Foreign Agricultural 
Service, United States 
Department of Agriculture
Côte d’Ivoire McGovern-Dole 
Project 
Midterm Evaluation
January 2019 
This publication was produced at the request of the United States Department of Agriculture. It was
prepared independently by IMPAQ International, LLC. 
DISCLAIMER: The author’s views expressed in this publication do not necessarily reflect the views of the 
United States Department of Agriculture or the United States Government.
Mid-Term Evaluation of “Support for the Integrated School 
Feeding Program” in Côte d’Ivoire
The United States Department of Agriculture (USDA) McGovern-Dole (MGD) project in Cote 
D’Ivoire was implemented by World Food Programme (WFP) from 2015 to 2020. The project’s 
primary objective is improving nutrition and health outcomes and increasing the literacy of 
125,000 school aged children in 613 rural public primary school in seven high-priority regions of 
Côte d’Ivoire.
Agreement Number: No. FFE-681-2015/006-00
Project Duration: 2015-2020
Implemented by: WFP
Evaluation Authored by: IMPAQ International, LLC
Individual Authors
Deichsel Sakari, Research Analyst
Maria DiFuccia, Qualitative Lead
Michaela Gulemetova, Team Leader
Edoxi Kindané, Field Manager
Marc Masson, Research Analyst
Elnaz Safarha, Quantitative Lead
Prepared 
Decentralized Evaluation
Mid-Term Evaluation of “Support for the Integrated School 
Feeding Program” in Côte d’Ivoire
September 2016 to June 2018
Evaluation Report
January 2019
World Food Programme Country Office in Côte d'Ivoire 
Evaluation Manager Moyabi Sylla
Prepared by IMPAQ International, LLC
Sakari Deichsel, Research Analyst
Maria DiFuccia, Qualitative Lead
Michaela Gulemetova, Team Leader
Edoxi Kindané, Field Manager
Marc Masson, Research Analyst
Elnaz Safarha, Quantitative Lead
DECENTRALIZED EVALUATION FOR EVIDENCE-BASED DECISION MAKING
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) i | Page 
Acknowledgements
The authors would like to thank the World Food Programme (WFP) Country Office in Côte 
d'Ivoire and the U.S. Department of Agriculture (USDA) for their financial support. The authors 
would also like to thank the Steering Committee and the Technical Committee of the project, 
as well as AVSI, the Ministry of National Education, Technical Education and Vocational Training 
(MENETFP), the Directorate of School Canteens (DCS), the Ministry of Agriculture and Rural 
Development (MINADER), and the National Agency for Support to Rural Development
(ANADER) for supporting and facilitating the rollout of data collection.
The authors have very much enjoyed their collaboration with WFP and would especially like to 
thank Moyabi Sylla, Alti Bema, and Adeyinka Badejo-Sanogo. The authors also highly 
appreciated the welcome provided by Elly Bahati, Lorenzo Manzoni, and the whole team at 
AVSI. The authors thank Rébéka Kakou for her support and expertise doing qualitative 
interviews, particularly with the women’s production groups. The authors thank Mont Horeb 
for their professionalism and flexibility with data collection. The authors also thank all the 
enumerators for their excellent work in the field. The authors also acknowledge Guy Kacou, 
the head of research at Mont Horeb, for providing technical insights on cognitive testing of the 
instruments and managing the fieldwork team. At IMPAQ, the authors thank their colleagues 
Drs. Sara Borelli and Kajal Gulati for their technical expertise during the evaluation design and 
analysis, and Jonathan Simonetta for providing guidance and quality control.
Disclaimer
The opinions expressed in this report are those of the authors, and do not necessarily reflect 
those of WFP or USDA. Responsibility for the opinions expressed in this report rests solely with 
the authors. Publication of this document does not imply endorsement by WFP or USDA of the 
opinions expressed.
The designation employed and the presentation of material in maps do no 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.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) ii | Page 
Table of Contents
Table of Figures...................................................................................iii
Executive Summary ............................................................................. v
Context..................................................................................................................................................... v
Methodology.......................................................................................................................................... vi
Key Findings ......................................................................................................................................... vii
Recommendations................................................................................................................................ x
1. Introduction.................................................................................... 1
1.1. Overview of the Evaluation Subject.............................................................................. 1
1.2. Context..................................................................................................................................... 4
1.3. Evaluation Methodology and Limitations..................................................................... 5
2. Evaluation Findings ........................................................................ 9
2.1. Evaluation Criteria 1 – Relevance.................................................................................. 9
2.2. Evaluation Criteria 2 – Effectiveness ..........................................................................11
2.3. Evaluation Criteria 3 – Efficiency .................................................................................15
2.4. Evaluation Criteria 4 – Impact......................................................................................17
2.5. Evaluation Criteria 5 – Sustainability .........................................................................32
3. Conclusions and Recommendations ........................................... 37
3.1. Overall Assessment/Conclusions..................................................................................37
3.2. Lessons Learned and Good Practices..........................................................................37
3.3. Recommendations..............................................................................................................38
Annex 1: Evaluation Matrix............................................................... 43
Annex 2: Evaluation Methodology....................................................48
Annex 3: Documents Reviewed / Bibliography .............................. 55
Annex 4: Stakeholders Interviewed................................................. 57
Annex 5: Data Collection Tools......................................................... 58
Annex 6: Additional Figures.............................................................. 59
Annex 7: Project Results Framework............................................... 76
Annex 8: Terms of Reference............................................................ 79
Annex 9: MGD Indicators ..................................................................80
Annex 10: Map of MGD School Feeding Program............................ 87
Annex 11: List of Abbreviations .......................................................88
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) iii | Page 
Table of Figures
Figure 1: Project Effects Using DID Method ................................................................................. 19
Figure 2: Reading Assessment Scores by Grade (Midline Only) ...................................................... 20
Figure 3: Reading Proficiency by Grade and Gender (MGD Only) .................................................... 20
Figure 4: Teacher Absences by Region (Midline Only) ................................................................... 21
Figure 5: Student-Reported Teacher Absences (Midline Only) ........................................................ 21
Figure 6: Access to School Materials by Region............................................................................ 22
Figure 7: Access to Reading Materials (Mobile Library, Reading Board, etc.) .................................... 22
Figure 8: Teachers Having Participated in AVSI Trainings by Region (Midline MGD) .......................... 22
Figure 9: Schools Having Received Training on Teaching of Reading............................................... 23
Figure 10: Number of Students out of 10 Described as Attentive by Teachers on a Typical Day (Midline 
MGD Only) .............................................................................................................................. 23
Figure 11: Student Attentiveness by Region (Midline MGD Only).................................................... 23
Figure 12: Students with Perfect Attendance (Midline Only) .......................................................... 24
Figure 13: Average Student Enrolment per Grade (All Schools)...................................................... 24
Figure 14: Ratio of Girls to Boys Enrolled in CM1 & CM2 (MGD Schools).......................................... 25
Figure 15: Frequency that Students Ate at the School Canteen...................................................... 26
Figure 16: Portion of Days that Canteen Provided Minimum Dietary Diversity .................................. 26
Figure 17: Teacher Training on Health Practices........................................................................... 27
Figure 18: By Region, Canteen Managers Know Three Food Storage and Food Preparation Best Practices
............................................................................................................................................. 27
Figure 19: Type of Access to Water in Schools............................................................................. 28
Figure 20: Distance to School’s Water Source, Where Available (Midline MGD Only) ......................... 29
Figure 21: Most Common Problems with Schools’ Water Source..................................................... 29
Figure 22: Access to and Quality of Sanitary Facilities .................................................................. 29
Figure 23: Sufficient Food Preparation Equipment........................................................................ 30
Figure 24: Sufficient Food Storage Equipment ............................................................................. 30
Figure 25: Stove Type in the Canteens ....................................................................................... 31
Figure 26: Stove Type in the Canteens, by Region ....................................................................... 31
Figure 27: Student Sample Distribution by Gender, Grade, and Type of Respondent ........................ 59
Figure 28: Reading Proficiency by Grade..................................................................................... 59
Figure 29: Teacher Attendance According to School Records and Self-Reported............................... 59
Figure 30: Self-Reported Teacher Absences by Region (Midline Only) ............................................. 60
Figure 31: Access to Reading Materials (Mobile Library, Reading Board, etc.) by Region (Midline MGD 
Only)...................................................................................................................................... 60
Figure 32: Usefulness of AVSI Trainings (Midline MGD Only) ......................................................... 60
Figure 33: Number of Students out of 10 that were Attentive, According to Teachers (Midline MGD 
Only)...................................................................................................................................... 61
Figure 34: By Region, Number of Students out of 10 that were Attentive, According to Teachers (Midline 
MGD Only) .............................................................................................................................. 61
Figure 35: By Region, Average Student Recorded Attendance (Midline Only) ................................... 61
Figure 36: Average Student Enrolment per Class (MGD Schools with Take Home Ration Schools 
Separated).............................................................................................................................. 62
Figure 37: Average Student Enrolment per Class (MGD Schools with Take Home Ration Schools 
Separated).............................................................................................................................. 62
Figure 38: Ratio of Girls to Boys Enrolled by Grade (MGD Schools with Take Home Ration Schools 
Separated).............................................................................................................................. 62
Figure 39: Ratio of Girls Enrolled to Boys Enrolled by Grade (MGD Schools with Take Home Ration 
Schools Separated) .................................................................................................................. 63
Figure 40: Number of Days Students Missed Due to Illness (Midline Only) ...................................... 63
Figure 41: Students Missing More than 10 Days Due to Illness (Midline Only).................................. 63
Figure 42: Foods Eaten in Household during the Previous Day ....................................................... 64
Figure 43: By Region, Foods Eaten in Household During the Previous Day (MGD Midline).................. 64
Figure 44: By Region, Foods Eaten in Household during the Previous Day (MGD Midline) .................. 65
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) iv | Page 
Figure 45: Coping Strategies Used by Household ......................................................................... 65
Figure 46: By Region, Days* Coping Strategies Used (Midline MGD)............................................... 66
Figure 47: By Region, Percent* Using Coping Strategies (Midline MGD) .......................................... 66
Figure 48: By Region, Percent* Using Coping Strategies (Midline MGD) .......................................... 67
Figure 49: Percent of School Days during which Canteen Operated by Region ................................. 67
Figure 50: By Region, Percent of School Days during which Canteen Operated ................................ 68
Figure 51: Parents’ Awareness of the School Canteen Project........................................................ 68
Figure 52: Students Eating through the School Canteen Program................................................... 68
Figure 53: Frequency that Students Eat at the School Canteen...................................................... 68
Figure 54: Foods Used to Prepare Canteen Meals (Midline Only) .................................................... 69
Figure 55: By Region, Canteen Managers Can Cite Three Health & Hygiene Practices ....................... 69
Figure 56: By Region, Canteen Managers Can Cite Three Health & Hygiene Practices ....................... 69
Figure 57: Teacher Training on Health Practices........................................................................... 70
Figure 58: Nutrition-Related Trainings at School .......................................................................... 70
Figure 59: Type of Access to Water of the School......................................................................... 70
Figure 60: Distance to School’s Water Source, Where Available (Midline MGD Only) ......................... 70
Figure 61: Most Common Problems with Water Source ................................................................. 71
Figure 62: Access to and Quality of Sanitary Facilities .................................................................. 71
Figure 63: Deworming Pills and Micronutrient Pills Distributed to Students ...................................... 71
Figure 64: Deworming Pills and Micronutrient Pills Distributed to Students ...................................... 71
Figure 65: Access to Food Storage and Equipment....................................................................... 72
Figure 66: Access to Food Storage and Equipment....................................................................... 72
Figure 67: Sufficient Food Preparation Equipment........................................................................ 72
Figure 68: Sufficient Food Storage Equipment ............................................................................. 73
Figure 69: Stove Type .............................................................................................................. 73
Figure 70: Stove Type by Region ............................................................................................... 73
Figure 71: Reading Assessment Scores by Treatment Group, Phase, and Grade............................... 74
Figure 72: Food Consumption Score Group by Gender of Head of Household ................................... 75
Figure 73: Household Dietary Diversity Score by Gender of Head of Household ............................... 75
Figure 74: Reduced Coping Strategies Index by Gender of Head of Household*............................... 75
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) v | Page 
Executive Summary
1. This evaluation report is for the mid-term evaluation of the “Support for the Integrated 
School Feeding Program” in Côte d’Ivoire, a school feeding and early grade reading 
project implemented by World Food Programme (WFP) and funded by the McGovern￾Dole (MGD) International Food for Education and Child Nutrition Program of the United 
States Department of Agriculture (USDA). Under MGD, the USDA awarded USD 
35,678,500 to WFP over five years, which includes an in-kind donation of 24,600 metric 
tons (MT) of food. The project’s primary objective is improving nutrition and health 
outcomes and increasing the literacy of 125,000 school aged children in 613 rural public 
primary school in seven high-priority regions of Côte d’Ivoire.
2. This mixed-methods evaluation is commissioned by the WFP Country Office in Côte 
d'Ivoire and covers the period from September 2016 to June 2018. The two main 
objectives of this mid-term evaluation are to: 
• Assess and report on the progress made in achievement of the results, compared 
with the baseline, for the school meals project as it was implemented during the 
first two years.
• Assess and identify key achievements and challenges, draw lessons, and identify 
best practices for learning. It provides evidence-based findings to enlighten 
operational and strategic decision-making, improve in partnership coordination, and 
inform sustainability. Findings from this evaluation will be actively disseminated, 
and lessons will be incorporated into relevant lesson sharing systems. 
Context
3. Since 1989, to encourage enrolment in primary school and address the challenge of 
lunchtime hunger at school, the Government of Côte d’Ivoire (GoCI) has administered 
a national school feeding program with the support of WFP. In 2000, the GoCI integrated 
nutritional dimensions in addition to educational objectives, aiming for sustainability by 
encouraging the production of local communities in the form of women’s production 
groups (WPGs). This national program, entitled the “Integrated Program for the 
Sustainability of School Canteens” (PIPCS), aims to address the problems of chronic 
child malnutrition, which is high at 23.2 percent for boys and 19.9 percent for girls 
under 5 years old, and poor performance in primary education, with 63 percent of the 
population being illiterate1 and only 75.1 percent of children finishing primary school in 
2016.2 It encourages girls' education and ensures their retention in school. This national 
program resulted in the operation of 5,708 school canteens across the country in 2014-
2015, providing hot and balanced meals to 1,104,138 elementary school children.
4. However, as indicated in the GoCI’s national school feeding policy (2018-2025) and 
strategy (2018-2022), the provision of food to the canteens remains a significant 
challenge. In 2016-2017, the government’s program was only able to deliver 18 days 
out of 100 school days of food to the canteens.3 WFP complements the GoCI’s efforts
through its MGD funding, which covers a total of 613 rural public primary schools with 
125,000 students benefiting from hot and balanced meals across seven priority regions 
 
1 MICS 2016: Enquête a indicateurs multiples, Côte d’Ivoire. Abidjan : Institut National de la Statistique 2 MENETFP/DSPS, 2017: Statistiques scolaires de poche 2016-2017, MENETFP. 3 MENETFP/DCS. Stratégie Nationale d’Alimentation Scolaire en Côte d’Ivoire (2018-2022), Côte d’Ivoire. Abidjan. P. 
13
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) vi | Page 
(Poro, Bagoue, Tchologo, Bounkani, Gontougo, Bafing, and Cavally). In addition, 10,000 
girls in upper grades (CM classes) in three regions (Bagoué, Poro, and Tchologo) benefit 
from take-home rations each year (a total of 50,000 girls throughout the project). 
Through this effort, MGD-supported schools were able to deliver on average 72 out of 
100 school days of food during the 2016-2017 school year.4 Concurrently, WFP provides 
technical assistance to the GoCI to enhance the sustainability of the national school 
meals program.
5. The main components of the MGD project are: 1) providing school meals to primary 
school students; 2) delivering take-home rations for girls in upper grades (CM1/CM2);
3) distributing deworming tablets and micronutrients, 4) improving student literacy 
education; 5) training canteen management staff in the use of good health and food 
practices; 6) providing cutlery and equipment for the preparation and storage of food 
to the canteens; 7) constructing improved stoves and training canteen staff in their 
use; and 8) building capacity (both technical and financial) of women’s production 
groups (WPGs) to contribute production to the school canteens of their villages or 
communities.
6. The two main planned outcomes of the intervention are improvement in the literacy of 
school-aged children and increased use of health and dietary practices. These two 
outcomes are expected to result from a combination of more consistent teacher 
attendance, better access to school supplies and materials, improved access to learning 
materials for reading, and increased skills and knowledge of teachers. The other 
outcomes that fall under the increase use of health and dietary practices include 
improved knowledge of health and hygiene practices, safe food preparation and storage 
practices, nutrition, and the use of improved stoves, as well as access to preventive 
health interventions. 
Methodology
7. The evaluation was designed to evaluate the “Support for the Integrated School Feeding 
Program”, financed by the USDA’s MGD Program, against the following evaluation 
criteria: relevance, effectiveness, efficiency, impact, and sustainability (the 
comprehensive set of evaluation questions are found in the conceptual framework in 
Annex 1). In order to respond to these questions, the evaluation team employed 
quantitative surveys (ultimately reaching 104 schools, among which 68 treatment [i.e. 
MGD] and 36 control [i.e. non-MGD]) and qualitative focus group discussions (FGDs) 
with beneficiaries, along with key informant interviews (KIIs) with top-level project 
stakeholders. As part of its FGDs with beneficiaries, the evaluation team met with 10 
WPGs supported by the MGD project since 2017 (designated as "established") and 10 
WPGs supported since 2018 (designated as "new"). For the impact evaluation, the 
evaluation team designed a difference-in-differences (DID) method to compare the 
changes in reading proficiency between the population of beneficiaries (the treatment 
group) and the population that is not benefiting from the project (the comparison 
group).
8. Limitations and challenges of the evaluation included inaccurate or missing data from 
baseline, a delay in data collection until the beginning of the school year 2018-2019, 
and difficulties following the same subsample over time. However, measures were taken 
to mitigate these issues as much as possible, including revising the quantitative tools 
 
4 Ibid.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) vii | Page 
developed by the baseline evaluator to more accurately capture the required MGD 
indicators; developing and collecting data with a teacher survey to improve the key 
performance indicator measurements; and measuring the project effect on literacy 
outcomes for at least one year, instead of two, at midline to maintain statistical power. 
Key Findings
9. The key findings of the evaluation team are summarised below, structured according to 
the evaluation criteria. Results related to relevance, effectiveness, efficiency, and 
sustainability are primarily informed by the qualitative analysis, while the results related 
to impact are primarily informed by the quantitative analysis.
Relevance
10. The project is highly integrated into and complements the government’s efforts. 
Beneficiaries agree that the food delivery, school improvements, and literacy activities 
aspects of the intervention are relevant to their needs.
11. While the project aligns with Government priorities and was developed in collaboration 
with the MENETFP, which concurred with the project proposal prior to submission, some 
representatives from the DCS said they were not consulted during the final stages of 
the project design phase. Because of this, these representatives believed that several 
activities were not realistically planned given the timeframe and resources.
12. By encouraging or allowing children to attend school with the promise of a lunchtime 
meal, the MGD project is also directly relevant to the needs of parents of the students 
by reducing the time they needed to devote to childcare during the school week–
increasing the amount of time they can devote to their livelihoods.
13. All the WPGs reported that receiving materials and training was relevant and highly
appreciated. Women in the more established groups appreciated that prior to 
distributing materials, representatives met with them to determine their needs. This 
ensured that the equipment received was directly targeted and relevant to the individual 
groups.
14. The literacy component of the project was welcomed by parents; they said that 
education is the key to future success for both their sons and daughters. Many talked 
about their own challenges not knowing how to read and write and hoped for something 
better for their children.
Effectiveness
15. The absence of a regularly provisioned school canteen, seen as a major challenge which 
previously discouraged or prevented children from attending school or staying for the 
full day, was addressed under WFP’s MGD project according to the school management 
committees (COGES) and parents.
16. Parents appreciated not having to prepare meals for their children, thus freeing their 
time for other livelihood activities. 
17. Members in the established5 WPGs stated that their production has grown substantially 
due in part to the technical support received through the project. All groups were able 
to increase their contributions to the school canteens from last year to this year. 
 
5 The sample of WPGs contains 10 groups supported by the MGD project since 2017 (designated as "established") and 
10 groups supported since 2018 (designated as "new").
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) viii | Page 
18. Overall, parents were happy with the quality of the teachers in MGD schools, the 
education their children receive, and the progress they are making in school.
19. Some planned activities are behind schedule, such as the support to WPGs and the 
training of COGES. These delays were explained by ambitious expectations in mobilizing 
additional funding. In addition, threats to implementation include the low financial 
means of beneficiaries, who report not being able to afford school-related costs, 
including school uniforms, textbooks, and the daily 25 FCFA6 canteen fee. Other factors 
influencing project effectiveness include poor school infrastructure (lack of running 
water or latrines and insufficient school table-benches and classrooms) and teacher 
absenteeism.
20. The WPGs still rely on men, as there are several agricultural activities that are labour 
intensive and typically done by males. For some of the established WPGs, integrating 
youths to do this work has been successful, especially with the support of the village 
chief. In other groups, however, the members said their efforts to incentivize younger 
community members has been unsuccessful so far.
Efficiency
21. Overall, the project is efficient as it provides beneficiaries with school meals, take home 
rations (THR), and literacy instruction. 
22. There are some activities where the complementary funds have not been mobilized as 
anticipated, specifically the capacity building activities for WPGs and COGES. As a result, 
WFP made strategic choices that benefited the primary components of the project at 
the expense of secondary components.
23. Some government stakeholders and local community members suggested that by 
providing meals every day, the project discourages the community from taking 
ownership of the canteens.
24. It is difficult to monitor the project using school records. Student and teacher 
attendance reports appear to be inflated, as the official reports show almost no 
absenteeism, conflicting with the qualitative results. Canteen records did not appear 
inflated, but some were missing or incomplete.
25. External factors influencing the efficiency of the WPGs include climate variation and lack 
of irrigation systems. Other factors include the low incomes, high-illiteracy rates, and 
lack of management skills of the community members, which prevent them from fully 
executing certain activities.
Impact
26. The impact analysis shows that the MGD intervention significantly improves the reading 
proficiency of students. Exposure to the MGD project increases the reading assessment 
literacy scores of students in treatment schools at midline by 1.2 points (p<0.01) 
relative to comparison group, which is approximately one-level increase in the reading 
assessment. When looking only at the students who have been exposed to the project
for two years (CP1 students from baseline who are now in CE2 in 2018-2019), the 
estimated impacts are somewhat larger (1.7 points).
 
6 The 25 FCFA canteen fee per child/day is mandated by the Government of Cote d’Ivoire to cover the salary of canteen 
workers, firewood, and condiments, with a portion also going to the IEP and COGES for activity monitoring. It is 
equivalent to roughly 0.05 USD.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) ix | Page 
27. The project effect is twice as great for male students, who show almost two levels of 
improvement in reading proficiency, compared to one level for girls. Girls and boys have 
similar levels of reading proficiency in CP1 and CP2, but starting in CE1, the gains for 
boys are much greater.
28. Enrolment increased at MGD schools for girls in CM grades where THR are offered. In 
the three regions with THR, the average enrolment increased by almost 10 girls in the 
upper grades (from 33.9 CM 1 and CM2 girls at baseline to 42.5 students at midline). 
In contrast, in regions where THRs are not offered, girls’ CM1 and CM2 enrolment did 
not increase as much (from 17.7 to 22.2). The ratio of girls to boys enrolled in these 
two grades increased from 0.88 to a near-parity ratio of 0.96 in THR schools, while the 
ratio increased from 0.7 to 0.81 with no THR.
29. The percentage of school days during which food was provided was nearly 100 percent
in each region, but varied slightly, ranging from to Poro (88 percent) to Bafing (100
percent). There were no substantive differences between the days of canteen services 
available to boys and girls. This was a large increase from baseline, where schools on 
average only provided meals on 57 percent of the days when school was in session. 
30. Almost all parents (94 percent) of children enrolled at MGD schools were aware of the 
school canteen project, and almost all (92 percent) reported that their children ate at 
the canteen in the previous school year. Over 80 percent of students reported eating at 
the canteen regularly (three days a week or more).
31. Most MGD canteens served food containing the minimum dietary diversity on more than 
80 percent of school days at midline, compared to 40 percent of days at baseline. 
32. While most canteen managers were able to identify three or more health and hygiene 
practices at midline (88 percent), the number had decreased slightly from baseline (93
percent).
33. Among canteen managers who reported that there was a canteen management 
committee at their school, 33 percent indicated that they received nutrition training, 
compared to 41 percent at baseline. Teachers in more than half of MGD schools (54.4
percent) received trainings on nutrition.
34. About half of both MGD (53 percent) and non-MGD (50 percent) schools did not have 
any adequate access to water, similar to the baseline numbers. In MGD schools, about 
one-third (32 percent) of those with water access had a source on the school grounds, 
while 43 percent had a water source less than a 15-minute walk. The most common 
problem preventing access to water in both MGD (59 percent) and non-MGD (50
percent) schools was broken water pumps.
35. Slightly more MGD schools (59 percent) than non-MGD schools (50 percent) have 
sanitary facilities on-site. Less than one-third (31 percent) of MGD schools and only 
one-quarter (25 percent) of non-MGD schools have gender-separated sanitary 
facilities.
36. The number of MGD schools receiving deworming treatments decreased from baseline 
to midline (from 72 to 64 percent), as did the number of schools receiving micronutrient 
supplements (53 to 26 percent).
37. The percentage of canteen managers who reported having sufficient food preparation 
equipment increased from baseline to midline (from 19 to 32 percent), as did sufficient 
food storage equipment (from 10 to 38 percent). 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) x | Page 
38. Across all regions, 46 percent of canteens had an improved stove with a chimney, 
instead of just a traditional stove or a furnace without a chimney. This includes 100
percent of canteens in Cavally and Tchologo where there is a stove. This is large 
increase from baseline, where only 7 percent of school canteens had an improved stove. 
Sustainability
39. While it is still early to fully assess the impact of the WPGs and whether this component 
of WFP’s sustainability plan is viable, the members of the most established WPGs which 
began receiving project support in 2017 have indicated that their work has started to 
have an effect. In the ten established groups, the members have substantially increased 
their canteen contributions from 2017 to 2018. Most of these groups reported that their 
generated income was now sufficient to purchase their own seeds, fertilizer, and 
equipment maintenance without needing external financial support from the project. 
40. In the established groups, women reported the material and trainings received through 
the project is making their groups stronger and therefore more likely to sustain 
contributions after the project is over. For example, the tricycles, grain mills, and scales 
save time and money, and the financial training and market access has allowed them 
to sell produce for a higher price at the markets.7
41. For all WPGs, social networks are created, and members rely on each other as a coping 
strategy mechanism in the case of adverse social shocks, such as an illness or death in 
the family. In the more established groups, this social support goes even further, thanks 
to the money gathered through income generating activities and to the financial 
management training received as part of the project. These groups are able to give 
small loans to their members. These auxiliary activities develop WPGs as community 
institutions, with functions beyond supporting their school canteen, increasing their 
sustainability along with the benefits they provide to their communities.
42. The factors preventing full effectiveness of the project’s WPGs in supporting their school 
canteen are the same as those that hinder sustainability. Lack of access to physical and 
natural capital prevent WPGs from becoming stronger and autonomous, while climate 
variability and the associated shocks threaten their production, both for their members 
and the school canteen they support. 
43. Turnover is a threat to sustainability at all levels–at both the local level (cooks, COGES, 
school canteen management committees, WPGs, and teachers), and at the regional and 
national level (regional advisors, elected government officials, and WFP/partner staff). 
Strengthening the capacities of local communities (COGES and school canteen 
committees, in addition to the WPGs) and government partners will increase the 
likelihood of the project achieving sustainable, long-term impact.
Recommendations
44. Based on the findings and conclusions of this evaluation, the recommendations of the 
evaluation team are outlined below. The target group for each recommendation is 
clearly identified. The opinions expressed in these recommendations are those of the 
authors, and do not necessarily reflect those of WFP or USDA.
45. Recommendation 1: Strategically Reduce the Proportion of Canteen Days 
Covered in MGD Schools. Currently, WFP provides meals throughout the entire 120-
 
7 The MGD project’s support to WPGs takes into account the specific needs of the groups and as such, the material 
and training WPGs receive may differ.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) xi | Page 
day school term. As heard from government stakeholders and local beneficiaries, this 
discourages parents and WPGs from contributing money and/or food to the canteens. 
At the same time, few WPGs reported being close to being able to adequately support 
a school canteen without the MGD project’s support and it is unclear when they will be 
able to.8 Considering the critical importance of WPGs in the project’s sustainability plan, 
WFP should immediately perform an assessment, informed by other projects supporting 
WPGs in similar contexts, to estimate the time needed for WFP-supported WPGs to fully 
support a school canteen and determine when WPGs could realistically be expected to 
graduate from the project and under what conditions. After evaluating what local 
communities can support, WFP should integrate these findings into the transition plan 
it is developing with the GoCI and indicate how, when, and in which schools and regions 
school feeding will be transitioned to the GoCI’s national program. WFP should start 
reducing the dependence of canteens on MGD commodities by the start of the 2019-
2020 school year, replacing these commodities with GoCI and WPG contributions, and 
find other beneficial uses for the surplus commodities, such as increasing the number 
of girls receiving THRs by expanding this offer to all 7 project regions. 
46. Recommendation 2: Reinforce DCS and DPFC M&E Capacity at the National and 
Local Level. The MGD project’s monitoring network relies heavily on the GoCI’s own 
network of local and regional government school feeding and teaching advisors to 
collect indicator data. However, the data collected through this system does not include 
key MGD indicators, particularly on student attendance and their absences due to 
sickness, teacher absences, and food diversity and security, which cannot be effectively 
calculated through evaluation. In addition, while the MGD project delivers annual M&E 
training to DCS and DPFC, several national-level DCS personnel requested that WFP 
provide them with additional M&E training and capacity building to help them better 
support the MGD project and other school feeding related projects. By the end of the 
2018-2019 school year, WFP personnel should host a workshop with M&E personnel 
from DPFC and DCS to jointly reassess their training needs, develop a training plan that
will respond to them, and consider strategies to improve MGD monitoring efficiency and 
effectiveness. Through this workshop, the MGD implementing partners should address 
the issue of accurately monitoring teacher and student attendance data, which remains
a challenge, and the possibility of consolidating government monitoring activities where 
relevant. By the start of the 2019-2020 school year, WFP should support DCS and DPFC 
in modifying existing data collection tools and methods for indicators relevant to project 
implementation and reporting, while also consolidating existing school-level data 
collection visits across the school feeding and literacy components to more efficiently 
and frequently visit schools.
47. Recommendation 3: Improve Field Presence and Beneficiary Interaction. There 
are currently only eight WFP field monitors who are responsible for over 600 schools. 
They are working with government school feeding advisors who, as best they can, 
mobilize communities, supervise day-to-day canteen management, and collect indicator 
data, but have limited capacity in actuality. As WFP is trying to build both local capacity 
and long-term sustainability, they should consider working with the appropriate
government stakeholders and local partners to increase the capacity of the local and 
regional school feeding advisors, as described in Recommendation 2. However, this 
 
8 Through WFP technical assistance, WPGs are expected to incrementally increase their production while continuing 
to contribute 1/3 to their associated school canteen, with the end goal of fully supporting the canteen in conjunction 
with the Government’s national school canteens program.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) xii | Page 
does not replace regular, direct interaction between WFP and its project beneficiaries,
which is essential for responding to local needs in a timely manner, more accurately 
capturing M&E data, and effectively disseminating innovations in the field. To efficiently 
improve its field presence and increase its touch-points with beneficiaries, the MGD 
project should consider using the data collection networks, agents, and activities
already being used for the literacy component, such as community focal points and 
school monitoring visits, to more frequently visit schools and communities for other 
components of the project. 
48. Recommendation 4: Increase Coordination, Communication, and Planning 
with the DCS. While the MGD project is deeply integrated with the GoCI, WFP could 
improve its advance planning of activities involving the DCS, with a clear and 
transparent view of the available funds and the project timeline. While DCS
representatives participated in meetings to develop WFP’s proposal for the MGD project, 
multiple DCS key informants indicated that they were not informed of some late stage 
meetings, where they could have shared threats to some of the proposal’s assumptions. 
For example, these informants thought that the WPGs, a key component of this project, 
are not yet supported at sufficient number to provide project sustainability partly due 
to insufficient funds stemming from underestimating the cost of supporting WPGs. 
Having supported WPGs for nearly two decades, the DCS and ANADER has an intimate 
knowledge of these groups. The DCS informants thought their feasibility, given the 
resources and timeline, could have been more accurately described if they were fully 
included in discussions during the entire planning phase. Several DCS representatives 
also said that they were forced to start activities later than anticipated due to delays or 
uncertainty in receiving funds and were sometimes unaware of what MGD funding is 
allotted for their work. WFP should more frequently include DCS, DPFC, and other GoCI 
partners in conversations on the design, budgeting, and implementation of MGD project 
activities. In turn, the GoCI should ensure representatives from all relevant 
departments within it participate in the MGD project meetings organized by WFP to 
review and approve the tasks and budgets pertaining to their work.
49. Recommendation 5: Reinforce Girls’ Education. The results of the reading 
assessment showed that male student improved twice as much as female students. 
AVSI should work with DPFC and its field staff to adjust its programming to ensure girls 
benefit from the project as much as boys. While no beneficiaries in the FGDs said that 
girls were discriminated against at school, the evaluation team recommends that, by 
the start of the 2019-2020 school year, AVSI should conduct a study to explain these 
results and subsequently adapt its approach to ensure girls participate, read, and learn 
at the same rate as boys. Considering that teachers reported that girls in their class 
were roughly as attentive as boys, the evaluation team suggests that AVSI and DPFC 
consider integrating or reinforcing a gender focus in their teacher training curriculum.
In addition, considering that THRs have shown to be effective in the three regions where 
it was offered in increasing girls’ attendance and keeping them in school, WFP should 
consider using a surplus MGD commodities to expand the THR offering to all seven 
regions by the start of the 2019-2020 school year.
50. Recommendation 6: Increase Female Representation in COGES. WFP targeted 
COGES to include 50 percent women as members, however, most in the evaluation 
team’s sample did not include this many women, and in fact, some COGES were 
exclusively male. By the start of the 2019-2020 school year, WFP should encourage 
women to participate in its anticipated basic literacy training for WPGs and canteen 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) xiii | Page 
management committees to also join their school’s COGES, as well as make additional 
efforts to sensitize COGES of the need to integrate women in their leadership.
51. Recommendation 7: Strengthen COGES Capacity to Support MGD Activities. The 
COGES help their school and its canteen in many ways which could be enhanced by 
expanded MGD training and support. All COGES agreed they could use additional 
training in managing COGES activities, including accounting, commercial farming, and 
fundraising. While the current MGD project is designed to offer only limited support to 
COGES, by the end of the project, to the extent feasible WFP should explore offering 
training and financial support to COGES in developing commercial activities or building 
infrastructure for the school, either directly or through partnerships with relevant 
Government structures or UN agencies and other organizations. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 1 | Page 
1. Introduction
52. This evaluation report is for the mid-term evaluation of the “Support for the Integrated 
School Feeding Program” in Côte d’Ivoire. This evaluation is commissioned by the World 
Food Programme (WFP) Country Office in Côte d'Ivoire and will cover the period from 
September 2016 to June 2018.
53. The two main objectives of this mid-term evaluation are for accountability and for 
learning purposes. In particular, this evaluation will: 
• Assess and report on the progress made in achievement of the results, compared 
with the baseline, for the school meals project as it was implemented during the 
first two years. 
• Assess and identify key achievements and challenges, draw lessons and identify 
best practices for learning. It will provide evidence-based findings to enlighten 
operational and strategic decision-making, improvement in partnership 
coordination, and sustainability. Findings from this evaluation will be actively 
disseminated and lessons will be incorporated into relevant lesson sharing systems.
54. The expected users for this Evaluation Report are the WFP Country Office and its 
decision-making partners, Regional Bureau, WFP Headquarters in Rome, Evaluation 
Bureau (OEV), WFP Executive Board (EB), Government (Ministry of Education, Technical 
Education and Vocational Training [MENETFP] and Ministry of Agriculture and Rural 
Development [MINADER]), United Nations country team, Steering Committee, 
Technical Committee, AVSI, and the United States Department of Agriculture (USDA) 
Foreign Agriculture Service (FAS).
1.1. Overview of the Evaluation Subject
55. In support of the national integrated school feeding program in Côte d'Ivoire, WFP was 
awarded funding from the McGovern-Dole International Food for Education and Child 
Nutrition Program of the United States Department of Agriculture (USDA). This funding 
supports the school feeding program over the period from September 2015 to July 2020 
in seven regions (Poro, Bagoué, Tchologo, Bounkani, Gontougo, Bafing, and Cavally), 
with the goal of improving nutrition and health outcomes and increasing the literacy of 
school aged children in 613 rural public primary schools these high-priority areas.
56. In addition to the school feeding component, the project provides support to literacy 
learning. The international NGO AVSI was selected to improve the reading skills of 
assisted students through the literacy component. In collaboration with the Ministry of 
National Education (MENETFP), AVSI designed reading improvement tools and trained 
teachers in new reading techniques.
57. The current project activities are to: 1) Provide daily hot lunches composed of rice 
(150g / per day / student), split peas (30g / day / pupil), and vegetable oil (10g / day 
/ student) to 125,000 primary school students at 613 schools in seven targeted 
departments; 2) Provide take home rations (THR) of 50kg of rice to 10,000 regularly
attending CM1/CM2 girls, three times per year; 3) Train canteen management staff 
once a year to enhance management and administration capacities, in the areas of food 
preparation, nutrition, reporting, filling out data collection tools, using the mobile data 
collection tool (ODK), and building improved fuel-efficient stoves; 4) Train 613 school
management committees (COGES) every two years in community mobilization, school 
canteen management, and relevant crosscutting issues affecting the community; 5) 
Provide food preparation and storage tools and equipment (cooking kits, energy efficient 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 2 | Page 
cook stoves, wooden pallets for commodity storage); 6) Strengthen women’s
production groups (WPGs) established around school canteens through enhanced 
agricultural knowledge, access to structured markets, and adult literacy (50/year); 7)
Distribute deworming tablets twice a year to all students; 8) Provide technical 
assistance, limited financial support, and trainings to the Government of Cote d’Ivoire 
in the development of a national school feeding policy and enhancement of monitoring, 
logistics and commodity management capacity; 9) Developing reading improvement 
toolkits for teachers, community members, school administrators and Ministry officials 
as a guide to quality reading instruction that combines best practices in literacy; 10)
Distributing suitable French language books and mobile libraries in targeted primary 
schools; 11) Improve distribution and use of Government’s existing literacy materials; 
12) Conduct four professional literacy instruction workshops; 13) Build capacity of 
Government and communities to improve literacy instruction through Reading 
Promotion Circles and an Early Grade Reading Symposium, in addition to implementing 
other reading-related interventions.
Timing
58. The duration of the current project is from September 2015 to July 2020. The evaluation 
team’s mid-term evaluation is timed to allow it to find early signals of the effects of the 
project and allows time for modifications or mid-course corrections.
Objectives and Activities
59. The two main objectives are to improve nutrition and health outcomes and to increase 
literacy of school aged children in high-priority rural areas. The project aims to reach—
directly or indirectly—more than half a million beneficiaries. The key project activities 
are described above. 
Outcomes
60. The two main planned outcomes of the intervention are improvement in the literacy of 
school-aged children and increased use of health and dietary practices. These two 
outcomes are expected to result from a combination of more consistent teacher 
attendance, better access to school supplies and materials, improved access to learning 
materials for reading and increased skills and knowledge of teachers. The other 
outcomes that fall under the increase use of health and dietary practices include 
improved knowledge of health and hygiene practices, increased knowledge of safe food 
preparation and storage practices, increased knowledge of nutrition, and also access to 
preventive health interventions. 
61. The WFP proposes to go beyond these two results to put into perspective the other 
equally important results such as ensuring children receive nutritious lunches that feed 
their bodies and minds, improving enrolment, retention, and full day attendance of boys 
and girls in schools, support to local production, protecting the environment through 
the construction of improved fuel efficient stoves, and the reinforcement of the capacity 
of the government for the management and implementation of the school feeding 
program.
62. The mid-term evaluation of the WFP “Support to a Sustainable School Feeding 
Programme (DEV 200465)” project, which operated from 2013 to 2016, highlighted that 
the planned transfer of responsibility for WFP-supported school canteens to the 
government did not occur, the food distributed was insufficient to generate the expected 
impact, and the limited support to WPGs reduced the potential sustainability of the 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 3 | Page 
project.9 This evaluation will pay particular attention to whether these stated shortfalls 
have been corrected in the MGD project.
Results Framework
63. The results framework developed for this intervention is based on the two main planned 
objectives and linking project activities in a chain of logical connections. For each 
projected outcome, there are foundational results that include increased capacity of 
government institutions and improved policy and regulatory framework as well as 
increased government support and engagement of local organizations and the 
community in general. The mechanisms of change are clearly presented through project 
activities, planned outputs, and planned outcomes. For example, providing school meals 
reduces short-term hunger, increases school attendance, and improves attentiveness, 
all which are relevant to the improved literacy of school aged children. The project 
results framework is presented in Annex 8. 
Partners
64. The key partners involved in the intervention are WFP, AVSI, and the GoCI, through 1) 
the MENETFP, including the School Canteens Directorate (DCS) and 2) the Ministry of 
Agriculture and Rural Development (MINADER) et l’Agence Nationale d’Appui au 
Développment Rural (ANADER).
65. These partners collaborate within two structures, the MGD Steering Committee, which 
oversees the implementation of the project and is chaired by the MENETFP, and the 
MGD Technical Committee, which includes technicians from all project stakeholders who 
provide technical support for the implementation of the project. 
Resource requirements and funding situation
66. The total award of the USDA MGD Food for Education Program to WFP for the 
implementation of the school feeding program and the literacy component, as well as 
the activities listed above, is USD 35,678,500 over five years, which includes an in-kind 
donation of 24 600 metric tons (MT) of food to be distributed by WFP over the life of 
the project.
Gender dimensions of the intervention
67. In order to address the gender disparity in girls’ school enrolment, the intervention 
provides THR to 10,000 girls from the CM classes (the last two grades of primary 
education). Several other WFP Country Offices in West and Central Africa have provided 
take-home rations to girls, which has improved girls’ enrolment and retention at the 
primary school level. Take-home rations act as an incentive for parents to keep girls in 
school. 
68. Another gender dimension is providing support to 50 WPGs per year who will allow their 
schools to graduate from MGD and transition to the GoCI’s national program. Anchored 
in the government’s vision for “One school, One canteen, One women’s group,” WPGs 
are community-based groups that receive technical support from the GoCI or its 
partners, and jointly cultivate food commodities from communal plots to contribute to 
the school canteens. These allow women to develop their own economic power through 
 
9 Côte d’Ivoire, Projet de Développement 200465 – «Appui au Programme Intégré de Pérennisation des Cantines 
Scolaires » Evaluation d’Opération du PAM Mi-Parcours (2013-2016). PAM: Décembre, 2015. 
https://www.wfp.org/content/c%C3%B4te-d%E2%80%99ivoire-dev-200465-support-sustainable-school-feeding￾programme-operation-evaluation-t
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 4 | Page 
which they can influence decisions in their communities. The groups are committed to 
share 30 percent of their production in order to supply schools. By the end of the 
project, WFP expects to support a total of 200 WPGs, including 10,000 women. WFP is 
mobilizing additional funding to provide technical support to these groups through 
ANADER and other local partners. 
69. Finally, WFP will build the capacity of the COGES to oversee the school canteens, 
expecting to train 12,500 COGES members, and target women as 50 percent of the 
members of these committees (6,250 women). By providing training in community 
mobilization, school canteen management, and relevant crosscutting issues affecting 
the community, WFP will strengthen COGES’ ability to provide leadership and oversight 
to the school and canteen with effective female representation.
1.2. Context
70. Since the end of the post-election crisis, Côte d'Ivoire's economic recovery has been 
notable, with the country experiencing one of the highest growth rates in sub-Saharan 
Africa. According to the World Bank's latest Living Standards Measurement Survey from 
2015, the poverty rate declined from around 51 percent in 2011 to 46 percent in 2015. 
This improvement triggered by recent economic recovery has affected both rural and 
urban areas. Nevertheless, poverty remains a predominant rural phenomenon, marked 
by inequalities in access to essential services and gender disparities. Approximately 30 
percent of children under 5 years old suffer chronic malnutrition and 13 percent of the 
population is food insecure.
71. Women face particularly acute challenges in schooling. In 2012, 63 percent of women 
in Côte d’Ivoire were illiterate, compared to 49 percent of men. Girls were also more 
often educationally disadvantaged than boys. Only 14 percent of girls reached 
secondary school, compared to 30 percent of boys. In primary school, 9 girls were 
enrolled for every 10 boys, and 34 percent of girls abandoned their education 
prematurely, compared to 28 percent of boys.10
72. As soon as it became independent, Côte d'Ivoire pledged to reach a schooling rate of 
100 percent. As a result, it has placed education as a national priority by allocating 
more than 40 percent of the budget in the sector. In September 2015, the GoCI passed 
law n°2015-635 to make school mandatory and free for children between 6 and 16 
years old. However, a number of factors slowed down the aspiration, including the 
thorny problem of noon hunger, which was soon faced by many children whose schools 
were located several kilometres from the family home. The adequate and 
comprehensive response to this important problem required the conduct of a social 
policy based in particular on school meals.
73. Since 1989, the GoCI has administered a national school feeding program with the 
support of WFP. In 2000, the GoCI integrated nutritional dimensions in addition to 
educational objectives, aiming for sustainability by encouraging the production of local 
communities in the form of WPGs. This national program, entitled the “Integrated 
Program for the Sustainability of School Canteens” (PIPCS), aims to address the 
problems of chronic child malnutrition, which is high at 23.2 percent for boys and 19.9 
percent for girls under 5 years old (MICS 2016), and poor performance in primary 
education, with 63 percent of the population being illiterate and only 75.1 percent of 
 
10 Bailly-Traore, Anna and Charles Kapie. December 2017. “Etude sur l’Impact du Programme d’Aide Alimentaire du 
PAM sur les Roles de Genre : Rapport Preliminaire.“ WFP Country Office in Côte d’Ivoire. Page 18.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 5 | Page 
children finishing primary school in 2016. This national program resulted in the 
operation of 5,708 school canteens across the country in 2014-2015, providing hot and 
balanced meals to 1,104,138 elementary school children.
74. In April 2012, the Government of Cote d’Ivoire, through the MENETFP’s Directorate of 
School Canteens (DCS), with the technical assistance of the WFP and the United Nations 
Development Program (UNDP), developed the strategy for the National School Feeding 
Program for the period 2012-2017. The priority areas of the intervention were defined 
through an analysis based on a composite indicator of the level of food insecurity, gross 
enrolment rate, gross completion rate, and prevalence of chronic malnutrition. Thus, 
the following regions have been identified as priority areas for school feeding 
interventions: priority 1 (Cavally, Guémon, Poro, Bagoué, Tchologo, Bafing), priority 2 
(Worodougou, Béré) and priority 3 (Gontougo and Bounkani).
75. Aligned with this national strategy, from September 2013 to December 2016, WFP 
implemented the previous iteration of the subject of this evaluation, entitled: "Support 
for the Integrated Programme for Sustainable School Feeding". With an expected 
number of 571,000 beneficiaries, this project targeted 29 percent of all school canteens 
and 15 percent of all public primary schools nationally. This project covered 1,634 
school canteens in the 10 priority regions. Building on the progress from this project, 
WFP conducted a vulnerability analysis and mapping (VAM) exercise to select the most 
food insecure communities within 7 of the 10 priority regions for the 2015-2020 MGD 
project, in coordination with the DCS, although WFP continues to provide school meals 
with funding from other donors in 6 other regions in Côte d’Ivoire (which are not 
included as part of this evaluation). 
1.3. Evaluation Methodology and Limitations
Evaluation Design
76. To address the research questions at midline, the evaluation team built on the 
quantitative methodology used at baseline to ensure consistency in the approach and 
to minimize the introduction of new biases. In addition, the evaluation team added a 
qualitative component to complement the quantitative analysis and shed light into the 
processes and mechanisms of change and the factors that may have affected 
implementation.
77. To assess the causal effect of the MGD project after two years of implementation on 
student’s reading proficiency11, the evaluation team designed a difference-in￾differences (DID) method at baseline. This method compares the changes in outcomes
between the population of beneficiaries (the treatment group) and the population that 
is not benefiting from the project (the comparison group) by following the same 
students over time. 
Evaluation Questions
78. The evaluation team organized the evaluation questions under the following criteria: 
relevance, effectiveness, efficiency, impact, and sustainability. Given the breadth of the 
questions for this evaluation, the evaluation team has compiled them into a 
comprehensive conceptual framework, in Annex 1. For each evaluation question, the 
evaluation team lists data collection strategy, data analysis method, and the responsible 
party for collecting and measuring them. In Annex 1, the evaluation team also maps 
 
11 Reading proficiency is defined as ability to achieve the grade-appropriate reading level by USDA. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 6 | Page 
each evaluation question to their corresponding indicators or assessment criteria, 
required by USDA, based on the MGD results framework. For example, the evaluation 
team listed the output indicators under efficacy (e.g., number of girls receiving take￾home rations as a result of USDA), and outcomes indicators under impact (e.g., 
proportion of students at the end of grade 2 who demonstrate reading proficiency at 
their grade level). 
Sampling Frame
79. As requested in the revised TOR and in compliance with the requirement for consistent 
DID methodology, at midline, the evaluation team followed the same 99 school impact 
evaluation sample used for the baseline evaluation and added 9 extra treatment schools 
which the evaluation team randomly selected, maintaining the regional distribution, 
from the 169 MGD schools sampled at baseline. 
80. The evaluation team targeted a sample of 1,753 students and their corresponding 
households, out of the 125,000 student beneficiaries, to ensure proportional 
representation of respondents within 104 schools12 across the 7 regions, among which 
68 treatment schools and 36 control schools. The evaluation team also ensured a 
balanced sample of girls and boys to disaggregate data by gender where applicable. In 
each school, the evaluation team also surveyed teachers in Grades CP1-CM2, the 
principal, and the canteen manager. Exhibit A2.1 in Annex 2 shows the distribution of 
sample respondents by region.
81. In collaboration with WFP, the evaluation team selected 20 project sites to visit. At each 
site selected, the evaluation team conducted focus group discussions (FGDs) with CM 
students, WPGs,13 parents, and COGES members. Where possible, the evaluation team 
separated the parent focus groups by gender to allow respondents to speak more freely 
and to compare mothers’ and fathers’ perceptions. Finally, the evaluation team
conducted key informant interviews (KIIs) with internal and external stakeholders, as 
described in Annex 4. In all mixed-gender FGDs and KIIs, the evaluation team actively 
encouraged women’s participation to ensure the evaluation team’s evaluation 
incorporated their experiences and recommendations.
Data Collection and Analysis Methods 
82. Surveys. To answer evaluation questions, the evaluation team reviewed instruments 
that were used during baseline and revised the tools to more accurately capture the 
MGD indicators and fill in the gaps from baseline (e.g., add or revise questions related 
to student attendance, teacher attendance, minimum acceptable diet, etc., particularly 
to distinguish between girls and boys). The evaluation team also developed a teacher 
survey that it calibrated into Ivorian context to capture teachers’ training needs and 
use of the new techniques.
83. Reading Assessment. To measure the impact on reading skills of school children and 
analyse any change from baseline, the evaluation team administered the same reading 
assessment tool, ASER. To avoid any possible bias in reading outcomes, the team
updated the test content. The evaluation team ensured that the updated test has the 
 
12 Among the 108 selected schools, the evaluation team learned that 4 control schools closed since baseline, which 
the evaluation team excluded from the evaluation.
13 The sample of WPGs contains 10 groups supported by the MGD project since 2017 (designated as "established") 
and 10 groups supported since 2018 (designated as "new").
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 7 | Page 
same level of complexity as the one used at baseline to be able to compare students’ 
reading skills between baseline and midline. 
84. Qualitative Tools. In addition, the evaluation team administered KIIs and FGDs 
protocols to conduct an in-depth stakeholder analysis, including at the beneficiary level. 
The tools assisted in interpreting the quantitative findings by contextualizing and filling 
in gaps from the baseline, as several of the MGD indicators are not directly comparable.
85. Data Collection. In collaboration with its data collection partner, Mont Horeb (MH), 
and its fieldwork managers, the evaluation team recruited and trained 40 experienced 
enumerators, with local language skills in addition to French, to collect the midline 
evaluation data in September–October 2018 (at the beginning of the school academic 
year, 2018-2019). MH organized these 40 enumerators into 10 teams of three 
individuals and one supervisor, including both men and women, to collect data in the 
seven target regions. The evaluation team and MH’s fieldwork managers closely 
followed the ten teams daily to oversee the quality of the data that enumerators 
collected and provide them with technical support. The evaluation team conducted a 
stakeholder analysis with internal and external stakeholders, and conducted qualitative 
interviews with local stakeholders and beneficiaries.
86. Evaluation Analysis. After the field activities were completed, the evaluation team 
conducted a final review of the collected data. The evaluation team then compiled the 
survey responses into a master file for the performance-evaluation analysis. Where the 
accuracy/completeness of the baseline quantitative data allow, the evaluation team 
conducted a cohort comparison to measure the project’s progress in reaching its target 
goals and provided descriptive analysis (percentages and averages) where the 
comparison was not feasible in order to establish base values for comparison later at 
endline. When possible, the evaluation team also conducted subgroup analyses by 
grade, student gender, highlighting emerging patterns. 
87. For the impact evaluation, the evaluation team compared the changes in reading 
outcomes between students who attended MGD schools (treatment group) with 
students who were enrolled in non-MGD project (comparison group) to estimate the 
two-year project’s effect on literacy growth. The evaluation team surveyed students 
from the same schools in baseline and midline so that it can construct the change in 
reading outcomes for each cohort of student. Since the evaluation team collected the 
midline data at the beginning of the 2018-2019 school year, to be able assess their 
reading outcomes at the end of their grade, it had to measure their reading proficiency 
retrospectively which excluded CM2 cohort from the analysis.
Validity and Reliability
88. To measure project impact between baseline and midline, the evaluation team collected 
surveys and reading assessments at the same schools selected by the baseline 
evaluator (INS). The evaluation team and WFP were unable to meet with INS to assess 
the validity of the baseline sampling and data collection. However, having designed and 
validated the reading assessment for the baseline evaluation, the evaluation team is 
confident in the validity of the related indicators at baseline and the evaluation team’s 
ability to infer causal relationships between the project and progress in literacy. 
89. Before midline data collection, the evaluation team conducted an adaptation workshop 
with a group of local reading, curriculum, and assessment experts from MENETFP with 
support from AVSI, as the evaluation team successfully did at baseline, to ensure that 
the updated ASER test is still culturally appropriate and consistent with Côte D’Ivoire’s 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 8 | Page 
learning standards for Grades 1–6. The long-form methodology in Annex 2 presents the 
structure of the ASER reading test in Exhibit A2.2, including the test’s levels and 
corresponding grades and reading skills.
90. In addition, the evaluation team held a validation workshop at midline with DCS and 
WFP to review all of the survey instruments. Based on the discussion, adjustments to 
the instruments were made prior to data collection. The evaluation team also shared its 
inception report, including the surveys and qualitative protocols, with WFP, AVSI, and 
Government stakeholders to address their comments prior to any field work.
91. Quantitative and qualitative data were compared to cross-validate findings across data 
collection methods and verify the reliability of beneficiary responses. To ensure inter￾rater reliability, an IMPAQ research analyst performed daily quality checks on key 
variables, including reading assessment and food consumption scores, and followed-up 
quickly as necessary to address any significant differences in survey responses by 
enumerator. In addition, qualitative data was independently collected, with the findings 
and transcripts reviewed by IMPAQ home office staff.
92. Modifications and Limitations. There were some limitations and challenges in the 
MGD midline evaluation. These include: inaccurate or missing data from baseline; delay 
in data collection until the beginning of the school year 2018-2019; change in the 
project implementation plan for training teachers; and following the same subsample 
over time. Exhibit A2.3 in Appendix 2 describes the limitations and challenges faced in
the evaluation team’s midline evaluation in further detail, along with the mitigation 
strategies it implemented and their possible impact on the findings.
93. Ethical Considerations. During the evaluation the following ethical issues were 
considered for the design, data collection, data analysis, reporting and dissemination: 
providing sufficient information about the evaluation to participants; obtaining informed 
consent from participants; collecting confidential data from participants, assuring them 
that their identity will not be revealed; and analysing and presenting de-identified and 
aggregate findings. The following safeguards and measures to manage these issues 
were in place: submitting the research protocol and data collection instruments to 
Advarra Institutional Review Board (IRB); training enumerators on respondents safety 
and confidentiality, with particular consideration given to safeguarding child 
respondents; obtaining teachers’ and principals’ consent to survey their students along 
with each child’s individual assent; keeping the data in a secured place to protect 
respondents’ personal information; and active monitoring of data collection for 
compliance with IMPAQ quality and ethical guidelines.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 9 | Page 
2. Evaluation Findings
94. The evaluation findings and the evidence to substantiate them are presented below. 
They are structured as a response to each evaluation criteria and subsequent evaluation 
questions in turn. 
2.1. Evaluation Criteria 1 – Relevance
95. The first set of evaluation questions aim to assess the relevance and design of the 
intervention. The findings in this section are informed from the evaluation team’s
document review and stakeholder analysis, and address the following questions:
• To what extent was the design of the interventions in line with the needs of the 
target population—women, girls, boys, and men?
• To what extent is the design of the intervention aligned with national policies, 
strategies, and programs?
• Do the project design and implementation arrangements complement other donor￾funded and government initiatives?
• Is the project designed to reach the right people with the right type of assistance?
Internal Stakeholders (WFP) and Partners/Sub-Recipients (Government and AVSI)
96. The MGD project was designed in collaboration with the GoCI to reach the most 
vulnerable population in Côte d’Ivoire and was guided by a national school feeding 
strategy that the GoCI developed in 2012 with the technical assistance of WFP and 
UNDP. This national strategy identified ten priority regions for school feeding 
interventions, based on their level the level of food insecurity, gross enrolment rate, 
gross completion rate, and prevalence of chronic malnutrition. Ensuring that its 
activities are relevant to national priorities, the WFP implements the current MGD 
project in seven of those regions and distributes THR for girls in three of the seven 
regions (Bagoué, Poro, and Tchologo) to reduce their high gender enrolment disparity. 
97. Key stakeholders from the GoCI confirmed that the MGD project is highly integrated 
into and complements the government’s efforts to provide nutritious school meals to all 
primary school children in the country. Representatives from the DCS described how 
the activities fit into the National School Feeding Program, which was developed in 1989 
in partnership with WFP, and appreciated WFP’s technical support in developing the 
country’s first national school feeding policy, enacted in March 2018 and covering 2018 
to 2025.
98. The project also complements other donor- and Government-funded initiatives. 
Specifically, WFP has partnered with UNICEF and the Global Partnership for Education 
to target infrastructure improvements at the same schools receiving MGD funding. 
However, as confirmed by the evaluation team’s quantitative and qualitative data 
collection at midline, these infrastructure improvements have not yet begun in the 
schools in the evaluation team’s sample.
99. While the project is highly aligned with Government priorities, some representatives
from DCS said they were not consulted during the final stages of the design phase. 
Because of this, stakeholders believed that several activities were not realistically 
planned. For example, the original proposal anticipated providing meals to 125,000 
students in 1,000 rural schools, but the average school enrolment in these regions is 
much higher. While the number of student beneficiaries remained as proposed, WFP 
had to reduce the number of schools they worked in, which government officials said 
could have been decided earlier had they been more involved in the planning phase. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 10 | Page 
Some government officials also thought that many proposed activities were not realistic, 
for example, the allocated resources and timeline for the number of improved stoves,
the training of COGES, and the capacity building of WPGs. The training of 200 WPGs in 
particular was mentioned by multiple stakeholders as being ambitious given the 
timeframe and resources.
100. Several GoCI key informants mentioned wanting to receive additional training on the 
monitoring and evaluation (M&E) practices and tools that WFP personnel use on the 
MGD project. This would not only increase their current technical capacity to implement 
MGD, but also empower them to improve management of the national school feeding 
program and other projects. The MGD project delivers annual M&E training to DCS and 
DPFC, suggesting that either these trainings may not be attended by all relevant GoCI 
personnel or do not pertain to M&E topics most relevant to the Government’s work.
External Stakeholders (Beneficiaries) 
101. At the beneficiary level, parents and members of the COGES agreed that aspects of the 
intervention are relevant to their needs, specifically food delivered to the canteen, 
school improvements, and literacy activities. According to beneficiaries, a critical
challenge for their communities before the MGD project began was attracting students 
to school and consistently serving meals at the school canteen.
102. By encouraging or allowing children to attend school with the promise of a lunchtime 
meal, the MGD project is also directly relevant to the needs of parents of the students 
by reducing the time they needed to devote to childcare during the school week and so 
increasing the amount of time they can devote to their livelihoods. Some parents also 
spoke of the psychological relief in knowing their children will not go hungry. 
103. The WPGs reported that receiving high-quality seeds, herbicides, and fertilizers through 
the project is relevant and highly appreciated. Prior to MGD support, financial access to 
these inputs was challenging, and they had to adjust their practice due to shortages 
(for example, reduce the amount of recommended fertilizer as they did not have 
enough). In the ten established WPGs, members received even more materials, 
including watering cans, gardening tools, scales, tricycles, corn mills, and oxen. Women 
in these groups appreciated that prior to distributing materials, Development Training 
and Consulting Office (BFCD) staff met with them to determine their needs.14 This 
ensured that the equipment received was directly targeted and relevant to the individual 
groups. 
104. In addition to addressing food insecurity, which the WPGs in the north said was 
particularly severe, members said that the project addresses school level outcomes 
such as the improvement of children’s attention, attendance, and concentration in class. 
The WPGs and students’ mothers said that because food is provided at the canteen, 
they save time and money as they no longer have to make food for school lunches and
their children stay in school for their afternoon classes.
105. The literacy component of the project was welcomed by parents, as they said that
education is the key to future success for both their sons and daughters. Many talked 
about their own challenges not knowing how to read and write and hoped for something 
better for their children. However, some parents were disappointed that some children 
 
14 For the 10 established WPGs, WFP is working with BFCD, an Ivoirian NGO, instead of ANADER to assess WPG 
needs and deliver technical assistance.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 11 | Page 
in their communities who earned diplomas are currently jobless. Some parents reported 
heavily investing in their children’s education, who are unemployed or took a long time 
to find work. This demotivates parents, with some reporting that there is no value in 
sending children far in school. 
Key findings and conclusions – Relevance
• The project is highly integrated into and complements the government’s efforts. 
Beneficiaries agree that the food delivery, school improvements, and literacy activities 
aspects of the intervention are relevant to their needs.
• While the project is highly aligned with Government priorities, some representatives 
from DCS said they were not consulted during the final stages of the design phase. 
Because of this, stakeholders believed that several activities were not appropriately 
realistically planned given the timeframe and resources.
• The MGD project is also directly relevant to the needs of parents of the students by 
reducing the time they needed to devote to childcare during the school week–
increasing the amount of time they can devote to their livelihoods.
• The WPGs reported that receiving materials and training was relevant and highly 
appreciated.
• The literacy component of the project was welcomed by parents; they said that 
education is the key to future success for both their sons and daughters. Many talked 
about their own challenges not knowing how to read and write and hoped for something 
better for their children. 
2.2. Evaluation Criteria 2 – Effectiveness
106. The second set of evaluation questions seeks to evaluate the project’s effectiveness. 
The results in this section are based on the qualitative data (for quantitative results, 
please see Evaluation Criteria 3 – Impact in section 2.4). The results below aim to 
address the following:
• What is the progress of project implementation—is the project on track to carry out 
all activities as planned?
• To what extent have the interventions so far responded to the needs of the 
beneficiaries—women, girls, boys, and men?
• What are the main factors influencing the achievement, whether or not of the 
results/ objectives, of the intervention?
• Are any changes required to increase the project effectiveness?
Progress of Project Implementation and Response to Beneficiary Needs
107. At the beneficiary level, parents and COGES members agreed that aspects of the 
intervention addressed community needs, particularly the food delivered to the school 
canteens. The absence of a regularly provisioned school canteen, seen as a major 
challenge which previously discouraged or prevented children from attending school or 
staying for the full day, was addressed under WFP’s MGD project according to the 
COGES and parents. One group of parents also shared that the MGD canteens even 
attracted students enrolled in other schools nearby. This information was confirmed by 
key informants at the national level, who also mentioned that children from nearby 
schools were enrolling in MGD project schools to benefit from the canteen.
108. According to parents, by encouraging or allowing children to attend school with the 
promise of a nutritious meal, the MGD project is effective by reducing the economic 
burden on parents and making their children’s lunch during the school week, which 
increases the amount of time they can devote to their livelihoods.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 12 | Page 
109. While all COGES agreed that the project helped their communities and the MGD food 
relieves their limited budgets, most COGES members and parents raised concern about 
the diversity of the meals served to students in the canteen. While respondents agreed 
that enough rice was provided, along with split peas and vegetable oil, the COGES and 
newer WPGs said they struggled to grow or buy enough vegetables and animal protein
to complement the MGD staples and create varied and nutritious meals. Some asked 
that different food to be provided in addition to split peas, rice, and vegetable oil, as 
they did not consider these foods diverse or plentiful enough to make a healthy meal. 
They brought up examples of previous projects which brought canned sardines or beef 
to the canteen. Most students reported that they enjoyed the school meals, but a few 
children said their meals were not nutritious because the sauce did not contain enough 
vegetables. 
110. However, in the ten established WPGs, dietary diversity was not raised as an issue. 
Women in these groups reported that since receiving project support they were able to 
provide vegetables to the canteens to create nutritious and balanced meals. They also 
reported receiving training on nutritionally diversifying the types of meals they could 
make with their crops. Women said that they learned how to identify symptoms of 
malnutrition (which they had previously attributed to witchcraft) and were given 
guidelines on nutrition. By following these guidelines, they stated that in their 
communities there are now fewer cases of malnutrition. 
111. From the discussions with the ten established WPGs, members stated that their 
production has grown substantially due in part to the technical support received through 
the project. All groups were able to increase their contributions to the school canteens 
from last year to this year, with one group increasing their contribution from one ton of 
food last year to four tons this year. Members credited their higher production yields to 
learning how to efficiently plot seeds and distribute fertilizer, as well as on preventing 
crops from being destroyed by insects and disease. Prior to MGD support, many 
members said they threw seeds in their field and mixed the crops. Learning how to 
plant seeds in rows and appropriately apply fertilizer and pesticides was cited by the 
women as an essential part of increasing their yields. 
112. According to parents, the project helped their children in lower grades decipher letters, 
particularly for their children at the CP levels. Overall, parents were happy with the 
quality of the teachers and the progress of their children, although some parents said 
the teacher training for the older grades (CE and beyond) was less effective. Even so, 
they said their children still benefited from attending class with a trained teacher.
113. Multiple stakeholders believe that WFP is behind on several of their planned activities 
because they did not accurately estimate the required involvement and cost. The 
activities related to the WPGs were most frequently cited, but others were mentioned 
as well, including those related to school infrastructure improvement and capacity 
building of the canteen management committees and COGES. Both internal and 
external stakeholders spoke of being behind on activities, and when asked, said that
underestimations of the WPG support budget were made at the design phase, including 
underestimating the amount of external funds that could be leveraged to help support 
some activities. 
114. Although the MGD project intends to train COGES members every other year in 
community mobilization and school canteen management, most COGES members 
interviewed for this evaluation reported not receiving any training. One COGES 
mentioned attending a school management training from WFP in 2017. Other COGES 
reported receiving training from government advisors on creating and managing a 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 13 | Page 
sustainable fund for their COGES. However, it was unclear if this training was related 
to the MGD project or not. 
115. In addition, the COGES were targeted to include 50 percent women as members, 
however, in the evaluation team’s focus groups, most did not include this many women, 
and in fact, some COGES were exclusively male. The COGES who did include women 
stated that the female committee members had the same rights as men, and even if 
they performed different work, women and men were equally involved. Some COGES 
had women in leadership positions. In other COGES, women were excluded from 
leadership, explained by members as being because women were naturally different 
from men, were culturally quieter, and/or were illiterate.
116. COGES also differed in their operational involvement in the MGD project, with those in 
some communities checking the quality of food delivered to the canteen while others 
not participating in the supervision of these deliveries.
Factors Influencing Achievement
117. As stated by parents and children, the main factor influencing project effectiveness is 
financial constraints. Some parents said that they are not able to afford the indirect 
costs of sending their children to school, and struggle to afford the daily 25 FCFA 
canteen fee15 and the annual costs of books, school uniforms, and COGES contributions. 
Many COGES indicated that they lost or received less government funding in recent 
years, although the official number of subsidized COGES in Cote d’Ivoire has not 
decreased. Instead, COGES have solicited parents for contributions to fund their 
activities for the school, creating tension in the community and draining parents’ 
resources. For example, last year, one COGES estimated that only 58 out of 178 
students paid the annual COGES contribution. Some parents threatened to pull their 
children out of school if they were forced to pay the COGES contribution. However, 
these contributions are critical to giving the COGES the means to repair school 
infrastructure and build additional classrooms, sanitary installations, and teacher 
housing.
118. In 2015, an independent operation evaluation of WFP’s support to the GoCI’s national 
school canteens program recommended that WFP advocate and support the DCS in 
implementing a special project to reduce the cost of students who were unable to pay 
to the canteen fee. In its management response, WFP agreed to perform a study of 
community participation in the school canteens and specifically mechanisms for 
free/subsidized school meals, and help the government implement the government￾accepted recommendations from the study.16 This action was to be completed by 
October 2016. 
119. While the daily canteen fee is heavily subsidized, according to parents, COGES, and CM 
students in the FGDs, it was cited as a significant financial barrier to some children 
accessing MGD meals. One group of parents said the 25 FCFA canteen fee was especially 
difficult to manage during the off season for the cashew harvest (generally September 
to January). After the harvest starts in February, parents are once again able to pay 
 
15 The 25 FCFA canteen fee per child/day is mandated by the Government of Cote d’Ivoire to cover the salary of 
canteen workers, firewood, and condiments, with a portion also going to the IEP and COGES for activity monitoring. 
It is equivalent to roughly 0.05 USD. 16 Côte d’Ivoire DEV 200465 Support to a Sustainable School Feeding Programme: A mid-term Operation Evaluation. 
https://www.wfp.org/content/c%C3%B4te-d%E2%80%99ivoire-dev-200465-support-sustainable-school-feeding￾programme-operation-evaluation-t 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 14 | Page 
the fee, and their children are able to eat at the canteen. In the FGDs, students in this 
community also mentioned sometimes not eating at the canteen because they could 
not afford the fee. Students in other communities gave similar reasons as to why they 
did not always eat at the canteen, although the majority of students in the focus groups 
indicated that they regularly ate at their canteens. In the communities with the 
established WPGs, women said that the fee was not a barrier for them or their children, 
but did acknowledge that as members of the WPG, they were more financially stable 
than other women in their community. 
120. According to parents and students, in some schools the canteen was only open 3-4 days 
a week. This is validated by the canteen managers surveyed, who indicated that on 
average across all regions the canteen was open only 93.1 percent of actual school 
days, with strong regional variation. According to WFP, potential reasons for why the 
canteen is not open every day include some schools not operating on a five-day 
schedule and frequent teacher strikes and demonstrations. At the beginning of the 
2018-2019 school year, when the evaluation team’s focus groups occurred, children at
one school said that their canteen had not yet opened for the year, and therefore they 
had to eat at home. Doing so often prevents them from returning to school on time in 
the afternoon. Some students indicated that when the canteen was closed they do not 
attend school at all, but others said they still came to school to be with their friends or 
because they valued their education.
121. According to parents, COGES, WPGs, and students, school infrastructure is another 
factor influencing effectiveness. Some parents reported that the classrooms are 
overcrowded, with four or more students sitting on benches designed for two students. 
Some canteens had no tables or seats, so children had to eat sitting on the floor. In 
many schools, COGES reported that teacher housing was unavailable or in reportedly 
unacceptable conditions. COGES members and CM students mentioned that access to 
water was also a key challenge that forced students or canteen workers to walk long 
distances to collect water for school meals and other school utilities, delaying meals or 
forcing students to miss class.
122. Another factor influencing project effectiveness as mentioned by parents and students 
is teacher absenteeism. In one focus group, the students said that last year they had 
to miss class on four separate occasions because their teacher left to conduct personal 
business in town. 
123. Some students said they did not take books from the mobile library because they were
worried that they or a family member would damage them. In at least one school, it is 
the policy (or at least the perception of the students) that they must pay to replace any 
damaged book. This policy dissuaded children from taking full advantage of MGD￾provided learning materials, and potentially decreases project effectiveness.
124. Some COGES’ members shared instances where the rice has been of poor quality, which 
they believe could be linked to improper or lengthy storage before the school received 
its delivery. This information was collaborated by a government stakeholder who 
mentioned an instance where rice was delivered in powdered form and also where oil 
was delivered within a month of the expiration date. To prevent food waste, in these
situations the rice was treated so it could still be used, and the oil was distributed to 
multiple schools who prioritized its use. WFP also mentioned that the slow food 
consumption pace in 2017 combined with annual importing of food from the United 
States led to longer storage times. However, WFP addressed this by re-scheduling food 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 15 | Page 
import to be done in tranches throughout the year and implemented capacity building 
for the DCS on proper food storage and distribution. The government stakeholder who 
raised this problem agreed that it has since been addressed.
125. The WPGs still rely on men, as there are several agricultural activities that are labour 
intensive, and typically done by males. For example, ploughing with oxen, soil 
preparation, and using farming equipment such as rice threshing machines and tillers 
can be particularly physically demanding or gender specific. This is especially true for 
WPGs that are composed of older women as their age does not allow them to physically 
support field work. As a result, these groups depend on male volunteers or paid labour 
services. One group reported that the cost of day labourers was not affordable, 
particularly in areas with more lucrative crops like sugar cane. As these members said, 
a plot of rice requires physical effort that their own members cannot provide, including 
clearing grass in the field, transplanting rice, cutting it, beating it, and loading it in the 
trucks. As a result, this group is required to hire wage labourers at a great expense to 
their overall profit. WFP has noted this and is currently advocating integrating youth as 
members of the WPG to support their activities. For some of the established WPGs, 
integrating youths has been successful, especially with the support of the village chief.
In two groups, though, the members said that their efforts to incentivize younger 
community members have been unsuccessful so far. 
Key findings and conclusions – Effectiveness
• The absence of a regularly provisioned school canteen, seen as a major challenge which 
previously discouraged or prevented children from attending school or staying for the 
full day, was addressed under WFP’s MGD project according to the COGES and parents.
• Parents appreciated not having to prepare meals for their children, thus freeing their 
time for other livelihoods activities. 
• Members in the established WPGs stated that their production has grown substantially 
due in part to the technical support received through the project. 
• Some planned activities are behind schedule, such as the support to WPGs and the 
training of COGES. These delays were explained by ambitious expectations in 
mobilizing outside funds. 
• Threats to implementation include the low financial means of beneficiaries, who report 
not being able to afford school-related costs, including school uniforms, textbooks, and 
the daily 25 FCFA canteen fee. 
• Other factors influencing project effectiveness include poor school infrastructure (lack 
of running water, latrines, and school benches) and teacher absenteeism.
• The WPGs still rely on men, as there are several agricultural activities that are labour 
intensive and typically done by males. 
2.3. Evaluation Criteria 3 – Efficiency 
126. This section explores whether objectives were achieved on time and what factors led to 
efficiency. Although a cost-benefit analysis is outside the scope of this evaluation, 
through using qualitative data the team endeavoured to address the following
evaluation questions: 
• How efficient is the targeting?
• Does the assistance reach the right beneficiaries (girls, boys, men, and women) in 
the right quantity and quality and at the right time?
• Is the project efficient in terms of costs and costs/beneficiary? 
• What are the external and internal factors influencing efficiency of the project?
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 16 | Page 
127. Overall, the project is efficient in that it generally provides beneficiaries with the outputs 
as described in the proposal, specifically providing school meals, THR, and literacy 
instruction.
128. There are some activities that have yet to fully materialize as proposed, as the 
complementary funds have not been mobilized as anticipated. For these activities, 
strategic choices were made that benefited the primary components of the project at 
the expense of secondary components. For example, the literacy and numeracy training 
for WPGs/COGES were not prioritized, as indicated by national and local stakeholders.
129. During the stakeholder analysis, some government stakeholders suggested that WFP 
should reduce the number of school days for which it provides MGD commodities and 
reorient the extra food to serve more schools. According to these officials, and 
corroborated through FGDs with COGES, WFP’s current strategy discourages parents 
and community members from contributing money and/or food to the canteens, as 
basic commodities are provided by MGD for every day school is in session 
130. It is difficult to accurately and regularly monitor student and teacher attendance 
(including absences due to sickness) and canteen outputs. When reviewing school 
records, the evaluation team found that attendance rates of teachers and students 
conflicted with the qualitative data. The school records showed almost perfect 
attendance for teachers and students, while in the evaluation team’s FGDs respondents 
reported frequent teacher absenteeism, and instances where children missed school. 
Teachers self-reported missing school at a substantially higher rate than what the 
school records indicated. As for the canteen records, in the evaluation team’s survey of 
MGD canteen managers, referring to March 2018, 39 percent did not know the expected 
number of meal recipients for the month; 10 percent could not disaggregate by gender 
the number of meal recipients; and 28 percent could not indicate the number of days 
the canteen served four food groups. 
131. External factors influencing efficiency include climate variation and lack of irrigation 
systems that affect the crop yield of the WPGs. Members from all groups spoke of their 
reliance on rain to water their crops. When rain is plentiful, groups can grow vegetables 
and rice to support the canteen, but during the dry season from April through July, they 
are unable to grow these crops.
132. Other external factors influencing efficiency include the low incomes, high-illiteracy 
rates, and lack of management skills of the community members, and therefore of the 
COGES, which prevent them from fully executing several activities to improve the 
infrastructure of the schools or to start taking ownership of the canteens. The 
remoteness of the schools also affected the efficiency of several activities, due to the 
cost of exporting materials and expertise to these schools. For example, the
construction of improved, fuel-efficient stoves was more difficult and expensive than 
originally planned, because the materials and expertise are limited in isolated rural 
areas. 
Key findings and conclusions – Efficiency
• Overall, the project is efficient as it provides beneficiaries with school meals, THR, 
and literacy instruction. 
• There are some activities where the complementary funds have not been mobilized 
as anticipated. For these activities, strategic choices were made that benefited the 
primary components of the project at the expense of secondary components.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 17 | Page 
• Some government stakeholders and local community members suggested that by 
providing meals every day, the project discourages the community from taking 
ownership of the canteens.
• It is difficult to monitor the project using school records. Student and teacher 
attendance reports appear to be inflated, as the official reports show almost no 
absenteeism, conflicting with the qualitative results. Canteen records did not appear 
inflated, but some were missing or incomplete.
• External factors influencing the efficiency of the WPGs include climate variation and 
lack of irrigation systems. Other factors include the low incomes, high-illiteracy 
rates, and lack of management skills of the community members, which prevent 
them from fully executing certain activities.
2.4. Evaluation Criteria 4 – Impact 
133. This section forms the largest part of the evaluation team’s evaluation findings, and is 
divided into two main sections: 
• Impact Analysis, where the evaluation team uses a difference in differences 
approach to measure the effects of the MGD intervention on students’ reading 
outcomes; and
• Performance Analysis, where the evaluation teams use a descriptive approach 
using the survey findings to measure changes in outcomes, including triangulating 
where relevant examples of project impact informed by the stakeholder analysis.
134. The evaluation team address the following evaluation questions to assess changes as a 
result of the intervention: 
• What are the midterm effects of the operation on the beneficiaries in terms of: (a) 
improvement of school indicators; (b) improvement of pupil reading skills; (c) 
capacity building of groups?
• What are the reasons for the observed effects?
• Are there unintended effects on the beneficiaries? What have been the gender￾specific impacts, particularly with regards to girls’ schooling?
• To what degree have the project outcomes made progress toward positive long￾term effects on targeted beneficiaries (girls, boys, men, and women), households, 
communities, and institutions?
2.4.1 Impact Analysis
135. As described in the methodology section in Annex 2, the evaluation team used a 
difference in differences (DID) method to measure the changes in reading outcomes 
over time between students who attend schools that receive the MGD interventions17
(treatment group) and students in schools that do not receive the project (comparison 
group/Non-MGD schools). 
136. Given the change in implementation of teacher training in 2017-2018, across different 
grades at midline, students have not been exposed to the teacher training consistently. 
The only cohort of students who have been fully exposed to the project interventions 
for two years, including two-years of trained teachers, for which the evaluation team 
also has their baseline values are CE2 students (who were in CP1 at baseline). 
Therefore, to maintain the power of the study, the evaluation team restricted its sample 
to compare the change in students’ reading skills from baseline (when they were in 
grades CP1, CP2 and CE1) to midline (when they are in grades CE2, CM1, and CM2), 
with the change that occur among similar students from comparison schools. The 
 
17 School canteen and literacy activities. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 18 | Page 
evaluation team’s key outcome of interest is the improvement in children’s reading 
proficiency at grade level due to at least one-year project effects. 
137. It is important to note that since the midline data collection was done at the beginning 
of the 2018/2019 school year, the midline reading skills of CE2, CM1, and CM2 students 
reflect their reading ability at the end of the previous school year (i.e., reading ability 
at CE1, CE2, and CM1 level, respectively). To estimate this effect, the evaluation team 
ran the following model:
ΔMGD= [(𝑌𝑌1MGD − 𝑌𝑌0MGD) − (𝑌𝑌1non-MGD − 𝑌𝑌0non-MGD)]
Where:
• 𝑌𝑌0 and 𝑌𝑌1 denote reading outcomes at baseline and midline respectively; and
• MGD and non-MGD denote schools that are receiving MGD interventions and 
comparison group that are not receiving any activities. 
138. To obtain more precise estimates, the evaluation team included other covariates, such 
as the gender and grade of the student, as well as availability of story books at their 
homes. The evaluation team also controlled for the head of household’s educational 
attainment and gender as significant differences between their values in treatment and 
comparison groups at baseline were observed. In addition, because students are nested 
in a particular grade from a given school, literacy outcomes are correlated to each other. 
To account for this correlation, the evaluation team clustered the standard errors at the 
school level. 
139. At midline, the evaluation team tried to assess the same students from baseline. When 
they were not available to be surveyed for reasons such as dropping out of school or 
moving away, the evaluation team sampled new students randomly to replace them. 
However, some of these replacements might not have been exposed to the project if 
they have just joined the school for the 2018-2019 school year. As a robustness check, 
the evaluation team estimated its model by restricting the sample to include only the
students who have been in the school for at least a year18 by academic year of 2018-
2019. 
140. Figure 1 summarizes the impact results. The second column, highlighted in grey, 
represents the holistic project’s impact on children’s literacy outcomes (with at least 
one-year exposure to the trained teachers). As shown in Figure 1, exposure to the MGD 
project increases the reading assessment literacy scores of students in treatment 
schools at midline by 1.2 points (p<0.01) relative to comparison group, which is 
approximately one-level increase in the reading assessment. The evaluation team also 
estimated the model by restricting the sample to include only the students who have 
been exposed to the project for two years (CP1 students from baseline who are now in 
CE2 in 2018-2019). The estimated impacts are somewhat larger (1.7 points) when the 
evaluation team limits the sample (fifth column). However, the project effect remains 
significant at all levels. 
141. In addition, the data show that gender plays an important role in students’ reading 
scores. As shown in the exhibit, the project teacher training is more likely effective for 
boys as compared to girls. The data show that the project has significant positive effects 
 
18 To identify these students, the evaluation team asked all students when they started in their school and excluded 
those who mentioned 2018-2019 as an answer (i.e., students who just joined the school and have not been exposed 
to the program yet). 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 19 | Page 
on both boys and girls, but the effect was doubled for male students (almost 2 levels 
improvement for boys compared to almost 1 level for girls). Also, the effect on girls is 
only statistically significant at the 10 percent level. 
Figure 1: Project Effects Using DID Method
Difference At least one-year exposure
Two-year 
exposure
Overall Sample Female Male Overall Sample
𝑌𝑌1
𝑀𝑀𝑀𝑀𝑀𝑀 − 𝑌𝑌0
𝑀𝑀𝑀𝑀𝑀𝑀 4.3*** (0.239) 4.1*** (0.271) 4.7*** (0.256) 4.45*** (0.357)
𝑌𝑌1
𝑛𝑛𝑛𝑛𝑛𝑛−𝑀𝑀𝑀𝑀𝑀𝑀 − 𝑌𝑌0
𝑛𝑛𝑛𝑛𝑛𝑛−𝑀𝑀𝑀𝑀𝑀𝑀 3.12** (0.336) 3.2*** (0.433) 2.9*** (0.360) 2.72*** (0.358)
∆ 𝑀𝑀𝑀𝑀𝑀𝑀 (Project 
effect) 1.2*** (0.417) 0.91* (0.511) 1.8*** (0.439) 1.7*** (0.503)
N 1,740 1,036 704 446
Source: Student Survey; Coefficient (Standard Error); *p-value < 0.1 ** p-value < 0.05 *** p-value < 0.01; 
Standard errors shown in parentheses are clustered at the school level.
142. While both boys and girls are positively impacted by the MGD intervention, this indicates 
that girls are not benefiting as much. As shown in greater detail in the performance 
analysis section below, girls and boys have similar levels of reading proficiency at CP1 
and CP2, but starting in CE1, more boys are reading proficient. This trend was similar 
at baseline, and also in the non-MGD schools, and this achievement gap between girls 
and boys appears to be exacerbated by the intervention. The qualitative data does not 
explain why, and in fact, parents spoke often of the importance of educating their 
daughters. Multiple groups of parents spoke of being inspired by national female role 
models and believed that by sending their girls at school, they can become ministers, 
members of parliament, and managers: “big ladies” who are holding “big positions.” 
Therefore, more investigation needs to be done to determine what is driving this gap 
(teachers’ attitudes, lower attendance by girls, parents’ selection of who to send to 
school, etc. are all ideas to be explored).
2.4.2 Performance Analysis
143. The performance evaluation was designed to measure changes in outcomes over the 
life of the project with regards to USDA’s two strategic objectives, improved literacy of 
school aged children and increased use of health and dietary practices.
144. The findings below are from the student reading assessment, and the household, 
student, school, canteen, and teacher surveys. The MGD indicators can be found in 
annex 9 and project results framework in annex 7. 19
MGD SO1: Improved Literacy of School Aged Children
145. Grade-level reading proficiency increased at each grade level between baseline and 
midline. The gains observed in MGD schools were larger than the gains observed in 
non-MGD Schools at each grade level. As seen in Figure 2, at every grade level (CP1 
through CM1), a larger share of students in MGD schools were at or above grade-level 
reading proficiency than in non-MGD schools. 
 
19 At baseline, to measure MGD 1.2.1.1 (Increased Access to School Feeding) the evaluators used composite scores 
for household food consumption, dietary diversity, and coping strategies as three of the indicators. These indicators 
were captured at midline to maintain consistency across the two evaluations (see Figure 72 through Figure 74 in 
Annex 6). However, being household scores capturing data unrelated to the scope of activities in the MGD program, 
the evaluation team does not believe that these indicators are an appropriate measure of this program, and are 
therefore not reported in the main body of the report.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 20 | Page 
146. While the rate at which girls were able to reach reading proficiency more than doubled 
at nearly every grade level, their gains were outpaced by boys, who more than tripled 
the level of reading proficiency of their counterparts at baseline in every grade except 
CE1, as show in Figure 3. This is similar to the findings of the evaluation team’s impact 
evaluation, which showed that while the MGD intervention significantly increased the 
reading proficiencies of all students, boys benefited to a greater extent. When looking 
at this trend by grade, it appears that this achievement gap between boys and girls 
starts particularly growing starting at the CE2 level. More detailed scores by grade level 
can be found in Figure 71 in Annex 6. 
Figure 2: Reading Assessment Scores by Grade (Midline Only)
Source: Student Survey. The vertical black line indicates the threshold for grade-level reading proficiency. Students 
to the left of the threshold are below proficiency. Students to the right of the threshold are at or above proficiency. 
The sample size was 964 at baseline and 1753 at midline. CM1 girls in MGD schools were oversampled at midline to 
gather sufficient observations to be able to analyse the effects of the take-home-rations project in those schools.
Figure 3: Reading Proficiency by Grade and Gender (MGD Only)
Mid MGD Base MGD
Grade Boys Girls All Boys Girls All
CP1 43.8% 46.8% 45.2% 3.6% 5.7% 4.6%
CP2 39.7% 35.8% 37.7% 12.9% 15.0% 13.9%
CE1 51.1% 45.7% 48.5% 24.6% 23.2% 23.9%
CE2 45.6% 27.3% 36.5% 6.9% 13.5% 10.0%
CM1 30.2% 17.4% 20.5% 10.0% 5.0% 7.5%
Total 41.0% 26.6% 32.2% 11.6% 12.5% 12.0%
Source: Student Survey. The sample size was 604 at baseline and 1149 at midline. CM1 girls in MGD schools were 
oversampled at midline to gather sufficient observations to be able to analyse the effects of the take-home-rations 
project in those schools.
More Consistent Teacher Attendance 
147. Based on records kept by principals, 95 percent of teachers attended school at least 90
percent of the time. However, based on teachers’ self-reported absences, only 80
percent of teachers were in attendance for 90 percent or more days of school. This 
ranged from a low of 61 percent of teachers in Cavally to a high of 97 percent of teachers 
in Bounkani (see Figure 4). No schools in Bounkani region were able to provide data 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 21 | Page 
collectors with records kept by school principals, so these self-reported figures cannot 
be corroborated. Figure 29 and Figure 30 in Annex 6 also illustrate teachers’ attendance.
148. While unable to say for certain, one hypothesis to explain the difference in official and 
self-reported absences is perhaps due to inflation of attendance at the school level. The 
evaluation team’s field-level data collection found that while most schools were able to 
show records, they rarely reported absences. This was also true for student attendance, 
with school records showing very low levels of student absenteeism. In the case of 
teacher attendance, the evaluation team tends to put more trust in the teachers’ 
reports, which not only appears more realistic, but also is corroborated by reports from 
parents and students in the FGDs about the prevalence of teacher absenteeism.
Figure 4: Teacher Absences by Region (Midline Only)
Source: Teacher Survey. The sample size was 268. Teaches were asked how many days they missed as a result of 
each of several causes.
149. When students were asked if they were prevented from attending class because their 
teacher was absent, 64 percent of all students said that this happened “sometimes.” 
This data matches the teachers’ self-reported attendance, as compared to the official 
school records. A small number (1.3 percent) of students in both MGD and non-MGD 
schools said that this happened “almost all the time,” though students reporting this 
were concentrated in a few schools. Overall, student-reported teacher absenteeism was 
about the same in MGD and non-MGD schools.
Figure 5: Student-Reported Teacher Absences (Midline Only)
How often could you not have school because 
the teacher was absent when you arrived? MGD Non-MGD Total
Never 37.3% 30.5% 35%
Sometimes 62.2% 66.7% 63.8%
Almost all the time 0.4% 2.8% 1.3%
Observations 1,149 604 1,753
Source: Student Survey.
Better Access to School Supplies and Teaching Materials and Improved Literacy 
Instructional Materials
150. Nearly all MGD schools (91 percent) were provided with school materials (books, school 
supplies, etc.), while a majority (72 percent) of non-MGD schools were also provided 
with school materials. These figures are about the same as from baseline. Large 
changes that can be observed in Bounkani MGD schools and Tchologo non-MGD schools 
(see Figure 6) may be attributable to the small number of schools within each region. 
0%
20%
40%
60%
80%
100%
Records: Attending 90% Self-reported: Attending 90%
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 22 | Page 
Figure 6: Access to School Materials by Region
Region Bafing Bag. Bounk. Cavally Gont. Poro Tcho. Total Obs.
MGD Midline 100% 100% 85.7% 100% 82.4% 94.7% 83.3% 91.2% 68
Baseline 83.3% 100% 28.6% 100% 100% 100% 83.3% 89.7% 68
Non￾MGD 
Midline 80% 100% 50% 83.3% 66.7% 60% 75% 72.2% 36
Baseline 80% 100% 100% 83.3% 83.3% 70% 25% 75% 36
Source: School Survey. Sample size was 104 at baseline and midline.
151. The majority of students enrolled in MGD schools had access to mobile libraries, and 
many students also had access to other materials, including reading boards and 
illustrated boards. In non-MGD schools, only 10 percent of students had access to such 
learning materials. Regionally, students in Bagoué (23 percent) and Tchologo (25
percent) were most likely to say that they had no access to reading materials (see 
Figure 31 in Annex 6).
Figure 7: Access to Reading Materials (Mobile Library, Reading Board, etc.)
Reading Material MGD Non- MGD
Sculpted plastic letters 14.7% 0.2%
Junior dictionaries 17.8% 0.7%
Illustrated board 43.3% 3.5%
Reading boards 47.4% 4.8%
None 13.8% 89.7%
Mobile library 67% 5.5%
Responses 2343 630
Observations 1149 604
Source: Student Survey. Respondents could select multiple choices.
Increased Skills and Knowledge of Teachers
152. The vast majority of teachers at every grade level said that they had participated in 
trainings provided by AVSI. Overall, this included 90 percent of all teachers, with a low 
of 84 percent at CE1 and a high of 96 percent at CE2. Of the 10 percent of teachers 
who did not participate in an AVSI training, 70 percent selected the “other” option. Of 
the 13 teachers who specified their “other” answer, 7 indicated they were either not 
assigned to a MGD school at the time of training, while the remainder were either not 
invited to participate in the training because of the training targeted CP teachers only 
or they were unaware of the training in their school. This reinforces the finding that 
nearly all targeted teachers received the MGD project’s training intervention. Although 
nearly all trained teachers (98 percent) found AVSI’s training to be useful or very useful, 
only a minority (22 percent) were utilizing four or more of the tools and materials 
provided and were also using the General Grid for Student Performance. Still, 96 percent
of those that had been trained said that they were putting into practice their AVSI 
training. (See Figure 32 in Annex 6.)
Figure 8: Teachers Having Participated in AVSI Trainings by Region (Midline MGD)
Region Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Percent 81.3% 93.3% 86.5% 95% 90.3% 90.9% 91.7% 90%
Obs. 32 45 37 20 72 99 24 329
Source: Teacher Survey.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 23 | Page 
Figure 9: Schools Having Received Training on Teaching of Reading
Region Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
MGD 
Percent 100% 100% 100% 100% 100% 100% 83.3% 98.5%
Obs. 6 8 7 5 17 19 6 68
Non-MGD
Percent 40% 66.7% 50% 16.7% 66.7% 70% 25% 50%
Obs. 5 3 2 6 6 10 4 36
Source: School Survey.
Improved Attentiveness
153. On average, every grade level of teachers said that more than 6 in 10 of their students 
were attentive on a typical day. Teachers of CM2 students said that their students were 
most likely to be attentive (7.5 out of every 10 students). There were only small 
differences in the reported attentiveness of girls and boys in class. Teachers in Bafing 
rated their students as the most attentive (7.8), while teachers in Poro rated their 
students as the least attentive (6.3). 
Figure 10: Number of Students out of 10 Described as Attentive by Teachers on a 
Typical Day (Midline MGD Only)
Grade CP1 CP2 CE1 CE2 CM1 CM2 Total
Girls 6.9 6.9 6.5 6.4 6.6 7.3 6.8
Boys 6.8 6.6 6.3 6.3 6.7 7.7 6.7
All 6.8 6.8 6.4 6.4 6.6 7.5 6.7
Observations 52 60 57 55 52 54 330
Source: Teacher Survey.
Figure 11: Student Attentiveness by Region (Midline MGD Only)
Region Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Girls 7.7 6.6 6.7 6.4 6.8 6.5 6.9 6.8
Boys 7.8 6.7 7.1 7.1 7.2 6 5.8 6.7
All 7.8 6.7 6.9 6.8 7 6.3 6.4 6.7
Obs. 32 45 37 20 72 100 24 330
Source: Teacher Survey.
Improved Student Attendance
154. On average, students were recorded as having attended school on 99% of school days. 
No students were recorded as having attended fewer than 80 percent of school days. 
According to school records, 52.7 percent of students in MGD schools did not miss a 
single day of school, compared to perfect attendance in 40.1 percent of students in non￾MGD schools. (See Figure 35 in Annex 6 for regional disaggregation.)
155. Like the teacher attendance records, the student attendance records appear to be 
artificially inflated, based on what the evaluation team heard during its FGDs with 
parents and students, who stated that they did sometimes miss school.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 24 | Page 
Figure 12: Students with Perfect Attendance (Midline Only)
MGD Non-MGD
Student did not miss a single day of school 52.7% 40.1%
Source: School Survey. The sample size was 916.
Increased Student Enrolment
156. In contrast to what the evaluation learned from the stakeholder analysis, overall 
enrolment decreased in MGD schools. However, when looking at the data more closely, 
this appears to be due to a large decrease in the enrolment of boys and girls in CP1 and 
CP2.20 All other grades saw increases in average enrolment of both girls and boys. In 
MGD schools, girls’ enrolment in CM1 increased from 16.4 girls per school to 17.7 and 
enrolment in CM2 increased from 12.8 to 15.8 per class from baseline to midline. 
Figure 13: Average Student Enrolment per Grade (All Schools)
CP1 CP2 CE1 CE2 CM1 CM2 All 
Girls
Mid MGD 23.1 21.8 21.6 19.9 17.7 15.8 114
Base MGD 29 25.2 21.3 18.9 16.4 12.8 113
Mid non-MGD 18.9 18 18.3 15.8 15.1 12.3 83.7
Base non-MGD 21.2 17.4 16.9 12.5 10.6 10.8 69.4
Boys 
Mid MGD 20.8 21.2 20.8 20.7 18.7 18.3 114
Base MGD 29 26.9 23.1 20.5 18.8 17.3 123
Mid non-MGD 18.7 18.4 17.5 17.4 15.7 13.4 86.3
Base non-MGD 21.7 18.6 18.4 17.3 14.5 14.5 80.2
All
Mid MGD 43.9 43 42.4 40.6 36.4 34.1 228
Base MGD 58 52.1 44.5 39.4 35.2 30.1 236
Mid non-MGD 37.6 36.4 35.8 33.2 30.3 25.7 170
Base non-MGD 42.8 35.9 35.3 29.8 24.5 25.3 150
Source: School Survey. Sample size was 207 schools.
157. Increases in girls’ CM1 and CM2 enrolment (from 33.9 to 42.5) in regions where THR
are offered contributed to the overall increases in girls’ enrolment. In contrast, in 
regions where THR are not offered, girls’ CM1 and CM2 enrolment did not increase as 
much (from 17.7 to 22.2). The ratio of girls to boys enrolled in these two grades 
increased from 0.88 to a near-parity ratio of 0.96 in take-home-rations schools, while 
the ratio increased from 0.7 to 0.81 in non-take-home-rations schools. (See Figure 36, 
Figure 37, Figure 38, and Figure 39 in Annex 6 for more details on enrolment in schools 
with THR.)
 
20 The evaluation team is unable to offer an evidence-based reason for this discrepancy, however, one hypothesis is 
that data was collected at the beginning of the school year at midline (compared to the end of the school year at 
baseline), and therefore enrolment records may not be complete. The team anecdotally heard from some program 
partners that children in rural areas may still be returning to their village after spending a summer working away 
from home. However, this topic was not explored in-depth at midline. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 25 | Page 
Figure 14: Ratio of Girls to Boys Enrolled in CM1 & CM2 (MGD Schools)
Source: School Survey. The red line indicates gender parity in enrolment. Sample size was 133 schools.
Reduced Health Related Absences
158. According to their parents, last year, 4.1 percent of students enrolled in MGD schools 
missed more than 10 days due to illness. Again, the number of self-reported absences 
are greater than what is reported through the official student attendance records at the 
school. On average, the number of days missed was slightly lower for girls (2.8) than 
boys (3.3), but more girls were also absent for more than 10 days at a higher rate (4.2
percent for girls and 4 percent for boys). Students at Non-MGD schools missed more 
days of school on average (3.3) and were more likely to have missed more than 10 
days of school due to illness (6.4 percent). Eight percent of boys enrolled in non-MGD 
schools missed more than 10 days of school due to illness. (See Figure 40 and Figure 
41 in Annex 6.)
MGD SO2: Increased Used of Health and Dietary Practices
159. The percentage of school days during which food was provided was nearly 100 percent
in each region, but varied slightly, ranging from to Poro (88 percent) to Bafing (100
percent). There were no substantive differences between the days of canteen services 
available to boys and girls. This was a large increase from baseline, where schools on 
average only provided meals on 57 percent of the days when school was in session.
(See Figure 43 and Figure 50 in Annex 6 for total statistics and regional disaggregation.)
160. Almost all parents (94 percent) of children enrolled at MGD schools were aware of the 
school canteen project, and almost all (92 percent) reported that their children ate at 
the canteen in the previous school year. While the majority of children ate at the 
canteen regularly (three days a week or more), only about half (53 percent) of children 
ate at the canteen every school day. (See Figure 51 and Figure 52 in Annex 6 for 
parents’ awareness and children’s participation in the feeding project.) Qualitative data 
from parents and students suggest this may be because students did not have the 25 
CFCA daily fee, or the canteen was not operational five days a week. WFP staff noted 
that while technically there are 120 school days in a year, it is rare for schools to actually 
have this many days in reality.
0.0
1.0
Mid THR Base THR Mid non-THR Base non-THR
CM1 CM2
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 26 | Page 
Figure 15: Frequency that Students Ate at the School Canteen
Source: Household Survey. The sample size was 1,143.
161. The percentage of school days during which canteens served food groups constituting 
minimum dietary diversity (MDD) varied by region. However, the low number of schools 
in some regions with sufficient canteen records on the number of food groups and 
ingredients served partly explains this variation. Most regions served MDD more than 
80 percent of the school days, but schools in Bounkani served a meal meeting MDD 
requirements only 50 percent of the school days and in Cavally, records regarding 
dietary diversity were not available. Schools in Tchologo served an MDD meal 100
percent of canteen days. There was no difference between the MDD meals available to 
boys and girls. Overall, the percentage of school days where MDD was provided 
increased greatly from baseline, where only 41 percent of the school feeding days met 
this requirement. For greater detail on the foods served by region, see Figure 54. 
162. The focus groups with the WPGs indicate that in some schools, the lunch provided is 
more nutritionally diverse. Specifically, the ten established WPGs said they were taught 
how to cook a balanced diet for children and how to derive nutrients from their local 
crops. The workers in the school canteens where they contribute their food (including 
vegetables) have been trained on how to cook a balanced diet for school children.
Figure 16: Portion of Days that Canteen Provided Minimum Dietary Diversity 
Mid MGD Base MGD
Region Percent Obs. Percent Obs.
Bafing 83% 2 33% 5
Bagoué 83% 6 45% 4
Bounkani 50% 2 53% 6
Cavally N/A 0 30% 3
Gontougo 83% 9 39% 15
Poro 83% 10 43% 17
Tchologo 100% 3 36% 1
Total 83% 32 41% 51
Source: Canteen Survey. Minimum dietary diversity is defined as a day’s meals consisting of at least four of the 
following food groupings: 1) cereals, 2) tubers and root vegetables, 3) legumes and nuts, 4) dairy products, 5) 
meat, fish, and offal, 6) eggs, 7) high vitamin A fruit and vegetables, and 8) other fruit and vegetables.
0%
20%
40%
60%
80%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Every day Four Three Two One Never
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 27 | Page 
Increased Knowledge of Health and Hygiene Practices
163. While most canteen managers were able to identify three or more health and hygiene 
practices at midline (88 percent), the number had decreased slightly from baseline (93
percent). In Cavally, less than half of canteen managers (40 percent) could cite three 
or more health and hygiene practices. (See Figure 55 and Figure 56 in Annex 6.) Since 
canteen managers also serve as teachers, these results could potentially be explained 
by teacher turnover in these schools.
164. Teacher training on health practices at midline is somewhat similar to what it was at 
baseline. While skills-based health education and deworming trainings were almost 
unchanged, the number of MGD schools receiving HIV/AIDS prevention trainings nearly 
halved from 22 percent to 11 percent. These trends were generally present in non-MGD 
schools as well, except for skills-based health education where the proportion of 
teachers receiving the training decreased from 28 percent to 17 percent. (See Figure 
57 in Annex 6.)
Figure 17: Teacher Training on Health Practices
MGD Non-MGD
Type of Training Midline Baseline Midline Baseline
Skills-based Health 
Education 38.8% 35.3% 17.1% 27.8%
Deworming 62.7% 67.6% 33.3% 36.1%
HIV/AIDS Prevention 11.8% 22.1% 8.3% 22.2%
Observations 68 68 36 36
Source: School Survey. 
Increased Knowledge of Safe Food Preparation and Storage 
165. At the time of the evaluation, WFP indicated the proportion of MGD canteen managers 
trained in safe food preparation and storage was 100 percent in only two regions (Bafing 
and Cavally) and about 50 percent in the five other regions, with the remainder to be 
trained in December 2018 and March 2019. Suggesting that the project effectively 
increased canteen manager knowledge, MGD canteen managers’ knowledge of safe food
preparation and storage practices increased modestly in all regions between baseline 
and midline (from 70 percent to 82 percent for food prep and from 76 percent to 83
percent for food storage). However, this was not a strong trend; while some regions’ 
canteen managers’ knowledge of safe food preparation and storage practices increased, 
others decreased.
Figure 18: By Region, Canteen Managers Know Three Food Storage and Food 
Preparation Best Practices
Region Bafing Bagoué Boun. Cavally Gont. Poro Tcho. Total
Midline
Food 
Storage 67% 75% 100% 60% 88% 84% 100% 83%
Food 
Prep 83% 75% 100% 0% 94% 90% 80% 82%
Baseline
Food 
Storage 83% 50% 86% 60% 94% 68% 83% 76%
Food 
Prep 83% 75% 86% 20% 81% 68% 50% 70%
Obs. 6 8 7 5 16 19 6* 67
Source: Canteen Survey. Only five schools were available in Tchologo at midline.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 28 | Page 
Increased Knowledge of Nutrition
166. Only 12 percent of canteen managers in MGD schools at midline reported that members 
of the canteen management committee received training in nutrition. Among canteen 
managers who reported that there was a canteen management committee at their 
school (79 percent at midline, 64 percent at baseline), only 33 percent indicated that 
they received nutrition training, compared to 41 percent at baseline. 
167. While the MGD project focuses on training canteen managers, these managers also 
serve as teachers and rotate out of the canteen manager position, meaning teachers’
knowledge of topics relevant to canteen management can help ensure the acting 
canteen manager has the required knowledge. Teachers in more than half of MGD 
schools (54.4 percent) received trainings on nutrition. Only a small number (8.3
percent) of non-MGD schools received the same trainings. Non-MGD schools also did 
not receive any student nutrition training or vegetable garden training, while a modest 
number of MGD schools did receive these trainings. 
Increased Access to Safe Water
168. Access to improved water sources is about the same between MGD (44 percent) and 
non-MGD schools (47 percent). But about half of both MGD (53 percent) and non-MGD 
(50 percent) schools did not have adequate access to water. In MGD schools, about 
one-third (32 percent) of those with access had water access on the school grounds, 
while 43 percent were less than a 15-minute walk to a water source, and one-quarter 
were 15 minutes or further from a water source. In the evaluation team’s qualitative 
data collection, beneficiaries mentioned lack of water on school grounds as being 
problematic, as either students (generally girls) or canteen staff would have to fetch 
water for the school canteen. For students, they left their lessons during the day for 
this task. This additional task for girls should be explored as a potential reason which 
may partially explain the gender disparity in reading proficiency, as it may result in girls 
having less classroom time than boys. 
169. The most common problem preventing access to water in both MGD (59 percent) and 
non-MGD (50 percent) schools was broken water pumps. (See Figure 59, Figure 60, 
and Figure 61 in Annex 6) 
Figure 19: Type of Access to Water in Schools
Source: School Survey. Sample size at baseline and midline was 104.
0%
20%
40%
60%
80%
100%
Mid MGD Base MGD Mid non-MGD Base non-MGD
No water Unprotected well Protected well Village pump Tap
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 29 | Page 
Figure 20: Distance to School’s Water Source, Where Available (Midline MGD Only)
Source: School Survey. Sample size was 68.
Figure 21: Most Common Problems with Schools’ Water Source
Source: School Survey. The sample size was 56 at baseline and 37 at midline.
170. Slightly more MGD schools (59 percent) than non-MGD schools (50 percent) have 
sanitary facilities on-site. Less than one-third (31 percent) of MGD schools and only 
one-quarter (25 percent) of non-MGD schools have gender-separated sanitary facilities. 
Less than one-third of either MGD schools (26 percent) or non-MGD schools (31
percent) have separate sanitary facilities for teachers and students. (See Figure 62.)
Figure 22: Access to and Quality of Sanitary Facilities
Source: School Survey. The sample size was 104 at baseline and midline.
0%
20%
40%
60%
80%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
In the school grounds Less than 15 min walk Over 15 min walk
0%
20%
40%
60%
80%
100%
Mid MGD Base MGD Mid non-MGD Base non-MGD
Broken pump Animals occupying water source
Dried up water source Water is used for agriculture
Other Distance
0%
20%
40%
60%
80%
100%
Mid MGD Base MGD Mid non-MGD Base non-MGD
Has sanitary facilities Separated by gender Separate for teachers
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 30 | Page 
Increased Access to Preventative Health Interventions
171. The number of schools that reported receiving deworming treatments decreased from 
baseline to midline in both MGD (from 72 percent to 64 percent) and non-MGD schools 
(from 64 percent to 56 percent).21 Distribution of micronutrient supplements fell by 
more than half in MGD schools (from 53 percent to 26 percent) from baseline to midline.
(See Figure 63 and Figure 64.)
Increased Access to Requisite Food Preparation and Storage Tools and Equipment
172. In general, canteen managers said they were not provided with enough food 
preparation equipment (68 percent). However, this is an improvement from baseline, 
where 81 percent of canteen managers said they did not have enough food preparation 
equipment. 
Figure 23: Sufficient Food Preparation Equipment
Preparation 
Equipment Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Midline 17% 38% 14% 20% 31% 53% 0% 32%
Baseline 17% 25% 14% 20% 19% 26% 0% 19%
Obs. 6 8 7 5 16 19 6* 67
Source: Canteen Survey. Only 5 schools were available in Tchologo at midline.
173. About the same percentage of MGD schools have access to improved food storage 
equipment at midline as had access at baseline. Access to stock cards was the only 
food-related equipment that changed in accessibility from baseline, increasing from 65
percent to 100 percent of MGD schools. The percentage of canteen managers with 
enough food storage equipment grew modestly from 10 percent at baseline to 38
percent at midline. No canteen managers in Bounkani, Cavally, or Tchologo said that 
they had enough food storage equipment. (See Figure 65 and Figure 66.) These 
numbers could be impacted by WFP’s delivery of more than 1000 storage pallets to 
DCS, along with pallet jacks, to equip schools for the 2018-2019 school year. 
Figure 24: Sufficient Food Storage Equipment
Storage 
Equipment Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Midline 50% 63% 0% 0% 44% 53% 0% 38%
Baseline 0% 13% 0% 0% 31% 5% 0% 10%
Obs. 6 8 7 5 16 19 6* 67
Source: Canteen Survey. Only 5 schools were available in Tchologo at midline.
174. Across all regions, 46 percent of canteens had an improved stove with a chimney, 
instead of just a traditional stove or a furnace without a chimney. This includes 100
percent of canteens in Cavally and Tchologo where there is a stove. This is large 
increase from baseline, where only 7 percent of school canteens had an improved stove.
(See Figure 69 and Figure 70.)
 
21 WFP reports that the Ministry of Health conducted a national deworming exercise in 2018, distributing over 1 
million deworming tablets nation-wide, including to all MGD supported schools. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 31 | Page 
Figure 25: Stove Type in the Canteens
Source: Canteen Survey. The sample size was 68 at baseline and midline.
Figure 26: Stove Type in the Canteens, by Region
Source: Canteen Survey. The sample size was 68 at baseline and midline.
Key findings and conclusions – Impact 
• The impact analysis shows that the MGD intervention significantly improves the 
reading proficiency of students. Exposure to the MGD project increases the reading 
assessment literacy scores of students in treatment schools by approximately one 
level in reading compared to the non-MGD schools. Students who have been 
exposed to the project for two years have larger estimated impacts (1.7 levels).
• The project effect is twice as great for male students, who show almost two levels 
of improvement in reading proficiency, compared to one level for girls. Girls and 
boys have similar levels of reading proficiency in CP1 and CP2, but starting in CE1, 
the gains for boys are much greater.
• Enrolment increased at MGD schools for girls in CM grades where THR are offered. 
In the three regions with THR, the average enrolment increased by almost 10 girls 
in the upper grades 
• The percentage of school days during which food was provided was nearly 100
percent in each region. This was a large increase from baseline, where schools on 
average only provided meals on 57 percent of the days when school was in session.
• Most MGD canteens served food containing the minimum dietary diversity on more 
than 80 percent of school days at midline, compared to 40 percent of days at 
baseline. 
• Among canteen managers who reported that there was a canteen management 
committee at their school, 33 percent indicated that they received nutrition 
training, compared to 41 percent at baseline. 
Midline
None/Unknown Traditional stove
Furnace without chimney Stove with chimney
Baseline
None/Unknown Traditional stove
Furnace without chimney Stove with chimney
0%
20%
40%
60%
80%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
None/Unknown Traditional stove
Furnace without chimney Stove with chimney
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 32 | Page 
• About half of both MGD (53 percent) and non-MGD (50 percent) schools did not 
have any adequate access to water, similar to the baseline numbers. 
• Slightly more MGD schools (59 percent) than non-MGD schools (50 percent) have 
sanitary facilities on-site. Less than one-third (31 percent) of MGD schools and only 
one-quarter (25 percent) of non-MGD schools have gender-separated sanitary 
facilities.
• The number of MGD schools receiving deworming treatments decreased from 
baseline to midline (from 72 to 64 percent), as did the number of schools receiving 
micronutrient supplements (53 to 26 percent).
• The percentage of canteen managers who reported having sufficient food 
preparation equipment increased from baseline to midline (from 19 to 32 percent), 
as did sufficient food storage equipment (from 10 to 38 percent). 
• Across all regions, 46 percent of canteens had an improved stove with a chimney, 
instead of just a traditional stove or a furnace without a chimney. This is large 
increase from baseline, where only 7 percent of school canteens had an improved 
stove. 
2.5. Evaluation Criteria 5 – Sustainability 
175. In assessing sustainability, the evaluation team specifically focused on WPGs, as these 
were proposed as the foundation of the project’s handover strategy. The first section 
presents a case study of these groups, addressing the following evaluation question:
• To what extent are WPGs contributing to school canteen supply, and is there 
evidence that their contributions will continue or scale up after the project ends?
176. The second section includes an overall assessment of the project’s progress towards 
sustainability considering the following questions:
• To what extent are the results of the activities likely to be sustainable?
• What are the key factors likely to affect sustainability of the project? 
2.5.1 Women’s Production Groups
Description
177. Twenty WPGs were supported by the project as of the data collection period for this 
midterm evaluation. The evaluation team’s case study involved in-depth discussions 
with members of all 20 WPGs, 10 of which were newly supported (since June 2018), 
and 10 of which were more established (supported by the project from 2017). These 
groups are mainly composed of women who account for 80 percent of the total number 
of members in the newer groups and over 90 percent of members in the established 
groups. The few men that are part of the group generally have a higher education level 
and more agricultural experience, and as such, they support women with administrative 
management and physically challenging work. For example, prior to the start of the 
agricultural season, men decide in relation with the technical support structures when 
to start land preparation and apply specific soil management techniques.
178. These groups have been being supported by the local ANADER staff prior to WFP support 
through the MGD project. This support stems from a government support that usually 
provides such groups with a minimum package of capacity building activities under the 
umbrella of the GoCI’s national program of agricultural development and other projects 
with technical and financial partners. The groups the evaluation team interviewed have 
been established for a minimum of three years, with some established over 10 years 
ago. The number of members ranged from 30 to around 100 or more, while most groups 
said their membership numbers fluctuated. The average age of the women for the 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 33 | Page 
majority of the groups was 35, although there was one group were the women were 
much older (the youngest member was 42 years old). A common characteristic of all 
the women is their illiteracy. 
179. The main crops grown by these groups depends on their geographic location. While 
maize and rice are typically grown in the northern part of the country, cassava and yam 
are cultivated in the eastern part. In the ten established groups, women spoke positively 
about learning how to grow and use vegetables including eggplant, okra, peppers, 
tomatoes, onions, and greens. At the household level, they have increased the variety 
of foods they serve to their families by involving vegetables and leaves in their cooking. 
The schools these women support have also diversified their lunch offering, due to the 
nutritional training and the donation of vegetables by the WPGs. For the ten newer 
WPGs in the evaluation team’s sample, many have initiatives to grow vegetables, but 
this is still in an early stage of development. As reported by four newer WPGs, the main 
barrier to growing vegetables is the lack of water or irrigation systems.
180. Each group has enrolment and internal contribution fees, ranging from 2,000 FCFA to 
10,000 FCFA. One group requires that members be married and follow specific rules 
pertaining to respect, punctuality, and participation. Some groups are required to give 
a portion of their production to the village chief as a reward for providing land to the 
group.
181. All groups started contributing to the canteens before the MGD project started their 
support. However, the importance placed on food donation and the production capacity 
varies from one group to the next. The ten established groups reported much higher 
production rates and reported being able to donate more food to the canteens thanks 
to the support received through the project. 
Early Indicators of Success
182. Even though it is too early to fully assess the impact of these groups, the members 
have indicated that their work has started to have an effect. For example, the indirect 
impact their support to the canteen is the increase in school attendance, attentiveness, 
and nutritional outcomes of school age children. Members also reported that it is likely 
that the project indirectly helps parents secure a balanced diet for their kids and reduces 
their food expenses. Depending on the leadership of the group, the activities went 
beyond providing food to the canteen. For example, in one group, thanks to the revenue 
yielded from their agricultural activities, they were able to buy hand washing devices 
for their school.
183. In the ten more established WPGs, the gains are more apparent. The members cited 
specific examples of how they benefited from materials received. For example, by 
having their own scales, women avoid being cheated as they know the true weight of 
their harvest. Women also mentioned that prior to being given watering cans, they used 
cups to water their crops. In addition to using the materials for their own purposes, 
these groups also rent out their equipment to generate extra income for their groups. 
Not only do they save time from using grain mills to grind maize (rather than using a 
mortar and pestle), they charge community members a daily fee (usually around 500 
FCFA) to use the mill. 
184. According to members in the ten established groups, there are community level effects 
greater than the growth of their individual WPGs. Although they charge community 
members to use the grain mills, women said that this ultimately saved money for 
community members, as they no longer have to travel long distances to process their 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 34 | Page 
crops. The mills also transform the grains into different types of eatable foods (attieke, 
couscous, etc.) for donation to the school canteens. 
185. Women in the ten established groups also appreciated receiving marketing and grouped 
sales training. Prior to the MGD project, they used to sell their harvest individually. After 
their training, they are now grouping their sales and bargain collectively for a more 
favourable price at the market. One group said that they have been put in contact with 
the Ivorian board for the purchase of their agricultural products. Members in all groups 
appreciated the local support of ANADER (for the newer groups) and BFCD (for the 
established groups) on the ground. The continuous technical assistance provided is a 
strength and asset for the project to achieve sustainability.
186. Transparency is another key to the success of WPGs. As acknowledged by participants, 
when all members are given the groups revenue figures by the leadership, it encourages 
the members to continue supporting the group activities. Beyond transparency, a clear 
commitment of an engaged president is also key to the WPGs’ success. For the newer 
groups, community sensitization on the role of the members is necessary according to 
participants. As these WPGs are still in their infancy, there have not been substantial 
income distributions for their members. There are some members who complained 
about the lack of returns to participating in the group, so it is important for strong 
leaders and other group members to motivate all participants to avoid the collapse of 
their WPG. 
187. Women mentioned that their participation in group activities in a primarily female 
environment strengthens the social capital among group members. Within these 
groups, social networks are created, and members rely on each other as a coping 
strategy mechanism in the case of adverse social shocks, such as an illness or death in 
the family. One group gave an example of paying for a life-saving surgery for a member 
of their WPG. A common refrain the evaluation team heard from the women was: “If 
someone is in trouble, the group helps her.” 
188. In the more established groups, this social support goes even further, thanks to the 
money gathered through income generating activities and to the financial management 
training received as part of the project. These groups are able to give small loans to 
their members. In addition, most of these groups reported that their generated income 
was now sufficient to purchase their own seeds and fertilizer 
189. The groups are also places where members learn soft skills that benefit them within 
their families and communities. As members of some groups stated, the older women 
participants provide social advice that members appreciate. 
Threats to Sustainability
190. The factors preventing groups’ effectiveness are the same as those that hinder 
sustainability. Lack of access to physical and natural capital are factors that could 
prevent groups from becoming stronger and autonomous. For example, there is an 
increasing scarcity of agricultural land in Côte d'Ivoire, and as groups grow, finding land 
for vegetable crop production might become challenging in a near future. While some 
groups have property rights or own the land, most of the groups interviewed were given 
access to land through a customary land rights whereby the village chief shares part of 
his relatives’ land to the group. Women members said that there is real risk that the 
owners will withdraw these lands to take advantage of a rise in the price of cash crops 
such as coffee and cocoa.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 35 | Page 
191. Some of the newer WPGs supported by the MGD project since June 2018 raised 
concerns about being able to sell the produce they plan to grow, and requested 
assistance in bringing their produce to market. In response to these concerns, WFP 
plans to deliver this kind of technical assistance in the coming years as part of the 
package of services it works with ANADER to provide to the WPGs it supports. The ten 
established WPGs in the evaluation team’s sample appreciated this aspect of their 
training from BFCD, specifically mentioning the economic benefits of grouping their 
produce in the markets.
192. Factors related to climate variability also threaten sustainability. Rain variability from 
one year to another represents a serious threat to all groups’ agricultural activities. This 
variability leads several groups to state that despite the provision of agricultural inputs, 
insufficient or too much rain limits their production. Group members said this could be 
partially alleviated if they gained access to water retention and irrigated facility 
systems, saying lack of infrastructure prevents their growth. 
2.5.2 Overall Assessment of Sustainability
193. In March 2018, with technical support from the WFP, the GoCI published the country’s 
first national school feeding policy, covering the period of 2018 to 2025. It sets an 
ambitious goal: “By 2025, all students in Côte d’Ivoire enrolled in the basic education 
structures take a warm and balanced meal every school day at the canteen.” It lists 
specific educational, health/nutritional, local production, and environmental objectives 
in line with UN Sustainable Development Goals (SDGs) 3, 4, 5, 8, and 15. Importantly, 
it emphasizes the multi-sector approach required to implement a national school 
feeding program, places specific responsibilities on the MENETFP and partnering 
government entities, and establishes the national structures for steering and 
coordinating the GoCI’s efforts towards reaching full school feeding coverage by 2025. 
This is a strong foundation for future technical assistance to support the GoCI in 
effectively taking over school feeding in all MGD and other WFP-supported school 
canteens, eventually ensuring the project’s national ownership and sustainability.
194. However, at the local level, turnover remains a threat to sustainability at all levels – 
cooks, COGES, school canteen management committees, WPGs, and teachers. For 
example, if a new cook is not trained to use an improved stove correctly, it will crack. 
New cooks might not know how to properly cook split peas and will therefore leave 
them in the warehouse instead of using them as part of the daily school meals. Turnover 
also affects the support groups affiliated with each school, specifically the school 
canteen management committees, the COGES, and the WPGs. The evaluation team’s
qualitative research found that these groups do not receive regular training or continual 
support frequently enough to counter these groups’ turnover rates (for example, from 
the survey data, only 33 percent of canteen managers reported receiving training in 
nutrition) 
195. Teacher turnover is also a threat to sustainability, as teachers in the MGD schools are 
taught pedagogical and literacy techniques but some are then assigned to a different 
school. Local stakeholders mentioned that they struggled to retain teachers in their 
community because they lack adequate teacher housing, sanitary installations, and 
water sources. In these situations, in addition to losing the benefits of the MGD teacher 
training, the school loses the services of a government-funded, certified teacher and 
may have to rely on untrained volunteer teachers.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 36 | Page 
196. In addition to the WPGs, additional groups such as the COGES, who are well acquainted 
with and already assist their schools’ needs, are key to the community taking ownership 
of the project activities. Strengthening their capacities and implicating them as partners 
in the project will increase the likelihood of the project achieving sustainable, long term 
impact. 
197. The primary sustainability plan of this project is to handover the schools to become fully 
part of the National School Canteens Program, where schools will benefit from both 
government and community support. To that extent, government stakeholders indicate 
that sustainability efforts should include additional training for government staff in 
monitoring and evaluation and commodity management and logistics. While WFP 
provides M&E and logistics training to DCS every year, several government stakeholders 
mentioned wanting to learn more about the monitoring and evaluation component, as 
this would not only increase their current technical capacity with the MGD project, but 
also empower them to improve other projects.
Key findings and conclusions – Sustainability 
• In the ten established WPGs, the members have substantially increased their 
contributions from last year to this. Most of these groups reported that their 
generated income was now sufficient to purchase their own materials without 
needing external financial support from the project. These women reported the 
material and trainings received through the project is making their groups stronger 
and therefore more likely to sustain contributions after the project is over. 
• Within the WPGs, social networks are created, and members rely on each other as 
a coping strategy mechanism in the case of adverse social shocks, such as an illness 
or death in the family. 
• The factors preventing full effectiveness of the project’s WPGs in supporting their 
school canteen are the same as those that hinder sustainability. Lack of access to 
physical and natural capital prevent groups from becoming stronger and 
autonomous, along with factors related to climate variability.
• Turnover is a threat to sustainability at all levels–at both the local level (cooks, 
COGES, school canteen management committees, WPGs, and teachers), and at the 
regional and national level (regional advisors, elected government officials, and 
WFP/partner staff).
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 37 | Page 
3. Conclusions and Recommendations 
198. 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 8 
recommendations of how WFP and its partners can take action to build on the lessons 
learned.
3.1. Overall Assessment/Conclusions 
199. Committed to providing 24 600 MT of daily hot meals and THR to a total of 125,000 
primary school students at 613 rural schools over five years, WFP is largely successful 
in reaching its operational objectives at midline while delivering significant, positive 
impact on beneficiaries in seven of the most vulnerable regions of Côte d’Ivoire. 
200. For MGD strategic objective 1 (SO1), Improved Literacy of School Aged Children, 
students in the MGD schools have significantly increased their reading levels at midline. 
Relative to the comparison schools, student in schools with school feeding and trained 
teachers increased their reading levels by approximately one grade-level since the 
project started. The data show that the effect of being exposed to the project for two 
years is even stronger. However, while there are significant positive effects for both 
boys and girls, the effect was doubled for male students (almost 2 levels of
improvement for boys compared to almost 1 level for girls). 
201. In addition, the evaluation team also found a disproportional number of men in the 
membership of the COGES, and in fact, some groups were exclusively male. The 
leadership in the COGES also skewed toward male. This risks side-lining female voices 
in the management of their children’s school and perpetuating patriarchal gender roles.
202. At the beneficiary level, parents and members of the COGES agreed that the MGD 
project was highly relevant in addressing the critical challenges of attracting students 
to school and consistently serving meals at the school canteen. The impressive reading 
gains measured at every grade level in MGD-supported schools confirm the 
transformative impact the approach designed and implemented by WFP and AVSI had 
on its student beneficiaries.
3.2. Lessons Learned and Good Practices
203. Good Practice 1: Local Recipes and Cookbooks for Split Peas. Distributing an 
alternative source of protein to school feeding project beneficiaries may result in 
significant cost savings which allow implementers to reach more beneficiaries with 
nutritious meals. However, beneficiaries frequently disliked the split peas or found it 
difficult to prepare it for optimal taste and nutritional value. Multiple key informants 
from the Government and WFP said they faced that difficulty with one of the selected 
MGD commodities, split peas, in Côte d’Ivoire. However, DCS with WFP support were 
able to address it by developing locally-inspired and validated recipes incorporating the 
commodity into dishes commonly served in each region in Cote d’Ivoire. These recipes, 
presented in a large-print cookbook with pictures, reportedly addressed the problem of 
children rejecting split peas, canteen workers failing to properly prepare them, and split 
peas piling up unwanted in storehouses. This was confirmed in FGDs with CM student 
beneficiaries, where many enjoyed the split peas and few disliked them. This allowed 
for an efficient allocation of USDA project funding which can be replicated in similar 
contexts.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 38 | Page 
204. Good Practice 2: Strong and Efficient Presence in Beneficiary Communities. The 
MGD project relies heavily on the GoCI’s own monitoring network of DPFC education 
advisors and DCS extra-curricular advisors in charge of school canteens. This allows for 
frequent data collection, reduces cost to the project, and facilitates the project’s future 
take-over by the GoCI. However, this layer of separation from end-beneficiaries may 
hamper communication between WFP and local beneficiaries on the latter’s needs. 
Supplementing the GoCI’s efforts, AVSI efficiently maintained a local presence by 
employing community members as focal points for supporting project field logistics and 
monitoring efforts. With this network of focal points, AVSI visited each MGD school at 
least three times per year to collect information on student and teacher progress, as 
well as monitor the use of project-provided teaching and learning equipment. While 
there are opportunities to consolidate and improve MGD monitoring efforts, the model 
being implemented in Cote d’Ivoire is a good example of both government integration 
and efficient field monitoring.
205. Good Practice 3: Operational Integration with Government. The MGD project is 
greatly integrated into the national school feeding structures. WFP’s alignment with 
Government strategies fostered strong partnership at the national and regional level 
which allowed the project to be implemented through a train-the-trainers model with 
the Government’s existing network of regional advisors and technicians. This helps 
practically build the capacity of DCS and DPFC civil servants to administer activities 
both independently and effectively at the end of the project.
206. Good Practice 4: National School Feeding Policy. In March 2018, with technical 
support from the WFP, the GoCI published the country’s first national school feeding 
policy, covering the period of 2018 to 2025. It lists specific educational, 
health/nutritional, local production, and environmental objectives in line with UN 
Sustainable Development Goals. Importantly, it emphasizes the multi-sector approach 
required to implement a national school feeding program, places specific responsibilities 
on the MENETFP and partnering government entities, and establishes the national 
structures for steering and coordinating the GoCI’s efforts towards reaching full school 
feeding coverage by 2025. 
3.3. Recommendations
207. Based on the findings and conclusions of this evaluation, the recommendations of the 
evaluation team are outlined below. The target group for each recommendation is 
clearly identified. The opinions expressed in these recommendations are those of the 
authors, and do not necessarily reflect those of WFP or USDA.
208. Recommendation 1: Strategically Reduce the Proportion of Canteen Days 
Covered in MGD Schools. Currently, WFP provides meals throughout the entire 120-
day school term. As heard from government stakeholders and local beneficiaries, this 
discourages parents and WPGs from contributing money and/or food to the canteens. 
At the same time, few WPGs reported being close to being able to adequately support 
a school canteen without the MGD project’s support and it is unclear when they will be 
able to.22 Considering the critical importance of WPGs in the project’s sustainability 
plan, WFP should immediately perform an assessment, informed by other projects 
supporting WPGs in similar contexts, to estimate the time needed for WFP-supported 
 
22 Through WFP technical assistance, WPGs are expected to incrementally increase their production while continuing 
to contribute 1/3 to their associated school canteen, with the end goal of fully supporting the canteen in conjunction 
with the Government’s national school canteens program.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 39 | Page 
WPGs to fully support a school canteen and determine when WPGs could realistically be 
expected to graduate from the project and under what conditions. After evaluating what 
local communities can support, WFP should integrate these findings into the transition 
plan it is developing with the GoCI and indicate how, when, and in which schools and 
regions school feeding will be transitioned to the GoCI’s national program. WFP should 
start reducing the dependence of canteens on MGD commodities by the start of the 
2019-2020 school year, replacing these commodities with GoCI and WPG contributions, 
and find other beneficial uses for the surplus commodities, such as increasing the 
number of girls receiving THRs by expanding this offer to all 7 project regions. 
209. Recommendation 2: Reinforce DCS and DPFC M&E Capacity at the National and 
Local Level. The MGD project’s monitoring network relies heavily on the GoCI’s own 
network of local and regional government school feeding and teaching advisors to 
collect indicator data. However, the data collected through this system does not include 
key MGD indicators, particularly on student attendance and their absences due to 
sickness, teacher absences, and food diversity and security, which cannot be effectively 
calculated through evaluation. In addition, while the MGD project delivers annual M&E 
training to DCS and DPFC, several national-level DCS personnel requested that WFP 
provide them with additional M&E training and capacity building to help them better 
support the MGD project and other school feeding related projects. By the end of the 
2018-2019 school year, WFP personnel should host a workshop with M&E personnel 
from DPFC and DCS to jointly reassess their training needs, develop a training plan that 
will respond to them, and consider strategies to improve MGD monitoring efficiency and 
effectiveness. Through this workshop, the MGD implementing partners should address 
the issue of accurately monitoring teacher and student attendance data, which remains 
a challenge, and the possibility of consolidating government monitoring activities where 
relevant. By the start of the 2019-2020 school year, WFP should support DCS and DPFC 
in modifying existing data collection tools and methods for indicators relevant to project 
implementation and reporting, while also consolidating existing school-level data 
collection visits across the school feeding and literacy components to more efficiently 
and frequently visit schools.
210. Recommendation 3: Improve Field Presence and Beneficiary Interaction. There 
are currently only eight WFP field monitors who are responsible for over 600 schools. 
They are working with government school feeding advisors who, as best they can, 
mobilize communities, supervise day-to-day canteen management, and collect indicator 
data, but have limited capacity in actuality. As WFP is trying to build both local capacity 
and long-term sustainability, they should consider working with the appropriate 
government stakeholders and local partners to increase the capacity of the local and 
regional school feeding advisors, as described in Recommendation 2. However, this 
does not replace regular, direct interaction between WFP and its project beneficiaries,
which is essential for responding to local needs in a timely manner, more accurately 
capturing M&E data, and effectively disseminating innovations in the field. To efficiently 
improve its field presence and increase its touch-points with beneficiaries, the MGD 
project should consider using the data collection networks, agents, and activities 
already being used for the literacy component, such as community focal points and 
school monitoring visits, to more frequently visit schools and communities for other 
components of the project. 
211. Recommendation 4: Increase Coordination, Communication, and Planning 
with the DCS. While the MGD project is deeply integrated with the GoCI, WFP could 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 40 | Page 
improve its advance planning of activities involving the DCS, with a clear and 
transparent view of the available funds and the project timeline. While DCS 
representatives participated in meetings to develop WFP’s proposal for the MGD project, 
multiple DCS key informants indicated that they were not informed of some late stage 
meetings, where they could have shared threats to some of the proposal’s assumptions. 
For example, these informants thought that the WPGs, a key component of this project, 
are not yet supported at sufficient number to provide project sustainability partly due 
to insufficient funds stemming from underestimating the cost of supporting WPGs. 
Having supported WPGs for nearly two decades, the DCS and ANADER has an intimate 
knowledge of these groups. The DCS informants thought their feasibility, given the 
resources and timeline, could have been more accurately described if they were fully 
included in discussions during the entire planning phase. Several DCS representatives 
also said that they were forced to start activities later than anticipated due to delays or 
uncertainty in receiving funds and were sometimes unaware of what MGD funding is 
allotted for their work. WFP should more frequently include DCS, DPFC, and other GoCI 
partners in conversations on the design, budgeting, and implementation of MGD project 
activities. In turn, the GoCI should ensure representatives from all relevant 
departments within it participate in the MGD project meetings organized by WFP to 
review and approve the tasks and budgets pertaining to their work.
212. Recommendation 5: Reinforce Girls’ Education. The results of the reading 
assessment showed that male student improved twice as much as female students. 
AVSI should work with DPFC and its field staff to adjust its programming to ensure girls 
benefit from the project as much as boys. While no beneficiaries in the FGDs said that 
girls were discriminated against at school, the evaluation team recommends that, by 
the start of the 2019-2020 school year, AVSI should conduct a study to explain these 
results and subsequently adapt its approach to ensure girls participate, read, and learn 
at the same rate as boys. Considering that teachers reported that girls in their class 
were roughly as attentive as boys, the evaluation team suggests that AVSI and DPFC 
consider integrating or reinforcing a gender focus in their teacher training curriculum. 
In addition, considering that THRs have shown to be effective in the three regions where 
it was offered in increasing girls’ attendance and keeping them in school, WFP should 
consider using a surplus MGD commodities to expand the THR offering to all seven 
regions by the start of the 2019-2020 school year.
213. Recommendation 6: Increase Female Representation in COGES. WFP targeted 
COGES to include 50 percent women as members, however, most in the evaluation 
team’s sample did not include this many women, and in fact, some COGES were 
exclusively male. By the start of the 2019-2020 school year, WFP should encourage 
women to participate in its anticipated basic literacy training for WPGs and canteen 
management committees to also join their school’s COGES, as well as make additional 
efforts to sensitize COGES of the need to integrate women in their leadership.
214. Recommendation 7: Strengthen COGES Capacity to Support MGD Activities. The 
COGES help their school and its canteen in many ways which could be enhanced by 
expanded MGD training and support. All COGES agreed they could use additional 
training in managing COGES activities, including accounting, commercial farming, and 
fundraising. While the current MGD project is designed to offer only limited support to 
COGES, by the end of the project, to the extent feasible WFP should explore offering 
training and financial support to COGES in developing commercial activities or building 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 41 | Page 
infrastructure for the school, either directly or through partnerships with relevant 
Government structures or UN agencies and other organizations. 
215. In addition to the above recommendations formed from the evaluation findings, below 
the evaluation team present suggestions provided directly from beneficiaries in the field. 
They represent the unprompted suggestions of project beneficiaries and should be 
addressed as and where possible. Nonetheless, these suggestions may not be feasible 
for WFP or its partners to implement considering the scope of the current MGD project 
and the resources available, but should be considered for future projects.
216. Suggestion 1: Fund and train COGES to be able to implement school 
infrastructure improvement and financing activities. Multiple COGES requested 
funding to help them manage schools. Many COGES members said that subsidies they 
receive from the government or funds they collect from parents are insufficient, 
preventing them from purchasing teaching materials and textbooks or making repairs 
and improvements to school latrines, benches, classrooms, water pumps, and teacher 
housing. Some also requested materials and training to be able to launch their own 
agricultural activities and sell their crops to fund the school. Others requested 
electrification for school buildings so that children can study at night, with the help of 
the study groups the COGES or parents organize after school. Finally, COGES members 
suggested that activity and record keeping training multiple times a year would be 
helpful, especially considering that many reported not having received any training.
217. Suggestion 2: Offer incentives for teachers to stay in the community. Several 
COGES addressed the issue of teacher turnover as the main barrier for the benefits of 
pedagogical training staying in the targeted community. Some mentioned plans to build 
better housing for teachers to encourage them to stay in the communities. Some 
parents suggested that the MGD project help build lodging rooms for teachers to 
encourage them to stay.
218. Suggestion 3: Provide additional materials. Parent groups mentioned that they 
need additional materials and tools to be able to adequately produce food for the school 
canteen. One parent group mentioned that their WPG received only 30 work boots for 
135 members. Another parent group mentioned that insufficient amounts of plates 
made it impossible to serve all students at once and so the canteen served lunch in 
waves. Finally, both COGES and parents would like the MGD project to provide more 
teaching and learning materials for the schools.
219. Suggestion 4: Provide additional capacity building for WPGs. All WPGs, both 
established and newer, requested literacy training, indicating that it would help them 
with business interactions and reduce their reliance on non-members. Some of the 
newer groups mentioned wanting training on animal husbandry, so they could provide 
meat to their canteens and sell the surplus for additional income. 
220. Suggestion 5: Support community study groups. Many parents and COGES in 
FGDs brought up the after-school study groups formed and run by community 
volunteers to help their children memorize the lessons of the day. In some villages, 
everyone participated in these study groups, but in others not all parents made their 
children join them (although the COGES were trying to sensitize them). In other 
communities, families sought out tutors to make their children repeat their lessons. 
These groups and tutors are important since many parents are illiterate and can’t 
supervise or check their child’s work. Some parents and COGES suggested that 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 42 | Page 
electrification of their schools would help children in their community participate in these 
after-school study groups.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 43 | Page 
Annex 1: Evaluation Matrix
Evaluation Questions Illustrative Indicators or Assessment Criteria
Data Source/ 
Collection 
Methods
Data Analysis 
Method
I. Relevance
 To what extent was the 
design of the interventions in 
line with the needs of the 
target population—women, 
girls, boys, and men?
 To what extent is the design 
of the intervention aligned 
with national policies, 
strategies, and programs?
 Do the project design and 
implementation 
arrangements complement 
other donor-funded and 
government initiatives?
 Is the project designed to 
reach the right people with 
the right type of assistance?
 Beneficiaries’ perception of usefulness (by gender)
 Perception of the project activities in alignment with other 
national activities (e.g., ministry policies, other WFP initiatives)
 Beneficiaries’ perception in applicability (by gender)
 FGD (Parents, 
SMC 
members, 
women’s 
agricultural 
groups)
 KII (USDA 
staff, 
government 
officials, 
project staff, 
partners)
 Qualitative
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 44 | Page 
Evaluation Questions Illustrative Indicators or Assessment Criteria
Data Source/ 
Collection 
Methods
Data Analysis 
Method
II. Effectiveness
 What is the progress of 
project implementation—is 
the project on track to carry 
out all activities as planned?
 To what extent have the 
interventions so far 
responded to the needs of 
the beneficiaries—women, 
girls, boys, and men?
 What are the main factors 
influencing the achievement, 
whether or not the results/ 
objectives, of the 
intervention?
 Are any changes required to 
increase the project 
effectiveness?
 Number of textbooks and other teaching and learning materials 
provided with the help of USDA
 Number of target schools where students have additional 
reading material with the help of USDA
 Proportion of students in targeted schools who regularly 
consumed a meal before or during the school day (by gender)
 Number of girls receiving take-home rations as a result of 
USDA assistance (by new/continuing)
 Number of teachers / educators / teaching assistants trained or 
certified as a result of using the USDA assistance (by type, by 
sex)
 Number of targeted schools with access to better food 
preparation and storage equipment.
 Beneficiaries’ perception on effectiveness of the program
 Teacher 
survey 
 School survey
 KII (project 
staff, 
partners)
 FGD (Parents, 
SMC 
members, 
women’s 
agricultural 
groups)
 Qualitative
 Quantitative
 Number of textbooks and other teaching and learning materials 
provided with the help of USDA
 Number of students enrolled in schools receiving assistance 
from the USDA (by gender)
 Gender ratio, primary
 Number of people who received health education and child 
nutrition as a result of using the USDA (by gender) 
 Number of students who received deworming (by gender)
 Number of school-age children receiving school meals daily 
(lunch) as a result of assistance from the USDA (by gender; 
new and continuing)
 Number of daily school meals (lunch) provided to school 
children as a result of assistance from USDA
 Number of rations provided following USDA assistance
 Number of daily school meals (lunch) offered to school children 
as a result of using the USDA
 WFP and AVSI 
monitoring 
records
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 45 | Page 
Evaluation Questions Illustrative Indicators or Assessment Criteria
Data Source/ 
Collection 
Methods
Data Analysis 
Method
III. Efficiency
 How efficient is the 
targeting?
 Does the assistance reach 
the right beneficiaries (girls, 
boys, men, and women) in 
the right quantity and 
quality and at the right 
time?
 Is the project efficient in 
terms of costs and 
costs/beneficiary?
 What are the external and
internal factors influencing 
efficiency of the project?
 Stakeholders’ perceptions of cost-effectiveness
 Percentage of stakeholders reporting that beneficiary targets 
are met (by gender)
 Type of other beneficiaries that can be reached; type of 
assistance that can be provided to them (be gender)
 List of factors influencing efficiency of the project
 FGD (parents, 
SMC 
members, 
women’s 
agricultural 
groups)
 KII 
(government 
officials, 
project staff, 
partners)
 Qualitative
IV. Impact
 What are the midterm 
effects of the operation on 
the beneficiaries in terms of: 
(a) improvement of school 
indicators; (b) improvement 
of pupil reading skills; (c) 
capacity building of groups?
 What are the reasons for the 
observed effects?
 Are there unintended effects 
on the beneficiaries? What 
have been the gender￾specific impacts, particularly 
with regards to girls’ 
schooling?
 Proportion of students at the end of two years of elementary 
school who demonstrate that they can read and understand the 
meaning of grade level text (by gender)
 Number of teachers in the target schools who demonstrate the 
use of new techniques and educational tools for quality after 
using the USDA
 Proportion of teachers in the target schools and teach attending 
school regularly (at least 90% of school days) per school year
 Number of teachers/educators/teaching assistants in targeted 
schools who demonstrate the use of new and good technical 
and pedagogical tools (by type, by sex)
 Proportion of pupils identified as attentive in classrooms by 
their teachers (by gender, class)
 Proportion of households with acceptable food consumption (by 
gender of household head)
 Student 
survey and 
ASER 
 Teacher 
survey
 School survey
 Household 
survey
 School 
management 
committee 
survey
 Women’s 
agricultural 
group survey
 Quantitative
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 46 | Page 
Evaluation Questions Illustrative Indicators or Assessment Criteria
Data Source/ 
Collection 
Methods
Data Analysis 
Method
 To what degree have the 
project outcomes made 
progress toward positive 
long-term effects on 
targeted beneficiaries (girls, 
boys, men, and women), 
households, communities, 
and institutions?
 Index of survival strategies (average) (by gender of household 
head)
 Dietary Diversity Score (by gender of household head)
 Proportion of pupils regularly (80%) frequent in classes/schools 
supported by USDA (by gender)
 Proportion of students who miss more than 10 school days a 
year due to illness (by gender)
 Number of members of the management committees and 
members of women’s production groups who are aware of the 
importance of education
 Proportion of school-age children receiving a minimum 
acceptable diet (by gender)
 Proportion of members of the school management committee 
and canteen management staff who can identify at least three 
health and hygiene practices (by gender)
 Proportion of members of the school management committee 
and the canteen management staff who can identify at least 
three practices for safe preparation and storage of food
 Proportion of schools using an improved water source
 Proportion of schools with improved sanitation facilities
 Number of people benefiting directly from interventions 
financed by USDA (by gender and date)
 Number of people indirectly benefiting from interventions 
financed by USDA
 Number of schools targeted leaders that demonstrate new and 
good technical and pedagogical tools (by type, by sex)
 Number of school officials trained or certified as a result of 
using the USDA (by sex)
 Number of beneficiaries of social safety nets who participate in 
productive safety nets following assistance from the USDA (by 
gender; new and continuing)
 WFP and AVSI 
monitoring 
records
V. Sustainability
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 47 | Page 
Evaluation Questions Illustrative Indicators or Assessment Criteria
Data Source/ 
Collection 
Methods
Data Analysis 
Method
 To what extent are the 
results of the activities likely 
to be sustainable?
 What are the key factors 
likely to affect sustainability 
of the program?
 To what extent are women’s 
agricultural groups 
contributing to school 
canteen supply, and is there 
evidence that their 
contributions will continue or 
scale up after the project 
ends?
 Value of public and private investment leveraged as a result of 
using the USDA (host government)
 Annual growth rate of the budget allocated by the government 
to the National Directorate of School Canteens
 Number of policies in the areas of health and nutrition of 
children, regulations and administrative procedures, according 
to their level of development, due to the help of the USDA (for 
stage)
 Number of policies in the education sector, regulations, and 
administrative procedures, according to their level of 
development, due to the help of the USDA (for stage)
 Number of parent–teacher associations or similar governance 
structures “school” sustained as a result of using the USDA
 Number of public–private partnerships set up as a result of 
using the USDA (nutrition, education, health, multi-sectoral, 
and other)
 Number of government staff who are trained in management of 
food and monitoring and evaluation (by gender)
 Number of policies in the areas of health and nutrition of 
children, regulations, and administrative procedures, according 
to their level of development, due to the help of the USDA 
(from steps 1, 2, and 5)
 Number of public–private partnerships set up as a result of 
using the USDA (female producing groupings) 
 WFP 
monitoring 
data
 KII (USDA 
staff, 
government 
officials, 
project staff, 
partners)
 FGD (parents, 
SMC 
members, 
women’s 
agricultural 
groups)
 Qualitative 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 48 | Page 
Annex 2: Evaluation Methodology
Evaluation Design
221. To address the research questions at midline, the evaluation team used a mixed￾methods evaluation design for collecting, analysing, and reporting both quantitative and 
qualitative data integrated into a coherent midline evaluation of the MGD project. The 
evaluation team built on the quantitative methodology used at baseline to ensure 
consistency in the approach and to minimize the introduction of new biases. In addition, 
the evaluation team added a qualitative component to complement the quantitative 
analysis and shed light into the processes and mechanisms of change and the factors 
that may have affected implementation. 
222. To assess the causal effect of the MGD teacher training project after two years of 
implementation on student’s reading proficiency23 that is the confirmatory outcome 
(i.e., primary outcome of interest) for the impact evaluation, the evaluation team 
designed a difference-in-differences (DID) method at baseline. In this method, the 
evaluation team intended to compare the changes in outcomes between the population 
of beneficiaries (the treatment group) and the population that is not benefiting from the 
project (the comparison group) following the same students over time. This method 
enabled the evaluation team to take into account any differences between the treatment 
and comparison groups that are constant over time. The evaluation team designed the 
DID method to collect the same outcomes from both the treatment and comparison 
groups at midline in a similar fashion as was done at baseline. 
Evaluation Questions
223. The evaluation team organized the evaluation questions under the following criteria: 
relevance, effectiveness, efficiency, impact and sustainability. Given the breadth of the 
questions for this evaluation, the evaluation team has compiled them into a 
comprehensive conceptual framework, in Annex 1. For each evaluation question, the 
evaluation team lists data collection strategy, data analysis method, and the responsible 
party for collecting and measuring them. In Annex 1, the evaluation team also maps 
each evaluation question to their corresponding indicators or assessment criteria, 
required by USDA, based on the MGD results framework. For example, the evaluation 
team listed the output indicators under efficacy (e.g., number of girls receiving take￾home rations as a result of USDA), and outcomes indicators under impact (e.g., 
proportion of students at the end of grade 2 who demonstrate reading proficiency at 
their grade level). 
Sampling Frame
224. At baseline for the overall project evaluation, the National Statistics Institute (INS) 
applied a proportional sampling to select 169 schools out of the 613 beneficiary schools. 
From a pre-established group of 200 comparison primary schools with similar socio￾economic and geographic characteristics, INS drew a sample of 56 comparison schools 
to be included in the evaluation. For the impact evaluation of reading outcomes at 
baseline, after adjustments to the selection of sites to accommodate for the field work, 
the final sample consisted of 94 primary schools out 225 (58 treatment and 36
 
23 Reading proficiency is defined as ability to achieve the grade-appropriate reading level by USDA. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 49 | Page 
comparison schools) across the 7 regions: Bafing, Bagoué, Boukani, Cavally, Gontougo, 
Poro, and Tchologo.
225. As requested in the revised terms of reference (TOR) and in compliance with the 
requirement for consistent DID methodology, at midline, the evaluation team followed 
the same 99 impact evaluation sample in baseline and added 9 extra treatment schools 
which it randomly selected them, maintaining the regional distribution, from 169 visited 
MGD schools at baseline. For the quantitative component, the sample size of students 
was determined using the following formula
𝑛𝑛 = 𝑍𝑍2 𝑝𝑝(1 − 𝑝𝑝)
𝑑𝑑2 𝑘𝑘
 Where: 
• 𝑛𝑛 is the sample size
• 𝑍𝑍 = 1.96 is the probability statistic associated with a confidence level of 95 percent
• 𝑝𝑝 = 0.13 is the baseline estimator of prevalence of food security provided in the TOR
• 𝑘𝑘 = 1.5 is the cluster parameter allowing adjustment based on the fact that 
households of students from the same schools share similar characteristics 
• 𝑑𝑑 = 0.05 is the tolerated margin of error. 
226. The evaluation team multiplied 𝑛𝑛, the sample size obtained by the above formula, by 6 
to ensure sufficient precision by gender (boys and girls) and school grade (CP, CE, and 
CM). Additionally, the evaluation team adjusted the sample size to allow for 95 percent
response rate. The evaluation team further increased the final sample size to 1,753
students and their corresponding households to ensure proportional representation of 
respondents within 104 schools24 across the 7 regions, including 68 treatment (i.e. 
MGD) schools and 36 control (i.e. non-MGD) schools. 
227. AVSI provided the evaluation team with the list of the enrolled students for the 
academic year of 2018-2019. In each primary school, the team surveyed 12 students 
(2 from each grade CP2, CE1, CE2, and CM1 and 4 from CM2)25 and their corresponding 
households from each school, and an additional 12 students from CM2 from the selected 
schools in Poro, Bagoué, and Tchologo regions26. The evaluation team randomly 
selected new students in Grades CP2 and CE1 and, and to the extent possible, tracked 
students in Grades CE2, CM1, and CM2 from baseline to survey. For any students in 
Grades CE2, CM1, and CM2 who could not be found at midline, the evaluation team 
mitigated this attrition by randomly replacing them with a new sample of students to 
meet the required sample size from the same grade/school (i.e., referred to throughout 
as replacements). In the remaining 9 schools, where reading assessment was not 
administered at baseline, the evaluation team also randomly sampled new students—1 
boy and 1 girl from CP2, CE1, CE2, CM1, and 2 boys and 2 girls from CM2. In each 
school, the evaluation team also surveyed teachers in Grades CP1-CM2, the principal, 
and the canteen manager. Exhibit A2.1 shows the distribution of sample respondents 
by region.
 
24Among the 108 selected schools, the evaluation team learned that 4 control schools got closed since baseline which 
the evaluation team excluded them from the evaluation.
25 The evaluation team oversampled CM2 students from 2018-2019 school year because the students who were CM2 
in the previous 2017-2018 school year are no longer in school and cannot be surveyed. 26 The additional 12 students in CM2 were all female and were added to account for the take-home ration indicator.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 50 | Page 
Exhibit A2.1: Student Sample Distribution by Region and Type of Respondent* MGD
Region CP1 CP2 CE1 CE2 CM1 Total
Bafing 12 11 13 12 24 72
Bagoué 16 17 18 18 127 196
Bounkani 14 14 16 16 26 86
Cavally 12 14 16 8 10 60
Gontougo 26 51 39 41 55 212
Poro 33 40 43 58 186 360
Tchologo 13 12 24 25 89 163
Total 126 159 169 178 517 1,149
Non-MGD
Region CP1 CP2 CE1 CE2 CM1 Total
Bafing 13 12 14 10 16 65
Bagoué 10 5 5 8 18 46
Bounkani 14 4 2 4 24
Cavally 14 17 10 12 19 72
Gontougo 12 12 25 8 20 77
Poro 25 18 21 31 123 218
Tchologo 26 45 8 12 11 102
Total 114 113 83 83 211 604
Source: Student Survey. The sample size was 1753. *Gender disaggregated frequencies are included in 
Figure 27. 
228. In collaboration with WFP, the evaluation team selected 20 project sites to visit. At each 
site selected, it conducted FGDs with CM students, parents, women’s production group 
(WPG) members, and school management committee (COGES) members. Where 
possible, the evaluation team separated the parent focus groups by gender to allow 
respondents to speak more freely and to compare mothers’ and fathers’ perceptions. 
Finally, the evaluation team conducted interviews with internal and external 
stakeholders, as described in Annex 4. In all mixed-gender FGDs and KIIs, the 
evaluation team actively encouraged women’s participation to ensure its evaluation 
incorporated their experiences and recommendations.
Data Collection Methods
229. Surveys. To answer evaluation questions, the evaluation team reviewed and closely 
examined instruments that were used during baseline and revised the tools to more 
accurately capture the MGD indicators and fill in the gaps from baseline (e.g., add or 
revise questions related to student attendance, teacher attendance, minimum 
acceptable diet, etc., particularly to distinguish between girls and boys). The evaluation 
team built on and improved baseline tools, including school, household, student, and 
school canteen management surveys. The evaluation team also developed a teacher 
survey and calibrated it into the Ivorian context to capture their training needs and their 
use of the new gained techniques. In addition, the evaluation team administered KIIs 
and FGDs protocols to interpret the quantitative findings by contextualizing and filling 
in gaps from the baseline, as several of the MGD indicators are not directly comparable.
230. Reading Assessment. Importantly, to measure the impact on reading skills of school 
children and analyse the change in skills over time, the evaluation team administered 
the same reading assessment tool, ASER. Because of the possibility that either students 
have access to the test from their older cohorts or teachers have become aware of the 
assessment and started preparing students for the test, the evaluation team revised 
the version of the test. To avoid any possible bias in reading outcomes, the evaluation 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 51 | Page 
team updated the test content of the existing versions. The evaluation team ensured 
that the updated test has the same level of complexity as the one used at baseline to 
be able to compare students’ reading skills between baseline and midline. Before 
midline data collection, the evaluation team conducted an adaptation workshop with a 
group of local reading, curriculum, and assessment experts from MENETFP with support 
from AVSI, as the evaluation team successfully did at baseline, to ensure that the 
updated ASER test is still culturally appropriate and consistent with Côte D’Ivoire’s 
learning standards for Grades 1–6. The final version of the test included 11 levels (A – 
K), which roughly correspond to the reading standards for each grade level (see Annex 
5 for the ASER tool). Exhibit A2.2 presents the structure of the ASER reading test, 
including the test’s levels and corresponding grades and reading skills.
Exhibit A2.2. ASER-Reading Test Structure
Level Corresponding Grade Reading Skills
Level 0 None None
Level A Grade 1 (CP1) – Lower level Identify letters
Level B Grade 1 (CP1) – Upper level Read simple sounds
Level C Grade 2 (CP2) – Lower level Read complex sounds
Level D Grade 2 (CP2) – Upper level Decode simple words (1-2 syllables)
Level E Grade 3 (CE1) – Lower level Decode complex words (2-3 syllables)
Level F Grade 3 (CE1) – Upper level Read simple sentences
Level G Grade 4 (CE2) – Lower level Read complex sentences
Level H Grade 4 (CE2) – Upper level Read simple stories
Level I Grade 5 (CM1) – Lower level Answer reading comprehension questions on simple stories
Level J Grade 5 (CM1) – Upper level Read complex stories
Level K Grade 6 (CM2) Answer reading comprehension questions on complex stories
Source: IMPAQ.
231. Qualitative Tools. In addition, the evaluation team administered KIIs and FGDs 
protocols to interpret the quantitative findings by contextualizing and filling in gaps from 
the baseline, as several of the MGD indicators are not directly comparable.
Evaluation Analysis
232. The evaluation team compiled the survey responses into a master file for the 
performance-evaluation analysis. Where the accuracy/completeness of the baseline 
quantitative data allow, the evaluation team conducted a cohort comparison to measure 
the project’s progress in reaching its target goals and provided descriptive analysis 
(percentages and averages) where the comparison was not feasible in order to establish 
base values for comparison later at endline. When possible, the evaluation team also 
conducted subgroup analyses by grade, student gender, highlighting emerging 
patterns. 
233. For the impact evaluation, the evaluation team compared the changes in reading 
outcomes between students who attended MGD schools (treatment group) with 
students who were enrolled in non-MGD project (comparison group) to estimate the 
two-year project’s effect on literacy growth. The evaluation team surveyed students 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 52 | Page 
from the same schools in baseline and midline so that it could construct the change in 
reading outcomes for each cohort of student. Since the evaluation team collected the 
midline data at the beginning of the 2018-2019 school year, to be able assess their 
reading outcomes at the end of their grade, it had to measure their reading proficiency 
retrospectively which excluded CM2 cohort from the analysis.
Validity and Reliability
234. To measure project impact between baseline and midline, the evaluation team surveyed 
the same sample selected by the baseline evaluator (INS). The evaluation team and 
WFP were unable to meet with INS to assess the validity of the baseline sampling and 
data collection. However, having designed and validated the reading assessment for the 
baseline evaluation, the evaluation team is confident in the validity of the related 
indicators at baseline and its ability to infer causal relationships between the project 
and progress in literacy. 
235. In addition to validating the reading assessment tool at baseline and midline, the 
evaluation team validated its other survey instruments at midline in a validation 
workshop with DCS and WFP, adjusting its instruments as necessary prior to data 
collection. The evaluation team also shared its inception report, including its surveys 
and qualitative protocols, with WFP, AVSI, and Government stakeholders to address 
their comments prior to field work.
236. Quantitative and qualitative data were compared to cross-validate findings across data 
collection methods and verify the reliability of beneficiary responses. To ensure inter￾rater reliability, an IMPAQ research analyst performed daily quality checks on key 
variables, including reading assessment and food consumption scores, and followed-up 
quickly as necessary to address any significant differences in survey responses by 
enumerator. In addition, qualitative data was independently collected by IMPAQ’s two 
fieldwork managers, with findings and transcripts reviewed by IMPAQ home office staff.
Modifications and Limitations
237. There are some limitations in the MGD midline evaluation. The evaluation team 
discussed some in the inception report, but it learned about the others during the 
fieldwork. Exhibit A2.3 describes the limitation of the evaluation team’s midline 
evaluation, the mitigation strategies it implemented, and their possible impact on the 
findings.
Exhibit A2.3: Challenges and Solutions
Challenges Solutions Possible Impact on the Findings
Inaccurate or missing 
data from baseline
Revised the quantitative 
tools developed by INS to 
more accurately capture 
the required MGD 
indicators
The changes helped fix the caveats of the baseline 
surveys, but at the same time made the baseline and 
midline outcomes incomparable for some of the 
indicators. However, the evaluation team does not 
consider this as an issue as this improved the tools to 
capture the indicators more rigorously at endline with 
establishing base values at midline.
Lack of teacher survey at 
baseline
Developed a teacher 
survey and collected data 
from teachers in all the 
treatment schools 
instead of asking the 
school principal to 
improve the key 
performance indicator 
measurements. 
Collecting data directly from teachers was a different 
methodology that was not taken at baseline. This change
improved the precision of the measurement, but made 
the key indicators related to teachers incomparable to 
baseline. The evaluation team reported findings 
descriptively at midline without comparing changes over 
time and will re-administer survey at endline to see 
differences from midline.
Delay in data collection 
until the beginning of the 
school year 2017-2018
The evaluation team used 
the reading skill 
information in the 
For the impact evaluation, the evaluation team intended 
to follow the same students in Grades CP1, CP2, CE1 and 
CE2 at baseline and survey them again at midline when 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 53 | Page 
Challenges Solutions Possible Impact on the Findings
beginning of the school 
year as a proxy for what 
the reading skills were 
during the end of the 
previous school year. The 
evaluation team also 
oversampled CM2 
students at midline to 
meet the target. 
they were at CE1, CE2, CM1 and CM2, respectively, 
excluding CM1 and CM2 student cohorts at baseline who 
were gone from the school at midline. The delay in 
midline data collection left the evaluation team with one 
less cohort as CE2 students at baseline graduated in 
2018-2019 from school. However, oversampling CM2 
students at midline helped the evaluation team reach its
target without losing any power even with one less 
cohort. 
Change in the 
implementation plan for 
training teachers in all 
grade 
Measured the project 
effect on literacy 
outcomes for at least one 
year, instead of two, at 
midline to maintain the 
power. 
Based on the original implementation plan at baseline, 
only teachers in CP1 and CP2 were supposed to receive 
the training in the project at baseline. In the 2016-2017 
school year, teachers in CP1 (a new cohort) and CP2 
received the training. However, a year after that, 
teachers in all Grades (CP1-CM2) received the MGD 
training. Therefore, across different grades at midline, 
students have not been exposed to the teacher training 
consistently. Therefore, the evaluation team was not 
able to show the impact results exclusively for the two￾year project effects (i.e., the effect of canteen and 
literacy activities both together for two years) given their 
small sample. The evaluation team assessed the effect of 
the project on students’ reading scores with at least one 
year of exposure to the project to maintain the power.
However, the evaluation team estimated its regression 
model also with restricting the sample to include only 
CE2 students at midline (who were in CP1 at baseline) 
and the findings stayed statistically significant, 
regardless. 
Follow the same 
subsample over time and 
include additional schools 
for the reading 
assessment
The evaluation team used 
a DID and followed the 
same cohort of students 
rather same individuals in 
the same schools. 
At midline, the evaluation team was able to track 244 out 
of the 1106 students from the baseline sample, i.e., 22 
percent. Therefore, the results will not be coming from 
the same students over time but same cohorts. The 
evaluation team estimated its model by excluding 
students who joined the school in academic year of 
2018-2019 and have not been exposed to the project. 
The regression robustness checks show no difference in 
the project effects. 
Ethical Considerations
238. WFP's decentralised evaluations must conform to WFP and UNEG ethical standards and 
norms. The contractors undertaking the evaluations are responsible for safeguarding 
and ensuring ethics at all stages of the evaluation cycle. This includes, but is not limited 
to, ensuring informed consent, protecting privacy, confidentiality and anonymity of 
participants, ensuring cultural sensitivity, respecting the autonomy of participants, 
ensuring fair recruitment of participants (including women and socially excluded 
groups) and ensuring that the evaluation results in no harm to participants or their 
communities.
239. During the evaluation the following ethical issues were considered for the design, data 
collection, data analysis, reporting and dissemination: providing sufficient information 
about the evaluation to participants, obtaining informed consent from participants, 
collecting confidential data from participants, assuring them that their identity will not 
be revealed, analysing and presenting de-identified and aggregate findings. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 54 | Page 
240. The following safeguards and measures to manage these issues were in place: 
submitting the research protocol and data collection instruments to Advarra 
Institutional Review Board (IRB); training enumerators on respondents safely and 
confidentiality, with particular consideration given to safeguarding child respondents; 
obtaining teachers’ and principals’ consent to survey their students along with each 
child’s individual assent; keeping the data in a secured place to protect respondents’ 
personal information; and active monitoring of data collection for compliance with 
IMPAQ quality and ethical guidelines.
241. Prior to collecting data, the evaluation team submitted protocol documents to Advarra
IRB on September 11, 2018 (Pro00029763). IRB approval is necessary to ensure that 
proposed evaluations comply with local and international rules and procedures. The 
evaluation team received approval from Advarra IRB on September 18, 2018. The 
approved documentation for this evaluation includes: 
• School, canteen manager, teacher, household, and student, including ASER, surveys 
• Research protocol 
• Informed consent and assent forms 
• Qualitative protocols for key informant interviews and focus group discussions 
242. The evaluation team used these IRB-approved instruments to collect data. Prior to 
administering the survey, enumerators were trained on procedures for interviewing 
respondents, protecting respondent privacy and confidentiality, and securing the data, 
with a particular focus on safeguarding children at school. During the midline data 
collection, the survey team first obtained teachers’ and principals’ consent to survey 
their students. Then they asked for students’ assent, assuring children that their 
participation was voluntary and that they could terminate the survey at any point. After 
data collection, the evaluation team protected the privacy and confidentiality of 
respondents by storing the data on secure servers and separating personally identifiable 
information from the survey data.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 55 | Page 
Annex 3: Documents Reviewed / Bibliography
Document Type Titles of documents received
Project related documents
Appraisal mission report SNAS 2018 – 2022
Project document (including Logical Framework in Annex) TOR
Project proposal FY2015 McGovern-Dole Proposal
Monitoring & Reporting
M&E Plan A3. Evaluation Plan
Food Distribution and Post-distribution Monitoring Reports Monthly hot meals reports; Monthly THR report
Donor specific reports Performance Monitoring Plan (USDA)
Partners
Annual reports from cooperating partners (i.e. AVSI)
Analysis of the Performance Evaluation of CP Students in 
Reading for the year 2016/2017; Analysis of the CP 
Teachers' Evaluation t on the Use of New Teaching-Learning 
Strategies; State of the Schools 2016-2017
List of partners (Government, NGOs, UN agencies) by location/ activity/ 
role/ tonnage handled TOR
Evaluations/ Reviews
Evaluations/ reviews of past or on-going operations/ interventions
Baseline Evaluation Report, Midline Evaluation Report of 
“Support to a Sustainable School Feeding Programme” 
(2013-2016), Preliminary Report on Impact Study of the 
WFP Food Assistance Program on Gender Roles
Maps
Map of the intervention TOR
Food/Cash/voucher Distribution Location Map TOR
Food Security Map TOR
Other documents collected by the team (including external ones)
Decentralized Evaluation Quality Assurance System (DEQAS) DEQAS Document
Country Poverty Profile Survey on Household Living Standards in Côte d’Ivoire
Baseline Survey Questionnaires Survey
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 56 | Page 
USDA Indicators 
USDA Food for Progress and McGovern-Dole Indicators and 
Definitions
Baseline Databases Baseline McGovern Dole Questionnaires
Baseline School Sample List of schools and school samples
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 57 | Page 
Annex 4: Stakeholders Interviewed
Dates / Tools Locations/sites Stakeholders
July 18
KII
USDA Headquarters
Washington, DC, United 
States
Representatives from USDA
September 20
KIIs
DCS (MENETFP) Office
Abidjan, Côte d’Ivoire
Director of DCS and 
14 Representatives from DCS, 
including those responsible for 
M&E, nutrition, stove 
amelioration, accounting, 
logistics/provisions, and 
communication 
September 21
KIIs
WFP Office
Abidjan, Côte d’Ivoire
4 Representatives from WFP, 
including project 
coordination/business 
support; nutrition, market 
access/ resilience/ 
livelihoods ; national projects
September 24
KIIs
AVSI Office
Abidjan, Côte d’Ivoire
Representatives from AVSI 
including project 
management, M&E
September 24-25
KIIs
WFP Office
Abidjan, Côte d’Ivoire
Representatives from WFP, 
including M&E; education; 
accounting; and supply chain 
October 5-24
FGDs; Questionnaires
Poro, Bagoué, Tchologo, 
Bounkani, Gontougo, Bafing, 
and Cavally, Côte d’Ivoire
School principals; Parents; 
Teachers; Students; School 
management committees; 
WPGs
October 29
KIIs
ANADER (MINADER) Office
Abidjan, Côte d’Ivoire
Representatives from ANADER
October 29
KII
DPFC (MENETFP) Office
Abidjan, Côte d’Ivoire
Representative from DPFC
January 3-7
FGDs
Korhogo, Côte d’Ivoire WPGs
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 58 | Page 
Annex 5: Data Collection Tools
Quantitative Tools
1. Student Questionnaire
2. Teacher Questionnaire
3. Canteen Manager Questionnaire
4. Household Questionnaire
5. School Questionnaire
Qualitative Tools
6. KII Protocol: USDA Staff
7. KII Protocol: Project Staff / Partners
8. KII Protocol: Government Stakeholders
9. FGD Protocol: Parents
10. FGD Protocol: School Management Committees (COGES)
11. FGD Protocol: WPGs
12. FGD Protocol: CM Students
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 59 | Page 
Annex 6: Additional Figures
Figure 27: Student Sample Distribution by Gender, Grade, and Type of 
Respondent
MGD
CP1 CP2 CE1 CE2 CM1 Total
Boys 64 78 88 90 126 446
Girls 62 81 81 88 391 703
Total 126 159 169 178 517 1,149
Non-MGD
CP1 CP2 CE1 CE2 CM1 Total
Boys 57 60 38 41 55 251
Girls 57 53 45 42 156 353
Total 114 113 83 83 211 604
Source: Student Survey. The sample size was 1753 at midline.
Figure 28: Reading Proficiency by Grade
Source: Student Survey. The sample size was 964 at baseline and 1753 at midline.
Figure 29: Teacher Attendance According to School Records and Self-Reported 
Bafing Bag. Boun. Cavally Gont. Poro Tcho. Total
Official 
Records
Average 
Attendance Rate 96.2% 94.5% NA 93.5% 96.5% 98.7% 96.7% 96.6%
Attending 90% 
or More 100% 95% NA 77.3% 93.5% 96.5% 96% 94.4%
Obs. 31 40 NA 22 46 86 25 250
Self￾reported
Average 
Attendance 94.8% 90.1% 97.8% 92% 92.8% 89.9% 87.5% 92%
Attending 90% 
or More 95.8% 65.7% 97% 61.1% 82.8% 75.6% 81.3% 79.9%
Obs. 24 35 33 18 64 78 16 268
Source: School Survey and Teacher Survey. Teacher attendance records for schools in Bounkani region are not 
available.
0%
20%
40%
60%
80%
100%
Mid MGD Base MGD Mid non-MGD Base non-MGD
CP1 CP2 CE1 CE2 CM1 Total
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 60 | Page 
Figure 30: Self-Reported Teacher Absences by Region (Midline Only)
Source: Teacher Survey. The sample size was 268. Teachers were asked how 
many days they missed as a result of each of several causes.
Figure 31: Access to Reading Materials (Mobile Library, Reading Board, etc.) by 
Region (Midline MGD Only)
Reading Material Bafing Bagoué Boun. Cavally Gontougo Poro Tchologo Total
None 11.1% 23.0% 1.2% 0% 11.8% 10.6% 25.2% 13.8%
Sculpted plastic 
letters 13.9% 13.8% 11.6% 26.7% 7.6% 20.8% 9.2% 14.7%
Junior dictionaries 18.1% 6.6% 10.5% 30.0% 11.8% 30.8% 9.8% 17.8%
Illustrated boards 29.2% 18.9% 51.2% 53.3% 55.2% 49.4% 41.7% 43.3%
Reading board 26.4% 33.7% 51.2% 63.3% 55.2% 49.2% 50.9% 47.4%
Mobile library 75.0% 56.1% 68.6% 98.3% 71.7% 71.1% 49.1% 67%
Responses 125 298 167 163 452 835 303 2343
Observations 72 196 86 60 212 360 163 1149
Source: Student Survey. Respondents could select multiple choices.
Figure 32: Usefulness of AVSI Trainings (Midline MGD Only)
Percent 
Found AVSI training useful 97.8%
Put AVSI training to work 96.0%
Using AVSI training materials 21.6%
Observations 296
Source: Teacher Survey.
0%
20%
40%
60%
80%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Average Attendance Rate Teachers Attending 90% or More
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 61 | Page 
Figure 33: Number of Students out of 10 that were Attentive, According to 
Teachers (Midline MGD Only)
Source: Teacher Survey. Sample size was 330.
Figure 34: By Region, Number of Students out of 10 that were Attentive, 
According to Teachers (Midline MGD Only)
Source: Teacher Survey.
Figure 35: By Region, Average Student Recorded Attendance (Midline Only)
Region Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Percentage 99% 99% 99% 97% 98% 98% 100% 99%
Obs. 88 166 63 78 230 149 142 916
Source: School Survey.
0
1
2
3
4
5
6
7
8
9
CP1 CP2 CE1 CE2 CM1 CM2 Total
Girls Boys All
0
1
2
3
4
5
6
7
8
9
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Girls Boys All
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 62 | Page 
Figure 36: Average Student Enrolment per Class (MGD Schools with Take Home 
Ration Schools Separated)
CP1 CP2 CE1 CE2 CM1 CM2 All
Girls
Mid THR 26.5 26.6 26.4 24.2 22.3 20.2 42.5
Base THR 32.8 28.2 24.6 22.1 19.4 15.6 33.9
Mid non-THR 19.1 17.1 16.5 15.7 12.6 11.2 22.2
Base non-THR 24.7 22.2 17.7 15.3 12.8 9.21 17.7
Boys 
Mid THR 23.5 24.5 25.3 24.1 23.2 21.2 44.3
Base THR 31.6 27.9 25.9 23 20.7 19.2 38.7
Mid non-THR 17.5 18 16 17.4 13.9 15.3 27.5
Base non-THR 26 25.9 20.1 17.7 16.6 14.8 25.1
Source: School Survey. Sample size was 133 schools.
Figure 37: Average Student Enrolment per Class (MGD Schools with Take Home 
Ration Schools Separated)
Source: School Survey. Sample size was 133 schools.
Figure 38: Ratio of Girls to Boys Enrolled by Grade (MGD Schools with Take 
Home Ration Schools Separated)
CP1 CP2 CE1 CE2 CM1 CM2
CM1 & 
CM2
Mid THR 1.13 1.09 1.04 1.00 0.96 0.96 0.96
Base THR 1.04 1.01 0.95 0.96 0.94 0.81 0.88
Mid non-THR 1.09 0.95 1.03 0.91 0.90 0.73 0.81
Base non-THR 0.95 0.85 0.88 0.87 0.77 0.62 0.70
Source: School Survey. Sample size was 133 schools.
0
5
10
15
20
25
30
35
Mid THR Base THR Mid non￾THR
Base non￾THR
Mid THR Base THR Mid non￾THR
Base non￾THR
Girls Boys
CP1 CP2 CE1 CE2 CM1 CM2
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 63 | Page 
Figure 39: Ratio of Girls Enrolled to Boys Enrolled by Grade (MGD Schools with 
Take Home Ration Schools Separated)
Source: School Survey. Sample size was 133 schools.
Figure 40: Number of Days Students Missed Due to Illness (Midline Only)
Days missed due to illness? MGD Non-MGD Total
Boys 3.3 4.5 3.8
Girls 2.8 3.5 3
Total 3 3.9 3.3
Observations 801 421 1222
Source: Household Survey. 
Figure 41: Students Missing More than 10 Days Due to Illness (Midline Only)
Missed more than 10 days due to illness MGD Non-MGD Total
Boys 4% 8% 5.4%
Girls 4.2% 5.3% 4.6%
Total 4.1% 6.4% 4.9%
Obs. 801 421 1222
Source: Household Survey. 
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Mid THR Base THR Mid non-THR Base non-THR
CP1 CP2 CE1 CE2 CM1 CM2 CM1 & CM2
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 64 | Page 
Figure 42: Foods Eaten in Household during the Previous Day
MGD Non-MGD
Foods Midline Baseline Midline Baseline
Cereals 89.9% 95.9% 93.9% 96.9%
Roots, tubers, etc. 34.7% 46% 25.9% 37.2%
Vegetables 79.7% 93% 81.3% 90.8%
Fruit 8.1% 73.7% 9% 81.9%
Meat 22.6% 48.9% 26.5% 47.8%
Eggs 3.4% 15.7% 4.2% 17.7%
Seafood 54.8% 63.6% 45.9% 56.5%
Legumes, nuts, seeds 45.1% 37.6% 41.1% 43.8%
Milk 19.8% 40.2% 19.6% 47.8%
Oils and fats 52.4% 84.3% 51.4% 83.5%
Sweets 34.8% 67.8% 35.3% 67.8%
Condiments 62.2% 85.8% 63.2% 88%
Responses 5833 5142 2998 3228
Observations 1149 681 603 425
Source: Household Survey. 
Figure 43: By Region, Foods Eaten in Household During the Previous Day (MGD 
Midline)
Food Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Cereals 98.6% 99.5% 73.3% 85.0% 65.1% 97.8% 100.0% 89.9%
Roots, 
tubers, etc. 30.6% 14.8% 79.1% 46.7% 88.7% 15.0% 6.1% 34.7%
Vegetables 90.3% 83.7% 98.8% 83.3% 90.6% 72.5% 60.7% 79.7%
Fruit 13.9% 2.0% 2.3% 15.0% 19.3% 5.3% 4.9% 8.1%
Meat 25.0% 24.5% 20.9% 18.3% 29.3% 16.1% 27.6% 22.6%
Eggs 0.0% 0.5% 0.0% 1.7% 3.3% 6.4% 4.3% 3.4%
Seafood 45.8% 66.3% 50.0% 65.0% 69.8% 43.6% 49.1% 54.8%
Legumes, 
nuts, seeds 29.2% 53.6% 11.6% 58.3% 17.9% 66.7% 42.3% 45.1%
Milk 5.6% 21.4% 5.8% 3.3% 10.9% 26.9% 33.7% 19.8%
Oils and fats 55.6% 69.9% 66.3% 33.3% 57.1% 39.7% 51.5% 52.4%
Sweets 20.8% 38.8% 24.4% 15.0% 21.7% 45.0% 43.6% 34.8%
Condiments 65.3% 61.7% 69.8% 93.3% 40.6% 73.1% 50.3% 62.2%
Responses 346 1052 432 311 1090 1829 773 5833
Observations 72 196 86 60 212 360 163 1149
Source: Household Survey. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 65 | Page 
Figure 44: By Region, Foods Eaten in Household during the Previous Day (MGD 
Midline)
Source: Household Survey. The sample size was 1149.
Figure 45: Coping Strategies Used by Household
Mid MGD Base MGD Mid non-MGD Base non-MGD
Coping Strategy Percent* Days Percent Days Percent Days Percent Days
Consuming less preferred foods for cheaper 28.5% 1.0 33.5% 0.8 31.2% 1.0 31.5% 0.7
Depend on aid from relatives or friends 11.8% 0.2 16.5% 0.3 15.3% 0.3 14.8% 0.3
Decrease the amount of food during meals 19.7% 0.6 24.7% 0.6 22.4% 0.6 19.1% 0.4
Reduced quantities consumed by adult / 
parent for the benefit of young children 22.4% 0.7 21.6% 0.5 23.2% 0.7 12.5% 0.4
Reduce the number of meals per day 
(skipping one or two meals in the day) 14.4% 0.4 13.1% 0.3 14.8% 0.3 6.4% 0.1
Buying food on credit 23.2% 0.6 18.9% 0.4 23.6% 0.6 21.7% 0.4
Eating wild foods / Culturally not allowed 2.4% 0.1 5.1% 0.1 2.5% 0.1 3.3% 0.1
Consuming seeds 12.1% 0.5 14.8% 0.4 15.6% 0.6 7.8% 0.2
Begging for food or money to buy food 1.9% 0.0 3.7% 0.1 4.8% 0.1 2.8% 0.1
Send household members to eat elsewhere 
or to live with relatives or friends 3.1% 0.1 4.9% 0.1 4.8% 0.1 2.6% 0.1
Spend 1 or more days without eating 3.9% 0.1 5.3% 0.1 4.8% 0.1 2.8% 0.0
Have the children work 10.4% 0.2 4.9% 0.1 10% 0.3 7.5% 0.1
Total 1766 1136 1042 564
Cases 1149 681 603 425
Source: Household Survey. Households could report using more than one strategy. *Percentage of households 
using the strategy at least one day per week; average number of days per week that households use the 
strategy (this includes households using the strategy 0 days per week). Bolded strategies are the strategies 
used to create the reduced index.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 66 | Page 
Figure 46: By Region, Days* Coping Strategies Used (Midline MGD)
Coping Strategy Bafing Bag. Boun. Cavally Gont. Poro Tcho. Total
Consuming less preferred foods for cheaper 1.5 0.7 1.1 1.5 0.2 1.3 1.4 1.0
Depend on aid from relatives or friends 0.7 0.1 0.3 0.4 0.1 0.3 0.1 0.2
Decrease the amount of food during meals 1.2 0.6 0.4 1.3 0.1 0.7 0.5 0.6
Reduced quantities consumed by adult / parent 
for the benefit of young children 1.4 0.3 0.5 1.2 0.1 0.9 0.9 0.7
Reduce the number of meals per day (skipping 
one or two meals in the day) 0.8 0.1 0.3 1.1 0.1 0.6 0.3 0.4
Buying food on credit 0.9 0.6 0.2 1.0 0.1 0.8 0.5 0.6
Eating wild foods / Culturally not allowed 0.2 0.0 0.1 0.1 0.1 0.1
Consuming seeds 1.6 0.3 1.0 0.0 0.3 1.3 0.5
Begging for food or money to buy food 0.2 0.0 0.1 0.0 0.0
Send household members to eat elsewhere or to 
live with relatives or friends 0.3 0.0 0.1 0.0 0.1 0.0 0.1
Spend 1 or more days without eating 0.3 0.0 0.1 0.0 0.1 0.1 0.1
Have the children work 0.9 0.0 0.3 0.3 0.2 0.2
Responses 270 153 112 102 92 791 246 1766
Observations 72 196 86 60 212 360 163 1149
Source: Household Survey. Households could report using more than one strategy. *Average number of days 
per week that households use the strategy (this includes households using the strategy 0 days per week). 
Where table shows 0.0 days, the strategy is used less than 0.05 days on average, but more than 0 days.
Bolded strategies are the strategies used to create the reduced index.
Figure 47: By Region, Percent* Using Coping Strategies (Midline MGD)
Coping Strategy Bafing Bag. Boun. Cavally Gont. Poro Tcho. Total
Consuming less preferred foods for cheaper 54.2% 15.3% 29.1% 28.3% 8.5% 40.3% 32.5% 28.5%
Depend on aid from relatives or friends 29.2% 6.6% 14% 16.7% 5.2% 17.2% 4.3% 11.8%
Decrease the amount of food during meals 45.8% 14.8% 16.3% 23.3% 3.8% 27.8% 17.2% 19.7%
Reduced quantities consumed by adult / 
parent for the benefit of young children
52.8% 6.6% 20.9% 28.3% 4.7% 33.3% 25.2% 22.4%
Reduce the number of meals per day 
(skipping one or two meals in the day)
33.3% 4.1% 14% 28.3% 3.8% 22.2% 9.8% 14.4%
Buying food on credit 40.3% 22.5% 10.5% 35% 3.8% 33.6% 20.9% 23.2%
Eating wild foods / Culturally not allowed 5.6% 0.5% 6.7% 3.9% 3.1% 2.4%
Consuming seeds 47.2% 6.6% 17.4% 0.5% 11.4% 21.5% 12.1%
Begging for food or money to buy food 8.3% 0.5% 3.3% 1.8% 1.9%
Send household members to eat elsewhere or 
to live with relatives or friends
15.3% 0.5% 3.5% 0.5% 4.2% 3.1% 3.1%
Spend 1 or more days without eating 18.1% 3.5% 3.3% 0.5% 5.6% 3.7% 3.9%
Have the children work 25% 1.2% 12.3% 16.9% 8% 10.4%
Responses 270 153 112 102 92 791 246 1766
Observations 72 196 86 60 212 360 163 1149
Source: Household Survey. Households could report using more than one strategy. *Percentage of households 
using the strategy at least one Day per week. Bolded strategies are the strategies used to create the reduced 
index.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 67 | Page 
Figure 48: By Region, Percent* Using Coping Strategies (Midline MGD)
Source: Household Survey. Households could report using more than one strategy. *Percentage of households 
using the strategy at least one day per week. The sample size was 1149. The first five strategies listed are the 
strategies used to create the reduced index.
Figure 49: Percent of School Days during which Canteen Operated by Region
Region Midline MGD Obs. Baseline MGD Obs.
Bafing 100% 6 34.5% 6
Bagoué 93.6% 7 64% 8
Bounkani 97.4% 6 55.1% 7
Cavally 99% 5 29.7% 5
Gontougo 95% 16 66.2% 16
Poro 87.9% 18 68.7% 19
Tchologo 98% 5 6% 6
Total 93.1% 63 56.9% 67
Source: Canteen Survey.
0%
10%
20%
30%
40%
50%
60%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Consuming less preferred foods for cheaper
Depend on aid from relatives or friends
Decrease the amount of food during meals
Reduced quantities consumed by adult / parent for the benefit of young children
Reduce the number of meals per day (skipping one or two meals in the day)
Buying food on credit
Eating wild foods / Culturally not allowed
Consuming seeds
Begging for food or money to buy food
Send household members to eat elsewhere or to live with relatives or friends
Spend 1 or more days without eating
Have the children work
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 68 | Page 
Figure 50: By Region, Percent of School Days during which Canteen Operated
Source: Canteen Survey.
Figure 51: Parents’ Awareness of the School Canteen Project 
Mid MGD Base MGD Mid non-MGD Base non-MGD
Parent is aware of the canteen program 94.1% 98.2% 30.7% 42.1%
Obs. 1148 681 590 425
Source: Household Survey. Sample size was 1106 at baseline and 1752 at midline.
Figure 52: Students Eating through the School Canteen Program
Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Student ate at 
the canteen 94.4% 93.3% 95.2% 98.3% 95.8% 88% 86.5% 91.6%
Total 72 195 84 60 212 357 163 1143
Source: Household Survey. 
Figure 53: Frequency that Students Eat at the School Canteen 
Frequency Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Never 5.6% 6.7% 4.8% 1.7% 4.3% 12% 13.5% 8.4%
One day a week 6.9% 2.1% 6% 8.3% 1.4% 1.1% 0% 2.3%
Two days a week 2.8% 16.4% 4.8% 3.3% 3.8% 2.5% 6.8% 6%
Three days a week 9.7% 29.7% 27.4% 8.3% 10.9% 14% 7.4% 15.6%
Four days a week 20.8% 19.5% 19.1% 11.7% 16% 10.4% 16.6% 15.2%
Every day of the 
school week 54.2% 25.6% 38.1% 66.7% 63.7% 59.9% 55.8% 52.6%
Total 72 195 84 60 212 357 163 1143
Source: Household Survey. 
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Baseline MGD Midline MGD
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 69 | Page 
Figure 54: Foods Used to Prepare Canteen Meals (Midline Only) 
Source: Canteen Survey. The sample size was 67. Figure shows the probability that a given food group would 
be included in a meal in the region. This calculation was based on the number of days the food group was used 
in preparation of meals by each canteen and the total number of days that canteen operated.
Figure 55: By Region, Canteen Managers Can Cite Three Health & Hygiene 
Practices
Region Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
Midline 83.3% 75% 100% 40% 93.8% 94.7% 100% 87.9%
Baseline 83.3% 75% 100% 100% 100% 94.7% 83.3% 92.5%
Obs. 6 8 7 5 16 19 6* 67
Source: Canteen Survey. Only five schools were available in Tchologo at midline.
Figure 56: By Region, Canteen Managers Can Cite Three Health & Hygiene 
Practices 
Source: Canteen Survey. Only five schools were available in Tchologo at midline.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Cereals Tubers/roots Pulses/nuts Dairy Meat, fish,
giblets
Eggs Vitamin A
rich fruits
and
vegetables
Other fruits
and
vegetables
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Midline Baseline
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 70 | Page 
Figure 57: Teacher Training on Health Practices
Source: School Survey. Sample size was 104 at baseline and midline.
Figure 58: Nutrition-Related Trainings at School
MGD Non-MGD
Type of Training Midline Baseline Midline Baseline
Teacher nutrition training 54.4% 48.5% 8.3% 5.6%
Student nutrition training 19.4% 27.9% 0% 19.4%
Vegetable garden training 7.4% 4.4% 0% 19.4%
Observations 68 68 36 36
Source: School Survey. Sample size was 104 at baseline and midline.
Figure 59: Type of Access to Water of the School
Mid MGD Base MGD Mid non-MGD Base non-MGD Total
No water 53% 50% 64% 69% 57%
Tap 16% 10% 17% 3% 12%
Village pump 19% 31% 17% 17% 22%
Protected well 9% 6% 0% 3% 5%
Unprotected well 3% 3% 3% 8% 4%
Observations 68 68 36 36 208
Source: School Survey.
Figure 60: Distance to School’s Water Source, Where Available (Midline MGD 
Only)
Bafing Bagoué Boun. Cavally Gont. Poro Tcho. Total
In the school grounds 17% 25% 29% 20% 35% 47% 17% 32%
Less than 15 min walk 50% 63% 57% 80% 29% 32% 33% 43%
Over 15 min walk 33% 13% 14% 0% 35% 21% 50% 25%
Total 6 8 7 5 17 19 6 68
Source: School Survey.
0%
10%
20%
30%
40%
50%
60%
70%
80%
Mid MGD Base MGD Mid non-MGD Base non-MGD
Skills Based Health Education Deworming
HIV/AIDS Prevention Malaria Prevention
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 71 | Page 
Figure 61: Most Common Problems with Water Source
Mid MGD Base MGD
Mid non￾MGD
Base non￾MGD
Broken pump 59% 59% 50% 37%
Animals occupying water source 0% 0% 0% 5%
Dried up water source 15% 26% 0% 26%
Water is used for agriculture 0% 0% 0% 5%
Other 11% 7% 40% 21%
Distance 15% 7% 10% 5%
Total 27 27 10 19
Source: School Survey.
Figure 62: Access to and Quality of Sanitary Facilities
Mid 
MGD
Base 
MGD
Mid non￾MGD
Base non￾MGD
Has sanitary facilities 59% 54% 50% 31%
Separated by gender 31% 44% 25% 19%
Separate for teachers 26% 29% 31% 14%
Observations 68 68 36 36
Source: School Survey.
Figure 63: Deworming Pills and Micronutrient Pills Distributed to Students
Mid MGD Base MGD Mid non￾MGD
Base non￾MGD
Deworming treatments 64% 72% 56% 64%
Micronutrient supplements 26% 53% 17% 14%
Observations 68 68 36 36
Source: School Survey. The sample size was 104 at baseline and midline.
Figure 64: Deworming Pills and Micronutrient Pills Distributed to Students
Source: School Survey. The sample size was 104 at baseline and midline.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Mid MGD Base MGD Mid non-MGD Base non-MGD
Deworming treatments Micronutrient supplements
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 72 | Page 
Figure 65: Access to Food Storage and Equipment
Midline Baseline
Ventilated food storage 59% 63%
Improved pallets 24% 22%
Stock cards available 100% 65%
Food prepared on-site 99% 96%
Have a kitchen 90% 82%
Source: School Survey. The sample size was 68 at baseline and midline.
Figure 66: Access to Food Storage and Equipment
Source: School Survey. The sample size was 68 at baseline and midline.
Figure 67: Sufficient Food Preparation Equipment
Source: Canteen Survey. Only 5 schools were available in Tchologo at midline.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Ventilated food
storage
Improved pallets Stock cards
available
Food prepared
on-site
Have a kitchen
Midline Baseline
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Midline Baseline
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 73 | Page 
Figure 68: Sufficient Food Storage Equipment
Source: Canteen Survey. Only 5 schools were available in Tchologo at midline.
Figure 69: Stove Type
Midline Baseline
None/Unknown 11.76 17.65
Traditional stove 39.71 73.53
Furnace without chimney 2.94 1.47
Stove with chimney 45.59 7.35
Observations 68 68
Source: Canteen Survey. The sample size was 68 at baseline and midline
Figure 70: Stove Type by Region
Bafing Bagoué Bounkani Cavally Gontougo Poro Tchologo Total
None/Unknown 0% 13% 0% 40% 6% 21% 0% 12%
Traditional 
stove 33% 25% 43% 0% 65% 47% 0% 40%
Furnace 
without 
chimney 0% 0% 29% 0% 0% 0% 0% 3%
Stove with 
chimney 67% 63% 29% 60% 29% 32% 100% 46%
Observations 6 8 7 5 17 19 6 68
Source: Canteen Survey. 
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Bafing Bagoue Bounkani Cavally Gontougo Poro Tchologo Total
Midline Baseline
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 74 | Page 
Figure 71: Reading Assessment Scores by Treatment Group, Phase, and Grade
Mid MGD Base MGD
Score CP1 CP2 CE1 CE2 CM1 CP1 CP2 CE1 CE2 CM1
0 7.9% 3.1% 1.8% 1.7% 0.4% 63.0% 35.3% 11.5% 7.3% 5.0%
A 46.8% 25.2% 16.6% 10.7% 6.6% 32.4% 32.8% 30.1% 19.1% 10.8%
B 31.8% 25.2% 13.6% 13.5% 6.8% 4.6% 13.1% 15.9% 14.6% 16.7%
C 3.2% 8.8% 7.1% 4.5% 2.5% 4.9% 5.3% 7.3% 4.2%
D 4.8% 12.6% 10.1% 7.3% 4.8% 5.7% 7.1% 11.8% 9.2%
E 1.6% 3.1% 2.4% 1.7% 2.5% 1.6% 6.2% 9.2%
F 4.0% 10.1% 17.2% 14.6% 23.4% 5.7% 15.0% 20.9% 23.3%
G 7.6% 9.5% 9.6% 14.3% 0.8% 6.2% 9.1% 7.5%
H 1.9% 3.0% 7.3% 5.8% 0.9% 3.6% 1.7%
I 2.5% 8.3% 14.0% 12.4% 1.8% 2.7% 5.0%
J 10.1% 9.0% 12.8% 2.7% 5.0%
K 0.6% 6.2% 7.7% 0.9% 2.5%
Obs. 126 159 169 178 517 108 122 113 110 120
Mid non-MGD Base non-MGD
Score CP1 CP2 CE1 CE2 CM1 CP1 CP2 CE1 CE2 CM1
0 26.3% 10.6% 2.4% 0.5% 57.1% 40.3% 19.2% 1.5% 6.9%
A 58.8% 46.0% 34.9% 18.1% 11.9% 39.3% 23.4% 39.7% 15.2% 10.3%
B 12.3% 23.0% 18.1% 16.9% 16.6% 3.6% 14.3% 11.5% 24.2% 15.5%
C 0.9% 3.5% 6.0% 3.6% 1.9% 6.5% 1.3% 3.0%
D 0.9% 5.3% 10.8% 12.1% 7.1% 3.9% 6.4% 9.1% 5.2%
E 0.9% 2.4% 4.8% 0.5% 2.6% 6.4% 6.1% 8.6%
F 3.5% 8.4% 13.3% 21.3% 7.8% 11.5% 21.2% 32.8%
G 2.7% 3.6% 12.1% 8.5% 1.3% 2.6% 6.1% 10.3%
H 1.8% 2.4% 2.4% 3.3% 3.0% 5.2%
I 1.8% 6.0% 9.6% 12.3% 10.6% 1.7%
J 1.8% 3.6% 6.0% 10.4% 1.3% 3.5%
K 1.2% 1.2% 5.7%
Obs. 114 113 83 83 211 112 77 78 66 58
Source: Student Survey. Shaded areas are below grade-level reading proficiency. Unshaded areas are at or 
above grade-level reading proficiency. The sample size was 964 at baseline and 1753 at midline. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 75 | Page 
Figure 72: Food Consumption Score Group by Gender of Head of Household
MGD Non-MGD
Gender of 
Head of 
Household
Food 
Consumption 
Score Group
Midline Baseline Midline Baseline
Man
Poor 2.1% 0.8% 1.3% 0.5%
Borderline 7.9% 2.8% 9.5% 4.3%
Acceptable 90% 96.4% 89.2% 95.2%
Observations 1062 602 547 376
Woman
Poor 3.5% 0% 1.8% 6.1%
Borderline 12.6% 3.8% 10.7% 2%
Acceptable 83.9% 96.2% 87.5% 91.8%
Observations 87 79 56 49
All
Poor 2.2% 0.7% 1.3% 1.2%
Borderline 8.3% 2.9% 9.6% 4%
Acceptable 89.6% 96.3% 89.1% 94.8%
Observations 1149 681 603 425
Source: Household Survey. Sample size was 1106 at baseline and 1752 at midline.
Figure 73: Household Dietary Diversity Score by Gender of Head of Household
MGD Non-MGD
Gender of Head 
of Household Midline Baseline Midline Baseline
Man 5.1 7.5 5 7.7
Observations 1061 602 547 376
Woman 5 7.4 4.6 7.1
Observations 87 79 56 49
All 5.1 7.5 5 7.6
Observations 1148 681 603 425
Source: Household Survey. The score is a sum of the number of food groups consumed during the previous 
day. It measures the dietary diversity available in the foods consumed in the home. Sample size was 1106 at 
baseline and 1751 at midline.
Figure 74: Reduced Coping Strategies Index by Gender of Head of Household*
MGD Non-MGD
Gender of Head 
of Household Midline Baseline Midline Baseline
Man 4.4 3.7 4.6 2.9
Observations 1062 602 547 376
Woman 5.4 4.4 4.7 2.9
Observations 87 79 56 49
Total 4.4 3.8 4.6 2.9
Observations 1149 681 603 425
* A higher index indicates more severe coping strategies necessary and higher levels of food insecurity. 
Source: Household Survey. The Reduced Coping Strategies Index is a context non-specific weighted sum of the 
number of days per week each household employs each of five strategies to cope with food scarcity. The 
maximum reduced index is 56. The reduced index is appropriate to measure less-severe coping measures, 
while the full index is more appropriate in more food-insecure contexts.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 76 | Page 
Annex 7: Project Results Framework
MGD SO1: Improved Literacy of 
School Aged Children
MGD 1.1: Improved 
Quality of Literacy 
Instruction
MGD 1.3: Improved 
Student Attendance
1.1.2: Better 
Access to School 
Supplies and 
Teaching 
Materials
1.1.3: Improved 
Literacy 
Instructional 
Materials 
1.3.1: 
Increased 
Economic and 
Cultural 
Incentives 
(decreased 
disincentive)
1.3.5: Increased 
Community 
Understanding 
of the Benefits 
of Education
1.3.4:Increased 
Student 
Enrollment
Distribution: 
school supplies 
and materials
(UNICEF)
Distribution: 
school supplies 
and materials
(WFP, UNICEF)
Take-home 
rations 
(WFP)
Provide school 
meals (WFP)
Raising 
awareness on 
the importance 
of education
(WFP, WCF)
Raising 
awareness on 
the importance 
of education 
(WFP, WCF, 
UNICEF) 
MGD 1.2.1: Reduced 
Short-Term Hunger
1.2.1.1: Increased 
Access to Food 
(School Feeding)
Provide school 
meals (WFP)
1.1.5: Increased 
Skills and 
Knowledge of 
Administrators
Training: School 
Administrators
(UNICEF)
1.1.4: Increased 
Skills and 
Knowledge of 
Teachers
Training: 
teachers 
(UNICEF)
Result supported 
throughpartner
Result achieved by 
WFP
WFP Côte d’Ivoire FY13-FY15 McGovern-Dole Proposal – Program-Level Results 
1.3.3: Increased 
School 
Infrastructure
Building/ 
Rehabilitation: 
Kitchens (DNC, 
WFP, WCF)
Building/ 
Rehabilitation: 
Schools (UNICEF)
Establish School 
Gardens (WFP)
1.3.2: Reduced Health 
Related Absences
Distribution: de￾worming medication, 
vitamins and minerals 
(WFP, NPSUH) 
WFP activities Partner activities
Key
1.1.1: More 
Consistent 
Teacher 
Attendance
Training: 
Teachers
(UNICEF)
Promote teacher 
attendance 
(WFP / UNICEF)
Establish Parent￾Teacher 
Associations
(WCF)
MGD 1.2: Improved 
Attentiveness
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 77 | Page 
MGD 1.4.1: Increased Capacity of Government 
Institutions (including schools)
MGD 1.4.3: Increased Government 
Support
MGD 1.4.4: Increased Engagement of Local 
Organizations and Community Groups
MGD 1.4.2: Improved Policy and 
Regulatory Framework
Training: Commodity Management (WFP)
Develop Partnerships with Farmer Groups to 
supply food to schools (WFP, DNC)
Capacity Building Local, regional and national 
level (WFP)
Training: Parent-Teacher Associations (WCF, 
WFP)
Establish Parent-Teacher Associations (WFP, 
WCF)
SO1 Foundational Results
Capacity Building Local, regional and national 
level (WFP)
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 78 | Page 
MGD 2.6: Increased 
Access to Requisite Food 
Preparation and Storage 
Tools and Equipment
MGD SO2: Increased Use of Health 
and Dietary Practices
MGD 2.4: Increased 
Access to Clean 
Water and Sanitation
MGD 2.5 Increased 
Access to Preventative 
Health Interventions
Building/ Rehabilitation: 
Latrines (UNICEF)
Building/ rehabilitation: 
Wells and water 
stations/ systems
(UNICEF)
Building/Rehabilitation: 
Warehouses and 
Storerooms; Kitchens 
(WFP/WCF)
Provide Energy Saving 
Stoves (WFP)
Distribution: de-worming 
medication, vitamins and 
minerals (NPSUH, WFP)
MGD 2.2: 
Increased 
Knowledge of Safe 
Food Prep and 
Storage 
MGD 2.3: Increased 
Knowledge of 
Nutrition
Training: Good 
health and 
nutrition 
practices (DNC, 
WFP)
MGD 2.1: 
Increased 
Knowledge of 
Health and 
Hygiene 
Practices
MGD 2.7.1: Increased 
Capacity of Government 
Institutions
MGD 2.7.3: Increased 
Government Support
MGD 2.7.4: Increased 
Engagement of Local 
Organizations and 
Community Groups
MGD 2.7.2: Improved 
Policy and Regulatory 
Framework
Capacity Building Local, regional and 
national level (WFP)
Capacity Building Local, regional and 
national level (UNICEF)
Capacity Building Local, regional and 
national level (UNICEF) 
Develop Partnerships with Farmer 
Groups to supply food to schools
(WFP, DNC)
SO2 Foundational Results
Result supported 
through partner
Result achieved by 
WFP
WFP activities Partner activities
Key
Training: Food 
Preparation and 
storage practices
(DNC, WFP)
Capacity Building 
Local, regional and 
national level (WFP, 
WCF)
Establish School 
Gardens (WFP/FAO)
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 79 | Page 
Annex 8: Terms of Reference
The terms of reference (TOR) for the “Support for the Integrated School Feeding Program" 
project decentralized mid-term evaluation are available in English and French on WFP’s 
website using the following link: https://www.wfp.org/content/cote-divoire-school￾feeding-mid-term-evaluation. 
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 80 | Page 
Annex 9: MGD Indicators
RESULTS INDICATOR Baseline Midline Target
MGD SO1:
Literacy Learning 
of school children 
is improved
Proportion of school-children who, 
at the end of two years of primary 
school, demonstrate that they can 
read and understand the meaning 
of the school grade text (boys).
CP1= 5%
CP2=16%
CE1=25%
CE2=11%
CM1=8%
CM2=8%
CP1= 44%
CP2=40%
CE1=51%
CE2=46%
CM1=30%
Proportion of school-children who, 
at the end of two years of primary 
school, demonstrate that they can 
read and understand the meaning 
of the school grade text (girls).
CP1= 4%
CP2=13%
CE1=19%
CE2=11%
CM1=6%
CM2=7%
CP1= 47%
CP2=36%
CE1=46%
CE2=27%
CM1=17%
CM2=27%
Number of individuals benefiting 
directly from USDA-funded 
interventions (new). 
0
Number of individuals benefiting 
directly from USDA-funded 
interventions (old). 
0
Number of individuals benefiting 
directly from USDA-funded 
interventions (men).
0
Number of individuals benefiting 
directly from USDA-funded 
interventions (women).
0
Number of individuals benefiting 
directly from USDA-funded 
interventions. 
0
MGD 1.1
Improved quality 
of literacy 
learning of school 
children
Number of teachers in target schools 
demonstrating the use of new 
techniques or quality teaching tools 
as a result of USDA assistance.
0
MGD 1.1.1
More consistent 
teacher 
attendance
Proportion of teachers in target 
schools who attend and teach at 
school regularly (at least 90% of 
school days) per school year.
94.2% 94.4%
MGD 1.1.2
Better access to 
school supplies 
and materials 
Number of textbooks and other 
teaching and learning materials 
provided with USDA assistance.
0
MGD 1.1.3
Improved access
to learning 
materials for 
reading 
Number of target schools in which 
school children have additional 
reading materials with USDA 
support.
0
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 81 | Page 
MGD 1.1.4
Increased skills 
and knowledge of 
teachers 
Number of teachers / teachers’ aides 
in targeted schools demonstrating 
the use of new and good techniques 
or teaching tools (per type, per 
gender).
0
Number of teachers / teachers’ aides
trained or certified as a result of 
USDA assistance (per type, per sex).
0
MGD 1.1.5
Increased skills 
and knowledge of 
school 
administrators 
Number of targeted school 
administrators demonstrating new 
and good techniques or teaching 
tools (per type, per gender).
0
Number of leaders trained or 
certified as a result of USDA 
assistance (by gender). 
0
MGD 1.2
Improved school 
children 
attentiveness
Proportion of school children 
identified as being attentive in class 
by their teachers (per sex, per 
class).
Girl = 78%
Boy = 78%
Girl = 68%
Boy = 67%
MGD 1.2.1
Reduced short￾term hunger 
Number of school lunches 
(breakfast, snack, lunch) offered to 
school children following USDA 
assistance.
0
Proportion of school children in 
targeted schools who ate a meal 
regularly before or during the 
school day (per gender).
0
MGD 1.2.1.1 
/1.3.1.1
Increased access 
to school feeding
Number of take-home rations 
provided as a result of USDA 
assistance.
0
Number of girls having received 
take-home rations as a result of 
USDA assistance.
0
Number of girls having received 
take-home rations as a result of 
USDA assistance (new).
0
Number of girls having received 
take-home rations as a result of 
USDA assistance (continue).
0
Number of school children receiving 
daily school meals (breakfast, snack, 
lunch) as a result of USDA assistance 
(girls).
0
Number of school children receiving 
daily school meals (breakfast, snack, 
lunch) as a result of USDA assistance
(boys).
0
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 82 | Page 
Number of school children 
receiving daily school meals 
(breakfast, snack, lunch) as a result 
of USDA assistance (new).
0
Number of school children receiving 
daily school meals (breakfast, snack, 
lunch) as a result of USDA assistance 
(continue).
0
Number of daily school meals 
(breakfast, snack, lunch) provided to 
school-children following USDA 
assistance.
0
Proportion of households with 
acceptable food consumption per 
sex of the household head. 
Female = 
96.2%
Male = 
96.3% 
Female = 
83.9%
Male = 
90.0%
Coping strategy index (average) per 
sex of the household head. 
Female = 
4.1
Male = 2.9
Female = 
5.4
Male = 4.4
Dietary diversity score per sex of 
the household. 
Female = 
5.8
Male = 6
Female = 
5.0
Male = 5.1
Number of social safety net 
beneficiaries participating in 
productive safety nets operations as 
a result of USDA assistance 
(ongoing).
0
Number of social safety net 
beneficiaries participating in 
productive safety nets operations as 
a result of USDA assistance (new).
0
Number of social safety net 
beneficiaries participating in 
productive safety nets operations as 
a result of USDA assistance (men).
0
Number of social safety net 
beneficiaries participating in 
productive safety nets operations as 
a result of USDA assistance 
(women).
0
MGD 1.3
Improved
Student 
attendance 
Proportion of students regularly 
(80%) attending classes / schools 
supported by USDA (boys).
98% 100%
Proportion of student regularly 
(80%) attending classes / schools 
supported by USDA (girls).
98.3% 100%
MGD 1.3.1
Increased 
Economic and 
Cultural 
Incentives (Or 
Number of girls receiving take￾home rations as a result of USDA 
assistance (new).
0
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 83 | Page 
Decreased 
Disincentives) Number of girls receiving take￾home rations as a result of USDA 
assistance (ongoing).
0
MGD 1.3.2
Reduced Health 
Related Absences
Proportion of school children who 
miss more than 10 days of school 
per year due to illness (boys).
2% 0.7%
Proportion of school children who 
miss more than 10 days of school 
per year due to illness (girls).
1.7% 0.7%
MGD 1.3.4
Increased 
Student 
Enrolment 
Number of school children enrolled 
in schools receiving USDA 
assistance (boys).
0
Number of school children enrolled 
in schools receiving USDA 
assistance (girls).
0
Gender Ratio, elementary school.
0.85 0.99
MGD 1.3.5
Increased 
Community 
Understanding of 
Benefits of 
Education
Number of members of 
management committees and 
members of women production 
groups sensitized on the importance 
of education.
0
Value of Public and Private 
Investments creating a leverage 
effect due to USDA Assistance (Host 
Government).
0
Annual increase rate of the budget 
allocated by the Government to the 
Directorate of School Canteens. 
0
Number of policies in the child 
health and nutrition sectors, 
regulations and administrative 
procedures, by level of development, 
due to USDA support (per stage).
0
Number of education sector 
policies, regulations and 
administrative procedures, by level 
of development, due to USDA 
assistance (per stage).
0
MGD 1.4.4
Increased 
Engagement of 
Local 
Number of parent-teacher 
associations or similar "school" 
governance structures supported as 
a result of USDA assistance.
0
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 84 | Page 
Organizations 
and Community 
Groups
Number of public-private 
partnerships put in place as a result 
of USDA assistance (nutrition).
0
Number of public-private 
partnerships put in place as a result 
of USDA assistance (education).
0
Number of public-private 
partnerships put in place as a result 
of USDA assistance (health).
0
Number of public-private 
partnerships put in place as a result 
of USDA assistance (multi sectoral).
0
Number of public-private 
partnerships put in place as a result 
of USDA assistance (other).
0
MGD SO2
Increased Use of 
Health and 
Dietary Practices
Proportion of school-aged children 
receiving a minimum acceptable 
diet (boys).
47% 83%
Proportion of school-aged children 
receiving a minimum acceptable 
diet (girls).
47% 83%
MGD 2.1
Improved 
Knowledge of 
Health and 
Hygiene Practices
Proportion of School Management 
Committee members and canteen 
management staff who can identify 
at least three health and hygiene 
practices(men).
89.6% 88%
Proportion of School Management 
Committee members and canteen 
management staff who can identify 
at least three health and hygiene 
practices (women).
89.6% 88%
MGD 2.2
Increased 
Knowledge of 
Safe Food 
Preparation and 
Storage Practices
Proportion of school management 
committee members and canteen 
management staff who can identify 
at least three safe food preparation 
and storage practices.
.1% good 
food 
storage 
practice
72% good 
knowledge 
of safe food 
preparation 
practices
83% good 
food 
storage 
practice
82% good 
food 
preparation 
practices
MGD 2.3
Increased 
Knowledge of 
Nutrition
Number of individuals receiving 
training in child health and nutrition 
as a result of USDA assistance 
(men).
0
Number of individuals receiving 
training in child health and nutrition 
as a result of USDA assistance 
(women).
0
MGD 2.5 Number of school children who 
received deworming.
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 85 | Page 
Access to 
Preventative 
Health 
Interventions 
Number of school children who 
received deworming (boys).
Number of school children who 
received deworming (girls).
Proportion of schools using an 
improved water source.
48% has a 
water point 
93% have 
an 
improved 
water 
source
54% has a 
water point 
94% have 
an 
improved 
water 
source
Proportion of schools with 
improved sanitation facilities.
53% 59%
MGD 2.6
Increased Access 
to Requisite Food 
Prep and Storage 
Tools and 
Equipment
Number of targeted schools with 
access to improved food preparation 
and storage equipment. 
MGD 1.4.1 
/2.7.1
Increased 
Capacity of 
Government 
Institutions
Number of government staff trained 
in food management, monitoring 
and evaluation.
0
MGD 1.4.2 
/2.7.2 
Improved Policy 
and regulatory 
framework
Number of policies, regulations, 
and/or administrative procedures in 
the child health and nutrition 
sectors in each of the following 
stages of development as a result of 
USDA assistance (stage 1).
0
Number of child health and 
nutrition policies, regulations, 
and/or administrative procedures in 
each of the following stages of 
development as a result of USDA 
assistance (stage 2).
0
Number of child health and 
nutrition policies, regulations, 
and/or administrative procedures in 
each of the following stages of 
development as a result of USDA 
assistance (stage 5).
0
MGD 1.4.3
Increased 
government 
support
Value of Public and Private 
Investments creating a leverage 
effect due to USDA Assistance (Host 
Government).
0
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 86 | Page 
Annual increase rate of the budget 
allocated by the Government to the 
Directorate of School Canteens. 
0
MGD 1.4.4
Increased 
Engagement of 
Local of Local 
Organizations 
and Community 
Group
Number of parent-teacher 
associations or similar "school" 
governance structures supported as 
a result of USDA assistance.
0
Number of public-private 
partnerships set up as a result of 
USDA assistance (women 
production groups).
0
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 87 | Page 
Annex 10: Map of MGD School Feeding Program
Mid-Term Evaluation Report: MGD Project in Côte d’Ivoire (2015-2020) 88 | Page 
Annex 11: List of Abbreviations
AEA American Evaluation Association
AFD Agence Française de Développement
ANADER Agence Nationale d'Appui au Développement Rural
AVSI Associazione Volontari Servizio Internazionale
DCS Directorate of School Canteens
DID Difference-in-difference
FAO Food and Agriculture Organization
FFE Food For Education
FGD Focus group discussion
IFAD International Fund for Agricultural Development
IRB Institutional review board
KII Key informant interview
MAD Minimum acceptable diet
MENETFP Ministère de l'Education Nationale, de l’Enseignement Technique, et de la 
Formation Professionnel
MGD McGovern-Dole
OEV Office of Evaluation
SOP Standard Operating Procedure
THR Take home rations
TOR Terms of Reference
UNDP United Nations Development Program
UNEG United Nations Evaluation Group
UNHCR United Nations High Commissioner for Refugees
UNICEF United Nations International Children's Emergency Fund
World Food Programme (WFP) Country Office in Côte d'Ivoire
www.wfp.org