Prepare Decentralized Evaluation Activity Evaluation of USDA McGovern-Dole Grants (FFE-442-2019-013-00) for WFP School Feeding in Cambodia 01 November 2019 to 30 October 2023 Final Report for Baseline Study January 2021 Commissioned by WFP Cambodia Evaluation Manager: Bunthang Chhe Prepared by Damayanti D. Soekarjo, Team Manager Prem B. Bhandari, Overall Team Leader Soledad Posada, International Evaluator Jean-Pierre Silvéréano, Subject Matter Expert Chandara Gnim, National Evaluator DECENTRALIZED EVALUATION FOR EVIDENCE-BASED DECISION MAKING Decentralized Evaluation Acknowledgements This evaluation report was developed through wide consultation with programme partners and stakeholders of the USDA McGovern-Dole Programme. The Evaluation Team would like to acknowledge the inputs from the WFP Cambodia Team, the WFP Regional Bureau in Bangkok, the Royal Government of Cambodia (including the Ministry of Education, Youth and Sport) at national, provincial, district and commune levels), other involved ministries and stakeholders, for their contributions towards the evaluation. Special thanks to key staff in the WFP Cambodia Country Office for their time and contributions, especially the Education Unit, Riguen Thorn, Bunthang Chhe and Emma Conlan. Also, thanks to the Head of WFP Sub-Office in Siem Reap, Um Nisith, and the WFP Field Office team for Siem Reap, Kampong Chhnang and Kampong Thom. The KonTerra Evaluation Team would also like to thank Indochina Research Limited team, especially Francis Palomares Jr, Pisei Im and the enumeration team, for their dedication to the survey. Special thanks are also due to Bunyeth Ho for quality control of the survey and Hans van Zoggel for his role in the qualitative field mission. Finally, the team gratefully acknowledges the community participants in the evaluation process – households, teachers, school administrators, cooks, parents and school children. Disclaimer The opinions expressed in this report are those of the Evaluation Team, and do not necessarily reflect those of the World Food Programme. Responsibility for the opinions expressed in this report rests solely with the authors. Publication of this document does not imply endorsement by WFP of the opinions expressed. The designation employed and the presentation of material in maps do not imply the expression of any opinion whatsoever on the part of WFP concerning the legal or constitutional status of any country, territory or sea area, or concerning the delimitation of frontiers. Annexes to this report are supplied in a separate file. Table of Contents Executive Summary......................................................................................i 1. Introduction ...........................................................................................1 1.1. Overview of the Evaluation Subject ....................................................................... 1 1.2. Context....................................................................................................................4 1.3. Evaluation Methodology and Limitations..............................................................8 2. Baseline Findings ................................................................................. 11 2.1. Evaluation Question 1: How appropriate is the design of the McGovern-Dole programme to the context and to what extent is it aligned with the needs of women, men, girls and boys and with the priority needs of schools, and the Government? ....11 2.2. Evaluation Question 2: What are the results of the programme?...................... 18 2.3. Evaluation Question 3: How likely is it that the programme is sustainable?....36 2.4. Evaluation Question 4: Why and how has the programme resulted in the observed results? ............................................................................................................39 3. Conclusions and Recommendations .....................................................41 3.1 Overall Assessment/Conclusions ......................................................................... 41 3.2 Recommendations ................................................................................................43 List of Acronyms ....................................................................................... 45 List of Tables Table 1: Planned numbers of schools and beneficiaries, 2020-2021.................................................2 Table 2: USDA Food for Education schools in school years 2020-2023......................................... 10 Table 3: Benefits of school feeding for the child, as perceived by parents of selected girls and boys ..................................................................................................................................................... 12 Table 4: Benefits of school feeding for the household, as perceived by parents of selected girls and boys ............................................................................................................................................. 12 Table 5: Annual household income from different sources (in US dollars) .................................... 13 Table 6: Household expenditure ....................................................................................................... 13 Table 7: Surveyed schools .................................................................................................................. 18 Table 8: Teaching staff in surveyed schools (SY 2019/20) .............................................................. 19 Table 9: Teacher attendance (SY 2019/20) ...................................................................................... 19 Table 10: School mapping information in SY 2019/20 .................................................................... 19 Table 11: Promotion, repetition and drop-out rates (SY 2018/19) .................................................. 19 Table 12: Literacy activity performance indicators and annual targets...........................................20 Table 13: Improved Literacy of School-Age Children – Annual targets .......................................... 21 Table 14: Food distribution activity - Annual targets .......................................................................22 Table 15: Percentage of students in target schools who indicate that they are “hungry” or “very hungry” during the school day (March 2020)...........................................................................23 Table 16: Reported inattentiveness of students (March 2020)........................................................23 Table 17: Average number of morning sessions, percentage of children receiving school meals and those who took some home ........................................................................................................23 Table 18: Percentage of children who had received deworming medication in the past six months .....................................................................................................................................................24 Table 19: Improved health promotion activity - Annual targets ......................................................24 Table 20: Number of classrooms in surveyed schools......................................................................24 Table 21: Number of functioning drilled wells/ water catchments..................................................25 Table 22: School access to year-round clean water ..........................................................................26 Table 23: Percentage of schools with fixed hand and mobile hand washing stations.....................26 Table 24: Schools with soap at handwashing stations .....................................................................26 Table 25: Latrines in surveyed schools, by functionality.................................................................. 27 Table 26: Presence of functioning latrines for teachers ...................................................................28 Table 27: Presence of functioning latrines for boys and girls ..........................................................28 Table 28: Presence of functioning latrines for disabled students ....................................................28 Table 29: Solid waste disposal...........................................................................................................29 Table 30: Promotion of improved nutrition activity - Annual targets.............................................29 Table 31: Schools with vegetable gardens .........................................................................................30 Table 32: Schools receiving training on food preparation and good health and nutrition practices ..................................................................................................................................................... 31 Table 33: Percentage of schools participating in community awareness raising events on the importance of education ............................................................................................................ 31 Table 34: Capacity building activity performance indicators and annual targets...........................32 Table 35: Schools with existing LSFCs and SSCs..............................................................................33 Table 36: Availability of school management tools ..........................................................................33 Table 37: Average amount of community contributions to each school in SY 2019/20 .................33 Table 38: Distribution of surveyed households by province, supporting programme, and gender of sampled child..............................................................................................................................33 List of Figures Figure 1: Linkage between McGovern-Dole and LRP programmes...................................................3 Figure 2: Condition of kitchens in USDA supported schools (n=70) ..............................................25 Figure 3: Condition of storerooms in USDA supported schools (n=46, in percentage) .................25 Figure 4: Condition of well-functioning drilled well/rainwater catchments...................................26 Figure 5: Student handwashing practices - frequency of handwashing with soap and water ........ 27 Figure 6: Available stock of soap supply ...........................................................................................27 Figure 7: Provider of soap..................................................................................................................27 Figure 8: Condition of functioning latrines ......................................................................................28 Figure 9: Condition of non-functioning latrines...............................................................................29 Figure 10: Latrine maintenance ........................................................................................................29 Figure 11: Reason for establishing vegetable gardens in schools .....................................................30 Figure 12: Vegetable planting in school gardens by month (Oct 2019-Sep 2020)..........................30 Figure 13: School garden maintenance ............................................................................................. 31 Figure 14: Awareness amongst school personnel on three types of food groups ............................ 31 Figure 15: Number of times that awareness-raising events were conducted during SY 2019/20..32 Figure 16: Timing of awareness raising events .................................................................................32 Figure 17: Household main source of water......................................................................................34 Figure 18: Household sanitation (toilet) facilities ............................................................................34 Figure 19: Percentage of households reporting consumption of each food group in last seven days .....................................................................................................................................................35 Figure 20: Household Food Consumption Scores............................................................................35 Figure 21: Household use of consumption coping strategies...........................................................35 Figure 22: Transition process............................................................................................................38 McGovern-Dole Programme Baseline Report – Cambodia – January 2021 i |Page Executive Summary 1. This report presents findings from a baseline study of the United States Department of Agriculture’s McGovern-Dole support to the World Food Programme School Feeding activities in Cambodia (USDA Project Agreement number: FFE-442-2019-013-00. This baseline is the first phase of a series and was undertaken between August and October 2020; it will be followed by a midline evaluation in 2021 and an endline evaluation in 2023. This decentralised activity evaluation series was commissioned by the WFP Cambodia Country Office, based on the Terms of Reference attached as Annex 1. 2. This baseline covers the period from September 2019 to August 2023. The McGovern-Dole programme is a continuation of the previous phases (2013-2016 and 2017- 2019), and is being implemented in three provinces of Cambodia: Siem Reap and Kampong Thom (which both received support in the previous phase of the programme), and Kampong Chhnang which was added in this phase. 3. The objectives of the overall evaluation series are both accountability (identification of results and performance) and learning (factors affecting the results, as well as highlighting lessons for future activities). The purpose of the baseline is to establish the current situation against which the next rounds of the evaluation (mid-term and endline) will measure progress towards the targets. 4. The main expected users for this evaluation report include USDA, the WFP Regional Bureau and Country Office, its main partner the Ministry of Education, Youth and Sport (MoEYS) as well as other ministries of the Royal Government of Cambodia, and the other implementing partners in the programme: World Vision, Plan International and World Education Incorporated. Context 5. Cambodia has improved primary school enrolment and attendance in recent years and there are reducing gender disparities in education. The main challenge now is primary school completion, in particularly in rural areas. Boys and girls drop out due to poverty or relocation, and because schools often lack the basic infrastructure needed to provide a suitable learning environment, particularly for adolescent girls and children with disabilities. Although both repetition and dropout rates have steadily declined in the last five years, they remain a key concern to Ministry of Education, Youth and Sport. Food security and undernutrition also remain important public health concerns, with a recent study1 finding that the overall dietary intake of school children did not meet the local recommended dietary allowances. Programme activities, beneficiaries, indicators and financial resources 6. Under the programme a range of activities will be implemented, covering capacity building, food distribution, literacy, promotion of improved health and nutrition, and establishing school gardens. A daily cooked breakfast will be provided to 151,700 primary and pre-primary children, and the programme will gradually transition to local procurement of produce before being handed over to the Government. Capacity building activities will target national and subnational authorities and other stakeholders to support the creation of an enabling environment for a national school feeding programme, increase stakeholder capacity to implement and monitor the school feeding programme, promote community engagement, increase literacy teaching skills and set up and maintain IT-based platforms and tools to support the programme. To support the literacy component, the 1 Horiuchi, Y et al (2019) Urban-rural differences in nutritional status and dietary intakes of school-aged children in Cambodia. Nutrients. 2019 Jan; 11(1):14. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 ii |Page training of teachers and production of non-fiction books for Grade 2 students on nutrition and healthy eating will be developed and distributed. To promote better health and nutrition, WFP will work with partners to ensure that minimum standards of health and hygiene infrastructure and technical know-how exist in as many schools as possible, and are embedded into national systems and institutions. In addition, education and Social Behaviour Change Communication materials will be made available. Vegetable seeds and training will be provided to selected schools to support school gardens. The indicators used to monitor and evaluate this programme are provided in Annex 5 (Results Frameworks) of this report. They reflect the programme activities. 7. The total programme budget for three years is almost US$19 million and will be used mainly for activities (70.5 percent), with the rest for administrative costs (11.7 percent), Internal Transportation, Storage and Handling (9.7 percent) and total indirect costs (8.1 percent). Methodology 8. The evaluation series, of which this forms the baseline, was designed to assess the McGovern-Dole grants against the following evaluation criteria: appropriateness, effectiveness, efficiency and sustainability. The main evaluation questions, as indicated in the Terms of Reference, were about the appropriateness of the programme design, and to what extent it is aligned with the identified needs; what are the results of the programme; how sustainable it is likely to be; and why and how has the programme achieved the observed results? It should be noted that the second and fourth questions cannot be fully answered in this baseline as activities have not yet been implemented under this round of the McGovern-Dole supported programme. Rather, the evaluation team set out to determine the baseline numbers and structures, against which progress regarding effectiveness and efficiency will be measured in the subsequent rounds of the series. 9. In order to respond to these questions, the evaluation team undertook a desk review of programme documents and other relevant reports and literature. A quantitative survey was undertaken in a random sample of 70 USDA supported schools and 70 non-supported comparison schools, and covered a total of 834 families of students from these schools. The same statistical approach and sample size calculation tools were used as in the previous evaluation (2017-2019), based on a 95 percent confidence level, 80 percent power and a margin of error term of eight percent. The calculated sample size is 70 matched pairs of intervention and control schools (a total of 140 schools), which is representative and has sufficient power to detect changes. The schools were selected using systematic random sampling sorted out using the percentage of poor households in the village from the list of programme schools provided by the Country Office. 10. Qualitative interviews and group discussions were held with key stakeholders at the national, subnational and local levels, as well as with representatives of USDA and WFP’s Regional Bureau, Country Office and Headquarters. All information was triangulated between source, method and evaluator, as well as compared with the results of the 2019 endline evaluation of the previous round of the programme. 11. Descriptive analysis of quantitative data was performed for case-control comparisons using SPSS and R software, and wherever appropriate, difference in means were tested using an independent t-test. Qualitative information was verified and corroborated through systematic triangulation by source, method and evaluator through intensive discussions among the team members. Feedback on the initial findings and conclusions were obtained from WFP and partner stakeholders, and incorporated in the report. 12. Limitations included the delay in the implementation of this baseline, COVID-19 related travel restrictions for the international evaluation team members as well as school McGovern-Dole Programme Baseline Report – Cambodia – January 2021 iii |Page closures and safety considerations for the survey team and the national evaluators. Measures were taken to mitigate against these risks as far as possible, including strict adherence to health guidelines, adjusting the timeframe of the questionnaire, use of secondary data, the use of online interviews and regular phone discussions amongst the evaluation team members. Key Findings 13. The key findings of the evaluation team are summarised below, structured according to the main evaluation questions and indicating the type and strength of evidence supporting each finding. Evaluation question 1: How appropriate is the design of the McGovern-Dole programme to the context and to what extent is it aligned with the needs of women, men, girls and boys and with the priority needs of schools, and the Government? 14. The programme design was considered to be appropriate for the needs of women, men, girls and boys, and aligned with Government priorities, as well USDA, WFP, United Nations and other partners’ activities and policies. This finding was based on strong triangulated evidence from the desk review, as well as in-depth interviews with a large number of stakeholders from different agencies, ministries and communities. 15. The programme, which has been implemented under previous agreements for many years, is well accepted by all stakeholders who, across the board, were aware of its benefits for children, schools and communities. While originating from in-depth interviews only, the consensus among informants makes this evidence strong. 16. Based on the programme documents and in-depth interviews, the programme aims to include schools with high poverty rates and poor educational outcomes based on Government databases alongside WFP’s vulnerability analysis and mapping data. Differences with regards to the criteria were observed between selected schools within the programme during the field mission. Interviews with the country office staff revealed these are likely because schools were selected during previous rounds of the programme, though it is unclear if criteria have since been reviewed and adjusted according to changed circumstances. 17. Equal gender participation is anticipated but there are no gender-specific outcomes in the Results Framework that would explain more clearly what and how gender issues are intended to be addressed, nor indicators against which to measure progress, and no evidence of a sound gender analysis was found. 18. The other programme indicators are considered appropriate and feasible for use in monitoring and evaluation of this programme. Evaluation question 2: What are the results of the programme? 19. Community acceptance of, and participation in, the school feeding programme is high. This was evidenced by positive feedback provided by parents and school staff, and the establishment of local school feeding committees and school committees during previous rounds, that are still active. 20. The quantitative school survey revealed that baseline attendance rates of teachers and students are already high, which forms a solid basis for the programme. 21. According to a partner’s baseline study, virtually all students who graduate from Grade 2 continue to struggle in all key reading domains. With this in mind, this programme’s target for strategic objective 1 has been adjusted to a lower and more feasible percentage of children who should reach the threshold after two years. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 iv |Page 22. Direct observation, quantitative and qualitative baseline data, validated against the 2019 endline data, revealed that all programme schools have kitchens, although more than half were not in a good condition and fewer than half were using energy-efficient stoves. Access to sources of clean water was available to the majority but not all of the schools, and on average there was one non-functional latrine per school. Not all schools currently have gardens. Improvement of these facilities is part of the programme design and will require extra attention. 23. In previous rounds, capacity building trainings had been provided to school staff, cooks, teachers and Government staff. Staff turnover and suspension of many programme activities during the school closures may lead to a need for additional (refresher) trainings. Evaluation question 3: How likely is it that the programme is sustainable? 24. The Royal Government of Cambodia is highly invested in a national school-feeding programme as evidenced by a multisectoral school meals policy, including budget allocation. During this programme period, 141 additional schools are expected to be handed over to the Government, in line with WFP’s new School Feeding Strategy 2020- 2030. 25. As per the programme agreement with USDA, participating schools are being transitioned from the traditional school meals programme model towards home-grown school feeding (HGSF) via a HGSF-hybrid approach – with some commodities being imported and others being procured locally. The ongoing transition plan of the in-school activities to national ownership has a current end date of 2025. 26. Stakeholders at national and sub-national Government levels, as well as at the schools, were almost unanimous in expressing their need for continued capacity building and mentoring by WFP to enable a smooth transition and sustained implementation of a school meals programme in Cambodia. This is in line with the country strategic plan and WFP’s global direction to transition from direct implementation towards a supporting role. Evaluation question 4: Why and how has the programme resulted in the observed results? 27. Interviews with stakeholders revealed that WFP has a strong relationship with the Ministry of Education, Youth and Sport, and there is a network of donors, United Nations and other organisations that leverage the school feeding programme. 28. Informants at national level concurred that the Royal Government of Cambodia is highly committed to a national school feeding programme, and this was confirmed by the existence of national policies and budget allocations. However, the same informants stressed that they continue to require support from WFP and other partners. Senior officials specifically requested WFP to facilitate the transition from ‘donorship’ via ‘partnership’ to ‘ownership’, through increased and continued capacity building and mentoring. 29. The original timeline for handover was very short and the current COVID-19 pandemic has led to delays. WFP and the Government have agreed on a new target date of 2025. Because of school closures, programme achievements for SY 2019/20 (and SY 2020/21) are likely to fall short of targets. 30. There is a need for more clarity on roles and responsibilities between implementing partners and WFP in monitoring activities. The involvement of the Government in monitoring also needs to be augmented. 31. The Government’s Decentralisation & Deconcentration (D&D) strategy will ultimately lead to improved coordination at the district level, but while it is being rolled McGovern-Dole Programme Baseline Report – Cambodia – January 2021 v |Page out, it poses additional challenges such as differences in the level and speed of D&D between ministries and their subnational capacity to implement policies and programmes. Overall conclusions 32. The evaluation team concluded that the programme is relevant and well received by all stakeholders. It is well aligned with needs of beneficiaries, priorities of the Royal Government of Cambodia, USDA and WFP. Areas for improvement include transparency of school targeting and gender mainstreaming. The evaluation team concludes there is a need to ensure targeting is based on updated needs assessments and to ensure community stakeholders understand the process. While the programme indicators are relevant and feasible, the absence of gender specific outcomes and indicators in the Results Framework makes it impossible to adequately monitor and evaluate progress in these areas. 33. This round of the programme has not yet started, and as this is a baseline, no results can be analysed. The report sets out the baseline situation and numbers against which the effectiveness and efficiency of programme implementation will be measured during subsequent evaluation rounds (midterm and end line). Nevertheless, the evaluation team concluded that the programme benefits from an enabling human, material and institutional environment established during previous rounds of the McGovern-Dole programme. 34. The programme has already shown adaptability to changing circumstances, as evidenced by adjustment of targets for literacy results; by reallocating available resources during the school closures caused by the COVID-19 pandemic; and by the adjustment of the timeline for programme transition. 35. In response to the third question, the evaluation team concluded that the fact that the programme is well established, the Royal Government of Cambodia is highly committed to taking over the programme, and there are strong supportive networks in the country as well as globally for the school feeding programme, all increasing the chances for a sustainable programme. Implementation of the 2015 Roadmap is on good track and has been adjusted to the unprecedented consequences of COVID-19 pandemic. The main area of concern is the readiness of stakeholders at all levels to directly manage and monitor the programme. The need for continued capacity building and mentoring is clear across the board, implying the need for the Country Office to focus on its transition to a different role, moving from direct implementation to capacity building and supportive mentoring. 36. Question 4 refers principally to the entire programme duration, and this baseline aimed only to assess enabling factors as well as barriers for the likely success of this programme. The evaluation team concluded that WFP is well established in Cambodia and has strong relationships with relevant stakeholders in the Government, United Nations organisations, implementing partners and schools. The Country Office has extensive experience with school feeding programmes, and the Royal Government of Cambodia is highly invested in the programme. Challenges to a successful handover, though, include the capacity of stakeholders to take ownership of the programme, and of WFP to transition to a new role of capacity building and mentoring, as mentioned before. 37. This baseline indicated the need for organisational learning related to the design and implementation of the SFP, and the lack of gender guidelines. While many aspects of the programme are well designed and implemented, there is room for improvement, in particular related to gender and coordination. Recommendations 38. The findings and conclusions of this baseline exercise lead to the evaluation team making the following recommendations to enhance the likelihoods of a successful programme implementation. They are all priorities, but timelines have been indicated: McGovern-Dole Programme Baseline Report – Cambodia – January 2021 vi |Page Recommendation 1: Within the next six months the WFP CO should develop a Theory of Change (as this was not developed at the start of the programme), and a joint workplan for the activities, including a monitoring workplan, in close collaboration with the Ministry of Education, Youth and Sport, and other relevant ministries (Health, Interior, Social Affairs, Veteran and Youth Rehabilitation, and Economy and Finance), as well as other implementing partners. This should include a transparent communication mechanism to regularly share challenges, experiences and lessons learnt with all levels and include regular capacity needs assessments in the annual monitoring and progress reports; Recommendation 2: Within the next 12 months, the WFP Country Office should, using an assigned focal point, design a coherent and structured capacity building plan based on thorough needs mapping, a review of the current situation, with clear key roles, responsibilities, time targets and indicators that measure how capacity is improved and enhanced over time. Government capacity building in programme monitoring should be a priority in this plan. Recommendation 3: The WFP Country Office, the Ministry of Education, Youth and Sport and other ministries should consider incorporating specific indicators measuring the progress of the transition in the monitoring plan to enable careful monitoring and adjustments during the transition process. As the transition process has already started, this should be prioritized within the next six months. Recommendation 4: The WFP Country Office, together with USDA, and in consultation with the Ministry of Women’s Affairs, should develop specific gender indicators to enable integration of gender considerations into the programme design and activities, and ensure that the programme contributes to gender equity and women’s empowerment. The Country Office should ensure these indicators measure the progress in transformative changes, for not only for women and girls, but also for men and boys. To be effective, this should be completed within six months. Recommendation 5: The WFP Country Office should assess whether the targeting of the schools can or should be adjusted. An assessment of the needs of each school will lead to effective and efficient programme implementation and help prepare the schools for the transition process to a nationally-owned and -implemented school feeding programme. This should be done within the next 12 months. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 1 |Page 1. Introduction 1. This report presents findings from a baseline study of the United States Department of Agriculture’s (USDA) McGovern-Dole support to the World Food Programme (WFP) School Feeding (SF) activities in Cambodia (USDA Project Agreement number: FFE-442- 2019-013-00). This baseline is the first phase of a series and will be followed by a midline evaluation in 2021 and an endline evaluation in 2023. This decentralized activity evaluation series was commissioned by the WFP Cambodia Country Office (CO), based on the Terms of Reference (ToR) in Annex 1. The Evaluation Matrix can be found in Annex 2. 2. As a grant condition, the baseline was to be implemented in the first quarter of 2020, prior to the start of programme activities. However, the COVID-19 pandemic led to nation￾wide school closures between March and September 2020. Data were therefore collected at almost the same time as schools were gradually reopening, with programme activities to resume imminently. 3. This study was carried out by an evaluation team (ET) of independent consultants from the KonTerra Group in partnership with Indochina Research Limited (IRL), a local Cambodian research company. It was conducted concurrently with the baseline study for the USDA-supported Local and Regional Food Aid Procurement (LRP) programme,2 which will produce a complementary set of reports. Country-level data collection was done between August and October 2020, partially remotely due to COVID-19 limitations. 4. Through the LRP, WFP is utilizing a USDA grant fund to complete the on-going transition of its School Meal Programme (SMP) model to a nationally-led and -owned school feeding programme where all commodities are procured locally, and to complement activities under the McGovern-Dole programme. 5. The evaluation series serves the two mutually reinforcing objectives of accountability (identification of results and performance) and learning (factors affecting the results, as well as highlighting lessons for future activities) and will provide evidence-based findings to inform operational and strategic decision-making and incorporate lessons into lesson-sharing systems. The purpose of this study was to determine the baseline numbers and structures, against which progress regarding effectiveness and efficiency will be measured in the subsequent rounds of this evaluation series. 6. The internal and external stakeholders, and the principal users of the results, include WFP CO and the MoEYS as well as other ministries at national and subnational levels, cooperating partners, other United Nations agencies, USDA and other donors (see also Annex 3). 1.1. Overview of the Evaluation Subject 7. The USDA McGovern-Dole School Feeding Programme agreement in Cambodia was signed on 27 September 2019 and amended on 18 December 2019. The start date of the programme was 01 November 2019 with an end date of 30 October 2023. This programme is an extension of previous USDA grants for 2013-2016 and 2017-2019 (see Annex 4 for a timeline). This round of the programme focuses on the preparation of the handover to the Government, by increasing the focus on local production and procurement (see below). 8. The McGovern-Dole programme is implemented in 18 districts across three central provinces of Cambodia (Siem Reap (SRP), Kampong Thom (KTM), and Kampong Chhnang (KCG)) (see country map in Annex 5). Targeting of provinces, districts and schools for all WFP school feeding programmes is based on findings from Vulnerability Assessment Mapping (VAM) conducted by WFP, the Royal Government of Cambodia and development partners. 2 USDA Programme LRP441-2019-011-00. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 2 |Page Planned Outputs and Beneficiaries 9. The USDA McGovern-Dole School Feeding Programme was designed to provide one school meal per day to 151,700 pre-primary and primary school students, initially in 599 schools in three provinces. According to the phase out plan, 77 were only included in the plan for school year (SY) 2019-20, after which they were handed over to the Government, leaving 522 for the SY 2020-2021.3, 4 A breakdown of the schools and the students, disaggregated by gender and province for SY 2020-2021, is provided in Table 1. Table 1: Planned numbers of schools and beneficiaries, 2020-2021 Province Number of targeted schools (2020-2021) Number of girls Number of boys Kampong Chhnang (4 districts) 97 12,096 15,269 Kampong Thom (6 districts) 157 15,172 15,596 Siem Reap (8 districts) 268 35,272 35,697 Total 522 62,540 66,562 Logic Model: Result frameworks and foundational results 10. The McGovern-Dole supported school meals programme (SMP) supports Strategic Objective (SO)1 of WFP Cambodia’s Country Strategic Plan (CSP): ‘Vulnerable communities in Cambodia have access to nutritious, safe, diverse, convenient, affordable and preferred foods by 2025’.5 It is also aligned with WFP’s Global Strategic Plan 2017-2021, supporting SO1 (end hunger by protecting access to food), SO2 (improve nutrition), SO3 (achieve food security) and SO4 (support the implementation of the Sustainable Development Goals (SDGs)).6 11. The McGovern-Dole programme is based on two Results Frameworks (RFs) (shown in Annex 6):  Literacy Results Framework (RF1): The Strategic Objective of this framework is the ‘Improved literacy of school-age children’ (SO1). Achievement of SO1 is dependent upon the achievement of three ‘result streams’ (outcomes) related to improved quality of literacy instruction, improved attentiveness, and improved pupil attendance.  Health and Dietary Practices Results Framework (RF2): This covers the SO2 ‘Increased Use of Health and Dietary Practices’ in the school environment. The achievement of SO2 will support the outcomes of reduced health-related absences in RF1. The RFs are clear and relevant for the programme and the indicators and assumptions remain applicable. The indicators are based on the USDA McGovern-Dole standard indicators.7 The one caveat (as discussed in para 67) is the lack of indicators and activities that are gender specific and promote gender equality and inclusivity. 12. Figure 1 below displays the activities and objectives identified for the McGovern￾Dole programme support, and also indicates (in the green boxes) its complementarity with the LRP programme. The latter is the subject of a synchronous but separate baseline study (and later evaluation). 3 Programme Agreement FFE-442-2019/013-00-A 4 ToR, page 6; 329 schools with 42,800 boys and 41,000 girls to receive SMP only, and 270 schools with 33,800 and 31,200 girls to receive a HGSF-hybrid programme. The performance indicators also refer to 599 schools being reached as a result of USDA assistance. The ET is aware that the numbers of beneficiaries do not add up. 5 WFP Cambodia: Annual Country Report 2019. 6 WFP Strategic Plan 2017-2021. 7 https://www.fas.usda.gov/sites/default/files/2019-06/fad_indicator_handbook_feb_2019_0.pdf McGovern-Dole Programme Baseline Report – Cambodia – January 2021 3 |Page Figure 1: Linkage between McGovern-Dole and LRP programmes Source: Evaluation Team 13. To increase the likelihood of achieving the SOs and sustainability of results once the McGovern-Dole assistance ends, USDA identified a set of foundational results common to both RFs. These are: a) increased capacity of Government institutions: b) improved policy and regulatory framework; c) increased Government support; and d) improved engagement of local organizations and community groups (see Annex 6). 14. A set of standard USDA performance indicators showing annual targets for each activity for the whole duration of the programme has been attached to the agreement between the USDA and WFP. These performance indicators provide the information against which the midline and endline evaluations will measure progress towards the achievement of the programme's objectives (Annex 7). Key Activities 15. A range of activities was designed to support the results under both SOs, including the provision of a cooked breakfast at school, development of books as part of a learning package for Grade 2 students, teacher trainings and mentoring, capacity building, improving water, sanitation and school-feeding related infrastructure at schools, as well as the promotion of health and nutrition at schools and in communities.8 Details are provided in Annex 8. Main Partners 16. The McGovern-Dole supported programme is part of the wider WFP school feeding programme, and is implemented in partnership with several ministries of the Royal Government of Cambodia, of which the Ministry of Education, Youth and Sport (MoEYS) is 8 No amendments to the initial programme design were made, except for a reconfirmation of beneficiary numbers. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 4 |Page the key partner.9 In addition, WFP works closely with the Provincial and District Offices of Education (PoE and DoE). Other ministries involved include the Ministries of Health (MoH), Interior (MoI), Economy and Finance (MEF), Social Affairs, Veteran and Youth Rehabilitation (MoSAVY) and the Council for Agricultural and Rural Development (CARD). 17. In addition, the SMP is implemented in partnership with three non-Governmental organizations (NGOs) - Plan International (Plan), World Vision (WV) and World Education Inc. (WEI). United Nations agencies such as UNICEF, UNESCO and the Food and Agriculture Organization of the United Nations (FAO), provide complementary support to the SMP in their own technical areas – education, agriculture (school gardens), transition to Government ownership, and health and nutrition. Resource requirements and funding situation 18. The donor for the McGovern-Dole programme is USDA through the Foreign Agricultural Service (FAS), for a total budget of almost US$19 million over the three-year period. As per the agreement with USDA, no cost share is planned for this project although other donors contribute to the entirety of WFP’s school feeding programme.10 This budget will be used mainly for activities (70.5 percent), with the rest for administrative costs (11.7 percent), Internal Transportation, Storage and Handling (9.7 percent) and total indirect costs (8.1 percent). The proposed budgetary breakdown indicates 29.9 percent for promoting improved health and improve safe food preparation and storage, 19.4 percent for promoting improved literacy, 15.8 percent for promoting improved nutrition, 13.2 percent for capacity building, 12.8 percent to food distribution and 8.9 percent to Monitoring and Evaluation.11 19. Under this agreement, a total of 6,280 metric tonnes (mt) of commodities, consisting of three annual batches of vegetable oil (a total of 260 mt) and fortified rice (6,020 mt), are planned to be provided as in-kind support (Annex 9).12 Yellow split peas and canned fish are provided by other donors, while salt and firewood are contributed by the community. Fresh vegetables from school gardens and community contributions are used to complete the meals. Previous evaluations and recommendations 20. The 2019 McGovern-Dole endline evaluation13 found the activities to be well implemented and well aligned with the SDGs, and with WFP and Government targets and policies. Although the School Feeding Roadmap (2015) towards national ownership aspired to a transition by 2021, the endline evaluation recommended extending the timeframe to 2023 for a stronger and more sustainable process, and highlighted limitations with the MoEYS’ monitoring capacity. This was the rationale for having this new round of the McGovern-Dole supported programme until 2023. The 2019 endline recommendations are provided in Annex 10. 1.2. Context 21. The Royal Government of Cambodia has established impressive economic growth over the past 20 years, bringing the country to lower middle-income status in 2016 with a Gross Domestic Product (GDP) per capita of US$1,205 in 2019.14 However, this growth rate 9 In particular its Primary Education Department/the Scholarship Office (which supports the implementation and monitoring of the programme), School Health Department, and Policy Department, as well as the Teacher Training Department and the Curriculum Department (which are involved in the literacy component). 10 McGovern-Dole Project Agreement. 11 While the total budget was taken from the Project Agreement, a breakdown of the budget by activity was not available to the ET. The information provided here was taken from the proposal. Because the total budget amount in the proposal is different from that in the project agreement, the ET provides the breakdown only in percentages, rather than in US$. 12 ibid 13 Dunn et al (2020). Endline Evaluation of United States Department of Agriculture (USDA) McGovern-Dole Grant Food for Education Programme for WFP Cambodia FY 2017-2019 14 https://tradingeconomics.com/cambodia/gdp McGovern-Dole Programme Baseline Report – Cambodia – January 2021 5 |Page has recently been seriously impacted by the global COVID-19 pandemic, with the Asian Development Bank projecting that the rate will contract by 4.0 percent in 2020, and revive to 5.9 percent in 2021.15 22. In 2019, Cambodia ranked 144 in the global Human Development Index (HDI) out of 189 countries reporting.16 Life expectancy at birth and the education index17 are also on a positive trend, although both health and education indicators remain lower than in neighbouring countries. 23. Demographics and poverty: The country has a population of over 15.3 million (2019)18 that is predominantly young with a median age of 25 years. Nearly three-quarters of the population resides in rural areas where approximately 90 percent of the country’s poor live.19 The poverty indicators declined significantly from 47.8 percent in 200720 to 12.9 percent in 2018.21 However, the current economic situation caused by COVID-19 threatens to thrust many into poverty, especially those (previously) working in agriculture, tourism and manufacturing. There remains a very limited social safety net system in the country. 24. Food security and nutrition: The national objectives set for Cambodia for the Millennium Development Goals (MDGs) were not met.22 The new SDG indicators covering undernutrition and dietary diversity (Goal 2) suggest that 14 percent of households (HHs) continue to consume less than the minimum dietary energy requirement, while 11.6 percent have inadequate dietary diversity.23 Undernutrition is most prevalent in rural areas. The 2014 Demographic Health Survey reported that 32 percent of children under the age of five years were stunted, 10 percent were wasted, and 24 percent were underweight.24 In addition, 14 percent of women aged 15-49 were reported to be underweight. 25 On the upside, the Royal Government of Cambodia is taking food security and malnutrition seriously, and CARD has developed a food security and nutrition framework.26 25. Gender inequality persists in Cambodia, ranked 114 out of the 162 countries in the Gender Inequality Index (GII = 0.474)(2018)27 and ranked 89 out of 153 countries in the Global Gender Gap Index (GGI) 2020.28 The United Nations Office of the High Commissioner for Human Rights in Cambodia reported in 2015 that 70 percent of women in employment were engaged at lower levels and on less pay than men,29 with estimates that on average, women are paid 30 percent less for commensurate work.30 Women are also underrepresented in the public sector where 77 percent of employees and 85 percent of decision-making positions are occupied by men.31 Nationally, 25 percent of women were illiterate compared to 13.5 percent of men (2015).32 15 https://www.adb.org/countries/cambodia/economy 16 Human Development Report 2020. UNDP. http://report.hdr.undp.org/ 17 Education Index calculated using Mean Years of Schooling & Expected Years of Schooling http://hdr.undp.org/en/content/education-index 18 https://www.nis.gov.kh/nis/Census2019/Provisional%20Population%20Census%202019_English_FINAL.pdf (page 6) 19 https://www.worldbank.org/en/country/cambodia/overview 20 https://www.worldbank.org/en/country/cambodia/overview 21 https://www.adb.org/countries/cambodia/poverty 22 Cambodia had an objective of reducing the prevalence of undernourished people to <10%. 23 Cambodia Socioeconomic Survey, 2014, National Institute of Statistics, Ministry of Planning; Available at: https://www.nis.gov.kh/nis/CSES/Final%20Report%20CSES%202014.pdf 24 https://dhsprogram.com/pubs/pdf/FR312/FR312.pdf 25Cambodia Socioeconomic Survey, 2014, National Institute of Statistics, Ministry of Planning; Available at: https://www.nis.gov.kh/nis/CSES/Final%20Report%20CSES%202014.pdf 26 National Strategy for Food Security and Nutrition (NSFSN), 2019-2023. 27 http://hdr.undp.org/sites/all/themes/hdr_theme/country-notes/KHM.pdf (page 6); Ratio of women to men HDI values. Gender Development Index scores range from 0 to 1 with a score of 1 indicating equality between men and women. 28 https://reliefweb.int/sites/reliefweb.int/files/resources/WEF_GGGR_2020.pdf (page 8) 29https://cambodia.ohchr.org/en/issues/gender-equality-and-human-rights 30 CSO report on Cambodian gender issues. 2009 31 https://cambodia.ohchr.org/en/issues/gender-equality-and-human-rights 32 https://www.cia.gov/library/publications/the-world-factbook/geos/cb.html McGovern-Dole Programme Baseline Report – Cambodia – January 2021 6 |Page 26. However, Cambodia’s Gender Development Index (GDI) has improved in recent years,33 moving from the lowest category to Group 4 (GDI 0.922 in 2019).34 Increasingly, more women are creating independent sources of income within the private sector through urban migration for work or by starting small businesses, particularly in the clothing sector.35 The Government is working on reducing violence against women as shown by its National Action Plan to Prevent Violence Against Women (NAPVAW), which includes school-based prevention interventions to change social norms around violence among young women.36 The SDG targets on gender equality in education and literacy (Goal 5) have been achieved at the primary school level.37 27. Gender disparity is higher for secondary education as only 40 percent of girls complete secondary schooling. Girls in rural areas are at higher risk than boys of dropout due to poverty, to care for younger siblings, to help their parents or to move to urban areas to work. 28. Cambodia has been improving primary education programmes and reducing gender disparity in education for many years. The net primary school enrolment rate (NER) for public schools has been above 90 percent for at least the last decade, with an average of 91.5 percent (91.5 for girls, 92 for boys) in SY 2018/2019.38 This is a significant improvement from 81 percent in 2001. However, repetition in primary years is still high, particularly in the first grade of schooling after children have not had sufficient education in their early years. School dropout is most problematic at the end of the primary school as students transition to secondary school, and boys are more likely to drop out than girls, as are students in rural areas.39 Girls in rural areas drop out mainly due to severe poverty and end up caring for younger siblings, working alongside their parents in the rice fields, or moving to urban centres to find work, while the boys are more likely to leave school in order to find employment.40 In addition, when parents move for better job opportunities, they often take the children with them, or leave them with grandparents who may not be able to ensure their school attendance. Although most of the national education performance indicators do not differ substantially between boys and girls in primary school, completion rates among girls were significantly higher than among boys in SY 2018/19 (90.4 vs 82.3).41 Other factors that influence parents to remove their children from school before completion include a lack of basic school infrastructure, making schools an unsuitable learning environment, particularly for adolescent girls and disabled children, and poor teaching quality due to a lack of sufficiently qualified teachers. Key education indicators in the McGovern-Dole targeted provinces and districts are provided in Annex 11. 29. Under the Education Strategic Plan (ESP) 2014-2018, the MoEYS developed a programme to support the improvement of Early Grade Reading Assessment (EGRA), in partnership with USAID, WFP (through support to WEI), UNICEF, UNESCO and a number of national and international implementing development actors. The current ESP 2019- 2023 stresses that the MoEYS must further support students to improve reading and writing in Khmer at primary education level. As part of its main strategy, MoEYS implements training and mentoring for primary teachers on teaching and learning methodologies for early grade reading and mathematics. 33 Human Development Report 2020. UNDP. http://report.hdr.undp.org/ 34 Gender Development Index groups: Countries are divided into five groups by absolute deviation from gender parity in HDI values. Group 5 comprises countries with low equality in HDI achievements between women and men (absolute deviation from gender parity of more than 10 percent). 35 Commune Database 2013, Ministry of Planning 36 NAPVAW 2014-2018. Available at https://drive.google.com/file/d/0By4DrHxo3C94Skx4VmZUN2JoMmc/view?pli=1 37 https://sustainabledevelopment.un.org/content/documents/23603Cambodia_VNR_PublishingHLPF.pdf 38 MoEYS, Education Statistics & Indicators 2018/2019. According to the Education Congress Report 2019, the overall national NER for public and private schools is 98 percent for total and 98.5 percent for girls. 39 ibid 40 https://www.unicef.org/cambodia/reports/why-are-boys-leaving-lower-secondary-school-early-cambodia. 41 MoEYS. Education Congress Report 2019 McGovern-Dole Programme Baseline Report – Cambodia – January 2021 7 |Page 30. Under a newly endorsed national school health policy, the school health department at MoEYS is collaborating with WFP and the MoH to support food safety and health in schools with school feeding programmes. 31. In Cambodia, short-term hunger is a key factor affecting educational results, such as literacy, attendance and concentration in schools. The McGovern-Dole supported programme aims to encourage students’ enrolment, attendance and completion of their primary education, reduce short-term hunger and improve the children’s concentration in the classroom. The SMP also invests in community and school-based partnerships aimed at providing an incentive for parents so that boys and girls continue education beyond the primary level.42 32. The first Cambodian patient with COVID-19 was diagnosed on 03 March 2020 and at the time of finalising of this report, a total of 463 cases had been reported.43 While the caseload is much lower than in most other countries, the economic impact of the pandemic affects many, due to the national economy’s heavy reliance on international sources of income, through tourism and trade.44 There are indications of increased poverty due to loss of livelihoods - in particular, the tourism, agriculture, manufacturing, construction and informal sectors are all hard hit.45 In the garment industry, for instance, tens of thousands have already been laid off and hundreds of thousands run the risk of (temporarily or permanently) losing their livelihoods. The full impact on food insecurity is as yet unknown and studies are ongoing, but pockets of food insecurity have been found, increasing the relevance of the school meal as a social safety net. 33. The Royal Government of Cambodia reacted in a timely manner by adopting several policy measures to control the outbreak including: (i) the establishment of task forces by the MEF to control the price and supply of strategic goods and plan for budget policy on financing and social assistance, (ii) health preventive measures, and (iii) border control measures.46 Government operational budgets across the board were cut by up to 50 percent - with the exception of the MoH and MoEYS. School meal food supplies and funds were reallocated to provide THR for children from vulnerable households during the school closure, in addition to cash distributions to vulnerable households through the ID-Poor system.47 34. The most recent WFP food price updates for Cambodia (May 2020) reported that most food prices have remained fairly stable in the COVID-19 context so far.48,49 However, WFP concluded that the main impact of the COVID-19 pandemic on food security is more likely to be related to a restricted ability to purchase a variety of nutritious foods due to a loss of livelihoods and income.50 This is consistent with broader concerns that food security 42 Although recent MoEYS data show drop-out rates are higher and completion rates are lower among boys, and this needs attention, this does not mean reducing attention for girls’ continued education is warranted 43 Source: https://www.worldometers.info/coronavirus/country/cambodia/ (accessed 29 January 2021) 44 World Bank, ‘Cambodia Economic Update November 2018: Recent Economic Developments and Outlook report mentioned that “Over the last decade, Cambodia ranked among the countries that attracted the most FDI in relation to is size worldwide, averaging an FDI inflow of 7.8 per cent of GDP between 2000 and 2016. Currently, around 90 per cent of (non-financial sector) inflows originate from China.” (MacAuslan I et al. Country-led Evaluation of the National Education Scholarship Programmes of the Ministry of Education, Youth and Sports in Cambodia (2015-2018) Final Report – Volume II August 2018 – March 2019) 45 Angkor Research. Covid-19 Economic Impact Study. Headline Results Brief Round 1 Data Collection. May 2020 46 Future Forum and Angkor Research Cambodia are undertaking an economic impact study on COVID-19, specifically on the economic impacts on Cambodian citizens expected to be hit the hardest by the pandemic (farmers, wage workers and families with micro enterprises). Preliminary data show that the majority of the decrease in household income from farming can be attributed to Siem Reap province (followed by Phnom Penh), where mean income dropped from US$351.44 in January to US$186.01 in April 2020. The first round of data collection has also identified the emergence of a general trend in the country whereby households are substituting the economic returns of their produce for self-consumption. Farming profits in Siem Reap fell by 53 percent. 47 To date, two rounds of US$25 million each have been distributed to 500,000 poor households. 48 Food prices including rice, fish, and vegetable oil remained fairly stable but that there was a spike in prices for vegetables (up to 60%), eggs (14.5%), and pork (5.7%) 49 https://docs.wfp.org/api/documents/WFP-0000116617/download/?_ga=2.260110477.1491633889.1591985730-285171340. 1424459362 50 ibid McGovern-Dole Programme Baseline Report – Cambodia – January 2021 8 |Page could be impacted by restricted or disrupted food commodities,51 especially fresh vegetables, 50 percent of which are imported.52 35. All schools in Cambodia were closed in mid-March and were only allowed to reopen – within strict guidelines – from early September 2020. Online learning poses many challenges, in particular for students from vulnerable households, many of whom have difficulty accessing online resources. This will likely impact learning outcomes, including for literacy, and will impact poor students disproportionately. In addition, many activities for the school feeding programme were halted during the school closure, and are yet to fully resume after the initial reopening of the schools in September.53 36. At the global level, WFP, FAO and UNICEF have developed a joint note54 to provide guidance to national and local authorities, school administrators and staff and implementing partners on short-term measures to support and adapt school feeding programmes to safeguard the food security and nutritional status of school-aged children during the COVID-19 pandemic. The recommendations include a specific section around homegrown school feeding. 37. A number of other United Nations and international organizations are supporting the MoEYS with programmes complementary to WFP’s SFP. These include UNESCO with a number of initiatives to improve the quality of teaching for better student learning.55 Together with MoEYS, USAID is implementing its education strategy (2016-21)56 through their partners, Kampuchea Action to Promote Education (KAPE) and WEI, focusing on improving early grade reading skills. With support from WFP, WEI has trained MoEYS in delivering EGR. Similarly, WFP school feeding aligns with the Global Partnership for Education (GPE) grant which combines complementary efforts from MoEYS, UNESCO and UNICEF, as well as other education development partners including the European Union, Sweden and USAID.57 More details are provided in the findings section (paras 81-865). 1.3. Evaluation Methodology and Limitations Approach 38. To be able to critically and objectively review whether the programme is likely to meet its stated goals and objectives, the baseline study assessed: (i) the current status of key performance indicators related to students’ EGR capacity, primary schooling (enrolment, attendance, retention, attentiveness), food security, hunger, water and sanitation facilities and health and nutrition, taking into account that WFP has been running a school feeding programme in some of the same areas for several years; (ii) remaining progress to be made to achieve the objectives in literacy and health and nutrition practices as outlined by USDA; (iii) the factors that need further investment to ensure that objectives are met; and (iv) what further investments need to be made to ensure the programme is sustainable after USDA assistance has ceased. 39. Representatives from each of the groups of stakeholders (see para 6), as well direct and indirect beneficiaries at the school and community levels, have been involved in key informant interviews (KIIs) and/or focus group discussions (FGDs) during the qualitative field mission and/or included as survey interviewees during the quantitative data collection. 51 https://www.ifad.org/en/web/latest/blog/asset/41875817 52 https://www.khmertimeskh.com/707080/covid-19-lockdown-highlights-cambodias-dangerous-reliance-on-imported-food/ 53 Subsequent COVID-19 related school closures in mid-November 2020 related only to schools in Phnom Penh and Kandal province (around the city); though later closures from 30 November covered the whole country once again. 54 WFP, FAO and UNICEF. Mitigating the effects of the COVID-19 pandemic on food and nutrition of schoolchildren. Interim Guidance Note. 7 April 2020. 55 UNESCO Phnom Penh website: https://en.unesco.org/fieldoffice/phnompenh 56 MoEYS, Mid-Term Review Report in 2016 of the Education Strategic Plan 2014-2018 and Projection to 2020, November 2016. https://www.globalpartnership.org/sites/default/files/2016-11-cambodia-mid-term-review-education-sector-plan.pdf 57 https://www.globalpartnership.org/news/global-partnership-education-approves-us206-million-grant-strengthen-cambodias￾education-system McGovern-Dole Programme Baseline Report – Cambodia – January 2021 9 |Page A more detailed stakeholder analysis is presented in Annex 3 and a detailed list of stakeholders interviewed can be found in Annex 12. Evaluation questions and criteria 40. The overall evaluation series aims to answer the evaluation questions (see Annex 2.1) as listed in the ToR. However, for this baseline study the ET focused on the following main evaluation criteria and questions: (i) Relevance: Is the programme design relevant to the needs of the target population (men and women, boys and girls) and communities? (ii) Coherence: To what extent is the programme aligned with Government policies, with WFP’s and wider United Nations strategies, policies and guidance? (iii) Sustainability: Is there any evidence that activities are likely to be sustained or scaled up beyond the programme’s life? 41. This baseline study used a theory-based, participatory, and gender-responsive evaluation approach employing a quasi-experimental case-control design.Quantitative and qualitative primary data were collected and analyzed and secondary data and documents provided by the CO were reviewed (see the list of documents reviewed in Annex 13). Data collection and analysis methods 42. The primary data collection was done between August and October 2020 via quantitative surveys of schools (principals and teachers), using tools and methodologies adjusted from those applied during the previous McGovern-Dole Food for Education evaluation (2017-2019). Quantitative survey data collection was done by IRL in close coordination with the ET. Details of the data collection methodology are provided in Annex 14 and the tools used are presented in Annex 15. 43. Quantitative analysis of cleaned and recoded data was done by the ET using SPSS and R software wherever appropriate. Descriptive statistics were performed for case-control comparisons. Wherever appropriate, difference in means were tested using an independent t-test. 44. Qualitative data were gathered through key informant interviews (KIIs) and focus group discussions (FGDs) with a range of key stakeholders at national (81 in total), and sub￾national and at school/community level (247 in total) (see Annex 16 for the interview guides). Due to the COVID-19 related travel restrictions, the qualitative interviews were conducted principally by the national ET members, accompanied by an external expert recruited for this phase of the work.58 The international ET members conducted remote interviews with stakeholders at the national level as well as international stakeholders. 45. All evidence collected has been verified and corroborated through systematic triangulation by comparing information collected by different methods, evaluators and sources. The ET also ensured data integrity and factual accuracy with regular team discussions. Triangulation of data was conducted using method and source triangulation, including by comparing qualitative and quantitative findings of this baseline study, and by confirming the baseline findings with those of the 2019 endline evaluation, given the short time frame between the two exercises and the fact that only limited activities had taken place in between. Team members also used evaluator triangulation during evening and weekend discussions and whilst travelling together. The ET members met (online) for a weekend of collective data analysis to identify themes and patterns to illuminate the findings. Initial findings and conclusions were shared through presentations to the WFP CO and RB teams at the end of the field mission, and their feedback obtained. 58 A foreign external expert, already resident in Cambodia, was recruited to work with the national team members during the field visits. He had a well-developed knowledge and understanding of Cambodia, its culture, official structures etc., previous evaluation experience a background with humanitarian agencies, and spoke and understood fluent Khmer. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 10 |Page Gender Equality and Women’s Empowerment (GEWE) Considerations 46. The ET analysed the extent to which GEWE objectives and mainstreaming principles were included in the intervention design and aligned with the SDGs and other system-wide commitments enshrining gender rights. Gender analysis assessed the extent to which different voices, vulnerabilities, capacities and priorities of women, men, girls and boys were reflected in the programme cycle and how these distinct groups might have benefitted from the programme socially and materially ensuring equitable access and power sharing. 47. The ET ensured that the data collection process included active participation of women and men to inform a better understanding of the programme from their distinct perspectives at school, household and Government/WFP levels. Site mapping 48. The school list for WFP’s school feeding activities in the three provinces totals 522 institutions. By the end of the programme cycle in 2023, 141 of these schools will have been handed over to the Government and 79 schools will have been transferred to full HGSF funded by other donors. Thus, the remaining 302 schools will be fully supported by the USDA grants: HGSF-Hybrid (McGovern-Dole+LRP - 104 schools) and HGSF-Hybrid (McGovern-Dole+other donors - 198 schools). To enable evaluation of the USDA-supported programme over the entire programme cycle, the sample for intervention schools was drawn from these 302 institutions (as highlighted in Table 2 below).59 Table 2: USDA Food for Education schools in school years 2020-2023 Status of SMP activities as planned by SY 2022-2023 Programme schools SY 2020-2021 HGSF-Hybrid (USDA +LRP) HGSF-Hybrid USDA (non￾LRP) and other donors Total Handover to Government 47 94 141 Transferred to full HGSF (other donors) 12 67 79 HGSF-Hybrid (USDA+LRP) 104 198 302 Total 163 359 522 Timeline 49. The quantitative data was collected between 31 August and 15 September 2020 in all schools and households. The full qualitative data collection phase took place over four weeks (21 September to 16 October). After baseline data collection, the ET conducted internal and external debriefings (19 and 20 October) to share preliminary findings with WFP staff, partners and key stakeholders, and used these sessions as an opportunity to discuss and validate results. Additional details are provided in the Fieldwork Timeline (Annex 17). Limitations, risks and mitigation strategies 50. Due to COVID-19, the ET (in close consultation with the WFP CO and in reference to WFP’s Technical Note60) anticipated and planned for various limitations and risks in implementing this delayed baseline study, and the work ultimately was completed satisfactorily without issue. To minimize the risks of exposure to COVID-19, the ET paid particular attention to health guidelines in force. For field level data collection, IRL designed safety protocols for field surveys, staff training and face-to-face KIIs and FGDs. 51. Most quantitative data (originally planned for Q1 of 2020) was collected when schools were still closed. This limitation was mitigated as much as possible by adjusting the timeframe of the questionnaires and relying on some secondary data. Memory bias caused by the long timeframe of data collection was mitigated by validating the baseline data through comparison with the 2019 endline data. Additional relevant questions were added 59 Further sampling details are given in para 88 below, footnote 87, and in Annex 14. 60 WFP OEV. Technical Note for Planning and Conducting Evaluations during COVID-19. April 2020. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 11 |Page into the quantitative questionnaires and the reference time-frame of data collection for some indicators was adjusted.61 52. It remained possible (with some advance planning) to have direct contact with communities and meetings with school and Government officials, with the Cambodian enumerators and evaluators being physically present and their international colleagues being online for many of the interviews, plus regular follow-up telephone debriefings. The ET also worked with WFP to establish a broader baseline that compares the 2019 data with the current situation. 53. While it is unclear when the pandemic will be under control, it is important to note that until that happens, the implementation of the programme (and potentially the implementation of the midterm and endline evaluation phases) will be impacted. However, estimating the extent of these impacts would be pure conjecture. 54. The absence of gender and inclusivity indicators and activities in the design of the McGovern-Dole programme, and hence in the RFs, makes it impossible to assess these aspects during the evaluation. While the ET has made a conscious effort to include questions related to gender in all interviews, many informants stated they thought these questions were not relevant to the programme. Ethical considerations 55. Interviews were carried out in accordance with UNEG’s 2008 Ethical Guidelines for Evaluation, notably to ensure informed consent, anonymity and confidentiality. The ET used UNICEF’s Procedure for Ethical Standards in research, evaluation, data collection and analysis related to child consultations.62 The field notes are kept confidential and will not be turned over to public or private agencies. In addition, steps were taken to ensure that men, women, boys and girls felt that interviews were conducted in appropriate locations so that they were able to freely express their views and concerns without fear of reprisal. Schools were notified in advance that they would be part of the survey. Permission to interview children at school was sought from their teachers, and from parents during household interviews. All identifiers were removed, so no data can be traced to any individual, after which the clean dataset was uploaded to a separate folder only accessible to selected ET members and WFP CO staff. No effects of the intervention on human rights or gender equality were identified. 2. Baseline Findings 56. The findings of the baseline study and the evidence to substantiate them are presented below. They are structured as a response to each evaluation question in turn through the evaluation criteria lenses of relevance, coherence and sustainability. All data presented in the following Tables and Figures is drawn from the results of the quantitative survey unless indicated otherwise. Detailed findings disaggregated by province are presented in Annex 18. 2.1. Evaluation Question 1: How appropriate is the design of the McGovern-Dole programme to the context and to what extent is it aligned with the needs of women, men, girls and boys and with the priority needs of schools, and the Government? 57. Design appropriate to the context and needs: The McGovern-Dole programme design is considered appropriate to the context and relevant to the needs of all categories of beneficiaries. The programme aims to improve literacy (SO1), which is still low in Cambodia, while the school meal component (1.2.1.1/1.3.1.1 in the McGovern-Dole 61 For example, questions related to attendance, attentiveness and hunger were specifically related to the period prior to the school closure in March 2020. 62 https://www.unicef.org/supply/files/ATTACHMENT_IV-UNICEF_Procedure_for_Ethical_Standards.PDF McGovern-Dole Programme Baseline Report – Cambodia – January 2021 12 |Page Results Framework; see Annex 5) functions as a social safety net. Without exception, teachers, school managers, parents and Government counterparts at all levels highly valued the programme and acknowledged its benefits. 58. The programme targets schools with a large proportion of poor households. Parents of children at the targeted schools stated that the school meal programme is likely to relieve their burden (as it did during the previous phases of the programme), by reducing both their expenditure on food and the time they need to spend to prepare breakfast for their child. This extra time can be used for household chores and income generating activities. 59. The main benefit of the programme in the eyes of the parents is that their children receive food, which makes them healthier and improves their ability to learn. Education￾related benefits were less frequently mentioned by the parents of both boys and girls as summarized in Table 3. Table 3: Benefits of school feeding for the child, as perceived by parents of selected girls and boys Benefits to the child Gets food Healthier Learns better More attentive in class Has more opportunities Illness-related absences reduced No benefit Girls 80% 76% 55% 47% 42% 22% 1.0% Boys 82% 77% 54% 44% 36% 17% 0.5% Total 81% 77% 55% 46% 39% 20% 0.7% 60. In addition, parents perceive multiple benefits to the household in general, in particular saving food, money and time (Table 4). Table 4: Benefits of school feeding for the household, as perceived by parents of selected girls and boys Saves food & money for the household Saves time for parents Saves morning cooking time No benefit Girls 84.3% 74.8% 61.9% 3.4% Boys 81.3% 77.4% 59.1% 1.9% Total 82.8% 76.1% 60.5% 2.6% 61. All officials stated that the programme is relevant to district/provincial priorities to increase school enrolment and attendance rates and to reduce the dropout rate. 62. Targeting of districts and schools. Districts have been strategically targeted using combined indicators of poverty and malnutrition rates and poor education performance using WFP’s Vulnerability Analysis and Mapping (VAM) (see map in Annex 4). Targeting the most vulnerable areas based on findings from VAM was relevant for the McGovern-Dole programme that aims to improve literacy and promote good dietary practices in poor communities. 63. Selection criteria for schools are based on combined indicators of education performance, poverty and malnutrition, and also consider existing infrastructure, capacity of communities to provide support, and from interviews with parents, teachers and school directors. Schools selected for this programme were the same as those identified in 2013 (under the earlier programme grant 2013-2016), subject to annual review. Commune council members reported that not all schools in their commune are included in the programme, the reasons for which they said they did not fully understand.63 Informants also reported that this caused migration of children from schools without the SFP to schools 63 The selection was validated at provincial and district school feeding committee levels and particularly at schools, by consulting with local school feeding committee, in which commune chiefs play a role as the committee chair, on existing infrastructure and the capacity of communities to provide support the programme. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 13 |Page that do provide food. These observations indicate a need to better involve (or at least inform) stakeholders regarding the targeting process. The ET could not establish whether the differences in school conditions were due to the implementation of previous rounds of the McGovern-Dole programme or that they suggest a need to better tailor support to the needs of each school. 64. The ET considers that the programme does target schools with a large proportion of poor households. The average income of USDA-supported households was lower (US$4,521 per annum/US$12.38 per day) than that of the comparison households (US$4,889/US$13.42), as shown in Table 5. The difference in income between the comparison and control groups (USDA-supported households) were not significant, and makes the two groups comparable The vast majority of the income of USDA households comes from work (80.8 percent), with 10.8 percent from sales, while 4.1 percent comes from remittances and 4.3 percent from other sources. Table 5: Annual household income from different sources (in US dollars) Income Source USDA supported schools Comparison schools Work 3,652.44 3,824.26 Sales 490.18 542.02 Remittances 183.15 339 Total 4,520.67 4,889.14 % Work 80.8% 78.2% % Sales 10.8% 11.1% % Remittance 4.1% 7.0% % Others 4.3% 3.7% 65. The average household expenditure of US$16.15 per day in USDA supported households exceeded their daily income, forcing households to use credit to meet their daily expenses (Table 6). Expenditure on food was approximately 50 percent for both groups, and expenditure on education slightly over eight percent of total household spending. Table 6: Household expenditure Household income Household expenditure % of expenditure spent on: Daily (US$) Daily (US$) credit food education USDA supported schools $12.38 $16.15 9.4% 49.5% 8.2% Comparison schools $13.42 $15.87 6.9% 53.5% 9.4% 66. Gender considerations and analysis in the design: The design of the McGovern-Dole programme identifies the equal participation of women, men, girls and boys by providing the support to schools with equal proportions of male and female students. The programme presents an opportunity to promote women’s empowerment by including many female beneficiaries such as parents, teachers and cooks as well as female representation on the local school feeding committees (LSFCs). 67. Although gender is an important programming priority for WFP, this has not been made explicit in the programme’s RFs and the programme design does not include gender￾specific activities, expected results or indicators to measure and analyze gender roles, behaviours and expectations.64 The USDA RFs have no gender-specific outcomes that would explain more clearly what and how gender issues are intended to be addressed through this McGovern-Dole programme intervention. 68. In addition, there appears to be a disconnect between WFP’s school feeding and gender policies: the WFP SF policies lack specific explanations of how gender is to be 64 The only and little developed link that is made in the McGovern-Dole proposal between school gardens and gender is that: "this activity ultimately contributes to achieving SO2. This activity also contributes to a well-nourished population, especially women and children (GFSS Objective 3)". McGovern-Dole Programme Baseline Report – Cambodia – January 2021 14 |Page considered in school feeding programmes. In the 2009 policy, there is only a broad gender outcome of ‘Increased Gender Equality in Education’ as part of its logic model. The 2020 WFP school feeding strategy65 notes that “for all age groups school feeding is an opportunity to address social cohesion, violence and gender issues, among others.” Similarly, no mention of ‘school feeding’ exists in WFP’s Gender Action Plan or in the Update on the Gender Policy (2015–2020).66 69. Although pursuing equal gender participation is to be recognized, the ET found no evidence that the programme was based on any sound analysis67 with which to seek to understand, and resolve with specific measures, the negative gender implications in the lives of the beneficiaries. The programme does not include any particular activities to promote gender equality, while traditional gender roles remain common in Cambodia. For instance, childcare, food preparation and household chores are considered women’s tasks, as was evidenced by the fact that only eight out of 57 participants of the FGDs with (grand)parents were men. Also, most of the visited communities experienced considerable difficulties in understanding that gender mainstreaming involves all members of a society and their needs, not just the women. In the schools, 76 percent of directors are men, while the majority of the teachers are women. 70. Moreover, as has been flagged during previous evaluations,68 the inadequate remuneration of cooks, the majority of whom are women,69 brings the risk of perpetuating gender inequalities and a high turnover, leading to increased needs for training of new cooks. In addition, as the cooks’ monthly incentives come from voluntary community contributions, income inequalities exist between cooks in different communes. However, for schools where the programme has been handed over to the Government, salaries for cooks are integrated into the annual Commune Investment Plan (CIP) and the three-year rolling plan. In contrast, as mentioned by informants in the schools, storekeepers are usually senior male teachers (as it is considered a physically demanding job and a large responsibility) and their turnover is due to quite different reasons, generally through being promoted to another position within the school. 71. The SMP is perceived by implementing partners to lack specific gender mainstreaming activities that would be of crucial importance before handover to the Government. The ET did not see evidence of uniform gender guidelines agreed upon with the partners for this programme. In fact, the partners prefer their own versions and some informants were not aware of the existence of WFP gender guidelines. For instance, Plan implements very gender-specific activities in schools which are not part of the USDA programme but of a complementary ‘Gender Socialization Project’.70 In subsequent rounds of this evaluation series, implementation aspects, such as the extent to which tackling longstanding perceptions, measuring progress in transformative changes for women, men, girls and boys or systematically promoting equity and equality ideals in schools is pursued, will be assessed. 65 WFP. (2020). A chance for every schoolchild: Partnering to scale up School Health and Nutrition for Human Capital. WFP School Feeding Strategy, 2020 – 2030. 66 For instance, the Update on the Gender Policy only exemplifies in a general way that ‘gender issues were integrated into the cost– benefit analysis of school meals in Bolivia (Plurinational State of); in Malawi, gender analyses were undertaken to identify gender-related programming gaps and opportunities in school meals’. 67 WFP's gender toolkit says that "a gender analysis is an examination and interpretation of quantitative data and qualitative information about people from a gender perspective". The toolkit indicates to "apply the analysis to planning, targeting, advocating, monitoring and evaluating to promote transformative changes for women, men, girls and boys". https://gender.manuals.wfp.org/en/ 68 Endline Evaluation report 2019 69 Currently, only 10 out of 555 cooks in Siem Reap schools are men 70 Plan provided a number of trainings on gender to selected teachers from 110 primary schools. Plan foresees that these trainings will enable those schools to “address gender issues in schools through identifying gender stereotypes, prioritizing issues for resolution, regularly following up gender issues in schools and allocating school budget for gender transformation activities, including ending gender￾based violence in schools. Significantly, those schools have developed action plans for promoting gender equality in schools”. Plan International, Quarterly report (April-June 2019). McGovern-Dole Programme Baseline Report – Cambodia – January 2021 15 |Page 72. Alignment with national policies. The ET considers that the programme is strongly aligned with strategic objectives and priorities of the Royal Government of Cambodia. The Government’s targets on improved nutrition, ending stunting and increasing income (by 20 percent for the poorest) have not yet been achieved,71 but with the country’s long-term development vision, the Rectangular Strategy for Growth, Employment, Equity and Efficiency Phase IV (2019–2023),72 the Government offers proof of strong commitment to education and child nutrition, which is viewed as a priority for “sustainable human resource development, economic growth, and social development”. Social protection and safety net programmes (including SMPs) are an increasingly important and recognized means of supporting food security and nutrition improvements by the Royal Government of Cambodia. 73. The transition to a nationally owned school feeding programme is coherent with the 2019-2023 Education Strategic plan, the National Strategy for Food Security and Nutrition (2019-2023), the National School Health Policy and the 2016-2025 National Social Protection Policy Framework (NSPPF). The NSPPF specifically includes a projected scale￾up of scholarships and school meals by 2021 to help achieve the Government’s human development objectives. The collaboration with WFP on school feeding and planned nationwide rollout is an integral part of the Government’s efforts.73 In recent years, with WFP’s support, the MoEYS has indeed been building their school feeding capacity and management systems at national, provincial and district levels. 74. The Government has already confirmed the continued absorption of WFP’s food and cash take-home ration beneficiaries into the national scholarship programme. To support this, since SY 2017/18, WFP has made some strategic adjustments to the use of the HGSF modality that involves a combination of USDA provided and locally procured commodities. This hybrid model is more coherent with the Government’s preference for programming with local commodities and acts as a transition step toward the full HGSF model where food commodities are locally procured. Therefore, WFP’s design of capacity strengthening activities aligns with the Government’s needs and priorities related to HGSF and local procurement in view of handing over the HGSF model to the authorities. 75. The Royal Government of Cambodia has several policies and programmes developed to end hunger, including: i) the National Fast Track Roadmap for Improving Nutrition (2014-2020);74 ii) The National Strategy for Food Security and Nutrition (NSFSN, 2014- 2018); and iii) the National Action Plan for Zero Hunger Challenge in Cambodia (2016- 2025).75 WFP and the Government are therefore in alignment on their commitment to zero hunger addressed mainly through the social protection and education sectors. 76. The Government has also committed to gender inclusivity across all its development actions, emphasized in the National Strategic Plan for Gender Equality and Women’s Empowerment (2019-2023). The HGSF programme contributes to GEWE through provision of equal income opportunities for women and men in producing and supplying food to schools and participating in food preparation.76 The country’s long-term development vision77 also emphasizes a strong commitment to strengthening gender equality and social protection. Furthermore, the Ministry of Women’s Affairs of Cambodia 71 https://sustainabledevelopment.un.org/content/documents/23603Cambodia_VNR_PublishingHLPF.pdf 72 RS-IV 2019–2023 – Rectangle 1 including 1) Improving the quality of education, science and technology; 2) Vocational training; 3) Improving public healthcare and nutrition; and 4) Strengthening gender equality and social protection. 73 http://inndec.com/library/docs/SPPF%20English%20-%20Final%20Ver.pdf 74 https://sustainabledevelopment.un.org/content/documents/23603Cambodia_VNR_PublishingHLPF.pdf 75 http://ocm.gov.kh/ocmwinwin20/wp-content/uploads/2018/12/6-National_Action_Plan_for_the_Zero.pdf 76 More details can be found in the LRP Baseline report 77 The Rectangular Strategy for Growth, Employment, Equity and Efficiency Phase IV (2019–2023) – Rectangle 1 including 1) Improving the quality of education, science and technology; 2) Vocational training; 3) Improving public healthcare and nutrition; and 4) Strengthening gender equality and social protection. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 16 |Page (MoWA), in partnership with UNDP Cambodia and the Swedish Government, is spearheading a comprehensive policy project aimed at narrowing the nation’s gender gap.78 77. Alignment with USDA’s priorities: USDA has been an important partner of WFP in Cambodia since 2001, as shown by the alignment between their policies and strategies. A key element of alignment are new outcome indicators in the WFP Revised Corporate Results Framework (2017-2021), including the USDA SO1 standard outcome indicator ‘percentage of students who, by the end of two grades of primary schooling, demonstrate the ability to read and understand grade level text’. 78. Complementarity with other WFP work in Cambodia: The McGovern-Dole food for education (FFE) programme is part of a larger school feeding programme implemented by the WFP CO in Cambodia. As mentioned, it is complemented by the USDA￾funded LRP programme through its focus on local procurement of the food for the school meals. 79. WFP implements complementary activities with other donor contributions, including from the Korea International Cooperation Agency (KOICA - US$10 million over five years), and various private sector entities (US$1 million per year). With these aligned contributions, WFP ensures complementarity and complete programme delivery.79 In particular, the KOICA funded programme, which started in 2019, is strongly aligned with the McGovern-Dole programme as it provides a similar SFP in schools not covered by USDA. The two programmes share outputs and outcomes, including improved literacy, improved knowledge of health and hygiene practices, and increased knowledge of safe food preparation and storage practices, as well as the provision of school meals.80 Moreover, the CO manages all SFPs with a single team, which ensures coherence of the programmes. 80. With all donor grants, WFP engages in policy development and institutional capacity strengthening activities with a knowledge management and peer-learning approach. WFP also shows alignment in supporting information management systems through field-based monitoring that capture critical programme information across all its grants. For example, WFP has supported Deutsche Gesellschaft für Internationale Zusammenarbeit GmbH (GIZ), UNICEF and MoEYS to improve monitoring of school-based water, sanitation and hygiene (WASH) programmes with mobile data collection and web-based mapping tools, leveraging the WFP’s Platform for Real-Time Information System (PRISM) initiative.81 81. Alignment with United Nations frameworks and other development partners: In seeking the improvement of literacy of school children, the programme fully aligns with SDG 4 (Ensure inclusive and equitable quality education and promote lifelong opportunities for all). The programme is aligned with SDG 2 (End hunger, achieve food security and improve nutrition and promote sustainable agriculture) and with SDG 17 (Strengthening partnerships), considering that school feeding programmes cover multiple sectors and engage various types of stakeholders. 82. The current United Nations Development Assistance Framework (UNDAF) (2019- 2023) in Cambodia pays particular attention to social protection and nutrition to accelerate the successful implementation of the UNDAF and the achievement of the SDGs in the country. The McGovern-Dole programme aligns with Outcome 2 - Social Development, 78 Leading the Way for Gender Equality (LWGE) programme, which began in 2017 and was on course to finalize in December 2020. Three project-pillars set the foundation for gender equality in Cambodia: 1- Institutional reforms through effective policy development and implementation; 2- Behavioural change through media mobilisation and youth engagement; 3- Promotion of women in decision￾making roles. https://www.kh.undp.org/content/cambodia/en/home/blog/leading-the-way-for-gender-equality--cambodia-s￾commitment-to-ad.html 79 LRP proposal. 80 Dunn S. (2020), Baseline assessment of the KOICA supported Home Grown School Feeding programme in Cambodia in Kampong Thom, Kampong Chhnang and Pursat Provinces 2020-2024, Draft Report. 81 McGovern-Dole proposal. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 17 |Page Social Protection and Human Capital, which aims to provide more people with quality social services. 83. UNESCO is currently supporting MoEYS through various education programmes, complementary to WFP’s SFP: Factory Literacy Programme (FLP); Basic Education Equivalency Programme (BEEP); and Strengthening Teacher Education Programmes in Cambodia (STEPCam) to improve the quality of teaching in order to improve student learning.82 84. Together with MoEYS, USAID is implementing its education strategy (2016-21)83 through their partners KAPE and WEI. The focus of the strategy is on improving reading in the early grades. Both WEI andWFP are working with UNICEF and USAID to support early grade reading (EGR) under the National Education Strategy and the Child Friendly School Policy Framework.84 85. With support from WFP, WEI has trained MoEYS in delivering EGR. This work has been complementary to the national EGR package adopted by the USAID-funded All Children Reading Cambodia (ACRC) project, which was developed and tested in schools supported by WFP through the USDA grants.85 Similarly, WFP school feeding aligns with the Global Partnership for Education (GPE) grant, which benefits children from disadvantaged groups, and seeks to foster effective teaching, leadership and management skills of local and national education staff. This grant brings together complementary efforts from MoEYS, UNESCO and UNICEF, as well as other education development partners including the European Union, Sweden and USAID.86 86. All three NGO partners are highly engaged through playing important roles in delivering support, which is complementary to the Government functions at the field level. Plan and WV help monitor the process of handover and provide technical support in procurement and bidding processes. They also support school infrastructure and provide refresher trainings. World Education Inc. is an essential partner in the literacy component. While the field level agreements (FLAs) outline their scope of work, the partners expressed the need for more clarity on the expectations upon them in the new McGovern-Dole programme in terms of their roles and responsibilities over monitoring, since they consider there is still some confusion between them and WFP, in areas like scheduling of monitoring visits, for example. Key findings and conclusions – Question 1 • Programme design was appropriate for the needs of women, men, girls and boys and is aligned with Government priorities, as well as those of USDA, WFP, United Nations and other partners; • The programme aims to include schools with high poverty rates and poor educational outcomes. Differences in these poverty rates and educational outcomes, infrastructure and management were observed between schools within the programme because these schools were selected during previous rounds; • The programme, which has been implemented under earlier agreements for many years, is well accepted by all stakeholders, who across the board are aware of its benefits for children, schools and communities; 82 UNESCO Phnom Penh website: https://en.unesco.org/fieldoffice/phnompenh 83 MoEYS, Mid-Term Review Report in 2016 of the Education Strategic Plan 2014-2018 and Projection to 2020, November 2016. https://www.globalpartnership.org/sites/default/files/2016-11-cambodia-mid-term-review-education-sector-plan.pdf 84 MoEYS, Child Friendly School Policy, December 2007. Available at: www.cncc.gov.kh/userfiles/image/download/Policies%20&%20Standards-E6%20Child-Friendly%20School%20Policy-En.pdf 85 McGovern-Dole proposal. 86 https://www.globalpartnership.org/news/global-partnership-education-approves-us206-million-grant-strengthen-cambodias￾education-system McGovern-Dole Programme Baseline Report – Cambodia – January 2021 18 |Page • Equal gender participation is anticipated but there are no gender-specific outcomes in the Results Framework that would explain more clearly what and how gender issues are intended to be addressed, nor indicators to measure progress against, and no evidence of a sound gender analysis was found. • The other programme indicators are appropriate and feasible for use in monitoring and evaluation of this programme. 2.2. Evaluation Question 2: What are the results of the programme? 87. While this overall evaluation question refers to the programme results, this baseline set out to identify areas for later study to answer the evaluation questions on efficiency, effectiveness and impact. This report provides the information on the situation at baseline, and while some analysis is provided, there have been no results of activities so far, with any results towards the objectives only becoming identifiable during the later rounds. This section provides the results from the baseline survey, disaggregated by USDA intervention/comparison schools. It should be noted that the McGovern-Dole programme has been implemented for several consecutive rounds in the majority of the intervention schools, which helps explain the differences found in these schools, in particular regarding training and infrastructure compared to the non-intervention schools. 88. As agreed with WFP CO during the inception phase, 70 USDA-McGovern-Dole supported schools were selected for detailed data collection.87 Most of these were located in Siem Reap (61 percent) followed by Kampong Thom (30 percent) and Kampong Chhnang (nine percent).88 Using the propensity score matching (PSM) technique, the same number of schools were selected from provinces without USDA support in Kampong Cham (49 percent), Battambang (27 percent), Kratie (13 percent) and Preah Vihear provinces (11 percent). Table 7 shows this breakdown. Table 7: Surveyed schools Province USDA supported schools Comparison schools Total # % # % # % Siem Reap 43 61.4 0 0.0 43 30.7 KampongThom 21 30.0 0 0.0 21 15.0 KampongChhnang 6 8.6 0 0.0 6 4.3 Battambang 0 0.0 19 27.1 19 13.6 KampongCham 0 0.0 34 48.6 34 24.3 Kratie 0 0.0 9 12.9 9 6.4 PreahVihear 0 0.0 8 11.4 8 5.7 Total 70 100.0 70 100.0 140 100.0 Teaching staff in surveyed schools 89. The survey collected information on teacher numbers in the surveyed schools to enable accurate planning of teacher-related activities such as training to support RF Outcome 1.1 (Improved Quality of Literacy Instruction). Table 8 shows that 57 percent of teachers in USDA supported schools, and 60 percent in comparison schools, are female. The USDA schools reported higher percentages of contract teachers (36 percent) than 87 For consistency, the same statistical approach and sample size calculation tools were used as in the previous evaluation (2017-2019), of using the most conservative approach to calculate the sample size based on a 95 percent confidence level, a power of 80 percent and a margin of error term of eight percent. The response distribution for grade promotion rate used is 84.61 percent as found by the previous endline evaluation for a total school population size of 522 schools. The calculated sample size is 70 matched pairs of intervention and control schools (a total of 140 schools). This will provide a power of 85 percent for total inattentiveness of students. Thus, this sample size is representative and has enough power to detect changes. Because the quantitative survey will be used as baseline for both the McGovern-Dole and the LRP evaluations, the ET allocated 50 percent of case schools to HGSFHybrid- USDA + LRP schools (n=35) and 50 percent to HGSF-Hybrid - USDA+other donors(non-LRP; n=35), proportionately allocated across the three provinces. The required number of schools was then selected using systematic random sampling sorted out using the percent of poor households in the village from the list of programme schools provided by WFP CO. Further details are given in the Inception Report. 88 This distribution by province is very similar to the overall group of USDA supported schools. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 19 |Page comparison schools (16 percent). The USDA supported schools had an average of seven full time teachers, while the comparison schools had eight teachers. Gender ratios were similar regardless of contract status and USDA support. Table 8: Teaching staff in surveyed schools (SY 2019/20) Teachers (Grades 1-6) Full time teachers Contract teachers Total Female % Total* Female % Total Female % USDAsupported schools 455 57% 6.6** 57% 165 65% Comparison schools 523 62% 8.2 60% 82 60% * number per school; **t-test: p<.05. 90. Output 1.1.1 of the McGovern-Dole programme is ‘more consistent teacher attendance’. All surveyed schools maintained teacher attendance sheets, although due to the school closure in March 2020, data on attendance were only collected on the November 2019 – March 2020 period. During the 104 school days in this period, mean teacher attendance was high at 102 days or 98 percent. This was similar in both USDA supported schools and comparison schools (Table 9). Table 9: Teacher attendance (SY 2019/20) Teachers USDA supported schools Comparison schools % of schools with attendance sheet 100% 100% Number of school days in a year (Nov 2019-March 2020) (mean) 104 104 Number of days teachers being absent (mean) 2.0 2.4 % teacher attendanceduring official SY 2019/20 98.1% 97.7% 91. Outcome 1.3 of the McGovern-Dole programme refers to improved student attendance. Student participation in education includes enrolment and attendance, as well as promotion, repeat and drop-out rates of students. Of the 14,787 children aged 6-11 years in the catchment area of the sampled USDA supported schools, 50 percent were girls (n=7,381) (Table 10). Net enrolment of children aged 6-11 years was 95 percent overall and 94 percent among girls, who constitute 49 percent of students in this age group. The USDA supported schools have an average of 235 students per school, compared to 325 in comparison schools.89 Table 10: School mapping information in SY 2019/20 # Children (6-11-years old) in the school catchment area # Students (6-11- years old) # Students (over 11 years old) Students per school Total Girls Total Girls Total Girls Total USDA supported schools 14,787 7,381 14,059 6,906 2,400 1,010 235 Comparison schools 19,534 9,596 18,251 9,139 4,511 2,421 325 92. The survey found that the overall promotion rate, repetition rate and drop-out rate of students in USDA supported schools were 86 percent, 11 percent and three percent respectively (Table 11).90 Girls were performing better according to all these indicators. No difference was found between the USDA supported schools and the comparison schools. Table 11: Promotion, repetition and drop-out rates (SY 2018/19) Promotion rate (%) Repetition rate (%) Drop-out rate (%) Total Girls Total Girls Total Girls USDA supported schools 85.83 89.94 11.25 7.88 2.92 2.18 89 This difference is related to differences in coverage areas of schools in the different provinces. Schools were matched based on poverty indicators, not on numbers of students 90 Data from SY 2018/19 as data on SY 2019/20 were not yet available during the survey. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 20 |Page Comparison schools 86.97 89.23 10.18 8.39 2.85 2.37 Literacy of school-aged children (SO1) 93. During the 2017-2019 period, WFP’s implementing partners for the literacy component of the McGovern-Dole programme (WEI and KAPE) conducted three Early Grade Reading Assessments (EGRA) in USDA supported schools in Kampong Thom and Siem Reap provinces, where activities aimed at improving the literacy scores among early grades children had been implemented.91 These EGRA exercises were designed to assess the reading proficiency of children at the end of Grade 2 and so verifying to what extent the programme had achieved one of its strategic objectives. 94. At endline, 30 percent of children at the end of Grade 2 were able to read and understand grade level text.92 As noted in the 2017-2019 endline evaluation report, the result showed a strong improvement over a short period of time, although it did not come close to meeting the programme target of 55 percent.93 95. World Education Inc. is also the primary field implementer of USAID's All Children Reading-Cambodia (ACRC) and All Children Learning (ACL) projects on literacy activities. In close cooperation with USAID, WEI (with support from WFP) will assist MoEYS in rolling out its new Early Grade Learning (EGL) package in USDA schools in Kampong Chhnang, a new implementation site for the McGovern-Dole programme. They will also provide follow-up mentoring and reinforcement of the EGL roll out in Kampong Thom and Siem Reap.94 96. Activities that will be developed in Kampong Chhnang under the literacy component of the programme will be similar to those implemented during the previous phase of the McGovern-Dole programme. They will cover Teacher Professional Development (TPD), support to MoEYS capacity building, technical assistance in training planning and delivery, and the development of non-fiction books for Grade 2 students. 97. Table 12 shows the performance indicators for the literacy activity and its annual targets, which will be used to measure progress during the midline and endline evaluations. Table 12: Literacy activity performance indicators and annual targets Annual target (FY) Indicator 2020 2021 2022 2023 Average teacher attendance rate 95 96 97 98 Number of school administrators and officials in target schools who demonstrate use of new techniques or tools as a result of USDA assistance 0 204 57 0 Number of school administrators and officials trained or certified as result of USDA assistance 0 255 71 0 Number of teachers/educators/teaching assistants in target schools who demonstrate use of new and quality teaching techniques or tools as a result of USDA assistance 0 327 79 0 Number of teachers/educators/teaching assistant trained or certified as a result of USDA assistance 0 837 99 0 No. of teaching & learning materials provided as a result of USDA assistance 0 0 837 0 Source: Attachment D - McGovern-Dole Programme Agreement; FY: fiscal year 98. In October 2019, WEI conducted a baseline literacy assessment in preparation for the implementation of the current round of the McGovern-Dole FFE programme. An EGRA was administered to students who had just completed Grade 2 at a participating school in a sample of 13 schools (18 percent of the planned target schools) in three districts of Kampong 91 USDA-supported schools in Battambang were not included as they had already been handed over to the Government. 92 World Education Inc. Quarterly Report - Year 3 Quarter 3: April 1, 2019 – June 31, 2019. 93 Dunn, S. et al (2019) Endline Evaluation of United States Department of Agriculture (USDA) McGovern-Dole Grant Food for Education Programme for WFP Cambodia (2017-2019). Final report. The KonTerra Group. 94 Agreement between the Government of the United States of America and World Food Programme for the provision of agricultural commodities through the Mc-Govern-Dole International Food for Education and Child Nutrition Program Act. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 21 |Page Chhnang province.95 As was done in the previous rounds, the ET will use findings from EGRA administrated by WEI to measure progress in this area. 99. The ET noticed that the EGRA tools used in the three provinces where the McGovern￾Dole programme will be implemented are not the same. For the present baseline, in Kampong Chhnang WEI used a tool that was developed in 2015 in cooperation with several departments from MoEYS and various NGOs. For Kampong Thom and Siem Reap provinces, WEI will use baseline data for Grade 2 students collected through the USAID￾funded ACRC project in June 2019.96 The agency undertook a comparison study between the two tools,97 the findings of which will allow them to "accurately compare baseline findings across the three provinces, as well as compare endline to baseline at the end of project even where different tools are used".98 100. For the final evaluation of the last FFE project, WEI used the 2015 tool. The results between that evaluation and the baseline for this new project are so different because WEI had to change the way the standard McGovern-Dole performance indicators (‘percentage of students who, by the end of two grades of primary schooling, demonstrate that they can read and understand the meaning of grade level text’) were defined. When planning the baseline, it was noticed that USAID’s ACRC project (which also uses the same indicator), defines this as the ‘percentage of students who can correctly answer 80 percent of comprehension questions (four out of five reading comprehension questions correctly answered)’.99 In order for full alignment with the results obtained by USAID under the current FFE, WEI will need to use the same method for defining this indicator, and all future evaluations will use the 2017 tool, as now adopted by the MoEYS. 101. The baseline data from Kampong Chhnang demonstrates that only 1.04 percent of students (1.9 percent of girls; 0 percent of boys) in the sampled target schools were able to correctly answer at least four of the five reading comprehension questions. While this is in line with the annual targets for this indicator set within the agreement between USDA and WFP, the original target of 55 percent of children being able to read and understand grade level text that had been prevailing for the previous McGovern-Dole programme round was adjusted to 15 percent, based on these findings (Table 13). Table 13: Improved Literacy of School-Age Children – Annual targets Indicator FY* 2020 FY 2021 FY 2022 FY 2023 Percentage of students who, by the end of two grades of primary schooling, demonstrate that they can read and understand the meaning of grade level text 0 0 15 15 Source: WEI Baseline Report Food for Education (FFE) 2019-2023 (February 2020); * FY: fiscal year 102. The results of the WEI baseline indicate that, despite having completed Grade 2, students are struggling in all key reading domains except for letter name identification. However, it is anticipated that the FFE project will be able to contribute to gains in students’ reading proficiency by providing the training to teachers and distributing teaching and learning materials to schools. Results presented at the end of the previous McGovern-Dole programme phase (FY 2017-2019) show a strong improvement over a short period of time (30 percent of students, by the end of two grades of primary school, demonstrated that they 95 The detailed findings from this baseline assessment can be found in the Baseline Report Food for Education (FFE) 2019-2023, prepared by World Education Cambodia for World Food Programme (WFP), February 2020. 96 The tool used was developed by USAID/RTI (Research Triangle Institute International) in 2017 under the ACRC project 97 Testing both tools on the same children revealed that children consistently scored higher on the 2017 tool (see EGRA Tool Comparison document). Statistical analysis produced a 95% confidence interval for the mean differences in scores generated by the different tests. (four out of five reading comprehension questions correctly answered). Therefore, WEI applies that confidence interval to the average obtained using the 2015 tool in order to compare with results from the 2017 tool. 98 Baseline Report Food for Education (FFE) 2019-2023, prepared by World Education Inc. Cambodia for World Food Programme (WFP), February 2020. 99The definition of this indicator in the last FFE project was ‘percentage of students achieving the reading standard of 45 corrected words per minute and one correct comprehension question.’ McGovern-Dole Programme Baseline Report – Cambodia – January 2021 22 |Page could read and understand the meaning of grade level). Although they may seem overambitious, targets in Table 13 are based on this very positive performance. Food distribution 103. The food distribution through the school meals aims to decrease short-term hunger and increase attentiveness in class (Output 1.2.1), as well as to provide an incentive for school attendance (Output 1.3.1) and reduce health-related absences (Output 1.3.2), all to support the objective of improved literacy. The performance indicators and annual targets for this activity are presented in Table 14. Table 14: Food distribution activity - Annual targets Indicator Annual target (FY*) 2020 2021 2022 2023 Average no. of school days missed by students due to illness 7 6 5 4 Average no. of school days per month on which multi-fortified or at least four food groups were provided 20 20 20 20 Average student attendance rate in USDA supported classrooms / schools 80 85 90 95 No. of daily school meals (breakfast, snack, lunch) provided to school-age children as a result of USDA assistance 7,416,667 19,750,000 14.416,660 10,750,000 No. of individuals benefiting indirectly from USDA-funded interventions 193,183 167,264 122,604 91,078 No. of individuals participating in USDA food security programs 152,365 132,966 97,452 72,165 No. of school-age children receiving daily school meals (breakfast, snack, lunch) as a result of USDA assistance 151,787 131,422 96,332 71,561 No. of schools reached as a result of USDA assistance 599 512 385 296 No. of social assistance beneficiaries participating in productive safety nets as a result of USDA assistance 151,787 131,422 96,332 71,561 No. of students enrolled in schools receiving USDA assistance 151,787 131,422 96,332 71,561 Percent of students in target schools identified as attentive by their teachers 88 89 90 91 Total quantity of commodities (mt) provided for school meals as a result of USDA assistance 890 2,370 1,730 1,290 Source: Attachment D - McGovern-Dole Programme Agreement; * FY: fiscal year 104. In USDA supported schools, only 3.1 percent of students were reported to be hungry in class every day, according to the teachers who participated in the survey.100 This was significantly lower than in comparison schools, where no school meals were provided and 14 percent reported daily hunger (Table 15). The survey found that this problem affects boys significantly more often than girls. In comparison schools,101 on average 33.8 percent of boys and 30.1 percent of girls were reported to be hungry. 100 These data are based on teacher reports pertaining to March 2020 (when schools were open and the school meal programme operational) during the baseline survey in August 2020. They were validated against the previous endline data, which were similar. 101 A t-test (paired two sample for means) was applied to test the hypothesis of equal percentages of hungry boys and hungry girls in comparison schools. The hypothesis was rejected (difference=7.3 percent, t-statistic=-4.4127, p-value (one-tail) =0.0000). Therefore, we can assume that boys are reported to be hungry significantly more often than girls in comparison schools McGovern-Dole Programme Baseline Report – Cambodia – January 2021 23 |Page Table 15: Percentage of students in target schools who indicate that they are “hungry” or “very hungry” during the school day (March 2020) Programme indicator USDA supported schools Comparison schools Percentage of female students reported to be hungry in class (daily+sometimes) 9.2%*** 30.1% Percentage of male students reported to be hungry in class (daily+sometimes) 12.8%*** 33.8% Total percentage of hungry children (daily + sometimes) 10.7%*** 31.8% Percentage of students reported to be hungry in class daily 3.1%*** 14.0% Percentage of all students reported to be hungry in class sometimes 7.6%*** 16.8% t-test: ***p<.001 t-test (two-tailed) for independent samples resulted in a significant difference between USDA-supported schools and comparison schools. 105. A small but statistically significant difference was found in the reported percentage of inattentiveness of students, with 13.5 percent of students in USDA supported schools noted as inattentive compared to 15.8 percent in comparison schools (Table 16). In both USDA-supported and comparison schools, inattentiveness was more frequently reported among boys. Table 16: Reported inattentiveness of students (March 2020) Programme indicator USDA supported schools Comparison schools Percentage of children who are inattentive (sleepy, inactive) during classes Everyday 4.5% 5.6% Sometimes 8.6% 10.0% Total 13.5%* 15.8% Girls 10.0% 10.8% Boys 17.7% 22.0% t-test: *p<.05. t-test (two-tailed) for independent samples resulted in a significant difference between USDA-supported schools and comparison schools for total only 106. One of the main outcomes under the literacy objective is school attendance. All students attending the morning sessions receive a daily school meal. According to their parents, before the school closure children in USDA supported schools attended morning sessions every weekday of the month.102 Attendance was high at 97 percent among girls and 95 percent among boys, as shown in Table 17. The remainder of the children (13 boys and seven girls) were absent from school on some days, and therefore they did not receive food every day. Table 17: Average number of morning sessions, percentage of children receiving school meals and those who took some home Average number of morning sessions % children who received a school meal every day Girls 21.8 97% Boys 21.9 95% Total 21.8 96% Improved health 107. Intestinal parasites interfere with nutrient absorption from food and regular deworming is therefore important. The MoH provides deworming tablets twice a year to schools at their request. The survey found that 75.6 percent of children at USDA supported schools had received this medication in the six-month period prior to March 2020. This was slightly higher than in comparison schools (Table 18). Most children had received their deworming medication at school. 102 Household survey participants were sampled among families of students who attended morning classes, as the school meal is provided in the morning. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 24 |Page Table 18: Percentage of children who had received deworming medication in the past six months % children receiving deworming medication Location School Health centre At home USDA supported schools 75.6% 85.8% 6.0% 8.2% Comparison schools 63.5% 79.2% 8.3% 12.5% 108. The McGovern-Dole programme includes a number of activities to improve the health of the students, including rehabilitation and construction of school infrastructure, ensuring adequate water and sanitation, and providing training and health education. These activities support the RF outcomes 1.3, 2.1, 2.2, 2.4 and 2.6. The performance indicators and annual targets are shown in Table 19. Due to the school closure, it was not possible to fully assess the condition of the infrastructure, but baseline data were corroborated with the endline data from 2019. Table 19: Improved health promotion activity - Annual targets Annual target (FY*) Indicator 2020 2021 2022 2023 Number of educational facilities (i.e. school buildings, classrooms, improved water sources, and latrines) rehabilitated/ constructed as a result of USDA assistance 294 349 297 194 Number of individuals reached through IEC** materials as a result of USDA assistance 119,800 102,400 77,000 59,200 Number of individuals trained in safe food preparation and storage as a result of USDA assistance 558 434 932 594 Number of individuals who demonstrate use of new safe food preparation and storage practices as a result of USDA assistance 391 304 652 416 Number of schools provided with kitchen utensils as a result of USDA assistance 100 200 164 135 Number of schools using an improved water source 512 472 361 283 Number of students benefiting from educational facilities (i.e. school buildings, classrooms and latrines, water infrastructure etc.) rehabilitated/constructed as a result of USDA assistance 36,750 43,625 37,125 24,250 Number of target schools that have at least one month’s supply of soap (hand and dish soap) 515 440 331 254 Percent of schools have soap at a functional hand washing station 80 85 90 95 Source: Attachment D McGovern-Dole Programme Agreement; *FY: fiscal year, ** IEC: information, education, communication School classrooms 109. On average, the sampled USDA supported schools had six classrooms per school (compared to seven in the comparison schools), 93 percent of which had sufficient teaching materials such as books and posters (Table 20). In the comparison schools, only 86 percent of classrooms had sufficient teaching materials. Table 20: Number of classrooms in surveyed schools # Grade 1-6 Classrooms # classrooms with sufficient material # classrooms per school % classrooms with sufficient material USDA supported schools 394 337 5.6 93% Comparison schools 465 427 6.6 86% School kitchens and equipment 110. All of the 70 USDA schools included in the survey had school kitchens, compared to none of the comparison schools. Of these school kitchens, 47 percent were in good condition and 33 percent had clean cooking and eating equipment, although 64 percent lacked utensils, 39 percent had leaking roofs, 11 percent were flooded during the rainy season, seven percent used rocks as stoves, and six percent of kitchens had no walls (these and additional details shown in Figure 2). McGovern-Dole Programme Baseline Report – Cambodia – January 2021 25 |Page Figure 2: Condition of kitchens in USDA supported schools (n=70) 111. The survey found that 46 percent of USDA supported schools (n=32) were using energy-efficient stoves. Of these, 66 percent (n=21) were in good condition, while the rest were still working despite being in a poor condition. No energy efficient stoves were found in comparison schools. Storerooms 112. Of the 70 surveyed USDA supported schools, 66 percent (n=46) had food storerooms, compared to none of the comparison schools. Of these, 67 percent stored food on pallets off the ground. Most were clean (74 percent), had a dry floor (87 percent), and could be locked (89 percent), while 44 percent was well ventilated. This and additional information shown here in Figure 3. Figure 3: Condition of storerooms in USDA supported schools (n=46, in percentage) Water and sanitation facilities 113. Access to clean water and sanitation is important to support healthy behaviour in schools. The survey found that virtually all (99 percent of USDA supported schools (n=69)) and 96 percent of comparison schools (n=67) had drilled wells or rain water catchments. In the USDA supported schools, 72 percent of the wells and 80 percent of the water catchments were functioning, compared to 65 percent and 81 percent respectively in comparison schools, as shown in Table 21. Table 21: Number of functioning drilled wells/ water catchments # drilled wells # water catchments % functioning drilled wells % functioning water catchments USDA supported schools (n=69) 121 111 72% 80% Comparison schools (n=67) 87 153 65% 81% 114. Over 80 percent of the school water sources were functional throughout the year (Figure 4). More USDA supported schools than comparison schools in the survey had clean 0% 10% 20% 30% 40% 50% 60% 70% Good condition Lacking kitchen utensils Clean cooking and eating equipment Leaking roof Flooded during rainy season Using rocks as stove No wall Old stove Old pots Unhygienic 0 10 20 30 40 50 60 70 80 90 Door is locked well Floor is dry Well cleaned Pallets for food storage Foods are stored in order Storeroom had ventilation Food was stored off ground Security guard at night time/ during school vacation Leaking rooms Broken windows/door No hygiene McGovern-Dole Programme Baseline Report – Cambodia – January 2021 26 |Page platforms (52 percent and 43 percent, respectively). Figure 4 below provides this and other comparisons. Figure 4: Condition of well-functioning drilled well/rainwater catchments 115. To support the cooking of the meals, as well as handwashing, year-round access to clean water at the schools is essential. Informants at 57 of the 70 USDA-supported schools in the survey reported that clean water was available , while 13 schools reported not, or not always, having access to clean water. (Table 22). Table 22: School access to year-round clean water % and # of schools with year-round access to a clean water source Yes (whole school year) No (some months not available) No (no water) USDA supported schools 81% [57] 16% [11] 2.9% [2] Comparison schools 72.9% [51] 10% [7] 17% [12] 116. Handwashing stations were present in 97 percent of USDA supported schools and 93 percent of comparison schools (Table 23) and most were in good condition and reported as functional throughout the year. Table 23: Percentage of schools with fixed hand and mobile hand washing stations No hand washing station Hand washing station Type Fixed only Mobile only Both USDA supported schools 2 (2.9%) 68 (97%) 28 (40%) 6 (8.6%) 34 (48.6%) Comparison schools 5 (7.1%) 65 (93%) 26 (37.1%) 4 (5.7%) 35 (50.0%) 117. Almost all USDA supported schools had soap available at the handwashing stations (99 percent), while this was the case in 94 percent of comparison schools as shown in Table 24. Students in the USDA supported schools reported better behaviour related to handwashing with soap than students in comparison schools. Most (89 percent) students in USDA supported schools reported they always (74 percent) or often (15 percent) used soap and water to wash their hands (Figure 5), but 14 percent said they never washed their hands with soap, though in the comparison schools all children reported washing their hands with soap. Table 24: Schools with soap at handwashing stations Schools with soap at handwashing stations # % USDA supported schools 69 98.6% Comparison schools 66 94.3% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Functioning well year-round Water is used for human consumption Platform is clean System to clean the wells No water out when pumping Dilapidated Red water Comparison schools (%) USDA supported schools (%) McGovern-Dole Programme Baseline Report – Cambodia – January 2021 27 |Page Figure 5: Student handwashing practices - frequency of handwashing with soap and water 118. More than two-thirds of USDA supported schools (69 percent) and comparison schools (73 percent) reported having soap supplies for at least one month (shown in Figure 6). The main sources of soap for the USDA supported schools were the MoEYS Programme Based (PB) budget (90 percent) and/or WFP and its partners (94 percent). Comparison schools received soap or funds for soap from the PB Budget (92 percent) or from the NGO Beautiful Cambodia (Figure 7). The vast majority of USDA supported schools (88 percent) reported buying soap every one to three months, or less frequently, while comparison schools purchased more often. Figure 6: Available stock of soap supply Figure 7: Provider of soap 119. To support a healthy environment in schools, and facilitate equal access to education, the presence of sufficient functioning latrines for girls and boys, disabled students and teachers is essential. All surveyed schools had latrines, of which about 87 percent were functional. On average, there is one non-functional latrine per school (Table 25). Table 25: Latrines in surveyed schools, by functionality % of functional latrines % of non￾functional latrines # of functional latrines per school # non-functional latrines per school USDA supported schools 87.4% 12.6% 5.3 1.0 Comparison schools 87.3% 12.7% 4.9 0.7 120. More USDA supported schools (64 percent) had separate and functional latrines for teachers compared to comparison schools (54 percent). They made up around 30 percent of all functional latrines in these schools. Almost 70 percent were functioning, providing one functioning staff latrine for five teachers in USDA supported schools and one for six teachers in comparison schools (Table 26). 0% 20% 40% 60% 80% 100% USDA supported schools (%) (n=69) Comparison schools (%) (n=66) Always Often Sometimes Never 0% 20% 40% 60% 80% 100% USDA supported schools (%) (n=69) Comparison schools (%) (n=66) Whole school year 4-6 months 1-3 months 3-4 weeks 1-2 weeks Less than 1 week No soap 0 10 20 30 40 50 60 70 80 90 100 PB Budget WFP/PLAN Angkor Children’s Hospital Other NGOs Charity persons Student bought it Community Comparison schools (n=66) USDA supported schools (n=69) McGovern-Dole Programme Baseline Report – Cambodia – January 2021 28 |Page Table 26: Presence of functioning latrines for teachers # Schools with separate functional latrine for teachers % of functional latrines for teachers # of functioning latrines for teachers # of teachers per functional staff latrine USDA schools 45 29% 68 5 Comparison schools 38 36% 69 6 121. Separate latrines for boys and girls were found in 74 percent of USDA supported schools, compared to 69 percent of comparison schools. The ratio of functioning latrines to number of students was high but acceptable in all schools,103 but the situation was better in USDA supported schools: 1:68 for boys and 1:64 girls, compared to 1:96 boys and 1:87 girls in comparison schools (Table 27). Table 27: Presence of functioning latrines for boys and girls # Schools with separate functioning latrines for boys and girls % of non-separate latrines for students104 # of boys per separate latrine for boys # of girls per separate latrine for girls USDA schools 52 28% 68 64 Comparison schools 48 32% 96 87 122. Functioning special latrines for disabled students were present in 64 percent of USDA supported schools compared to only 23 percent of comparison schools (Table 28). In the USDA schools, the specially designed latrines for students with disabilities accounted for 20 percent of all latrines, compared to only 10 percent in comparison schools. All of the specially designed latrines for students had ramp access and most were suitably large (78 percent), while 31 percent had a handle for students with disability. Table 28: Presence of functioning latrines for disabled students # of schools with latrine for disabled students # latrines for disabled students % latrines for disabled students44 # disabled students per latrine for disabledstudents USDA schools 45 56 20% 0.9 Comparison schools 16 34 10% 1.0 123. The latrines in the USDA supported schools were generally clean and well maintained (84 percent), while this was only the case in 59 percent of the comparison schools (Figure 8). Most common problems in non-functioning latrines were a broken or full pit latrine (26 percent), broken door, commode or washbasin (21, 20 and 10 percent, respectively), as shown in Figure 9. Figure 8: Condition of functioning latrines 103 The minimum ratio is 100 boys per 2 latrines and 3 urinals, 100 girls per 3 latrines. Minimum Requirement Guidelines on Water, Sanitation and Hygiene in School (Wins), School Health Department, November 2016 104 Out of total student latrines 0% 20% 40% 60% 80% 100% Clean and well maintained Dirty, not well maintained Broken but still usable Unusable Latrine is old Comparison schools USDA supported schools McGovern-Dole Programme Baseline Report – Cambodia – January 2021 29 |Page Figure 9: Condition of non-functioning latrines 124. The latrines were generally maintained by the students (Figure 10). Schools also keep the latrines locked during holidays, and ensure there are hand washing stations in the vicinity and that there is enough water in washbasins. Still, about one third of the schools implement the negative practice of making the students leave shoes outside the latrine. Figure 10: Latrine maintenance 125. The vast majority of both USDA supported and comparison schools have bins and other equipment for solid waste management (Table 29). Slightly more USDA supported schools (93 percent) have daily waste collection than in comparison schools (90 percent). Table 29: Solid waste disposal Schools with bins or other equipment for solid waste management Schools with daily solid waste collection # % # % USDA supported schools 64 91% 65 93% Comparison schools 62 89% 63 90% Improved nutrition 126. Under SO 2, the McGovern-Dole programme includes a number of activities to improve the nutrition of the students, including training in child health and nutrition, rehabilitation and construction of school gardens, and social behaviour change communication (SBCC) activities. The performance indicators and annual targets are shown in Table 30. Table 30: Promotion of improved nutrition activity - Annual targets Annual target (FY*) Indicator 2020 2021 2022 2023 Number of individuals trained in child health and nutrition as a result of USDA assistance 20 18 18 10 Number of individuals who demonstrate use of new child health and nutrition practices as a result of USDA assistance 14 13 13 7 Number of people reached through interpersonal SBCC approaches 220,781 191,159 140,119 104,089 Number of school gardens at target schools rehabilitated or constructed 296 265 267 230 Source: Attachment D - McGovern-Dole Programme Agreement; * FY: fiscal year, 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Train students and take turn to clean latrines Keep soap/hand washing facilities within or near the toilets Keep soap/hand washing facilities within or near the toilets Ensure washbasin is full of water Propose users to leave shoes outside of latrine stalls Hire someone to maintain it Comparison schools USDA supported schools 0% 5% 10% 15% 20% 25% 30% Pit latrine is broken or full Door is broken Commode is broken Washbasins are broken Locked Far from the well Broken, can't be used Comparison schools USDA supported schools McGovern-Dole Programme Baseline Report – Cambodia – January 2021 30 |Page 127. While the school survey could not assess the state of the school gardens as they had not been maintained during the school closure, 77 percent of USDA supported schools had school gardens, compared to 27 percent of comparison schools (Table 31). According to the school staff, most (83 percent) of these gardens were rehabilitated and were functional in March 2020, while this was the case in only 32 percent of comparison schools. Table 31: Schools with vegetable gardens Total # of schools with garden % of schools with vegetable garden % of rehabilitated vegetable gardens USDA supported schools 70 54 77% 83% Comparison schools 70 19 27% 32% 128. Cambodia has a long-standing practice of using school gardens to teach students life skills as well as gardening and agricultural concepts and skills, while the produce can be used to supplement commodities provided for the school meal programme. The percentages of schools giving these reasons are shown in Figure 11. Figure 11: Reason for establishing vegetable gardens in schools 129. Figure 12 shows that vegetables are planted in the school gardens throughout the school year, except during the school closure, when the gardens were mostly abandoned. In a limited number of schools, principals or teachers visited the school and took care of the garden.105 Figure 12: Vegetable planting in school gardens by month (Oct 2019-Sep 2020) 130. Both USDA supported schools and comparison schools reported that the school gardens were maintained by the school children, under the guidance of teachers and community members (Figure 13). Different grades were assigned to separate plots of land in 69 percent of USDA supported schools and 26 percent of comparison schools. Most schools also fence the area to prevent animals from entering. 105 During the school closure, directors and teachers in some schools maintained the gardens and utilized the produce. 0% 10% 20% 30% 40% 50% 60% 70% Supplementing SMP recipe Practicing life skills Both reasons Comparison schools (n=19) USDA supported schools (n=54) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% October December February April June August 2019 2020 Comparison schools (n=19) USDA supported schools (n=54) McGovern-Dole Programme Baseline Report – Cambodia – January 2021 31 |Page Figure 13: School garden maintenance 131. The McGovern-Dole Programme includes training for teachers, storekeepers and cooks on a variety of subjects, under Outcome 2.4. Table 32 shows that during the SY 2019/20 almost all the USDA supported schools had received training on food preparation and on good health and nutrition. Of the comparison schools, seven and 17 percent received training in food preparation and good health/nutrition practices (respectively) from other parties (mostly MoEYS) unrelated to the USDA support. These trainings were conducted during previous rounds as well as at the beginning of SY 2019/20. As a result, almost all personnel in USDA-supported schools were aware of the three main food groups: foods that provide energy (carbohydrates), building foods (protein sources) and protective foods (containing vitamins and minerals) (Figure 14). Table 32: Schools receiving training on food preparation and good health and nutrition practices Schools receiving training on food preparation Schools receiving training on good health and nutrition practices # % # % USDA supported schools 65 93% 64 91% Comparison schools 5 7% 12 17% Figure 14: Awareness amongst school personnel on three types of food groups 132. The programme also includes awareness raising events on the importance of education, again as a continuation of activities in the previous rounds. Almost all USDA supported schools were engaged in these activities during SY 2019/20, compared to 73 percent of comparison schools (Table 33). In addition, the USDA supported schools conducted these events more often than the comparison schools (Figure 15). Table 33: Percentage of schools participating in community awareness raising events on the importance of education Percentage of schools USDA supported schools 90% Comparison schools 73% 0% 20% 40% 60% 80% Children are mentored by trained teachers and community Children from grade to grade are assigned to each plot of land Prevent animals from entering school compound by repairing schools’ fence once per year Comparison schools (n=19) USDA supported schools (n=54) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Energy Building Protective Don't know Comparison schools (n=19) USDA supported schools (n=54) McGovern-Dole Programme Baseline Report – Cambodia – January 2021 32 |Page Figure 15: Number of times that awareness-raising events were conducted during SY 2019/20 133. Most community awareness events were conducted at the start of the school year, during religious ceremonies or village meetings, as shown in Figure 16. Figure 16: Timing of awareness raising events Capacity building 134. An important component of the McGovern-Dole programme is capacity building of various stakeholders. The related performance indicators are shown in Table 34. Table 34: Capacity building activity performance indicators and annual targets Indicator Annual target (FY*) 2020 2021 2022 2023 Number of Parent-Teacher Associations (PTAs) or similar "school" governance structures supported as a result of USDA assistance 599 512 385 296 Number of policies, regulations or administrative procedures in each of the following stages of development as a result of USDA assistance 2 2 1 1 Value of new commitments, and new public and private sector investments (USD) leveraged by USDA to support food security and nutrition 70,000 60,000 51,000 37,000 Source: Attachment D - McGovern-Dole Programme Agreement; • FY: fiscal year 135. The survey collected various information on school management including information on the Parent-Teacher Associations (PTAs), the presence of LSFCs and/or School Support Committees (SSCs), and the availability of various school management tools. During the previous award, the target for PTAs or similar governance structures was already reached. 136. Table 35 shows that not only do all surveyed USDA supported schools (100 percent) have an LSFC or an SSC, but these are also well functioning. This is likely to be due to the work done by WFP in the previous awards. In both USDA supported and comparison schools the majority of members are male (72 percent). 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% USDA supported schools Comparison schools None Once Twice More than twice 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Start school year Ceremony pagoda Village meeting Rice distribution Scholarship End school year Middle of school year Online education campaign Comparison schools USDA supported schools McGovern-Dole Programme Baseline Report – Cambodia – January 2021 33 |Page Table 35: Schools with existing LSFCs and SSCs LSFC/SSC Membership # Schools % Functioning # % Female % Male USDA supported schools 70 100% 302 28% 72% Comparison schools 54 73% 448 28% 72% 137. The majority of both the USDA supported and comparison schools had school development plans, a safe school environment and a record and reporting system in place. While all USDA supported schools reported having at least one of these management tools, three of the comparison schools (four percent) had none, as shown below in Table 36. Table 36: Availability of school management tools USDA supported schools Comparison schools School development plan 99% 91% Safe school environment 83% 87% Record and reporting system 63% 84% Don’t have any of the above 0% 4% 138. The McGovern-Dole programme actively encourages community participation and financial support to the schools. Table 37 below shows that in SY 2019/20 USDA supported schools received an average of US$288.13 in cash and the equivalent of US$270.97 as in￾kind contributions from the community. Although receiving less cash contributions, the USDA supported schools received more in-kind contributions, in line with the presence of the school meal programme. Table 37: Average amount of community contributions to each school in SY 2019/20 In Cash (US$) In Kind (US$ equivalent) USDA supported schools 288.13 270.97 Comparison schools 701.87 199.50 Findings from the household survey 139. The household survey was conducted with 418 households from USDA supported schools (258 in Siem Reap, 124 in Kampong Thom and 36 in Kampong Chhnang), as well as with 416 households from comparison schools. More details are provided in Annex 18. Table 38 shows the distribution of the households by province and USDA supported/comparison schools, as well as the gender of the surveyed children in those households. Table 38: Distribution of surveyed households by province, supporting programme, and gender of sampled child Province USDA supported schools Comparison schools Total (%) Siem Reap 258 0 258 (30.9) Kampong Thom 124 0 124 (14.9) Kampong Chhnang 36 0 36 (4.3) Battambang 0 114 114 (13.7) Kampong Cham 0 200 200 (24.0) Kratie 0 54 54 (6.5) Preah Vihear 0 48 48 (5.8) Gender of surveyed children Boys Girls Boys Girls Boys Girls 208 210 208 208 416 418 (49.8) (50.2) (50.0) (50.0) (49.9) (50.1) Sub-Total (%) 418 (50.1) 416 (49.9) Total 418 416 834 McGovern-Dole Programme Baseline Report – Cambodia – January 2021 34 |Page 140. The average household size (five members) and the dependency ratio (average of one adult to 1.1 children or elderly) were the same among USDA supported and comparison HHs. Most heads of household were married and a quarter were women. The majority of heads of the USDA supported HHs had no secondary education (82.8 percent) and slightly more than half (54.8 percent) reported being able to read and write, while this was 67.5 percent among heads of HHs from the comparison schools. 141. Surveyed households had on average one school-aged child, who started school at six years old. Only three children in USDA supported HHs and five children in comparison school HHs did not attend school. Reasons included the child’s learning ability, the distance to school, poverty and safety issues. 142. The main reasons for sending children to school were because of proximity to the school, and the desire of parents for their children to get an education (Annex 18). The latter reason was more commonly reported by parents in the USDA supported schools (21.6 percent), compared to 17.7 percent of parents from comparison schools. In addition, the quality of the teachers was mentioned by 22.9 percent of parents in USDA supported schools compared to 17.1 percent in the comparison schools. These findings indicate a higher awareness of the importance of good education in the USDA supported households. Household water and sanitation 143. Half the households in the survey had improved sources of water106 in their homes (Figure 17). The vast majority (approximately 80 percent) of HHs used flush toilets or ones that used water in some way, while approximately one fifth of HHs (20 percent) used open sources such as bush, fields, rivers or ponds (Figure 18). Figure 17: Household main source of water Figure 18: Household sanitation (toilet) facilities Household food security and nutrition 144. On average, respondents reported they consumed an average of just under three meals per day (2.7 for adults and 2.9 for children) (more details also in Annex 18). The HH food consumption score (FCS) (see Figure 19) shows the overall food consumption pattern over the previous seven days, with most HHs in all groups reporting consumption of cereals (rice), meat or eggs, vegetables and tea/spices on the day before the survey. Notably, 100 percent of households reported consuming sources of animal protein. Overall, HH food consumption was similar across all groups, with the average FCS for each group (56.8 for USDA supported households and 58.4 for comparison households) being classified as 106 Improved water sources include piped water into the dwelling or yard, public tap or standpipe, tube-well or borehole, protected dug well, protected spring or rainwater. Unimproved sources include surface water, tanker-truck, cart with tank/drum, unprotected dug well and unprotected spring. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% USDA supported schools Comparison schools Improved Unimproved 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% USDA supported schools Comparison schools Flush latrine/toilet with water Traditional pit latrine Partly open pit (no roof or wall) River/pond side Bush/open field McGovern-Dole Programme Baseline Report – Cambodia – January 2021 35 |Page ‘acceptable’.107 Moreover, based on 24-hour dietary recall, 56.2 percent of students in USDA supported schools and 51.4 percent of students in comparison schools consumed four or less (out of eight) food groups the day prior to the interview. Figure 19: Percentage of households reporting consumption of each food group in last seven days 145. More than 90 percent of households had an acceptable FCS, with none of the households in either group being classified as having a poor FCS (Figure 20). Figure 20: Household Food Consumption Scores 146. When there is not enough food - or money to buy food - approximately 50 percent of all households were relying on less preferred and less expensive foods at the time of the survey (Figure 21). Considerable numbers of HHs also used more severe coping strategies such as reducing the number of meals eaten each day, reducing their portion size or relying on help from family and friends. The Reduced Coping Strategies Index (rCSI)108 shows food security stress in both USDA supported (rCSI=10.6) and comparison households (rCSI=9.6).109 Figure 21: Household use of consumption coping strategies 107 The “Food consumption score” (FCS) is a score calculated using the frequency of consumption of different food groups consumed by a household during the seven days before the survey. There are standard weights for each of the food groups that comprise the food consumption score. Categories for FCS for Cambodia are poor (<25), borderline (25 – 38.5) and acceptable >38.5) 108 rCSI is calculated by summing the number of days a household resorts to each of the coping strategies during a 7 day period, after applying a standard weight to each strategy. Then cut-offs are used to determine the severity of the famine (0-3=minimal stress; 4- 18=stressed, 19 and above=crisis) 109 https://fscluster.org/handbook/Section_two_rcsi.html https://fews.net/sectores-t%C3%B3picos/abordagem/classifica%C3%A7%C3%A3o-integrada-de-fases 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Cereals Root and tubers Legumes Vegetables Fruits Meat Condiments (plus prahok) Dairy (milk) products Oil and fat Sweets Insects Comparison schools USDA supported schools 4% 5% 96% 95% 0% 20% 40% 60% 80% 100% USDA supported schools (n=418) Comparison schools (n=416) Poor (FCS 1-24.9) Borderline (FCS 25.0-38.5) Acceptable (FCS >38.5) 0% 10% 20% 30% 40% 50% 60% 70% Less preferred, less expensive food Borrowed food or relied on help Reduced the number of meals Reduced portion size of meals Reduced amount eaten by adults Comparison schools USDA supported schools McGovern-Dole Programme Baseline Report – Cambodia – January 2021 36 |Page 147. A full set of data results from the baseline is presented in Annex 18. Key findings and conclusions – Question 2 • Baseline attendance rates of teachers and students are high, which forms a solid basis for both the literacy and the school meal components of the programme. • According to the WEI baseline study, virtually all students who graduate from Grade 2 struggle with all key reading domains. The target for SO1 has been adjusted accordingly to a lower and more feasible percentage of children who reach the threshold after two years. • Different EGRA tools were used to measure literacy in the three provinces. • At baseline, all schools already have kitchens; however, more than half were not in a good condition and less than half were using energy-efficient stoves. Access to sources of clean water was possible for the majority but not all of the schools, and on average there was one non-functional latrine per school. Not all schools currently have gardens. Improvement of these facilities are part of the programme design and will require extra attention. • Community acceptance of, and participation in, the school feeding programme is high. • In previous rounds, capacity building trainings had been provided to school staff, cooks, teachers and Government staff. Staff turnover and suspension of many activities of the programme during the school closures may lead to a need for additional (refresher) trainings. 2.3. Evaluation Question 3: How likely is it that the programme is sustainable? 148. In order to assess the sustainability of the school feeding programme after handover to the Royal Government of Cambodia, a number of factors were considered, including the enabling environment, funding stability, partnerships, capacity of stakeholders at all levels, programme evaluation, programme adaptation and strategic planning.110 149. School feeding programmes are an important component of WFP’s work and global policies are in place. Dedicated staff and experts at HQ, RB and CO levels support the programme as a means to increase enrolment and attendance, as well as improve the nutritional status of school children. Champions exist within WFP, as well as in other organizations and the Government, to promote the school feeding programme. National ownership 150. WFP Cambodia is currently implementing its CSP 2019-2023 in five out of Cambodia’s 26 provinces, with a strong focus on national and sub-national level systems support. The CSP aims to strengthen the Government’s capacities at the lower levels to enable national ownership of programmes, including school feeding. This transition of the SFP to national ownership aligns with WFP’s new School Feeding Strategy 2020-2030 which aspires to handover SFPs to governments in stable low-income and lower-middle￾income countries by 2030. 151. The 2019 endline evaluation of the previous USDA-funded McGovern-Dole programme111 provided an overview of the status of the MoEYS’ capacity to implement school feeding activities according to the five policy pillars of the Systems Approach for Better Education Results (SABER) (School Feeding) which is the internationally recognized 110 Adapted from: Washington University, St Louis MO. The Program Sustainability Assessment Tool 111 Dunn, S. et al (2019) Endline Evaluation of United States Department of Agriculture (USDA) McGovern-Dole Grant Food for Education Programme for WFP Cambodia (2017-2019). Final report. The KonTerra Group. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 37 |Page system for measuring government capacity in relation to school feeding.112 The evaluation identified several areas that needed improvement in relation to institutional capacity, and made some relevant recommendations (see Annex 9). The Operational Plan for the present McGovern-Dole programme indicates that WFP will build on existing capacity strengthening initiatives initiated under former USDA supported projects, and provides a detailed description of activities that will be undertaken accordingly. 152. The development of a school meals policy was initiated in 2019.113 It will be an inter￾ministerial and multi-stakeholder process involving the MoEYS, MEF, MoH, MoI, MAFF, the National Social Protection Council and WFP among others. This is due to the recognition of the multiple benefits of school feeding114 and that a multi-sector approach with a common objective is required for optimal implementation. The development of a Joint Action Plan for the implementation of school feeding was scheduled for 2020 but has been delayed due to COVID-19; the draft policy is currently expected in May 2021. 153. The midline and endline evaluations of the new cycle of USDA-funded school feeding programmes (this McGovern-Dole FFE programme and the LRP programme) will include assessment/evaluation of the MoEYS’ capacity to implement school feeding and document the progress against the School Feeding Roadmap initially endorsed in 2015. Transition and handover 154. WFP began transitioning schools from the traditional SMP model towards HGSF in SY 2017/18 after an officially declared preference for this model by MoEYS. As an interim modality, WFP used a HGSF hybrid approach – with some commodities being imported and others being procured locally. The ongoing transition plan of the in-school activities to national ownership has a current end date of 2025. 155. Efforts to start the Royal Government of Cambodia’s direct financing and implementation of HGSF were accelerated with the inclusion of HGSF targets in the 2019- 2023 Education Strategic Plan and the allocation of domestic finances by the MEF for fiscal years 2019-2021. From the start of SY 2019-2020, the Government assumed responsibility for the implementation of HGSF in 205 schools, 123 of which were transitioned from WFP’s prior support; WFP will provide technical support in 82 new schools. The Government allocated US$1.8 million as its initial direct investment, scaling up to more than US$4 million over the course of the next five years. 156. Monitoring and evaluation (M&E) data from WFP CO provide detailed information about the expected transition process within USDA supported schools during the 2020- 2023 period. They illustrate both the progressive shifting from SMP to HGSF model and progressive handing over of activities to the Government. As of SY 2022/23 all USDA supported schools will be implementing either the HGSF-hybrid model or a full HGSF modality with all foods procured locally. This development is summarised in Figure 22. 112 According to the SABER approach, in general terms a quality school feeding programme has the following in place: (1) a national policy framework, (2), sufficient institutional capacity for implementation and coordination, (3) stable funding, (4) sound design and implementation, and (5) community participation. 113 WFP Cambodia Annual Country Report 2019. 114 The WFP Cambodia Annual Country Report 2019 stresses that educational outcomes continue to be a primary Government objective. According to the Government’s education management information system, in the school year 2018-2019, the districts with school feeding support had a stronger net enrollment rate (99.2 percent against 91.8 percent), completion rate (85.3 percent against 81.9 percent) and dropout rate (4.5 percent against 4.8 percent) than those without the programme. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 38 |Page Figure 22: Transition process Source: Evaluation Team 157. However, to support smooth handover and implementation by the Government, continued capacity building and mentoring by the CO is planned. Stakeholders across the board confirmed a strong need for further capacity building. At the national level this is mostly related to policy development and knowledge management, while at the subnational levels and the schools there is a need to strengthen skills in programme implementation, including monitoring and reporting. 158. In addition, not all schools have adequate infrastructure yet, including the appropriate conditions for the school kitchens, availability of functional energy-efficient stoves, functioning water and sanitation facilities and year-round access to clean water. The support to the schools that the McGovern-Dole programme can provide is essential in the preparation of the handover. Key findings and conclusions – Question 3 • The Royal Government of Cambodia is highly invested in a national school-feeding programme as evidenced by a multisectoral school meals policy, including budget allocation. During the period of the new grant, 141 additional schools are expected to be handed over to the Government in line with WFP’s new School Feeding Strategy 2020-2030. • Participating schools are being transitioned from the traditional SMP model towards HGSF via a HGSF hybrid approach – with some commodities being imported and others being procured locally. The ongoing transition plan of the in-school activities to national ownership has a current end date of 2025. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 39 |Page • Stakeholders at national and sub-national Government levels, as well as at the schools, require continued capacity building and mentoring by the CO to enable sustained implementation of a school meals programme. This is in line with the CSP and WFP global direction to transition from direct implementation towards a supporting role. • In addition, adequate food preparation facilities and water and sanitation are not yet available in all programme schools. The McGovern-Dole programme includes activities to improve these and this support is essential in the preparation of the handover. 2.4. Evaluation Question 4: Why and how has the programme resulted in the observed results? 159. As this is the baseline study, the programme has not yet produced any measurable results, and these will be assessed at later stages of the evaluation series. In this section, internal and external factors are discussed that may either enable positive results and in particular a successful handover to the Government, or provide barriers for the same. 160. WFP is widely acknowledged as a strong partner in school feeding programmes. Programme management and technical capacity, including in logistics and supply chain, at the CO is strong, and the CO also has access to technical support from the Regional Bureau in Bangkok and from Headquarters in Rome. WFP provides highly appreciated and strong support to the Government in terms of strengthening monitoring systems, evidenced by the integration of WFP’s PRISM into MoEYS’ systems to improve programme targeting for vulnerable households. In addition, WFPs cooperating partners (Plan, WV and WEI) all have extensive experience and high capacity in their fields of expertise. 161. Moreover, the CO has built a long-term engagement and strong relationship with the MoEYS during the decades of implementing previous SFPs in Cambodia. The current shift in focus from direct implementation to capacity building and technical assistance under the CSP provides a suitable environment to support the handover process of the current SFP to Government ownership. To support this, the CO is engaging with other ministries as well such as MAFF and MOI. 162. The Royal Government of Cambodia is highly supportive of school feeding, and has shown its commitment through the development of new policies and by integrating HGSF activities into the Commune Investment Plans. The MoI has included a budget for salaries of cooks in schools where the SFP is Government owned. In addition, there is a strong network of supporting organisations and donors for the SFPs (including USAID and KOICA) 163. Community support for the SFP is also strong. Parents are highly aware of the benefits of the programme for their children and their household, and voluntarily provide in-kind contributions such as vegetables and firewood. Primary education is free in Cambodia and schools are not allowed to request financial contributions from parents, but in some communes, the School Committee does raise funds in the community. 164. However, there are a number of challenges and areas that need further attention. These include the ability of WFP CO and its staff to transition to technical assistance and capacity building only. For instance, the monitoring of the programme activities in the schools is mostly done directly by WFP and its cooperating partners at present, with limited involvement of Government staff. However, the provincial level authorities expressed the desire to be more involved in monitoring activities. They also highlighted the need for WFP monitoring visits to be more frequent, and in the spirit of coaching/mentoring in order to reorient actions, rather than merely to measure against targets. Across the board, stakeholders indicated concerns in their own ability to implement and manage the programme. High staff turnover, especially in the schools, including cooks, requires extra McGovern-Dole Programme Baseline Report – Cambodia – January 2021 40 |Page trainings in addition to those routinely implemented. The short time frame allocated to transition from ‘donorship’ via ‘partnership’ to ‘ownership’ (in the words of one key informant in the national Government), further aggravated by the school closures and postponing of programme activities caused by the COVID-19 pandemic, provides an additional challenge. In addition, Cambodia also remains prone to natural disasters such as floods and droughts which may impact programme implementation. A new timeframe of handover in 2025 has been agreed on by WFP and the Government. 165. Not only the transition process was affected by the COVID-19 pandemic, but also the education of the students (and thus their literacy achievements), as well as the other aspects of programme implementation. It can be expected that this was not only the case during SY 2019/20 but continuing into SY 2020/21, and possibly beyond. 166. It should be noted that the management of school meals, in particular HGSF, puts an extra burden on teachers and school management, whose main task is to ensure the quality of education. While they can see and experience the benefits of SFP, and do not object to the programme, they cannot be overloaded with non-teaching duties when educational outcomes are still far from desirable. 167. The Royal Government of Cambodia is currently rolling out a series of decentralization and deconcentration (D&D) reforms that will place greater responsibility with sub-national authorities to plan and deliver basic services. While aiming to streamline implementation of programmes at the sub-national level, for instance by targeting commune councils to jointly oversee daily SMP implementation represented through the LSFC, this is a long process that requires restructuring and reorganizing Government institutions. On one hand, the varying levels of D&D pose a challenge for smooth coordination and collaboration,115 especially because not all line ministries will roll out the D&D at the same pace and to the same extent, which may lead to more complicated decision making and coordination. On the other hand, the envisioned integration of district level education and health officials into the district governance will ultimately enable better coordination at the district level. In addition, increased decision-making power and programming budgets at the commune council level provides an opportunity for WFP capacity building to strengthen local participation and ownership. 168. Based on the interviews with a variety of stakeholders, it is clear that further capacity building is essential for a successful handover of the school feeding programme to the authorities. Government staff at national and sub-national levels expressed a need for capacity building in enabling environment and information management. Not all teachers and school managers have the skills and access to adequate technical infrastructure to use digital reporting, or the capacity for implementation and management of the school meals programme. 115 In the McGovern-Dole proposal, WFP explains that “by working with district-level education and health officials and leveraging the active district and local school feeding committees, WFP will strengthen capacities on school feeding operations and strengthen awareness on literacy and good nutrition practices”. ‘We need to go from Donorship via Partnership to Ownership’ Senior Government Official McGovern-Dole Programme Baseline Report – Cambodia – January 2021 41 |Page Key findings and conclusions – Question 4 • WFP has built extensive experience with school feeding programmes but will need to adapt to a different role, focused on capacity building rather than direct implementation. • WFP has a strong relationship with the MoEYS, and there is a network of donors, United Nations and other organisations that leverage the school feeding programme. • The Royal Government of Cambodia is highly committed to a national school feeding programme but is not ready to implement this without support from WFP and other partners. • The original timeline for handover was very short and the current COVID-19 pandemic has led to delays. WFP and the Government have agreed on a new timeline for 2025. Because of school closures, programme achievements for SY 2019/20 (and SY 2020/21) are likely to fall short of targets. • Stakeholders at all levels stressed their need for capacity strengthening, in particular related to information management and the management of the school feeding programme. • There is a need for more clarity on roles and responsibilities between implementing partners and WFP in monitoring activities. The involvement of the Government in monitoring also needs to be augmented. • The D&D will ultimately lead to improved coordination at the district level, but while it is being rolled out, it poses additional challenges such as differences in the level and speed of D&D, and subnational capacity to implement policies and programmes. 3. Conclusions and Recommendations 169. Based on the findings presented in the previous section, an overall assessment that responds to the evaluation questions is provided below. This is followed by four recommendations of how the WFP CO, the MoEYS and other stakeholders can take action to build on the lessons learned to date and address some outstanding factors to ensure the longer-term viability of this important programme. 3.1 Overall Assessment/Conclusions 170. The programme is relevant. The design of the McGovern-Dole programme, which has recently started its third round in Cambodia, is appropriate for the needs of women, men, girls and boys and aligned with Government priorities, as well as with USDA, WFP, United Nations and other partners. In particular, it is relevant to fulfil the targets of the ESP, social protection, food security and nutrition frameworks, and the commune development plans. The high prevalence of food insecurity among households of students indicates the school feeding programme is highly relevant. Overall, stakeholders expressed satisfaction with the programme, and the Government is invested in taking over the programme eventually. Parents feel the SMP relieves their burden by reducing both their expenditure for food and the time they need to spend to prepare breakfast for their child. 171. The indicators are feasible. The baseline collected data based on the indicators in the RFs and found that all indicators area feasible for use in monitoring and evaluation of this programme. However, the ET was not able to assess the extent to which the design of this round of the programme was based on a thorough updated needs assessment and stakeholder consultations, as well as a sound gender analysis. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 42 |Page 172. Appropriateness of targeting is not entirely clear. Inclusion criteria (poverty, poor education outcomes, high rates of malnutrition) are appropriate, and this baseline confirmed that mostly students from vulnerable households were targeted. The CO reported that they reviewed and selected the schools based on the combined criteria in 2013 for the earlier phase of the programme, and the same schools have been retained under the current grant. The selection was validated at provincial and district school feeding committee levels and at schools through consultations with the local SFCs. However, the school targeting process is apparently not clear to all stakeholders, indicating a need to ensure they are informed and aware of how and why certain schools are selected. It could not be assessed whether the observed differences between schools in terms of infrastructure and management capacity are due to targeting errors related to the use of outdated field assessment, or evidence of successful implementation of previous rounds of the programme, or of other interventions by the Government or other agencies. 173. Gender does not receive much attention in the design. Equal gender participation is clearly envisaged in the design and the programme presents opportunities to promote women’s empowerment (many female beneficiaries are included in all activities and there is female representation on the LSFCs). However, the programme design does not include gender-specific activities, expected results or indicators in the RF to measure and analyse gender roles, behaviours and expectations. In addition, the WFP SF policies lack specific explanations of how gender is to be considered in school feeding programmes. The ET concludes that the design of the programme was not based on any solid gender analysis and without agreeing uniform gender guidelines with the partners. Specific gender mainstreaming activities in the SMP, of crucial importance before handover to the Government, need to be done, and this omission from the design represents a missed opportunity for WFP and the Government’s learning. 174. Low literacy among Grade 2 students has led to downward revision of the targets. This illustrates the relevance of a programme that aims to improve literacy education, as well as its adaptability. The use of different EGRA tools to measure literacy in the three provinces is a matter of concern and indicates a need to improve the consistency of the approach. 175. The programme benefits from an enabling environment that encompasses favourable human, material and institutional aspects. Building on the foundation of two previous rounds of the McGovern-Dole programme, the baseline found that the SFP was familiar to and well received by stakeholders in the Government at all levels, as well as by the communities it serves. Overall, teacher and student attendance are high and already meet the programme targets, and supporting infrastructure is in place in most schools. The WFP CO enjoys a strong relationship with the MoEYS, and the Royal Government of Cambodia is committed to roll out a national school feeding programme. Several donors, United Nations organisations and NGOs implement complementary programmes that leverage the school feeding programme. 176. The partners have shown a flexible approach to feedback and problem solving. As the results of the EGRA in the WEI baseline were very low, the target for SO1, which was 55 percent in the previous round, was reduced to a more attainable 15 percent. Also, the handover timeline has been adjusted in response to the COVID-19 pandemic. This ability to adjust to changing circumstances is important to ensure effective and sustainable programme implementation, while maintaining good relationships with all stakeholders. 177. Implementation of the 2015 Sustainability Roadmap is on the right track. The Royal Government of Cambodia is invested in a national school feeding programme, as evidenced in national strategy. Since 2019, handover of the programme has started and is planned to continue over the course of this grant. A shift in modality to the HGSF-hybrid model or a full HGSF modality is planned for SY 2022/23 in preparation of the handover. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 43 |Page 178. However, the original timeline for handover is very short and the COVID-19 pandemic has impacted implementation in SY 2019/20 and will likely impact in SY 2020/21 as well, and programme achievements for these school years are likely to fall short of targets. The CO has taken appropriate action to discuss a transition plan that will supersede the current roadmap and aim for handover by 2025. 179. Stakeholders at all levels of Government, as well as at the commune and school levels, stressed their need for capacity strengthening in particular related to information management, and the management of the SFP. Improved coordination with regard to monitoring activities and agreement on divisions of roles and responsibilities were also mentioned as areas for particular focus. 180. This baseline indicated the need for organisational learning related to the design and implementation of the SFP, and the lack of gender guidelines. While many aspects of the programme are well designed and implemented, there is room for improvement, in particular related to gender and coordination. 3.2 Recommendations 181. Based on the findings and conclusions of this baseline study, five recommendations from the evaluation team are outlined below, structured by theme and timeframe. The target group for each recommendation is clearly identified. The recommendations are intended to offer suggestions to enhance the success of the programme. The ET considers all recommendations to be of high priority but each with a different timeframe. Recommendation 1: WFP CO should develop a Theory of Change and joint workplan for the programme, in close collaboration with MoEYS and other relevant ministries (MoH, MoI, MoSAVY and MEF), as well as other implementing partners. There is a need for increased coordination of efforts for projects operating in schools, with more comprehensive direction and coordination of programming by MoEYS and supporting funds from MEF. As a ToC was not developed at the outset, in order to support the implementation of the ongoing programme, WFP CO should develop one within six months, also considering the following components: • A Theory of Change to guide the implementation and monitoring of the programme activities; • A joint workplan and monitoring plan for all levels of all implementing agencies including the Government, with clearly indicated roles and responsibilities; • A clear and transparent communication mechanism to regularly share challenges, experiences and lessons learnt with all levels; • Regular capacity needs assessments in the annual monitoring and progress reports. Recommendation 2: The WFP CO should put more focus on Government capacity building and strengthening. Within the next 12 months, the CO should design a coherent and structured capacity building plan for the entire project cycle and based on thorough needs mapping, a review of the current situation, with clear key roles, responsibilities, time targets and indicators that measure how capacity is improved and enhanced over time. The plan should include strengthening Government monitoring capacity as a key priority. As this has a high priority in view of the limited time available, a focal point should be identified for the development of this plan and to oversee its implementation. 182. Recommendation 3: WFP CO, the MoEYS and other ministries should consider incorporating specific indicators measuring the progress of the transition in the monitoring plan. This will enable careful monitoring and adjusting during the transition process. As the transition process has already started, this should be done within the next six months. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 44 |Page 183. Recommendation 4: WFP CO, together with USDA, and in consultation with the Ministry of Women’s Affairs, should give greater emphasis on gender analysis to strengthen gender considerations in both design and implementation, by developing specific gender indicators to be integrated as part of the programme design (RFs). This will enable the CO to ensure that gender considerations are integrated into the programme activities and that the programme contributes to gender equity and women’s empowerment. The CO should ensure these indicators measure the progress in transformative changes, for not only for women and girls, but also for men and boys. This should be completed within the next six months so amendments to the design and targets can be identified and put in place. 184. Recommendation 5: The WFP CO should assess whether the targeting of the schools should be adjusted and harmonize/balance the support given (including the community contributions), based on a clear needs assessment. By providing a more transparent targeting approach, the CO will improve understanding and acceptance of the design by all stakeholders. The assessment of the needs of each school will lead to effective and efficient programme implementation and help prepare the schools for the process of transition to a nationally owned and implemented school feeding programme. To implement this recommendation, in-depth discussion with relevant Government departments and other stakeholders is essential. This should be done within the next 12 months. McGovern-Dole Programme Baseline Report – Cambodia – January 2021 45 |Page List of Acronyms ACL All Children Learning ACRC All Children Reading Cambodia Project BEEP Basic Education Equivalency Programme CARD Council for Agricultural and Rural Development CIP Commune investment plan CO Country office CSP Country Strategic Plan D&D Decentralization and deconcentration DEQAS Decentralized evaluation quality assurance system DoE District office of education DTMT District training and monitoring team EGR Early grade reading EGRA Early grade reading assessment ESP Education strategic plan ET Evaluation team FAO Food and Agricultural Organization of the United Nations FAS Foreign Agricultural Service FFE Food for education FGD Focus group discussions FLP Factory literacy programme GDI Gender development index GDP Gross domestic product GEWE Gender equality and women’s empowerment GGI Global gender gap index GII Gender inequality index GIZ Deutsche Gesellshaft für Internationale Zusammenarbeit GmbH GPE Global Partnership for Education HDI Human Development Index HGSF Home-grown school feeding HH household HQ Headquarters IRL Indochina Research Limited KAPE Kampuchea Action to Promote Education kg kilogramme KCG Kampong Chhnang KII Key informant interviews KTM Kampong Thom LRP Local and Regional Food Aid Procurement LSFC Local school feeding committee MAFF Ministry of Agriculture, Forestry and Fisheries MEF Ministry of Economy and Finance MGD Millennium Development Goals MoEYS Ministry of Education, Youth and Sport MoH Ministry of Health MoI Ministry of Interior MoSAVY Ministry of Social Affairs, Veterans and Youth Rehabilitation MoWA Ministry of Women’s Affairs mt metric tonne NER Net enrolment rate NGO Non-governmental organization McGovern-Dole Programme Baseline Report – Cambodia – January 2021 46 |Page NSFSN The National Strategy for Food Security and Nutrition NSPPF National Social Protection Policy Framework OEV Office of Evaluation (WFP) PBB Programme Based Budget Plan Plan International PMP Performance Monitoring Plan PoE Provincial Office of Education PRISM Platform for Real-Time Information System PSM Propensity Score Matching PTA Parent-Teacher Association QA Quality assurance RBB Regional Bureau Bangkok (WFP) rCSI Reduced Coping Strategies Index RF Results framework SABER Systems Approach for Better Education Results SBCC Social behaviour change communication SDG(s) Sustainable development goal(s) SF School feeding SMP School meal programme SO Strategic objective SRP Siem Reap SSC School Support Committee STEPCam Strengthening Teacher Education Programmes in Cambodia SY School year THR Take home ration ToR Terms of reference ToT Training of trainers TPD Teacher professional development UNDP United Nations Development Programme UNEG The United Nations Evaluation Group UNESCO The United Nations Educational, Scientific and Cultural Organization UNICEF The United Nations Children's Fund USDA United States Department of Agriculture US$ United States Dollar (currency) VAM Vulnerability Analysis and Mapping WASH Water, sanitation and hygiene WEI World Education Inc. WFP World Food Programme WHO World Health Organization WV World Vision McGovern-Dole Programme Baseline Report – Cambodia – January 2021 47 |Page WFP Cambodia Country Office https://www.wfp.org/countries/cambodia [ Phnom Penh, January 2021, Report number ]