme: A Case Study, forthcoming

18 Mill s estimated price of maize meal is NAD 4,700 per metric ton. The cost is due to the milling cost, a levy charged by the Agronomic Board, etc. MOE is therefore able to procure maize meal for NSFP at a cheaper rate than Namib Mill s going rate. The Recommendations section describes that a further investigation into price efficiency of the supply chain for the NSFP blend is advised as it can reveal if there are opportunities to minimize expenditures on commodity procurement. 7. Future and Better A number of recommendations emerged from the operational review and subsequent validation workshop, held in Windhoek on July 4-5, Among those many recommendations, the minimal and immediate recommendations that emerged were for: increasing staffing at both national and regional levels, paying an incentive to cooks, providing cooking utensils including pots and fuel efficient stoves, and supplying children with plates and spoons for eating. This section reviews the options and associated costs for enhancing the program by implementing these changes. The costs presented are estimates only, based on the recommendations from the operational review, cost estimates from suppliers, the distribution of beneficiaries across schools, and importantly assumptions on the number of cooks, pots, and stoves necessary based on a minimal number of school visits. The cost projections should be considered a rough estimate of actual necessary expenditures pending a baseline survey into school-level school feeding requirements. An estimate of the increase in staffing cost was done for two new posts at the central level, one at the level of Inspector and one at the level of Chief Control Officer. In addition, thirteen new posts at the regional level were assessed, at the level of hostel officers. 14 The total anticipated cost for the increase in staffing (using current rates for the same grades) is NAD 3,328,000 annually of which approximately NAD 548,000 a year is for the additional central level staff persons and NAD 2, for regional level staff increases. This assumes that these posts are 100 percent dedicated to school feeding. Using the number of beneficiaries in 2011 of 269,585, the cost per child of these additional staff comes to approximately NAD 12 per year. This represents approximately a 200 percent increase in staffing costs from 2011 and a five percent increase in MOE total costs (not including MOE costs at the school level). Cooks spend anywhere from three to eight hours of their day on school feeding. Currently, most cooks are paid in-kind with maize blend bags. This form of compensation is 14 Personal communication with Ms. Edda Bohn, MOE Director, on July 5,

19 considered inadequate. A key recommendation that emerged from the operational review is the need to increase compensation and to compensate cooks in cash for their labour. There is no current legal standard for the minimum wage paid to cooks. The only minimum wages which have just been inscribed into law are for labourers / general workers / nonskilled workers in the construction, security and agricultural sectors who are paid NAD 4 to 5 per hour. There is a domestic wage bill currently in parliament and when passed, this will cover the hourly rate for cooks, domestic workers, etc. Whereas the hourly rate for cooks is based on their negotiated wage, the Ministry of Labour (MOL) recommends that the hourly rate for cooks be NAD 5 per hour. Assuming a four hour work day for a 195 day term, the annual wage to cooks would be NAD 3,900. The current value of cooks compensation in in-kind donations at the five schools considered in this study was on average NAD 7.16 per hour. Therefore, the compensation rate recommended by the MOL represents a cost savings for MOE. Note however that wages for cooks tend to be highly differentiated by whether the school is located in an urban or rural area. The average rural compensation rate in the schools visited is NAD 4.15 per hour and the average compensation rate in urban areas is almost three times higher, at NAD per hour. Given that cooks at the schools visited worked on average four hours per day, five days a week, at an average compensation rate in in-kind donations of NAD 7.16 per hour, their monthly compensation is equivalent to NAD 573 per month or NAD 5,155 per year per cook for a nine month school year. Therefore, the decrease in expenditure per cook per year represents a twenty five percent cost savings to MOE. However, this price differential is an estimate because actual cooking time will vary according to school needs. 15 The cost savings for paying cooks in cash is presented as a negative additional cost in the table on the projected cost of school feeding below (Table 7). The actual number of cooks needed is not known. An analysis is needed that considers the distribution of cooks at schools and the proximity of schools. For example, it may be the case that if schools are located nearby and the number of learners at the schools is small, one cook may be able to serve more than one school. The estimate presented in this analysis is based on an analysis of the distribution of beneficiaries across schools in Namibia in In 2011, MOE reported that there were 8,350 cooks in the NSFP. The cook to learner ratio was therefore approximately 1 to 30 in this year. As a lower bound assumption for the future cost of cooks, this analysis assumes that one cook is needed for every 100 learners. As an upper bound, this analysis assumes that one cook is needed for 15 Other recommendations such as adequate cooking utensils and fuel efficient stoves- should contribute to contain food preparation time. 19

20 every 70 learners. This is the average ratio of cooks to learners in Haiti, Chad, and Sri Lanka s programs. 16 Although compensating cooks in cash will generate a cost saving for MoE, budgeted figures will not reflect the change in compensation as a savings. This is because currently MoE budgets only two bags per term for compensating cooks. In the schools visited for this study and the case study, actually compensation was far higher usually in the order of two bags per week per cook. Table 7 shows this discrepancy between budgeted and actual cost figures. The current planned compensation for cooks has a value of approximately NAD 4,500,000, assuming that each of 8,350 cooks is paid two bags per term. Actual expenditures to compensate cooks are far higher. Therefore, compensating cooks yields a cost savings even though the planned expenditures would be between NAD 10 million and NAD 15 million, which is higher than the budgeted amount. Many schools lack proper cooking equipment, particularly pots and stoves. The operational review recommended that schools be supplied with #25 cast iron 3-legged pots which are long-lasting and stand up to the hot flames generated by wood-fired cooking. However, not all schools need such large pots and not all schools need cast-iron pots because not all schools cook on an open flame. Table 7 below presents cost estimates for different sizes and types of pots taking account of the distribution of beneficiaries across schools in 2011 and assuming, based on interviews with cooks, that every litre of a pot can provide 2-3 learners with meals. An additional assumption was made that 60 percent of schools cook on an open flame and need cast iron pots, whereas 40 percent would use aluminium pots. However, the actual number of schools cooking with wood needs to be verified. Cast iron pots and fuel-efficient stoves can be considered to be an investment. The cast iron pots will last for a long time, in fact, some of the original WFP pots from 1991 are still in use today. Switching to cast-iron pots may in some cases generate a cost-savings as they will minimize the frequency of pot purchases. In addition, fuel-efficient stoves may generate a cost-saving over the long term as they will minimize the necessary expenditure on firewood. WFP is conducting analyses on procurement of fuel efficient stoves. There are many different types of stoves available at different cost points. Many of the stoves are quite affordable, starting at only USD 4 per stove, and can create significant savings in fuel use. For example, the JIKO stove used in several African countries, uses between 30 and 40 percent less fuel than traditional stoves. The WFP country office can provide more information on these and other stoves. 16 The cook to learner ratio in these countries is: Sri Lanka 2,159 cooks for 166,436 children, Chad 1,418 cooks for 148,231 children, and Haiti 14,400 cooks for 1.04 million children. 20

21 Many schools lack cooking shelters and are exposed to the sun and rain while cooking. The store Game offers a rudimentary structure (shown on the right) that is easy to move and provides shelter from the sun and rain although the sides are exposed. One shelter is suitable for one school. However, given that many schools already have cooking shelters this analysis assumes that only twenty percent of schools would benefit from these structures. Cost for cooking shelters may be underestimated and should be interpreted with caution. Finally, children at many schools do not have plates or spoons to eat their meal with. This analysis estimates the price of a one-time expenditure for providing all children with plates and spoons. Table 7 summarizes the projected cost for implementing NSFP with the addition of these minimal improvements based on the assumptions and estimates outlined above. 21

22 Table 7. Future Costs for Implementing NSFP with Minimal Improvements Item Current Costs Unit Cost/Initial Investment Additional Improvements Unit Cost per Year Estimated Quantity Additional Cost Total Future Cost Logistics & Storage N$ 9,900,000 N$ 9,900,000 Management & Administration N$ 869,000 N$ 869,000 Commodity N$ 49,000,000 NAD N$ 4,527,948 N$ 44,472, /cook/year Capital N$ 729,000 Fuel efficient stove N$ 1,600,000 #25 70 liter Cast-iron 3-legged Cooking Pot N$ 1,800,000 # liter cast-iron 3-legged Pot N$ 978,000 # liter cast-iron 3-legged Pot N$ 174, liter Aluminium Pot N$ 1,300,000 Cooking shelter N$ 91,000 Kango Enamel soup plate 22cm N$ 2,400,000 Ladle 8oz. 225ml. B/C N$ 88,000 Spoons N$ 2,700,000 Projected New Capital Costs with Additional Inputs N$ 11,900,000 Staff N$ 2,460,000 Central level - new staff at level of Inspector N$ 320,000 Central level - new staff at level of Chief Control Clerk N$ 228,000 Regional level - new staff at level of hostel officer N$ 2,780,000 Cook salary - estimate of cook to learner ratio of 1:100 N$ 5/hour N$ 10,600,000 Cook salary - estimate of cook to learner ratio of 1:70 N$ 5/hour N$ 15,100,000 Projected New Staff Costs with Additional Inputs - LOWER BOUND N$ 16,400,000 Projected New Staff Costs with Additional Inputs - UPPER BOUND N$18,100,000 TOTAL PROJECTED COST - LOWER BOUND N$ 83,541,052 TOTAL PROJECTED COST - UPPER BOUND N$ 85,241,052 22

23 8. Recommendations There are two issues identified in this analysis that MOE should consider in order to improve the efficiency of delivery of quality meals to learners: (1) Supply Arrangements: There are regional differentiations in the supply price of the NSFP maize blend components that correspond to suppliers. It is recommended that an assessment be conducted of the MOE current supply arrangements and potentials for modifying supply channels to improve efficiency. A food technologist should be identified that can provide technical assistance on food quality assurance. The technologist can determine whether the components of the NSFP blend are being supplied for the lowest available price for the quantity and quality of the product. In addition, the technologist can assess opportunities for adopting new supply arrangements. Specific new supply arrangements that the technologist could consider are decentralizing procurement and contracting with different suppliers. The food technologist could also investigate the potential for diversifying the food basket while maintaining the nutritional quality of the meals and the associated cost of such changes. Appendix II describes an initial assessment of two such options for basket differentiation. (2) Survey of school needs: There is little available data on the assets available at school-level for school feeding. It is therefore difficult to make a recommendation on the cost of supplying basic products for NSFP without an assessment of the inventory at schools. The projected costs presented in this report are estimates largely based on the assessment of school needs from six school visits. Therefore, these estimates should be approached with caution. There is a high level of variation in how NSFP operates in different schools, and as a consequence, what MOE costs will be for incorporating the minimal improvements recommended by the operational review (i.e. remuneration for cooks, provision of cooking pots, utensils, etc.). It is recommended that a school level survey be conducted to determine the cooking arrangements at different schools. Primary among the many aspects of school feeding that the analysis should identify is which schools are using wood-burning stoves and how many schools already have proper cooking equipment. This information will inform MOE on how many schools would benefit from the provision 23

24 of fuel efficient stoves and different types of pots (in particular whether they need cast iron or aluminum pots). In addition, in order to determine the number of cooks that are needed, it is necessary to determine how many beneficiaries there are per school and the distance between schools. 9. Conclusion The estimates provided in this study can be used as a guide by MOE in planning the future directions of NSFP. It provides estimates of the current costs of running NSFP on a per-child per-year basis as well as the budget that should be allocated per year if minimal improvements and expansions are made. Many of the expenditures on NSFP currently are quite minimal, namely on staffing. However, the cost of NSFP incurred by MOE (excluding school level costs which are borne by the community) is more expensive than other countries in which WFP delivers similar food rations (CSB) and modalities (central procurement and meals cooked in schools). There may be opportunities for further efficiencies in procurement. This analysis makes a recommendation for further investigation into the current supply arrangements. This cost analysis provides estimates for minimal improvements to the NSFP, as outlined in the operational review and validation workshop. Specifically, it presents estimates of the current costs of running NSFP on a per-child per-year basis as well as the unit costs that should be allocated on a per-year basis if minimal improvements and expansions are made. Although the analysis presents these improvements as costs, they could in fact constitute a net cost-saving. Fuel efficient stoves, for example, will save schools on their purchases of firewood. However, the projected costs should be considered to be estimates only. A baseline survey of school feeding needs at school level is essential in order to assess the actual quantities that should be delivered. The improvements considered in this analysis do not incorporate the many largescale recommendations that were made in the workshop and operational review, that would greatly enhance the quality of NSFP. Specifically, this analysis does not consider expansion scenarios (only maintaining the beneficiary numbers at 2011 levels), amendments of staff structures, ration differentiation, alterations to monitoring and evaluation systems nor many of the other large-scale changes considered in the operational review and subsequent validation workshop. Annex II considers the cost for incorporating two new food items into NSFP: dairy drink and mahangu biscuits. The development of the manufacture of these two items is currently being supported by MOA. Further investigation into other food 24

25 items as well may reveal other nutritious and cost-effective means to provide learners with their dietary needs. Annex I. Procurement Costs by Region and Term Table 8. Average Supplier Price by Region Maize and blending Protein blend, salt and sugar Transport Region Supplier Price/MT Supplier Price/MT Supplier Price/MT Caprivi Halutusane 4400 Halutusane 7212 Halutusane 464 Erongo Alason 3750 Alason 7057 Ehupo 425 Hardap Alason 3750 Alason 7057 Ehupo 430 Karas Alason 3750 Alason 7057 Ehupo 548 Kavango Tyno 4388 Alason 7057 Hakaranda 552 Khomas Alason 3750 Alason 7057 Hakaranda 352 Kunene Tyno 4388 Alason 7057 Ehupo 614 Ohangwena Pena 3900 Pena 8389 FU Rd 1049 Omaheke Alason 3750 Alason 7095 TAA Mokondjo 395 Omusati Pena 3900 Pena 8389 FU Rd 1045 Oshana Pena 3900 Pena 8389 ARC 924 Oshikoto Pena 3900 Pena 8389 TAA Mokondjo 914 Otjozondjupa Alason 3750 Alason 7057 Hakaranda

26 Annex II. Ration Differentiation Options Many of the recommendations that emerged from the operational review and validation workshop were for substantial changes in the way school feeding is being implemented. Such changes include: decentralizing the procurement and transportation structure, differentiating rations, implementing a new monitoring and evaluation system, and expanding the number of beneficiaries. Most of these modifications need to be thoroughly researched to determine the necessity and feasibility of such changes. As such, an analysis of the cost of these modifications is outside the scope of this study. This analysis examines the cost for two potential changes to the school feeding diet which are locally produced and would increase dietary variety. Many schools complained about the lack of variety in school feeding, where maize meal is provided to learners every day. In addition, the operational review and workshop participants expressed an interest in increasing the sourcing of locally produced food to stimulate local production within the context of school feeding. Namibia Dairies conducted an analysis of the potential for expanding NSFP to include dairy products. 17 They proposed providing a drinking yoghurt called nammilk Oshikandela TM. The product can be provided in a fortified form and is available in both fresh and long life versions. Namibia Dairies can produce a 200ml or 250ml dairy based long life product with a shelf life of at least six months. The nutritional value of all dairy based products (example of nammilk Oshikandela TM below) is high and based on a 200ml or 250ml pack, as shown in Table 9. The nutritional value ensures that an individual will receive his / her recommended dietary allowance of protein, calcium and Vitamin A. Further, fortification can be enhanced to ensure addition of Vitamin D and B12 as well as Iron, Iodine, Magnesium and other micro nutrients to deliver all basic nutritional elements on a daily basis. 17 School Nutrition Programme Namibia Dairies Report, October

27 Table 9. Nutritional info for Dairy Product Oshikandela sweet without fortification with fortification Current Maize blend Per 100ml Per 200ml * % NRV s Per 100ml * % NRV s Per 125 grams Protein (g) Fat content (g) Carbohydrate (g) Energy (kj) Calcium (mg) Vitamin A (µg) Vitamin D (µg) Crude fibre (g) 2 Sodium (mg) 625 Ash (g) 3 Iron (mg) 3 Zinc (mg) 2 Namibia Dairies has done a preliminary cost estimate based on a 200ml portion of long life fortified Oshikandela. This product, including distribution cost, can be provided to the NSFP for approximately NAD 2.20 per pack. This calculation is based on the purchase of at least 10 million portions per annum. However, pricing may vary depending on the fortification requirements, vitamins, supplements and nutrients as well as the number of packs ordered and supplied. Increasing the nutritional content of the packs is likely to also result in an increased price. Currently, NSFP expenditures are approximately 0.79 NAD per learner per day for the commodity and 0.98 NAD per learner per day for the commodity plus its transport and storage. Therefore, supplying the dairy product would necessitate over a two-fold increase in the budget for food and transport purchase. In addition, switching to a dairy drink would bring about other challenges in ensuring the environmentally friendly disposal of the milk packs. The Ministry of Agriculture is implementing a Mahangu Marketing plan for 2010 to Mahangu is an indigenous Namibian local produce that is traditionally used in porridge or fermented beverages. In cooperation with GB Mahangu Enterprises, MAWF and NAB have supported GB Enterprises in developing and producing several types of cookies. 27

28 Currently Mahangu Enterprises, founded last year (in 2011), is in the nascent stages of its development. The company is expanding to scale up capacity to produce kg per month, which it hopes to accomplish in less than two months time (by September, 2012). The viability of the product for use in NSFP is questionable. The shelf life has not yet been identified. As shown in Table 10, the nutritional content does not yet meet WFP standards not the levels of the current maize blend -- although the company is more than willing to work with a nutritionist / food technician to amend its recipe to boost the nutritional value. Table 10. Nutritional info for Mahangu cookies Mahangu cookie Current Maize blend Ingredient Per 100 grams Per 125 grams Protein and Carbohydrates (g) Fat (g) Crude fibre (g) 2 Calcium (mg) Sodium (mg) Ash (g) 3 Iron (mg) 3 Zinc (mg) Mahangu Enterprises expects that its wholesale price will be NAD 60 per kilogram. As one ration is 100 grams, the cost per child is thus NAD 6 per child per day, or more than a six-fold increase in the current cost. Diversifying NSFP rations can be done on a one, two, or three days per week basis, which may help contain costs. However, it seems likely that the alternatives analyzed in this study are cost prohibitive for MOE, and if possible, future analysts should look into other foods and suppliers. 28