1 
Emergency Food Security and Recovery Program for 
the vulnerable people affected by the residual effect 
of Hurricane Irma in the North-East and Central 
departments of Haiti
 
USAID EFSP
Final Evaluation Report
Submitted by:
Amisial LEDIX, Economist-Statistician/Monitoring & Evaluation Specialist
In collaboration with:
Amos MONTREUIL JEAN, Agronomist/Food Security and Nutrition Specialist
December 1, 2020
2 
Table of Contents
Acknowledgements 3
List of Tables, Figures and Graphs 4
List of Acronyms 6
Executive Summary 7
1) Background 10
2) Final Evaluation Objectives 12
3) Methodology 12
4) Constraints and Limitations of the Study 22
5) Analysis of the Final Evaluation’s Findings 23
5.1 General Characteristics of Beneficiary Households 23
5.2 Analysis of Project Relevance 29
5.3 Analysis of Project Effectiveness 37
5.4 Analysis of Project Efficiency 46
5.5 Analysis of Project Impact 49
5.6 Analysis of Project Sustainability 51
5.7 Analysis of Project Linkages and Exit Strategies 53
5.8 Analysis of the Local Market in the Intervention Areas 55
6) Conclusion and Recommendations 60
7) Bibliographical References 64
8) Appendix I: FCS Indicator 65
9) Appendix II: HHS Indicator 66
10) Appendix III: Reduced Coping Strategy Index (rCSI) 67
11) Appendix IV: Data Collection Tools (in separate files) 69
3 
Acknowledgements 
The team of consultants in charge of the present study wishes to express their gratitude, in a 
special manner, to all those who contributed to the completion of this work. Despite the many 
challenges and difficulties encountered, we were able to succeed through your active and 
effective participation. 
First, we wish to thank the Monitoring and Evaluation team, within the World Vision Haiti 
Central Office, for their open and fruitful collaboration. Interacting with our interlocutors was 
essential to the successful completion of the consultation. The support provided for the use 
of tablets, a few scales and measuring poles, and the reference documentation made available 
to us made our work much easier.
Our gratitude, then, goes to the EFSP project manager with whom we also interacted during 
the scope of work meeting for this study. We also thank World Vision’s Regional Manager in 
the North for his support in lending us the measuring poles, and the project’s Monitoring and 
Evaluation key person for their respective support in providing the necessary information to 
facilitate the progress of field operations. 
We would also like to thank the enumerators and supervisors who went to the field to collect 
data from beneficiaries in the project’s various intervention communities.
The consulting team remains convinced that the study’s findings and recommendations will 
effectively serve the EFSP project team in addressing the challenges and achieving expected 
results on behalf of the beneficiaries living in the organization’s areas of intervention. 
4 
List of Tables, Figures and Graphs
Table 1: Breakdown of the number of beneficiary households to be surveyed by 
commune............................................................................................................................................ 18
Table 2: Estimated sub-population of breastfeeding women beneficiaries by commune 19
Table 3: Breakdown of the sample of breastfeeding women to be surveyed by commune 
in the Northeast ................................................................................................................................ 19
Table 4: Breakdown of Recipients Sample with Children 0-59 Months of Age to be 
Surveyed.............................................................................................................................................. 20
Table 5: Number of Key Interviews by Department................................................................ 21
Table 6: Distribution of Respondents by Department by Housing Type (n = 1358)......... 27
Table 8: Distribution of respondents by department according to the second activity of 
the head of household...................................................................................................................... 28
Table 9: Main source of water for household domestic use.................................................... 31
Table 10: On the management and maintenance of water points ....................................... 32
Table 11: Distribution of beneficiary households by department according to the water 
conservation method used at home............................................................................................. 34
Table 12: Level of satisfaction by activity in which respondent took part .......................... 37
Table 13: Dietary Diversity Score by Department ................................................................... 38
Table 14: Food consumption score for project beneficiaries by department...................................... 39
Table 15: Hunger score for project beneficiaries by department ........................................................ 40
Table 16: Reduced Coping Strategies Index (rCSI) .............................................................................. 41
Table 17: Percentage of households where adults and children eat at least 2 meals a day (adults and 
children), by age .................................................................................................................................... 42
Table 18: Percentage of food use by type (household consumption, sale, exchange, livestock feed) by 
department............................................................................................................................................ 42
Table 19: Percentage of households reporting having received information to cope with economic 
shocks for the intervention areas and its communes........................................................................... 43
Table 20: Major shocks that caused hunger in percentage................................................................... 43
Table 21: Prevalence of acute malnutrition by department .................................................. 45
Table 22: Key indicators measured, and their progress compared to targets................... 45
Table 23: Comparison of baseline and final evaluation results .............................................. 50
Table 24: Level of Satisfaction with Criteria Achievement................................................................... 54
Figure 1 : Age pyramid of the communes where the EFSP Centre project is active. Error! 
Bookmark not defined.
Figure 2 : Age pyramid of the intervention communes of the EFSP Nord-Est project....................Error! 
Bookmark not defined.
Graph 3: Distribution of Respondents by Department by Gender (n=1,258) .................... 24
Graph 4: Distribution of Respondents by Department by Household Status (n=1,258).. 24
Graph 5: Distribution of heads of household by department according to their education 
level (n=1,258).................................................................................................................................... 25
Graph 6: Distribution of heads of household by department according to marital status 
(n=1,258)............................................................................................................................................. 26
Graph 7: Distribution of respondents by department according to whether or not there is at least 
one orphan in the household................................................................................................................ 26
Graph 8: Distribution of respondents by department according to whether or not there 
is a pregnant woman in the household ........................................................................................ 26
5 
Graph 9: Distribution of respondents by department according to whether or not there 
is at least one person with a disability in the household.......................................................... 26
Graph 10: Distribution of respondents by department by monthly household income 
group.................................................................................................................................................... 28
Graph 11: Distribution of respondents by department by first annual household 
expenditure item .............................................................................................................................. 29
Graph 13: Distribution of beneficiary households by department according to whether 
or not they received messages about newborn care in the health centers (n=618)......... 31
Graph 14: Distribution of beneficiary households according to whether they participate 
in training provided by mother leaders in the community..................................................... 31
Graph 15: Distribution of beneficiary households by department according to the amount of water 
available for use per day........................................................................................................................ 32
Graph 16: Distribution of beneficiary households by department according to the 
walking time (round trip) for water collection.......................................................................... 32
Graph 17: Distribution of beneficiary households by department according to whether 
or not they usually treat water at home ..................................................................................... 33
Graph 18: Distribution of beneficiary households by department according to the fact 
that they do not usually treat water at home............................................................................ 33
Graph 19: Distribution of beneficiary households by department according to the water 
treatment method used .................................................................................................................. 34
Graph 20: Distribution of beneficiary households by department according to whether 
or not they received handwashing messages ............................................................................. 35
Graph 21: Range for reduced coping strategies index (rCSI)................................................. 41
Graph 22: Prevalence of mothers who practice exclusive breastfeeding among infants 
six months of age or less ................................................................................................................. 44
Graph 23: Distribution of respondents according to their perception of products prices 
on the vouchers compared to prices for the same products at other sellers .................... 56
Chart 24. Distribution of respondents by who decided when to redeem and what products to get . 57
Graph 25: Distribution of respondents by use of products after voucher redemption.... 57
Graph 26: Distribution of respondents according to their satisfaction with the quality of 
food products after the voucher was redeemed ....................................................................... 58
Graph 27. Distribution of respondents by how they would like to receive assistance for a 
similar project.................................................................................................................................... 59
6 
List of Acronyms 
ASEC Assemblée de la Section Communale
CASEC Conseil d’Administration de la Section Communale
CEPAM Center for Promoting Breastfeeding (Centre pour la promotion de l’allaitement 
maternel)
CNSA Coordination Nationale de la Sécurité Alimentaire
COVID-19 Coronavirus Disease 2019
DINEPA Direction Nationale de l’Eau Potable (National Water Authority)
DPC Civil Protection Office (Direction de la protection civile)
EFSP Emergency Food Security Program
FCS Food Consumption Score
FEWSNET Famine Early Warning Systems Network
FFP Food for Peace
FNM Female No Adult Male
F&M Female and Adult Male
GAM Global Acute Malnutrition
GDP Gross Domestic Product
GPS Global Positioning System
HDDS Household Dietary Diversity Score
HHS Household Hunger Scale
HNO Humanitarian Needs Overview
HTG Haitian Gourde
IPC Integrated Phase Classification
LA Local Authorities
MAST Ministry of Social Affairs and Labor (Ministère des Affaires Sociales et du Travail) 
M&E Monitoring and Evaluation
M4P Making Markets Work for the Poor
NGO Non-Governmental Organization
OECD Organization for Economic Cooperation and Development
PDM Post-Distribution Monitoring
rCSI Reduced Coping Strategies Index
S4T Savings for Transformation
SOP Standard Operational Procedures
SPSS Statistical Package for the Social Sciences
SRS One-stage Simple Random Sample
TED Home Water Treatment (Traitement d’eau à domicile)
TEPAC WASH Communal Technician (Technicien d’eau potable et assainissement 
communal) 
USAID U.S. Agency for International Development
UVT Unconditional Voucher Transfer
VFW Voucher for Work
WV World Vision
WVI World Vision International
7 
Executive Summary 
The Emergency Food Security Program (EFSP), implemented by World Vision in six communes 
within the Northeast Department (Fort-Liberté, Carice, Mombin Crochu, Sainte Suzanne, 
Ferrier and Vallières) and two communes in the Central Department (Cerca-la-Source and 
Cerca Carvajal), seeks to provide immediate access to food to communities in crisis and 
emergency situations (IPC 3 and 4) and help communities recover from the residual effects of 
Hurricane Irma. For over 18 months this program has been implemented with USAID’s 
financial support, targeting more than 6,700 vulnerable households (33,500 people) who 
needed help mitigating food insecurity in times of crisis due to reduced harvests and 
livelihoods. In relation to initial objectives, the program included five main intervention areas: 
Distribution of food vouchers, Nutrition, Drinking Water Infrastructure, Agriculture, Financial 
support (Training and S4T).
As program interventions have come to an end, the present evaluation was conducted to 
assess the achievement of the goal, objectives, and outcomes and review the means by which 
they were achieved. For this work, a double quantitative and qualitative approach was used in 
collecting the required information for the evaluation. 678 beneficiaries per department and 
other project stakeholders (project managers, government entities, community leaders, etc.) 
were interviewed. The quantitative survey was conducted among program beneficiary 
households and was focused on reviewing their situation using the phase classification 
indicators in terms of food security, and reviewing the situation of the markets serving these 
households. The qualitative survey targeted key project stakeholders whose interviews were 
to be used to validate the initial information collected. In addition to this series of interviews, 
there are more evaluation questions based on OECD criteria.
Overall, the findings of this evaluation show that : 
 There are more people from the 5-19-year-olds in the population structure of the 
project’s beneficiary households and the female population is much more represented 
in the Northeast than in the Center, respectively with 71% and 42% of beneficiaries. 
 Most respondents from the Northeast are women, while the reverse is true in the 
Central Department. Overall, there are more households headed by adult women and 
men (F&M) in this survey.
 In both departments, most beneficiaries have no education at all (43% for the 
Northeast and 32% for the Center) or have not completed their primary education 
(31% for the Northeast and 48% for the Center). 
 Agriculture and trade are the main sources of income for most beneficiary households 
and, in most cases, these households earn a monthly income of about 5,500 HTG or 
less. 
 Poor households are more inclined to spend their scarce economic resources in food.
Taking into account the identified issues and analyzing the outcomes from the five intervention 
areas, the program’s relevance becomes obvious. From a strategic perspective, the nutrition 
component’s interventions are in line with the National Nutrition Policy (NNP) that seeks to 
improve the nutritional situation of communities based on a preventive approach to all forms 
of malnutrition. Outcomes show that the activities from this component have reached a large 
8 
number of beneficiaries who were facing this problem. At least 90% of respondents indicated 
having received messages related to newborn care, and at least 40% of respondents (women) 
state they attend trainings conducted by mother-leaders in their community. 
During the implementation period, the Haitian Gourde rapidly lost value and inflation rose 
steadily. To deal with this situation, monthly adjustments were made to voucher values, in line 
with the increase in food prices in the local markets. In this activity, the most vulnerable 
segments of the population, consisting of women leading single-parent households, are the 
main beneficiaries. 
According to outcomes from the infrastructure component, at least 60% of beneficiary 
households in the Central Department have access to an improved water source, compared 
to 35% of beneficiary households in the Northeast who use uncollected water sources. 
Training and awareness activities on TED and handwashing have reached a large number of 
beneficiaries, since (59% in the Northeast and 71% in the Center) most of them reported 
treating water at home (with chlorine and Aquatabs) and more than 80% reported having 
received handwashing messages. After the interventions, about 50% of beneficiaries can now
use sufficient water and have easy access to water, according to DINEPA standards. The 
outcomes show that these interventions are in line with the program’s goal of increasing or 
restoring access to safe drinking water for communities whose water systems were damaged 
after Hurricane Irma. 
In the agriculture component, actions were aligned with the National Adaptation Action Plan 
(NAPA) and directed towards reducing vulnerability to climate shocks and post-harvest losses, 
two major issues faced by farmers in the program’s intervention areas. To that effect, training 
and technical support on best land conservation practices were provided to smallholder 
beneficiaries in target communities. In addition, farmers also received training in post-harvest 
management.
The setting up of credit unions is also a relevant element in strengthening the households 
affected by Hurricane Irma. Through these groups, households were able to save, in spite of
their small income, thanks to the distribution of food vouchers. Thus, they were able to set 
up a small trade or strengthen an existing business activity thanks to the loan contracted in 
the community groups.
Through analyzing the effectiveness of program interventions, we saw the progress of key 
indicators against their targets. Only 3 of them did not reach 50% of the targets! However, in 
our interview with the project’s M&E team, we found out that 4 output indicators did not 
reach 50% of their targets, namely the “number of stone lines rehabilitated”, “number of check 
dams rehabilitated”, “number of soil dams rehabilitated”, “number of IEC materials 
distributed”. Overall, this implies that almost all project indicators have reached at least 50% 
of their target.
The efficiency of the project was assessed by comparing the cost of component activities to 
their effectiveness. In this analysis, the food voucher, nutrition, agriculture and financial support 
(training and S4T) components were taken into account. According to the analysis carried out, 
the program has had efficiency concerns, especially in terms of meeting deadlines with vendors 
and, in agriculture, due to a lack of human resources to monitor activities. Since we had no
financial report, we could not conduct further analysis on the program’s efficiency. 
9 
The program was able to reach all the targeted beneficiaries, and significant changes were 
observed in beneficiary households and communities, such as: improvement in the 
beneficiaries’ food insecurity and living conditions, improvement in the diet of nursing mothers, 
savings opportunities, and opportunities to diversify households’ business activities through 
savings groups. These changes are an evidence of the program’s impact on communities and 
households. 
The adoption of a strategy based on the project’s ownership by beneficiaries is a key element 
that can ensure the sustainability of project interventions. The building of infrastructure, 
training on farming techniques, and the setting up of credit unions are sustainable activities 
that will bring benefits to the communities, beyond the project’s duration. The involvement of 
local government actors in the implementation of activities is also a guarantee of sustainability 
for most project interventions. However, the exit strategy was not adequately implemented 
because the links with system partners (the sectors related to the project) are not clearly seen 
in the implementation process. 
The additional market research conducted to enrich the assessment confirmed that all major 
markets in the region were operating. Contracted vendors did not highlight any challenge
during the selection process, and also reported that the WV team provided critical advice that 
would enable them to meet the defined selection criteria (such as guidelines on licensing their 
small businesses). The project played no role in the fluctuation of product prices. Community 
leaders talked about complaints from non-program vendors about the loss of customers to 
registered vendors. The amount of money coming in benefits the communities, but the money 
flow and transactions only benefits a smaller number of vendors. The WV team emphasized, 
however, that an inclusive and participatory approach was used to engage vendors in the 
program, using clear selection criteria.
Considering this assessment’s findings, it becomes necessary to review the intervention 
strategy in a future program, especially in the food security and farmers’ capacity building 
components, so that activities can have greater impact. 
10
1) Background 
In 2018, Haiti had a per capita gross domestic product (GDP) of $870 and a human 
development index ranked 168th out of 189 countries, for the same year (World Bank1
, 2019). 
More than 6 million Haitians live below the poverty line of less than US$2.41 per day, and 
more than 2.5 million live below the extreme poverty line of US$1.23 per day. Political 
instability has severely hampered Haiti’s economic and social development in recent months, 
and the country has experienced a fast currency depreciation (around 30 percent) followed 
by a fast appreciation without much change in economic fundamentals, high levels of inflation 
(close to 20 percent), and a GDP contraction (projected at 0.5 percent) in fiscal year 2019.
In addition to political and economic insecurity, Haiti’s vulnerability to recurring climate shocks 
has increased over the past five years. The country remains highly vulnerable to natural 
hazards, mainly hurricanes, floods, and earthquakes, and more than 96 percent of the 
population is at risk due to chronic environmental degradation, poor infrastructure, and 
erosion-prone farming practices, among others. Natural and man-made shocks exacerbate 
Haiti’s many development challenges and underlying poverty drivers. One of the latest major 
natural disasters, Category 4 Hurricane Matthew, that hit Haiti in 2016, killed approximately 
600 people, affected 2.1 million people and caused extensive damage in the southern peninsula 
and northwestern part of the country. It caused a 32 percent loss of the country’s GDP. Prior 
to that, in 2010, an earthquake killed approximately 230,000 people, displaced 1.5 million 
people, and caused a loss of 120 percent of the GDP2
. 
As reported by FEWSNET and CNSA, crop shortages and livelihood disruptions caused by 
hurricane Irma and recent heavy rains have led parts of Haiti’s northeastern and central 
departments into a crisis situation (IPC 3) that continued until May 2018. Damage to 
infrastructure, homes and livelihoods caused by hurricanes Matthew and Irma was estimated 
at US$2.7 billion nationwide, including US$573.5 million for the agricultural sector3
. 
The spring harvest, which normally accounts for half of the country’s annual agricultural yield, 
fell due to recurrent droughts (from 2016 to 2018) that have affected almost all departments 
and consequently worsened the already precarious food security situation. In November 2019, 
the National Food Security Coordination (CNSA) highlighted a significant increase in the food 
basket’s value with monthly and annual rates of 14% and 40% respectively. Markets in the 
south (Jérémie, Cayes, Fond des Nègres, Jacmel, and Croix des Bossales) were the most 
affected, mainly due to the rise in grain prices (corn, rice, and flour). An estimated 430,468 
people (367,038 in the Northeast and 63,430 in the Center) are still under shock. “The strike 
of hurricane Irma and the heavy rains that followed in the Northeast hampered farming
activities, especially in the lowland areas. Excessive soil moisture in Ouanaminthe, Ferrier and 
Fort Liberté prevented farmers from planting winter crops. As a result, there were no 
1 See the revised Haiti Country Partnership Framework for the FY16-19 period (English) 
http://documents.worldbank.org/curated/en/673911467986337757/Haiti-Country-partnership-framework-for-the-period￾FY16-FY19 2 See the World Bank document: Haiti, providing opportunities for all Haitians
3 FAO Haiti: Hurricane Matthew Situation Report, 2017
11
harvests, except in the mountain areas of communes such as Sainte Suzanne, Mont Organisé, 
Carice and Vallières, where crop yields were still below average as expenses dropped due to 
reduced economic resources with local farmers”. On the other hand, dry periods in the Upper
Plateau, mainly in the communes of Thomassique, Cerca-la-Source and Cerca Carvajal, 
disrupted winter crops, resulting in poor or even non-existent harvests, especially for beans, 
which are one of the main cash crops produced at this time of year. According to the CNSA 
and FEWSNET surveys, very few farmers were able to do the winter plantings, indicating that 
below-average harvests are expected. FEWSNET predicts low yields due to reduced 
investment by reason of reduced economic resources4
. 
According to the October 2019 IPC analysis, focusing on the March to June 2020 period, 12% 
(1,203,000 people) of the total population will be in an emergency phase and 28% (2,898,000 
people) in crisis, which represents about 40% of the population (4.10 million people) facing 
food insecurity. Both in the current and forecast period, the situation seems to be worse in 
rural areas with 38% and 42% of the population in phase 3, respectively, compared to 28% and 
31% in urban areas. The drought led to a 12% drop in agricultural production in the Northwest, 
the Upper Artibonite, the Northern and Southeastern border regions, and the mountainous 
interior of the South and Nippes departments, which significantly reduced food access for 
many of Haiti’s poorest households. Many rural areas are particularly vulnerable due to a 
combination of risks related to natural disasters (drought, hurricanes, landslides) and current 
rates of food insecurity.
This is the situational context that led World Vision to propose this Emergency Food Security
Program (EFSP) to provide immediate access to food to IPC 3 communities and to help them
recover from the residual effects of hurricane Irma. This 18-month program received funding 
from USAID and reached the Northeast and Central Plateau departments. Nearly 6,700 
vulnerable households (33,500 people) were targeted with the goal of alleviating food 
insecurity in times of crisis due to reduced livelihood harvests. This program involved six 
communes in the Northeast (Fort-Liberté, Carice, Mombin Crochu, Sainte Suzanne, Ferrier, 
and Vallières) and two communes in the Central Plateau (Cerca-la-Source and Cerca Carvajal) 
that were in crisis and emergency situations (IPC3 and 4).
Program activities included :
● Unconditional electronic voucher transfer (UVT)
● Conditional voucher transfer-Food Voucher for Work (VFW)
● Essential complementary activities: agriculture, livelihoods, savings for transformation
(S4T) and nutrition.
The present work was carried out as part of the program’s final evaluation, which took stock 
of the main findings. 
4 CNSA/MARNDR bulletin 16_ footbasket_securite_alimentary-Jan-2018 
12
2) Final Evaluation Objectives 
The overall objective of this final evaluation is to analyze the level of achievement of the goal, 
objectives, and outcomes of the EFSP program and how these were done. 
Specifically, the final evaluation considers the following objectives:
● Evaluate program’s achievements with respect to goal, objectives, outcomes and 
targets.
● Assess program’s effects on local markets and certain interest groups (women and 
men; youth population; boys and girls, etc.).
● Evaluate the effectiveness and relevance of the modality, transfers and complementary 
interventions to achieve program’s outcomes. 
● Identify best practices, lessons learned, strengths and challenges in the program’s 
design, including the logical framework and implementation to achieve project’s 
achievements. 
● Recommended strategies for other projects or new interventions.
3) Methodology 
To achieve the objectives of this final evaluation, a dual quantitative and qualitative approach 
was considered. This approach combined the collection of project-related secondary data and 
primary data among project partners and participants in the intervention communities of the 
North-East and Central departments. 
The evaluation questions mentioned in the ToRs are based on OECD criteria. The following 
tables clarify the approach for addressing the different questions according to the Relevance￾Effectiveness-Efficiency-Impact-Sustainability criteria:
Criteria Evaluative Questions Data sources/Collection 
methods
Relevance What are the opinions of stakeholders on the nature and quality of 
implementation? Under what circumstances and/or in what contexts would 
the program be replicable, or could it be scaled up? Were relevant MAST 
officials involved?
Has the project implementation strategy been adjusted to take into account 
the realities on the ground? If so, in what way?
Were program activities and outputs consistent with the expected impacts and 
effects?
Was the program consistent with the needs and priorities of the most 
vulnerable and targeted PLWs?
What lessons have been learned?
-Quantitative survey of 
beneficiaries
-Key interviews with project 
manager and/or project 
component managers
-Key interviews with 
project’s key partners 
(MAST, AL, ...)
13
Efficiency Did the project have adequate and appropriate resources (human, financial 
and capital) for its implementation?
If there was a lack/problem of resources/capacity, how was it resolved?
Assess the communication structure in place and its effectiveness in supporting 
program implementation. 
Were quality control and accountability measures in place and consistently 
applied during the review, approval, funding disbursement, monitoring, and 
reporting phases? 
Do recipient comments indicate widespread instances where funds (vouchers) 
were taxed or stolen, or where receiving a voucher represented a protection 
risk?
-Key interviews with project 
manager and/or project 
component managers
-Key interviews with female 
and male beneficiaries
-Consulting Project financial 
reports
-Key interviews with 
project’s key partners 
(MAST, AL, ...)
Effectiveness To what extent were the objectives achieved?
What were the main factors that did or did not influence the achievement of 
the objectives?
Did the M&E system provide appropriate and reliable quality information to 
measure the planned indicators? 
How effective was the project model in terms of design, relevance, 
management and accountability?
How effective was the program in terms of implementation (coordination, 
cooperation, effectiveness, standardization)?
To what extent were strategies used to carry out project activities? 
Have humanitarian standards been met and humanitarian principles complied 
with (SPHERE, HAPs, Codes of conduct)?
What measures were taken to reduce the negative effects?
-Quantitative survey of 
beneficiaries
-Key Interviews with 
Project Manager and/or 
Project Component 
Managers/M&E Manager
-Key interviews with 
project’s key partners
of the project (MAST, AL,)
-Consulting periodic M&E 
tools/reports
Criteria Evaluative Questions Data sources/Collection 
methods
Impact Did the project reach the expected number of beneficiaries and territorial 
coverage?
To what extent has the project contributed to reducing beneficiaries’ 
vulnerability the level?
Did the response reduce future vulnerabilities?
What are the unintended positive and negative impacts of project 
implementation? 
To what extent do the interventions improve the condition of the affected 
communities?
How satisfied are the communities with the response?
Did the program require more time from the women?
-Quantitative beneficiary 
surveys
-Key interviews with 
Project Manager and/or 
Project Component 
Managers/M&E Manager
-Key interviews with 
project’s key partners
(MAST, AL, ...)
-Consulting Beneficiary 
Monitoring Database
14
What do men and other household members (in-laws, etc.) think about 
women’s participation in the programs?
What gender-specific issues were addressed?
What positive changes are observed in the lives of the target group as a result 
of project implementation?
Did the voucher project affect the market and context in any way (did the
voucher assistance have an impact on inflation?
Criteria Evaluative Questions Data sources/collection 
methods
Sustainability To what extent will project benefits continue after donor funding ends? Are 
positive effects sustainable?
To what extent did the project take into account factors that, in experience, 
have a major influence on sustainability such as economic, ecological, social and 
cultural aspects?
What sustainability drivers are obvious (local ownership, partnership, 
transformed relationships, household and family resilience)?
-Quantitative survey of 
beneficiaries
-Key interviews with 
project manager and/or 
project component 
managers 
-Key interviews with key 
project partners (MAST, 
AL, ...)
-Key interviews with 
beneficiaries
Linkages, 
stratification 
and exit 
strategies 
To what extent did the project take advantage of other USG and non-USG 
investments in the same area to facilitate linkages with complementary services, 
overlaying previous investments and implementing exit strategies to minimize 
reliance on external support?
Are there other non-USG investments on which the project has relied in its 
exit strategy to reduce reliance on external support?
To what extent has the project aligned and integrated with the Haitian 
government’s recent national policy for social protection and promotion? 
-Consulting government 
policy document related to 
the issues addressed by the
project
-Key Interviews with 
Project Manager and/or 
Project Component 
Managers/M&E Manager
-Key interviews with key 
project partners (MAST, 
AL, ...)
-Key interviews with 
beneficiaries
Quantitative survey
A quantitative sampling survey was conducted among project beneficiaries in the Central Plateau
intervention communes (Cerca-la-Source and Cerca Carvajal) and Northeast intervention communes 
15
(Fort-Liberté, Carice, Mombin Crochu, Sainte Suzanne, Ferrier and Vallières). As mentioned in the 
ToRs, the same sampling methodology used for the baseline was applied. Thus, a multi-step
Proportional Probability by Size (PPS) Cluster Sampling was used. Although beneficiary lists exist, which 
means according to the FFP protocol, simple random sampling in one step would apply here. However, 
for the purposes of comparing initial and final indicator values, the same sampling method was used. 
However, the interviewers had the lists as a reference for accessing households in the program 
intervention communities.
The structured questionnaire used in the project baseline for the collection of quantitative data is used 
in this study with additions taking into account the evaluative questions based on OECD criteria. 
The questions take into account the calculation method for project’s key indicators as presented :
Key indicators Definition of indicators Data collection method
Percentage of targeted 
households with an 
acceptable Food 
Consumption Score (FCS)
The frequency-weighted dietary diversity score is 
calculated from the consumption frequency for the 
various food groups consumed (Appendix 1) by a 
household during the 7 days preceding the survey.
Quantitative survey of project 
beneficiary households
Prevalence of households 
with little or no hunger 
(Household Hunger Scale - 
HHS)
It is essentially a behavioral measure that tends to 
capture more serious behaviors such as: 
- Was there never anything to eat in your house? For
lack of resources to get food?
- Did you or any family member go to bed hungry 
because there wasn’t enough food?
- Did you or any family member go a whole day and 
night without eating anything because there was not 
enough food (Appendix II)?
Quantitative survey of project 
beneficiary households
Proportion of households 
consuming at least 6 food 
groups in the previous 
month 
Household Dietary Diversity Scale (HDDS): Dietary 
diversity represents the number of different foods or 
food groups consumed in a given reference period.
- Similar to the FCS, but usually with a 24-hour recall 
period with no information on frequency or weighted 
categorical thresholds 
- It is an indirect measure of household access to food.
Quantitative survey of project 
beneficiary households
Percentage of food use by 
type (household 
consumption, sale, 
exchange, livestock feed)
This indicator takes into account food generally 
available in the household, regardless of its origin and 
quantity. It analyzes the different modes of use, 
namely: 
a. Self-consumption
b. Sale 
c. Exchange
d. Livestock feed
The priority use of food can inform the economic 
situation of the household. 
Quantitative survey of project 
beneficiary households
16
Key indicators Definition of indicators Method of data collection
Percentage of households 
where adults and children 
consume at least 2 meals a 
day (adults and children), 
by age
The average number of meals consumed per day in a 
household is an indicator of the level of food security or 
food insecurity. 
The number of meals in food insecure households is 
lower than in food secure households. 
In fact, this indicator records the number of households 
where adults and children eat at least 2 meals a day.
Quantitative survey of project 
beneficiary households
Prevalence of mothers 
who exclusively breastfeed 
their infants six months of 
age or younger
This indicator takes into account the proportion of 
mothers whose children aged six (6) months or less 
were exclusively breastfed. 
In the Quantitative Household Survey, 
a subgroup of questions will be asked 
of households with mothers who have 
infants 6 months of age or younger. 
Reduced Coping 
Strategies Index (rCSI)
rCSI measures behavior: What people do when they 
don’t have access to enough food by answering the 
question: What do you do when you don’t have enough 
to eat and you don’t have enough money to buy food?
- Measures the adjustments households make in 
consumption and livelihoods. These may be changes in 
consumption, reduced spending, or income growth;
- rCSI tends to measure less severe coping behaviors.
- rCSI uses the five most common strategies with
standardized weights:
1- Focus on less preferred and less expensive foods?
2- Borrow food or rely on the help of a friend or 
relative? 
3- Limit portion size at mealtime? 
4- Restrict adult consumption so that young children can 
eat? 
5- Reduce the number of meals consumed per day? 
(Appendix III)
Quantitative household survey
Percentage of households 
reporting greater ability to 
cope with economic 
shocks.
This indicator takes into account households with 
information on shocks. The households most 
vulnerable to risks are those without information. 
The measurement of this indicator therefore has to do 
with two (2) essential elements:
a. The identification of shocks (floods, wind, drought, 
plant disease, erosion, hurricanes and others) and the 
level of participation of each type of shock in food 
insecurity; 
b. Information received by households for managing 
shocks.
Quantitative survey of project 
beneficiary households
Prevalence of global acute 
malnutrition (GAM)
To measure GAM, anthropometric measurements
(weight-for-height or MUAC) of children aged 0-59 
months will be taken to calculate wasting. All children 
with a weight-for-height Z-score of less than -2 
standard deviation and/or edema are classified as 
having global acute malnutrition.
Anthropometric data from 
quantitative surveys of beneficiaries 
with children aged 0-59 months.
- Calculation of sample size
The indicators are measured by department, i.e. for all the communes of a department 
taken together. Thus, the previously presented indicators will be established for the North-
17
East by considering the 6 project intervention communes and, for the Center, by considering 
the 2 intervention communes. 
- Calculation of sample size
Since we use the same baseline methodology, i.e. a 2-step Proportional Probability of Size 
(PPS) Cluster Sampling, to determine the size of the beneficiary sample, the calculation formula 
is based on the FFP/USAID protocol in that the minimum required number of 678 beneficiaries 
per department to be surveyed is met for the food security indicators (FCS, HHS, rCSI). 
The formula for calculating the survey’s sample size is the following: 
𝑛𝑛𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝐷𝐷𝑒𝑒𝑒𝑒
⎣
⎢
⎢
⎡𝑍𝑍1−∝ �2𝑃𝑃�1 − 𝑃𝑃� + 𝑍𝑍1−𝛽𝛽�𝑃𝑃1,𝑒𝑒𝑒𝑒 �1 − 𝑃𝑃1,𝑒𝑒𝑒𝑒 � + 𝑃𝑃2,𝑒𝑒𝑒𝑒 �1 − 𝑃𝑃2,𝑒𝑒𝑒𝑒 �
𝛿𝛿
⎦
⎥
⎥
⎤
2
Where 
𝑛𝑛𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = is the initial sample size required by surveys for each one of the two-stage points
𝛿𝛿 = 𝑃𝑃1,𝑒𝑒𝑒𝑒 - 𝑃𝑃2,𝑒𝑒𝑒𝑒 = minimum effect size to be achieved over the period specified by both
surveys; 
𝑷𝑷𝟏𝟏,𝒆𝒆𝒆𝒆𝒆𝒆 = 𝟎𝟎. 𝟒𝟒𝟒𝟒 proportion of the population with an acceptable FCS baseline. FCS is 
chosen over the other 2 as the one that yielded the largest sample, although insufficient for 
the measurement of food insecurity indicators.
𝑃𝑃2,𝑒𝑒𝑒𝑒 = 0.69 Proportion of the population with an acceptable FCS in 2nd MDP 
achieved
𝑃𝑃1,𝑒𝑒𝑒𝑒 = represents a survey estimate of the true proportion of the P1 population at baseline
𝑃𝑃2,𝑒𝑒𝑒𝑒 = represents a survey estimate of the true proportion of the P2 population at the end of 
the survey.
𝑃𝑃 =
𝑃𝑃1,𝑒𝑒𝑒𝑒𝑒𝑒+ 𝑃𝑃2,𝑒𝑒𝑒𝑒𝑒𝑒
2 =
0.46+0.69
2
=0.575 
𝑍𝑍1−∝ is the value of the normal probability distribution corresponding to a confidence level of 
1-𝛽𝛽. For 1-𝛽𝛽 =0.95, the corresponding value is 𝑍𝑍0,95 = 1,64.
𝑍𝑍1−𝛽𝛽 is the value of the normal probability distribution corresponding to a confidence level of 
1-.𝛽𝛽. 
For 1-𝛽𝛽 =0.80, the corresponding value is 𝑍𝑍0,80 = 0,84.
𝐷𝐷𝑒𝑒𝑒𝑒 The estimated Design Effect (DEFF) of the survey is equivalent to 2 here. 
Hence
18
𝑛𝑛𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 2 �
1.64 ∗ �2 ∗ 0.575(1 − 0.575) + 0.64 ∗ �0.46(1 − 0.46) + 0.69(1 − 0.69)
0.46 − 0.69 �
2
𝑛𝑛𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 123
There is indeed a non-response rate (generally set at 1.1, corresponding to 10%, but may 
change depending on the context).
𝑛𝑛𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 123 ∗ 1.1 = 135 𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏.
As with each department, to measure FCS, HHS, rCSI in this final evaluation, at 
least 678 units are required, so the sample size per department is adjusted to 678. 
Thus, overall, 1,356 beneficiary households in the project’s two intervention 
departments were to be surveyed in this final evaluation.
Table 1: Breakdown of the number of beneficiary households to be surveyed by commune 
Departments Communes Number of
households Demographic 
weight
Sample 
allocation by 
commune
Quantity of 
surveys 
carried out
Center 
Cerca-La 
Source 1,847 63% 427
425
Cerca￾Carvajal 1,086 37% 251
255
Total 2,933 100% 678 680
Northeast
Fort-Liberté 229 8% 52 96
Ferrier 506 17% 115 121
Carice 251 8% 57 51
Mombin 
Crochu 783 26% 178
121
Sainte 
Suzanne 705 24% 161
178
Vallières 503 17% 115 111
Total 2,977 100% 678 678
Grand Total 5,910 1,356 1,358
For the indicator on the prevalence of mothers who exclusively breastfeed infants six months 
of age or younger, an attempt was made to identify from the beneficiary list the number of 
households with mothers who breastfeed children from 0 to 6 months of age. This helped us 
extract a random sub-sample to have enough sample units for a more valid measurement of 
this indicator.
In the first PDM carried out in August by the project in the 2 departments, the findings showed 
a difference between the communes because, in Cerca-Carvajal, 14% of households have a 
pregnant woman, 13% for the commune of Carice, 12% for Vallières, 9% for Cerca-la-Source, 
5% for Ferrier, 4% for Mombin Crochu, 3% for Fort-Liberté and 2% for Sainte-Suzanne. 
Assuming these pregnant women have had live births in the meantime, we estimate the 
following sub-population of breastfeeding women per commune:
19
Table 2: Estimated sub-population of breastfeeding women beneficiaries by commune 
Departments Communes Number of
households
% of households 
with nursing 
women
Sub￾population of 
households 
with nursing 
women
Center Cerca-la-Source 1,847 9% 61
Cerca-Carvajal 1,086 14% 95
Total 2,933 156
Northeast
Fort-Liberté 229 3% 20
Ferrier 506 5% 34
Carice 251 13% 88
Mombin Crochu 783 4% 27
Sainte Suzanne 705 2% 14
Vallières 503 12% 81
Total 2,977 264
Grand Total 5,910 420
Considering the number of beneficiaries to be surveyed in the Center department in the 2 
communes (427 for Cerca-la-Source and 251 for Cerca-Carvajal, respectively), it was possible 
to exhaustively survey the estimated number of nursing women within the sample of 
beneficiaries to be surveyed per commune in this department. 
However, for the North-East department, the sub-population of nursing women in some 
communes exceeded the sample of beneficiaries to be surveyed (for example, 57 beneficiaries 
would be interviewed in Carice while the sub-population of nursing women was estimated at 
88 beneficiaries). For that purpose, we determined a sample of nursing women to be surveyed 
in the Northeast. Hence, with a margin of error set at 7% and the proportion of women 
practicing exclusive breastfeeding in the Northeast estimated at 54% (project baseline), we 
determined a number of n=123 breastfeeding women beneficiaries. Hence the breakdown by 
commune becomes:
Table 3: Breakdown of the sample of breastfeeding women to be surveyed by commune in the 
Northeast 
Departments Communes Sub-population of 
households with 
nursing women
Demographic 
weight
Sample 
allocation
Quantity 
of 
surveys 
carried 
out
Center
Cerca-la-Source 61 - 61 319
Cerca-Carvajal 95 - 95
148
Total 156 156 467
Northeast
Fort-Liberté 20 8% 9 60
Ferrier 34 13% 16 51
Carice 88 33% 41 35
Mombin Crochu 27 10% 13 81
Sainte Suzanne 14 5% 6 53
20
Vallières 81 31% 38 60
Total 264 100% 123 340
We noted that more nursing women were interviewed than the sub-population estimated 
from the PDM data. This may be explained by the fact that new nursing mothers were added 
to the list.
GAM indicator
To measure the GAM (Global Acute Malnutrition Prevalence) indicator, one of the 3 
approaches proposed by the FFP protocol is used (approach B). We looked at the official 
statistics in each of the 2 departments on the proportion of the population of children under 
5 years of age: 0.10 in the Northeast and 0.11 in the Center. Data from the PDM carried 
out in the project confirm an average size of 6 persons per household in the 2 departments. 
Using these parameters for each of the departments in the USAID sample size calculator5
, 
we have the respective sizes of children under 5 to consider for the GAM indicator 
measurement:
- n (Northeast) = 1,270 children 0-59 months old
- n (Centre) = 1,173 children 0-59 months old
These samples of children under 5 years of age were sought within the surveyed household 
samples. It should be noted that within the same household, more than one child under 5 
years of age could be considered. Therefore, the child samples were therefore broken down 
according to the demographic weight of households by commune. Hence the following table:
Table 4: Breakdown of Recipients Sample with Children 0-59 Months of Age to be Surveyed 
Departments Communes Demographic 
weight of 
households 
Sample 
allocation of 
children to be 
considered by 
commune
Quantity of 
surveys 
carried out
Center Cerca-la-Source 63% 739 331
Cerca-Carvajal 37% 434 169
Total 100% 1,173 500
Northeast
Fort-Liberté 8% 98 63
Ferrier 17% 216 54
Carice 8% 107 39
Mombin Crochu 26% 334 82
Sainte Suzanne 24% 301 73
Vallières 17% 214 61
Total 100% 1,270 372
Grand Total 2,443 872
5 https://www.usaid.gov/documents/1866/population-based-survey-sample-size-calculator
21
More than 50% of children could not be observed in the two departments during the 
interviewers’ visits for the calculation of GAM. This is therefore a considerable limitation to
this study (see details in the Constraints and Limitations of the Study section).
In order to quickly get quantitative data, the Ona digital survey platform was used. The survey 
questionnaire is designed using the xlsform syntax loaded on the platform. Each interviewer 
was therefore provided with a digital tablet on which the data collection tool was installed. It 
was therefore possible to follow, in real time, the entire collection process from the Ona 
platform. The interviewers were also required to enter the GPS coordinates of the locations 
visited; this reduced the risk of an interviewer filling out the tool him/herself without actually 
visiting the beneficiary household. 
Qualitative survey
The qualitative approach of the final evaluation is essential not only to complement the 
quantitative data collected from project beneficiaries, but also to validate the study’s findings 
through the triangulation of information. It is also important to help better identify the changes 
induced by project interventions. Thus, individual interviews were addressed to project
key actors:
Table 5: Number of Key Interviews by Department 
Key Interviews Center Northeast Total
Project team - Agriculture 
component
- - 1
Project team - Food 
vouchers component
- - 1
Project team - VSLA/S4T 
component
- - 1
Project team - M&E 
component
- 1 1
Project team – Manager
Local Authorities 
(CASECs-Mayors) -
Community Leaders
7 6 13
Departmental Office –
MARNDR
1 1 2
Departmental Office –
MAST
2 1 3
CEPAM Manager 1 1
Sellers 6 7 13
Beneficiaries 29 42 71
Other organizations in 
charge
- - 0
Since COVID-19 is still present in Haiti, focus groups with project stakeholders were 
avoided to prevent the spread of the disease. For this reason, key informant interviews were 
advocated instead of focus group meetings. 
22
Assessing the effects of the program on markets
An analysis of the program’s impact on the markets was carried out in a qualitative manner, 
through interviews with the project’s main wholesalers and retailers, and with beneficiaries 
through separate key interviews (with men and women as their perceptions may be different). 
The key issues that were discussed were the following: 
• Has the market experienced significant variations (in availability, demand, prices, 
players present) since the program’s inception? And if so, what are the causes? How 
has the program responded to seasonal price fluctuations, volatility and possible 
differences? 
• Have challenges/constraints/distortions that could be attributable to the program been 
identified? Has information about the program been properly communicated? What 
market effects/impacts persist beyond the program’s closure? 
• Is the impact of the program on traders noticeable? Has it helped boost business 
activities and strengthen the capacity of traders to develop their business? What 
changes, if any, in marketing strategies/practices were observed among consumers and 
traders before and after the program? What are the main lessons learned from this 
program and what are the recommendations for the continuation of similar activities 
in the context surveyed? 
Evaluation of the effectiveness and relevance of the transfer modalities
The team conducted a qualitative analysis of the effectiveness of transfer modalities that helped
the project achieve or not achieve expected objectives. Discussions were conducted with 
program beneficiaries, stakeholders, the implementation team, wholesalers and retailers. This 
evaluation considered the following pillars:
1) Relevance: To what extent does the transfer modality used meet beneficiary needs 
while being gender-sensitive and based on a gender analysis? 
2) Effectiveness: To what extent were the project outcomes affected by the chosen 
transfer modality?
3) Efficiency: Was the transfer modality cost-efficient and implemented in a timely 
manner? What is the cost-effectiveness of the respective transfer modalities compared 
to other alternatives?
4) Constraints and Limitations of the Study 
The data collection operations took place in a difficult context where interviewers had to 
work under heavy rains. This particularly delayed data collection activities in the Center 
department. Interviewers had to end working days prematurely because of the long rains.
The methodology of the quantitative survey indicated the measurement of the weight and 
height of children aged 0-59 months to calculate the malnutrition indicator (prevalence of 
23
global acute malnutrition (GAM)). This required a heavy logistical burden where interviewers 
had to carry a scale and a height gauge. In the consultation proposal, we were counting on 
World Vision and health centers in the intervention communes to support us with these tools,
given the difficulty of getting them easily on the market, for the sole use of a survey. Although 
we were able to access the required number of height gauges for interviewers, this was not 
the case for scales. This had a negative impact on the survey, especially on the number of 
children to be observed for the sample calculated for this indicator. Therefore, one scale was 
used per commune in the North-East department. In the Center, the leadership of our field 
supervisors helped us borrow some scales from health centers.
It was also not possible to identify in the sampling frame, i.e. the lists of beneficiary households, 
all those with children aged 0-59 months. This also explained this limitation of the study in 
terms of reaching the children sampled to be considered by department to calculate the GAM.
Because the project financial report was not available, the efficiency analysis did not touch on 
aspects related to the percentage of the budget actually spent versus what was planned. 
5) Analysis of the Final Evaluation’s Findings 
In this section, the findings of the final evaluation are analyzed by first highlighting the general 
characteristics of the beneficiary households, then, OECD criteria (Relevance-Efficacy￾Efficiency-Impact-Sustainability), links, and exit strategies by answering evaluation questions.
5.1 General Characteristics of Beneficiary Households 
It is essential to characterize the universe of the study’s surveyed beneficiary households. In 
order to get an idea of the structure of beneficiary households in the project intervention 
departments, an age pyramid was designed for each one of them (Figures 1 to 2). The 
visualization of both graphs reveals almost the same structure for the beneficiary population 
in the two departments, i.e., very broad at the bottom and very tapered at the top. This is 
therefore an evidence of two young populations, which then reflects the same situation of the 
Haitian population in general. The most represented age groups are the 5-19 year old. In both 
departments, there tends to be a stronger presence of the female population. This reflects the 
reality of the Haitian population, which has more women than men. 
24
Graph 1 : Age pyramid for the intervention communes 
of the EFSP Center project 
Graph 2 : Age pyramid for the intervention communes 
of the EFSP North-East project 
There is a huge difference between the 2 departments in terms of the gender of respondents 
to the quantitative survey. Indeed, most respondents in the Northeast are women (71%) while 
this is the reverse in the Central department, where more than 50% of respondents are men 
(Graph 1). This difference shows slightly at the household status level where, in the 
Northeast, there are more female-headed households (FNM), i.e. 26% (Graph 2). Adult 
female- and male-headed households (F&M) are dominant in this survey, 93% and 64% 
respectively in the Center and Northeast.
Graph 1: Distribution of Respondents by 
Department by Gender (n=1,258) 
Graph 2: Distribution of Respondents by 
Department by Household Status (n=1,258) 
15.00% 10.00% 5.00% 0.00% 5.00% 10.00% 15.00%
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85+
F H
10.00% 5.00% 0.00% 5.00% 10.00%
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85+
F H
42%
71%
58%
29%
Center Northeast
Woman Man
4%
26%
3% 9%
93%
64%
Center Northeast
Adult female and non-adult male (FNM)
Male adult Female non adult (MNF)
Adult men and women(F&M)
Non-adult child led household (CNA)
25
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
If we now analyze the education level of project beneficiaries, we identify the same trend in 
the 2 departments, namely a predominance of “no education” and “incomplete primary 
education”, for between 30% to 50% of respondents (Graph 3). This corresponds to this 
known reality in Haiti where a large portion of the population does not always have access to 
education. It is often this population group that is abandoned in the provinces, especially in 
rural areas. Today, these rural areas still lack, both in quantity and quality, educational 
institutions that should help people develop their potential. It can even be argued, from this 
reality, that this somehow confirms these respondents come from the most vulnerable groups 
in their communities, with such a low overall level of education. Generally, the lack of 
economic resources is the reason why parents are unable to provide their children with 
education. This is a still ongoing situation in Haiti, particularly in rural areas.
Graph 3: Distribution of heads of household by department according to their education level 
(n=1,258) 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
When considering marital status in analyzing the characteristics of project beneficiary 
households, cohabitation and marriage are the most represented in the 2 departments. This 
does not contradict the trend in marital status we see at the national level. 24% of respondents 
with a widow(er) status is a large proportion among beneficiaries in the Northeast (Graph
6). And we see the consistency with the presence of orphans in households in the Northeast, 
which has slightly more orphans compared to the Central department (Graph 7). 
32%
43%
11%
9%
48%
31%
8%
13%
2%
1%
Center
Northeast
Incomplete academic
Full academic
Incomplete secondary
Secondary incomplete
Professional
Incomplete Primary
Full Primary
None
26
Graph 4: Distribution of heads of household by 
department according to marital status 
(n=1,258) 
Graph 5: Distribution of respondents by 
department according to whether or not there 
is at least one orphan in the household
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
We found virtually no household with pregnant women (Graph 8). 10% or less respondents 
only mentioned it in the Northeast or Center. This is also the case for the presence of persons 
with disabilities in beneficiary households, where only less than 15% of respondents reported 
this, in both departments (Graph 9). This must have been one of the vulnerability criteria for 
the selection of beneficiaries. This aspect is further analyzed in the evaluation criteria section. 
Graph 6: Distribution of respondents by 
department according to whether or not 
there is a pregnant woman in the household 
Graph 7: Distribution of respondents by 
department according to whether or not 
there is at least one person with a disability in 
the household 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
2%
7%
47%
31%
1%
5%
41%
34%
9%
24%
Center
Northeast
Widow(er)
Married
Divorced/Separ
ated
Concubinage
Single
86%
80%
14%
20%
Center Northeast
No Yes
89% 92%
11% 8%
Center Northeast
No Yes
90% 86%
10% 14%
Center Northeast
No Yes
27
The majority of the households benefiting from the project live in tin-roof housing (Table 6) 
(38% in the Center and 73% in the Northeast) and simple low-rise houses (33% in the Center 
and 6% in the Northeast). The type of housing is a good indicator of the household’s level of 
vulnerability. It is evident here that we do not see the worst types of housing such as cottages
and hovels/wooden huts. 
Table 6: Distribution of Respondents by Department by Housing Type (n = 1358) 
Type of habitat Center Northeast Grand Total
Hovels 4% 2% 3%
Other (wooden with tin 
roof, wooden with mud, ...) 1% 4% 2%
Two-story house 0% 1% 1%
Simple low-rise house 33% 6% 19%
Masonry house 0% 1% 1%
Mud house 14% 5% 10%
Tin-roof house 38% 73% 56%
Hut/Wooden hut 9% 7% 8%
Grand Total 100% 100% 100%
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Agriculture and trade are the main sources of income for most rural households in Haiti 
(FEWS NET6
, March 2015). The households benefiting from this World Vision program fall 
into these same livelihood categories. 85% and 61% of respondents, respectively in the Center 
and Northeast, reported agriculture as their main activity (Table 7). Trade is the second most 
important livelihood activity, although in relatively small proportion compared to agriculture.
Table 7 : Distribution of respondents by department according to the head of household’s main 
activity 
Main activity
Department
Center Northeast
Grand 
Total
Agriculture 85% 61% 73%
None 4% 17% 11%
Charcoal (production) 0% 4% 2%
Trade 7% 11% 9%
Breeding 1% 0% 1%
Mason 1% 1% 1%
Teacher 0% 1% 1%
Transportation 1% 1% 1%
Other profession (ironwork, cabinet making...) 1% 4% 2%
Grand Total 100% 100% 100%
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
As a second activity, in addition to trade and agriculture, there is animal breeding, which 
appears to be very present as a means of livelihood in the program intervention areas (Table 
8). This activity has long been considered the “piggy bank” of rural households. 
6 Haiti, Rural Livelihoods Profile, FEWS NET, March 2015.
28
Table 7: Distribution of respondents by department according to the second activity of the head 
of household 
Second activity
Department
Center Northeast
Grand 
Total
Agriculture 11% 12% 12%
No second activity 39% 63% 51%
Other profession 2% 2% 2%
Charcoal (production) 4% 7% 5%
Trade 12% 8% 10%
Breeding 28% 6% 17%
Mason 1% 1% 1%
Professor 1% 1% 1%
Transportation 1% 0% 1%
Grand Total 100% 100% 100%
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
The previously presented economic activities provide monthly income brackets from 1,500 to 
5,500 HTG. In the Rural Livelihoods Profile study cited above, the annual income of poor rural 
households is estimated at 50,000 HTG, or about 4,167 HTG monthly (FEWS NET7
, March 
2015). But households in the Northeast show lower income brackets that come close to being
unrealistic, less than 1,500 HTG (Graph 10). Given that beneficiary households are supposed 
to be among the most vulnerable in the project’s intervention areas, this may explain these 
lower income brackets.
Graph 8: Distribution of respondents by department by monthly household income group 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
The trend in expenditure items for beneficiary households is the same in both departments. 
Three main expenditure items emerged for beneficiary households: food, education (school 
fees) and health care (Graphs 11 and 12). This is, in fact, the same reality at the national 
level where the State is unable to protect households from these recurring costs related to 
7 Haiti, Rural Livelihoods Profile, FEWS NET, March 2015.
33%
42%
40%
29%
19%
12%
2%
2%
1%
3%
3%
8%
1%
5%
Center
Northeast
More than 11500 Gdes
9500 - 11500 Gdes
7500 - 9500 Gdes
5500 - 7500 Gdes
3500 - 5500 Gdes
1500 - 3500 Gdes
0 - 1500 Gdes
29
household size. Poor households are inclined to spend their meager economic resources on
food.
Graph 9: Distribution of respondents by department 
by first annual household expenditure item 
Graph 12: Distribution of respondents by department by 
second annual household expenditure item 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
5.2 Analysis of Project Relevance 
In this section, the project is analyzed based on the various evaluation questions related to the 
relevance criterion. The overall idea of this criterion is to see the extent to which the project’s 
objectives have remained valid while considering the needs or priorities of target communities. 
In other words, it was essential to look at whether activities and outputs were consistent with 
expected impacts and effects. The broad lines of planned activities for the project were linked 
to these five main components, namely :
- Distribution of Food Vouchers 
- Nutrition
- Drinking Water Infrastructure
- Agriculture
- Financial Support (Training and S4T)
a) Distribution of Food Vouchers
The interview with the project team revealed that while implementing the Distribution of 
Food Stamps component, everything was set up to reach the most vulnerable. Thus, 
beneficiaries were selected according to a participatory process, using criteria identified by 
the communities themselves, namely: 
• Children under five years of age suffering from malnutrition. 
• Households with physical disabilities. 
• Pregnant and lactating women. 
1%
2%
0%
9%
9%
0%
1%
87%
85%
3%
3%
0%
0%
Center
Northeast
Other (energy,...)
Health Care
Food
Rent
Tuition fees
Toiletries and
laundry items
Agriculture
11%
3%
3%
3%
60%
54%
0%
2%
10%
13%
13%
21%
1%
1%
2%
3%
0%
1%
Center
Northeast
Other (home repairs,
etc.)
No other activity
Clothing
Health Care
Food
Rent
Tuition fees
Toiletries and laundry
items
Agriculture
30
• Households headed by elderly people. 
• Households caring for orphaned children. 
• Households headed by a woman. 
• Households that are hit and struggling until exhausted, following hurricane Irma. 
It was not clarified whether the project had conducted a needs assessment prior to evaluating 
the voucher value. However, the implementation team made monthly adjustments to 
vouchers, in line with the increase in food prices in local markets, due to the depreciation of 
the Gourde and rising inflation. 
Women heading single-parent households have been one of the project’s and the food voucher 
component’s main targets to ensure these vulnerable households have access to a reliable 
food source. 
b) Nutrition
The Nutrition component was implemented by CEPAM (Center for the Promotion of 
Breastfeeding), a World Vision partner. The interview conducted with a representative from
this CEPAM revealed that the identified geographic areas of implementation are those with 
the highest proportions of vulnerable people affected by hurricane Irma. However, CEPAM 
states that the WV team conducted local discussions to select the nutrition theme since it was 
noted there was limited knowledge about good nutritional practices within communities and 
more specifically with pregnant and lactating women and for cases of malnutrition. 
Interventions under this component are in line with the National Nutrition Policy (NNP) in 
the sense that they aim to improve the nutritional situation of communities based on a 
preventive approach to all forms of malnutrition through the promotion of healthy, nutritious 
and adequate food (MSPP8
). According to CEPAM, the communities were well informed of 
the activities and interventions on nutrition, the mother leader strategy, the expected 
deliverables and expected behaviors. 
Indeed, the quantitative survey data confirm that beneficiary households in both intervention 
departments indicated they received messages related to newborn care from at least 90% of 
respondents (Graph 13). In addition, between 42% and 47% of respondents (women) said 
they attended trainings conducted by mother-leaders in their community (Graph 14).
8 https://scalingupnutrition.org/wp-content/uploads/2013/06/Haiti_Plan-Strategique-Nutrition-2013-2018.pdf 
31
Graph 10: Distribution of beneficiary 
households by department according 
to whether or not they received 
messages about newborn care in the 
health centers (n=618) 
Graph 11: Distribution of beneficiary households 
according to whether they participate in training 
provided by mother leaders in the community 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
c) Drinking Water and Sanitation Infrastructure
In this component, the project’s objective was to increase or restore access to drinking water 
for communities whose water systems were damaged after the passage of hurricane Irma. 
Thus, the quantitative survey showed a much better situation in the Center compared to the 
Northeast in terms of the percentage of households with access to an improved water source 
(private connection, private kiosk, public water point, etc.). Nearly 35% of beneficiary 
households in the Northeast use uncollected water sources (Table 9).
Table 8: Main source of water for household domestic use 
Household water source
Department
Center Northeast
Grand 
Total
Private connection 3% 1% 2%
Private water sales kiosk 0% 2% 1%
Public water point (DINEPA) 62% 25% 44%
Public pump 4% 31% 17%
Uncollected water source 29% 35% 32%
Impluvium/Rainwater 0% 0,3% 0%
Protected wells 1% 2% 2%
Unprotected wells 0% 2% 1%
Truck/Tanker (Untreated Water) 0% 0,1% 0%
River 0% 1% 1%
Grand Total 100% 100% 100%
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020 
According to DINEPA (Direction Nationale d’Eau Potable et Assainissement) standards, 
access to drinking water also takes into account the amount of water available per person per 
10%
9%
90%
91%
Center
Northeast
Yes
No
58%
53%
42%
47%
Center
Northeast
Yes
No
32
day and the time taken to travel to and from distribution points. Thus, a person is said to have 
access to a Potable Water Supply/Water Point (PWS/WP) in the sense that he or she can 
access it in less than 30 minutes round-trip. The system provides at least 15 liters of water 
per person per day at water points, standpipes and kiosks, and at least 25 liters at private 
connections. To get water volumes in gallons, simply multiply the liter amount by 0.264 gallons. 
So, the quantitative data shows that in the Central and Northeast regions, respectively 57% 
and 45% (Graph 15) of respondents indicate they have between 10 and 20 gallons of water 
available per day for use. In addition, more than 50% of respondents in the two departments 
only take 5 to 15 minutes (Graph 16) for the round trip when collecting water.
Graph 12: Distribution of beneficiary 
households by department according to the 
amount of water available for use per day. 
Graph 13: Distribution of beneficiary 
households by department according 
to the walking time (round trip) for 
water collection 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 
2020.
To ensure the sustainability of water service, the project should also work to strengthen the 
management and maintenance of water points in the communities. The quantitative survey 
revealed that very few beneficiary households are aware of the creation of water committees 
(Table 10).
Table 9: On the management and maintenance of water points 
Center Northeast
Number of 
people %
Number of 
people %
Aware of the creation of water committees 
194 29% 59 9%
24%
35%
57%
45%
19% 21%
Center Northeast
More than 10
gallons
Between 10
and 20 gallons
Less than 10
gallons 53% 52%
19% 26%
13%
11%
8% 7% 7% 3%
Center Northeast
More than 1
hour
45 mns to 1 hr
30 to 45 mns
15 to 30 mns
5 to 15 mn
33
Member of the water committee?
23 3% 7 1%
Participating in pump repair activities
36 5% 20 3%
Member of the technician groups 
7 1% 13 2%
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
This component also sought to strengthen the hygiene conditions of the project’s target 
households by working primarily on home water treatment but also on handwashing. Indeed, 
in the quantitative survey for this final evaluation, the majority of respondents stated they treat 
water at home (59% in the Northeast and 71% in the Center (Graph 17)). Those who don’t 
usually treat water at home say they don’t have water treatment products. In the Northeast 
and Center, more than half say they already have clean water (Graph 18).
Graph 14: Distribution of beneficiary 
households by department according 
to whether or not they usually treat 
water at home 
Graph 15: Distribution of beneficiary 
households by department according to the 
fact that they do not usually treat water at 
home 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Chlorine and a derivative of this product, Aquatabs, are widely used as home water 
treatment methods both in the Northeast and Center (Graph 19). In the Center, they use
another home water treatment product that is manufactured in Haiti. It should be verified 
whether there is a production unit for this product in this department; this would ensure its 
availability in a sustainable way for households. 
29%
41%
71%
59%
Center
Northeast
Yes
No
51%
9%
47%
77%
1%
9%
Center Northeast
Clean water
Does not know
how to treat
water
No water
treatment
products
Products or
processing
equipment too
expensive
Other( discomfort
with treated
water...)
34
Graph 16: Distribution of beneficiary households by department according to the water 
treatment method used 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Treating water at home is one thing, but storing to keep it clean is, in many cases, a real 
challenge for rural households. In this study, the water storage containers reported by 
beneficiary households were a covered bucket and a gallon in both departments (Table 11). 
Table 10: Distribution of beneficiary households by department according to the water 
conservation method used at home 
Water storage container
Center Northeast
Number % Number %
Covered Drum 101 15% 56 8%
Non-Covered Drum 53 8% 15 2%
Covered Bucket 585 86% 493 73%
Non-Covered Bucket 131 19% 109 16%
Gallon 555 82% 527 78%
Bowl 47 7% 24 4%
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Under the COVID-19, more than 80 percent of respondents to the quantitative survey 
reported receiving handwashing messages (Graph 20). The project took advantage of the 
need brought about by the pandemic to get handwashing messages across. Interviews with 
local leaders acknowledged that hygiene training and awareness raising have increased 
community understanding of the essential concepts of good health practices related to water 
storage and treatment, the use of latrines to reduce the spread of cholera or eradicate it, and 
handwashing, which helped better appreciate and apply these principles under COVID. 
27%
5%
36%
54%
24%
22%
16%
2%
10%
Center
Northeast
Jif/Aquajif
Reverse Osmosis
Snowshoe
Filter
Aquatab
Boiling water
SODIS (in the sun)
Chlorine grain
Lemon
Gadyen dlo
35
Graph 17: Distribution of beneficiary households by department according to whether or 
not they received handwashing messages 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
d) Agriculture
In this project’s agricultural component, training and technical support on best land 
conservation practices were provided to smallholder beneficiaries in target communities. For 
the project implementation team, these best practices corresponded to the needs of farmers 
as the farms in project areas are vulnerable to climate shocks and changes. 
According to the National Adaptation Action Plan (NAAP), agricultural production in the 
Center and Northeast is particularly vulnerable to climate shocks, mainly drought, high winds 
and adverse rainfall, given the significant increase in land aridity in the area. The weather 
changes, with regard to the seasons’ characteristics and manifestation period, disrupt farmers 
in their monitoring methodology for better production-related decision-making and make the 
agricultural system even more vulnerable. Crops are lost and create a deficit in the 
communities’ ability to meet the food needs of its people (NAAP9
). Farmers have also received 
post-harvest management training because, without adequate large-scale storage 
infrastructure and conditions for the conservation of agricultural commodities, they 
periodically experience losses ranging from 30% for cereal crops to 50% for fruits and 
vegetables (University of Florida10). Considering that the farms are small estates of less than 1 
hectare, these losses are huge and affect the farmers’ economic life. Therefore, WV, by 
facilitating these trainings, seed distribution and post-harvest support, supports farmers in
adapting their production to recurring climatic hazards. Indeed, the respondents (beneficiaries) 
stressed the importance of the technical training received to tackle the water deficit affecting, 
in 2019, the departments of Northeast, North and Northwest (FAO, 202011).
9 https://unfccc.int/resource/docs/napa/hti01f.pdf 
10 http://blogs.ifas.ufl.edu/area/files/2018/05/fr-factsheet-postharvest-loss-04-25-18-final.pdf 11 http://www.fao.org/fileadmin/user_upload/FAO-countries/Haiti/docs/Apercu_de_la_réponse_de_la_FAO.pdf 
5%
12%
95%
88%
Center
Northeast
Yes
No
36
e) Financial support (training and S4T)
In this component, the project sought to strengthen community savings groups, which are 
solidarity-based financial tools based on trust between the people from the intervention 
localities. According to the program team, the actions related to this component were in line 
with the needs and priorities of the beneficiaries who were part of the savings groups, meaning 
that they were able to save part of their meager income among themselves (as we saw earlier, 
almost all of which is spent on food), since they received a monthly food voucher as part of 
the food voucher component.
The credit aspect worked well in the operation of community savings groups, according to the 
project team. Beneficiaries were able to either set up a small trade or strengthen an existing 
business, breaking away from the local system of usurious loans (called ponya in Creole, 
meaning “stab”) charging monthly interests of 25-50%. This, in many cases, does not offer a 
significant profit margin to the lenders. On the contrary, within savings groups, interest rates 
are set by the group through consultation and hardly exceed 5% a month, allowing more 
flexibility, less pressure and a double return on investment for members. In addition, according 
to beneficiaries, due to the lack of banking services available nearby, these groups help them 
save and have their money available in a timely manner for activities ranging from agricultural 
production, livestock breeding, payment of school fees and care for the sick. But, with the 
arrival of the COVID-19 pandemic, the process stopped as groups could no longer meet.
Overall, one can conclude that the program was relevant, since it was a direct response to 
support the communities affected by Hurricane Irma, which left tens of thousands of families 
suffering, mainly in the larger North, and caused partial flooding by runoff in twenty-two (22) 
communes of the North, North-East, North-West, Center, Artibonite and West departments, 
bringing about floods in major rivers or marine submersion (Haiti en marche)
12. This is what 
community leaders and local authorities admitted in interviews conducted with them in the 
various communes where the project is operating. However, they are not very clear on the 
process to identify the project’s intervention pillars, since they were not contacted by the 
program team nor did they take part in participatory needs assessment exercises. However, 
they emphasized that program interventions and pillars responded to the needs of the 
vulnerable and Irma-affected populations. 
Overall, WVI did a good job communicating program’s objectives and implementation 
strategies with leaders and communities during the start-up phase through meetings, field 
agents and officers in various public spaces such as churches and local organizations’ premises. 
Community representatives and local elected officials took part in these start-up meetings, but 
in the Northeast (specifically in Fort-Liberté and Mombin-Crochu) leaders are advocating for 
greater inclusion of local leaders and the population. There were not enough information￾sharing meetings conducted during the project’s implementation in the two departments, 
contrary to the plans, and this considerably limited local actors’ understanding of the project 
and, specifically, they were unable to share their advice, grievances and recommendations 
together. 
12 https://www.haitienmarche.com/index.php?option=com_content&view=article&id=2001:bilan-encore-partiel-des-degats￾causes-par-le-par-irma-en-haiti&catid=18&Itemid=268 
37
According to community and local leaders, the project team did not contact or liaise with 
other organizations that conducted similar projects in the intervention areas. They even took 
as an example a CARE-implemented response project to hurricane Irma. This limited the WVI 
team’s ability to avoid duplication and overlaps, while capitalizing on lessons learned. Thus, 
during project implementation, one of the coordination-related guiding principles was not 
applied by WV. According to the OECD, coordination within the overall humanitarian cash 
response is essential to avoid aid duplication or gaps, ensure information sharing and 
standardize approaches. Coordination must also happen among all stakeholders, including local 
and national actors and governments13. 
Clearly established criteria allowed for the participation of different vulnerable groups in the 
affected communities, according to the leaders interviewed in Cerca-la-Source. Some stressed 
the fact that many people with disabilities were not able to participate in the project. They are 
not clear why. No case of payment for participation in the project was noted by local leaders.
Data from the quantitative survey show satisfaction levels ranging from satisfied to very 
satisfied for operational components such as: agriculture and seed voucher, food voucher, 
community work in exchange for food vouchers, mother-leader nutrition and Savings and 
Credit Group (S4T) (Table 12).
Table 11: Level of satisfaction by activity in which respondent took part 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
In order to score compliance with this criterion, following the previous analysis, we consider 
for each criterion this satisfaction scale: 1=Not satisfied at all 2=Not satisfied 
3=Moderately satisfied 4=Satisfied 5=Very satisfied. Thus, for relevance, the project got a 
score of 4: satisfied for this criterion.
5.3 Analysis of Project Effectiveness 
Analyzing the effectiveness of a project during its evaluation helps see the extent to which 
project objectives were achieved. For that purpose, impact and effect indicators were analyzed 
13 https://www.oecd.org/development/humanitarian-donors/docs/cashbasedresponse.pdf
Project activities Number
More or 
less 
satisfied Satisfied
Very 
satisfied
Agriculture and seed voucher Center 56 2% 57% 41%
Northeast 103 0% 23% 77%
Food voucher Center 298 1% 45% 54%
Northeast 423 2% 48% 49%
Community work in exchange for a food 
voucher
Center 436 0% 60% 40%
Northeast 263 0% 40% 60%
Mother Leader-Nutrition Center 10 0% 70% 30%
Northeast 57 5% 42% 53%
Savings and Credit Group (VSLA/S4T)
Center 17 0% 47% 53%
Northeast 7 0% 57% 43%
38
through the quantitative survey conducted among beneficiary households, namely:
- Household Dietary Diversity Score (HDDS), 
- Household Food consumption score (FCS), 
- Household Hunger scale, 
- Reduced Coping Strategies Index (rSSr/rCSI)
- Percentage of households where adults and children eat at least 2 meals a day 
(adults and children), by age,
- Percentage of food utilization by type, 
- Percentage of households reporting greater ability to cope with economic 
shocks 
- Prevalence of mothers who exclusively breastfeed infants six months of age or 
younger, 
- Prevalence of global acute malnutrition (GAM)
a) Household Dietary Diversity Score
Household dietary diversity for this assessment is calculated from the number of different food 
groups consumed (nine groups in total) by households in project intervention area over the 
four weeks prior to the survey. According to survey results, over the whole project 
intervention area (the two departments), nearly 88% of surveyed households are in the high 
dietary diversity, i.e. 1,193 out of 1,358 households surveyed for 6 or more food groups during 
the 4 weeks prior to this study. Only 1.6% of them have a low or poorly diversified diet. At 
the departmental level, the findings related to dietary diversity for surveyed households in the 
Center are more satisfactory than those in the Northeast. Indeed, in the Center, more than 
90% of surveyed households have a diversified diet compared to 84% in the Northeast. 
However, it should be noted that the percentage of households with average dietary diversity 
in the Northeast is twice as high as in the Center. In both departments, the fat-based food 
group is the most consumed, followed by the cereal-based food group, followed by the 
beverage food group.
Table 12: Dietary Diversity Score by Department 
Indicator: Dietary Diversity
Center Northeast All Departments
Num14 Den % Num Den % Num Den % 
Low dietary diversity 5 680 0.70% 17 678 2.5% 22 1358 1.6%
Average dietary diversity 48 680 7.1% 95 678 14.0% 143 1358 10.5%
High dietary diversity 627 680 92.2% 566 678 83.5% 1193 1358 87.9%
Average Dietary Diversity 
Score 8 7 8 
Standard Dietary Diversity
Deviation 1.6 1.62 1.60
14 Num=Numerator 
 Den=Denominator 
39
Confidence Interval at 95% of 
the Average [7.88; 8.12] [6.88; 7.12] [7.91;8.09 ]
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
b) Household Food Consumption Score (HCS/FCS)
For this study, the household food consumption score takes into account dietary diversity, 
frequency of food consumption and the relative nutrient intake of different food groups. It is 
calculated based on the consumption frequency for the different food groups consumed by a 
household during the 7 days prior to the study. Taking into account the threshold set to
interpret the results for this indicator, over the whole project intervention and among the 
1,358 surveyed households, 195 or 14.3% of them have a low food consumption, 381 or 28.1% 
of them have a borderline-acceptable food consumption, and 782 or about 58% of them have 
an acceptable food consumption. As observed in analyzing the dietary diversity score, the same 
picture emerges at the departmental level for the household food consumption score. The 
Center, once again, shows better results than the Northeast. Nearly one-third of surveyed 
households in the Center have food consumption that is either poor or borderline acceptable, 
while, in the Northeast, only half of them are in this situation.
Table 13: Food consumption score for project beneficiaries by department 
Indicator: Dietary 
Diversity
Center Northeast All Departments 
Num15 Den % Num Den % Num Den % 
Poor food consumption 
(Percentage of households 
with FCS of 0-21)
44 680 6.5% 151 678 22.3% 195 1358 14.3%
Borderline Food 
consumption (Percentage 
of households with FCS of 
21.5-35)
183 680 26.9% 198 678 29.2% 381 1358 28.1%
Acceptable food 
consumption (Percentage 
of households with SCC> 
35)
453 680 66.6% 329 678 48.5% 782 1358 57.6%
Average Food 
Consumption Score 43.68 35.82 38.76
Food consumption
standard deviation 17.19 16.82 16.96
Confidence interval at 95% 
of the average [42.39 ; 44.97] [34.55 ; 37.09] [37.86 ; 39.66 ]
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
Situation of beneficiary households in relation to hunger 
To analyze the situation of households in relation to hunger, three indicators were selected. 
These are mainly the Household Hunger Index (HHS), survival (coping) strategies-rCSI and 
whether the household uses their production or not.
15 Num=Numerator 
 Den=Denominator 
40
c) Hunger Scale Score/Household Hunger Index (HHI/HHS)
This indicator is used to provide information on an important dimension of food security, 
namely access to sufficient food in the household. Thus, according to the findings of this study, 
80% of surveyed households are either severely or moderately hungry in both departments of 
the project’s target areas. At the departmental level, there is no real difference in the hunger 
scale score for both departments. However, contrary to the results for the two previous 
indicators, the Northeast shows better results for this indicator. In the Northeast, 22.3% of 
surveyed households have access to sufficient food or experience no or mild hunger; however, 
in the Center, these people represent about 18%. In the Northeast, about 48% experience a 
situation of severe hunger, while, in the Center, 53% experience severe hunger.
Table 14: Hunger score for project beneficiaries by department 
Indicator: Hunger Scale 
or Household Hunger 
Score (HHS)
Center Northeast All Departments 
Num Den % Num Den % Num Den % 
No or mild hunger in 
households (scores 0-1) 121 680 17.8% 151 678 22.3% 272 1358 20.0%
Moderate household 
hunger (scores 2-3) 199 680 29.3% 198 678 29.2% 397 1358 29.0%
Severe household hunger 
(scores 4-6) 360 680 52.9% 329 678 48.5% 689 1358 51.0%
Average Household Hunger 
Index
4 5 4 
Household Hunger Index 
standard deviation
2.4 2.43 2.44
Confidence interval at 95% 
of the average
[3.82;4.18] [4.82;5.18] [3.87;4.13]
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
d) Reduced Coping Strategies Index (rCSI)
The average score of the reduced coping strategy index calculated is 16.84 and 12.57, for 
beneficiaries in the Center and Northeast, respectively (Table 16). The higher the index 
score, the more food insecure the household in question is, meaning that the household uses 
negative strategies more frequently and/or that the strategies used are more severe compared 
to a household with a lower score. The maximum score for the index is theoretically 56. The 
average found for the index for the whole is in the very low rCSI modal class (4-18), i.e. 43% 
of households at the national level, from the last National Emergency Survey on Food and 
Nutrition Security (ENUSAN, MARNDR/CNSA, 2019). It is evident that the upper limits of 
the confidence interval for the estimate of the average calculated by department do not reach 
19. More beneficiary households use negative strategies in the Center compared to those in 
the Northeast.
41
Table 15: Reduced Coping Strategies Index (rCSI) 
Reduced Coping Strategies
Index (rCSI)
Center Northeast All Departments
Average 16.84 12.57 14.70
Standard deviation 12.84 13.13 12.98
Confidence interval of the 
average [15.87;17.81] [11,58 ; 13,56] [14.01 ; 15.39]
Median 12 12 12
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
Graph 18: Range for reduced coping strategies index (rCSI) 
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
e) Percentage of households where adults and children eat at least 2 meals a 
day (adults and children), by age
The average number of meals a day in a household is considered an indicator of the food 
security or food insecurity level. The percentage of households with children aged 6-23 
months who ate at least two meals a day is 17.8% in the Center compared to 15.8% in the 
Northeast and 16.8% in the whole project intervention area. For households with children 
aged 24-59 months, 35.6% of them eat at least two meals a day in the Center, compared to 
25.2% in the Northeast, and 30.4% in all target areas. Households with children aged 5 to 18 
years who eat at least two meals a day account for 54.7% in both departments (70.3% in the 
Center and 44.4% in the Northeast). As for adults within households, for the 19 to 59 years
age group, Table 17 shows a percentage of 70.3%, 50.0% in the Northeast and 60.2% for 
those who eat at least two meals a day. Overall, for both children and adults, the percentage 
of households eating at least two meals a day is 72.9% in the Center versus 49.6% in the 
Northeast, or 61.3% overall. 
41%
25%
33%
51%
25% 23%
Center Northeast
Less severe(<4)
Moderate(4-18)
Severe(>=19)
42
Table 16: Percentage of households where adults and children eat at least 2 meals a day (adults 
and children), by age 
Number of meals a day by age 
category
Center Northeast Department Set
Num Den % Num Den % Num Den %
Percentage of children aged (6-
23 months) who eat at least 2 
meals a day
121 680 17.8% 107 678 15.8% 228 1358 16.8%
Percentage of children aged 
(24-59 months) who eat at 
least 2 meals a day
242 680 35.6% 171 678 25.2% 413 1358 30.4%
Percentage of children aged (5-
18 years) who eat at least 2 
meals a day
478 680 70.3% 301 678 44.4% 779 1358 60.2%
Percentage of adults aged (19-
59 years) who eat at least 2 
meals a day
478 680 70.3% 339 678 59% 817 1358 60.2%
Percentage of adults (60 years 
and older) who eat at least 2 
meals a day
167 680 24.5% 166 678 24.5% 333 1358 24.5%
Percentage of children and 
adults who eat at least 2 meals 
a day
496 680 72.9% 336 678 49.6% 832 1358 61.3%
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
f) Percentage of food utilization, by type (household consumption, sale, 
exchange, livestock feed)
This indicator helps analyze the different types of food use by surveyed households. Over the 
whole project area, nearly 100% of households use agricultural production for household 
consumption, 46.0% of them use it for sale because it helps raise their income, and 4.3% said 
they use it to feed their livestock. At the departmental level, the trend is almost the same for 
households using agricultural production for their own consumption, for exchange and to feed 
their livestock, but it is not the same for sale. In the Center, twice as many households 
reported selling their agricultural production as in the Northeast department.
Table 17: Percentage of food use by type (household consumption, sale, exchange, livestock feed) 
by department 
Indicator: Production use
Center Northeast All Departments 
Num Den % Num Den % Num Den % 
Household consumption 677 680 99.6% 678 678 100.0% 1,355 1,358 99.8%
Sale 425 680 62.5% 200 678 29.5% 625 1,358 46.0%
Exchange 20 680 2.9% 14 678 2.1% 34 1,358 2.5%
Livestock feed 28 680 4.1% 31 678 4.6% 59 1,358 4.3%
g) Percentage of households reporting greater ability to cope with economic 
shocks 
Among the 680 surveyed households in the Center, only 32 of them (4.7%) reported having 
received training in shock management, while in the Northeast more households (50) (7.4%) 
received training. And overall, 82 out of 1,358 or 6.0% have received training in shock 
43
management. For those who implemented shock management methods, they represent about 
98% overall (100% in the Center versus 96.0% in the Northeast).
Table 18: Percentage of households reporting having received information to cope with 
economic shocks for the intervention areas and its communes 
Economic Shock
Center Northeast All Departments 
Num Den % Num Den % Num Den % 
Percentage of households 
reporting having received 
information to manage 
economic shocks
32 680 4.7% 50 678 7.4% 82 1358 6.0%
Percentage of households 
reporting having used 
information to manage 
economic shocks
32 32 100.0% 48 50 96.0% 80 82 97.6%
Based on household reports, the two main shocks that have caused hunger are drought 
(73.1%) and the hurricane (16.3%) over the whole target areas. The trend is similar at the 
departmental level.
Table 19: Major shocks that caused hunger in percentage 
Indicator: Major shocks that 
caused hunger in percentage
Center Northeast All Departments
Num Den % Num Den % Num Den % 
Drought 468 680 68.8% 525 678 77.4% 993 1,358 73.1%
Erosion 1 680 0.15% 12 678 1.8% 13 1,358 0.96%
Plant disease 29 680 4.3% 15 678 2.2% 44 1,358 3.24%
Hurricane 116 680 17.05% 105 678 15.5% 221 1,358 16.3%
Landslide 0 680 0% 0 678 0% 0 1,358 0%
Other problems 66 680 9.7% 21 678 3.1% 87 1,358 6.4%
h) Prevalence of mothers who exclusively breastfeed infants six months of age 
or younger 
The prevalence of exclusive breastfeeding among infants aged six months or less is 65.2% in 
both departments combined. This practice appears to be more prevalent among mothers in 
the Central department (72.2%) than those in the Northeast (55.6%).
44
Graph 19: Prevalence of mothers who practice exclusive breastfeeding among infants six months 
of age or less 
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
i) Prevalence of global acute malnutrition (GAM)
To measure and assess the nutritional status of children, data on their height or length, weight 
and age were collected so that anthropometric parameters such as height-for-age, weight-for￾height and weight-for-age could be calculated. In this study, the main focus is on low weight￾for-height or wasting, which is considered a measure of acute undernutrition and the 
consequence of inadequate nutrition over the period just prior to the survey. It measures the 
body mass in relation to height or length and describes the current nutritional status. Children 
whose Z-score for weight-for-height is below minus two standard deviations (-2SD) from the 
median of the reference population are considered lean (wasted) or acutely undernourished. 
According to the study’s findings, over the whole project intervention area, among the children 
aged 0-59 months surveyed, 5.4%16 suffer from acute malnutrition. Gender-wise, boys (5.6%) 
show a slightly higher percentage than girls (5.2%). At the departmental level, the Center 
stands out considerably from the Northeast with a prevalence of acute malnutrition three 
times lower (3.3% versus 9.3%). In the Northeast, where the situation seems critical, again for 
the 0-59 month age group, out of every 100 girls surveyed, about 10 are too lean for their 
height compared to 8 for boys (Table 21).
For the 0-23 months and 24-59 months age groups, there is no real difference in the overall 
prevalence of acute malnutrition among children. However, at the department level, the 
findings for the Center seem more acceptable than those for the Northeast. As the prevalence 
of wasting is influenced by seasonal variations, it is difficult to interpret these variations over 
time.
16 According to the findings of the latest EMMUS-VI survey (2016-2017), 4% of children aged 0-59 months are wasted, in the 
10 departments. The prevalence of wasting in the Center is 3.2%, and the Northeast is 1.5% according to the findings of 
EMMUS-IV (0-59 months). 
72.2%
55.6%
65.2%
Center
Northeast
Department set
Figure 20: Prevalence of mothers who exclusively breastfeed 
infants six months of age or younger
45
Table 20: Prevalence of acute malnutrition by department 
Indicator: Prevalence of acute 
malnutrition in children by 
department (6-59 months)
Center Northeast All Departments 
Num Den % Num Den % Num Den % 
% of male children under five years (0-
59 months) with a weight/height Z￾score lower than -2 standard deviation
and/or edema 10 235 4.3% 10 120 8.3% 20 355 5.6%
% of female children under five years 
(0-59 months) with a weight/height Z￾score lower than -2 standard deviation 
and/or edema 6 243 2.5% 14 139 10.1% 20 382 5.2%
% of children less than 5 years of age 
(0-59 months) with a weight-for-height 
Z-score lower than 2 standard 
deviation and/or edema 16 478 3.3% 24 259 9.3% 40 737 5.4%
Percentage of children 0-23 months of 
age with a weight-for-height Z-score 
Lower than -2 standard deviation
and/or edema 4 175 2.3% 12 127 9.4% 16 302 5.3%
Percentage of children 24-59 months 
of age with a weight-for-height Z-score 
lower than -2 standard deviation
and/or edema 12 303 3.9% 12 132 9.1% 24 435 5.5%
Source: EFSP Central and Northeast Final Evaluation Survey, World Vision, September 2020.
Data from the quantitative survey therefore helped establish the values for these key project 
indicators in this final evaluation. We saw the key indicators progress against their targets. 
Only 3 did not reach 50% of their targets! However, in the interview with the project’s M&E 
team, they mentioned that 4 output indicators did not reach 50% of their targets, namely the 
number of stone lines rehabilitated, the number of check dams rehabilitated, the number of 
earth dams rehabilitated, the number of IEC materials given out. This suggests that, overall, 
many indicators have reached at least 50% of their target. In the impact analysis section, the 
values for key indicators are compared with those at the project baseline.
Table 21: Key indicators measured, and their progress compared to targets 
Indicators
Final Evaluation Values
Center (%) ± CI Northeast (%) ± CI
Target % of target 
met Target % of target 
met
Percentage of targeted 
households with an 
acceptable food 
consumption score 
(FCS)
66,6 ± 3,55% 75 88,8% 48,50 ± 3,75% 75 64,67% 
46
Prevalence of 
households with mild
or no hunger 
(Household Hunger 
Scale - HHS)
17,8 ± 2.9% 70 25,43% 10.9 ± 2,35% 70 15,57% 
Proportion of 
households eating at 
least 6 food groups in 
the previous month
92,2 ± 2.0% 70 131% 83,5 ± 2,8% 70 119% 
Percentage of food use 
by type (sale, trade, 
feeding)
62,5 ± 3.6%
70 89,28% 
29,5 ± 3,4%
70 42,14% 
Percentage of 
households where 
adults and children eat 
at least 2 meals a day 
(adults and children), 
age 0-59 months
72,9 ± 3.35%
70 104% 
49,6 ± 3.75%
70 70,85% 
Prevalence of mothers 
who exclusively 
breastfeed infants six 
months of age or 
younger 
72,2 ± 6.9%
60 120% 
55,6 ± 9.0%
60 92,67% 
Reduced coping 
strategy index (% of 
households with a less 
severe or moderate 
index) 
16,84 ± 0.95 (58%)
70% 82,85% 
12,6 ± 1.0(74%)
70% 108% 
Percentage of 
households reporting 
greater ability to cope 
with economic shocks.
4,7 ± 1.6%
60% 8% 
7,4 ± 1.95%
60% 12,33% 
Global prevalence of 
acute malnutrition 
(GAM) 
3.3 ± 1.65%
- - 
9.3 ± 3.55%
- - 
Based on the previous analysis of project indicators, it would therefore be tempting to assign 
a score of 4=Satisfactory on the previously established scale to determine the extent to 
which the OECD criteria are met in the project.
5.4 Analysis of Project Efficiency 
To determine the efficiency of the project, we looked at the information available to see how 
cost-effective activities were in each component. The cost-effectiveness analysis criterion is 
used wherever possible. We also examined whether the objectives set for the components 
were achieved on time, based on the resources made available for their implementation. 
Compliance with provisional/contractual deadlines is also considered. 
47
For this analysis, the food vouchers, nutrition, agriculture and financial support (training and 
S4T) components are considered. 
a) Food Vouchers
WV set a joint team of 4 people on this program component to lead the operational steps and 
design the vouchers using a system, to design the distribution plan, negotiate prices with the 
vendors and ensure the quality control of food products. At the time of vouchers’ distribution 
and redemption, WV set up a distribution committee to support beneficiaries and collect their 
grievances and an OSDM (One Site Distribution Monitoring) focal point to ensure compliance 
with humanitarian standards. WV conducted post-distribution monitoring every 3 months and 
WV’s internal accountability team managed complaints. In addition, the WV team considered 
humanitarian and security principles in determining distribution points for food vouchers to 
beneficiaries in order to allow easy access and timely interventions respecting the rights of 
participants. However, the team felt the need to identify other people in households with 
reduced mobility to retrieve the vouchers and food and keep them from coming to distribution 
centers. Waiting times for beneficiaries during voucher distributions and redemption from
vendors were reasonable for most surveyed communes. However, delays were recorded, and 
catch-up sessions were set up to ensure all beneficiaries were served.
In the Center during interviews with local authorities, in Cerca-la-Source for example, on 
many occasions the project team experienced delays and distributions were held in late hours; 
this posed security risks for beneficiaries, mainly pregnant and nursing women and people with 
reduced mobility.
Communication between the WV team, vendors and beneficiaries was more or less smooth
with respect to the planning of voucher distributions, the adjustment of their purchase value 
and their redemption with vendors. However, misunderstandings were recorded in both 
departments between beneficiaries and vendors, which led to tensions on the price of 
products with vendors. In fact, WV carried out quality control measures for food products 
and monitored prices on local markets, monthly price-fixing meetings were organized with 
vendors considering the inflation and the depreciation of the Gourde against the dollar, and 
restitution meetings were held with beneficiaries. In some areas, vendors worked with WV to 
mobilize and travel to meet beneficiaries in remote areas. Thus, they had to readjust the costs 
of products without notifying WV, which increased discussions and tensions with vendors.
The issue of contractual delay was mainly discussed with a few vendors who pointed out that 
WV needed more than two weeks to pay, following the presentation and validation of 
products delivery sheets; which created a challenge, for many of them, to replenish their 
stocks. This is mainly the case for vendors in Mombin-Crochu, Fort-Liberté and Cerca￾Carvajal.
b) Nutrition
CEPAM, which was fully in charge of all the activities under this nutrition component, 
coordinated the selection of mother leaders, nutrition training, and dissemination and 
awareness campaigns on care and better hygiene practices. Their leaders believe WV made 
adequate resources available to the project within the desired timeframe to complete the 
activities. Collaboration and coordination with WV was efficient, as much as the partnership 
was well defined. WV provided CEPAM with training materials so that they could better 
48
understand their role in disseminating information to the population. The project also 
provided them with cooking materials to continue teaching better cooking practices.
The unit cost for infant and young child nutrition counseling ranges from US$5 to US$7 in 
sub-Saharan Africa (A Framework for Investment in Nutrition, World Bank 2017). It would be 
interesting to see here the cost to raise the awareness of mother-leaders, especially knowing 
how much they complain about the accessibility of the areas and transportation costs. To 
travel from one community to another, people use motorcycles in some communes. 
Transportation costs can vary from 1,000 to 1,500 HTG (15-20 USD). Since we do not have 
all the information related to the costs involved, we were unable to make a comparison. These 
kinds of analysis are important for projects to conduct to see how sustainable these practices 
are that still create good dynamics in communities. 
c) Agriculture
In the interviews with community leaders and local authorities, they pointed out a lack of 
human resources in implementing the agricultural component. In fact, the training and plot 
monitoring (after the planting of seeds) activities were not carried out within the allocated 
time. This caused considerable delays for example on the availability and subsequent 
distribution of seeds. Therefore, this was a negative efficiency score!
Although WV had community endorsement for the implementation of the project, there were 
no formal partnerships per se, but short collaborations between WV and local leaders at 
specific moments in the project, such as during the start-up meetings, mobilization of 
beneficiaries, managing of security aspects during the distribution of food vouchers and seeds 
by CASEC and ASEC. 
So the lack of agricultural staff was a major challenge. The simultaneous management of several 
farms by the same person resulted in burnouts. This may also lead to a loss of focus and affect 
the planning. Under this component too, the project is looking for clear data on sowing 
activities, plot management and subsequent harvesting. For beneficiaries, supporting 
agricultural production is essential to ensuring food availability within the communities and 
especially to equipping farmers with smart conservation practices and techniques to adapt 
production to climate hazards. They claim to be very satisfied and in agreement with the 
trainings received but think that more efforts can be made under this component. For example: 
WV could establish community nurseries to multiply fruit and forest seedlings, they could train
people on agroforestry, they could manage firewood lots to reduce the cutting of trees and 
increase biodiversity and ecosystem services. 
d) Financial support (training and S4T)
The necessary human and financial resources were put in place by World Vision for this 
project component. For local leaders from the 2 project’s intervention areas (departments), 
the technical trainings on the setting up of new SLA groups and the strengthening of existing 
groups were a success on which the communities capitalized and with which they continued 
after the project. Indeed, according to them, the savings and credit groups not only had an 
economic objective (savings and credit) to help vulnerable people access additional funds for
investment in income-generating activities such as agriculture, animal husbandry and small 
businesses, but also they helped strengthen social ties within the communities. For them, the 
savings and credit groups helped renew community ties, self-help and family management. 
49
In addition, the project team estimated the cost of setting up and strengthening community 
savings and credit groups at US$420 per unit, and an agent costs US$150 per unit. Through 
discussions with the Mercy Corps and CARE Haiti specialists in group formation and 
strengthening, they consider the $400-$500 range as the unit cost to set up and strengthen 
savings and credit groups when considering the start-up materials to be provided, 2-3 trainings 
to be given to a 30-member group. They estimate the unit cost for the agent who supports 
the groups at a US$100-200 range when they consider 3 trainings provided, for example. Based 
on these figures, it can be said that the project has been financially efficient in this specific 
component. 
To summarize, the project has had efficiency concerns especially in terms of meeting deadlines 
with vendors and in agriculture, due to the lack of human resources for monitoring activities. 
On the satisfaction scale presented above, it deserves a score of 3=Average satisfaction on 
the basis of available information. We could not get the project’s financial report to know the 
percentage of budget expenditures, and which budget item was under or over evaluated. 
5.5 Analysis of Project Impact 
The idea behind this criterion is to establish the level of major informed (reported) and/or 
observable changes in the living conditions of the populations (or target groups) directly or 
supposedly related to the project outcomes. According to the M&E team, the project did 
indeed reach the expected number of beneficiaries and territorial coverage. The project’s 
interventions virtually led to an overall improvement in key indicators (Table 22). A strong 
downward difference is noted in the Central Department with the indicator related to 
exclusive breastfeeding. And yet, everyone knows the project mainly worked with mother￾leaders to act on these aspects. This is probably due to a sampling bias at baseline where the 
percentage was too high for exclusive breastfeeding, especially when we refer to the national 
average of 39% (EMMUS, 2016).
The difference test explains whether or not there is a statistically significant difference between 
the values found for the indicators at baseline and the final evaluation. Where the test indicates 
there is no difference, it is highlighted in red, meaning the values may be different but are not 
statistically significant to indicate there has been a change in the indicator considered. That 
way, we may note the result is slightly better in the Center than in the Northeast.
50
Table 22: Comparison of baseline and final evaluation results 
Indicators
Baseline values Final Evaluation Values Difference Test
Center (%) ± CI Northeast (%) 
± CI
Center (%) 
± CI
Northeast 
(%) ± CI Center Northeast
Percentage of targeted 
households with an 
acceptable food 
consumption score 
(FCS)
46.0 ± 6.0% 49,9 ± 5.3% 66,6 ± 3.55%
p= 0.00< 
0.05, 
there is a 
difference 
p= 0.62>0.05, no 
difference 
Prevalence of 
households with mild
or no hunger 
(Household Hunger 
Scale - HHS)
8.9 ± 3.45% 6.0 ± 2.5% 17,8 ± 2.9% 10.9 ± 2.35%
p= 0.002< 
0.05, 
there is 
difference 
p= 0.00< 0.05, 
there is a 
difference 
Proportion of 
households eating at 
least 6 food groups in 
the previous month
78,1 ± 5.0% 76,6 ± 4.5% 92,2 ± 2.0% 83,5 ± 2.8%
p= 0.000< 
0.05, 
there is a 
difference 
p= 0.00< 0.05, 
there is a 
difference 
Percentage of food use 
by type (sale, trade,
feeding) 
Sale
62,5 
± 
5.85%
63,7 ± 5.1% 62,5 ± 3.6% 29,5 ± 3.4%
p= 
0.5>0.05,
no 
difference 
p= 1>0.05, no 
difference 
Exchange
7,55 
± 
3.15%
4,0 ± 2.05% 2,9 ± 1.25% 2,1 ± 1.05%
p= 
0.99>0.05,
no 
difference 
p = 0.96>0.05, 
no difference 
Livestock 
feeding 
2,3 ± 
1.8% 4,2 ± 2.15% 4,1 ± 1.5% 4,6 ± 1.55%
p= 0.03< 
0.05, 
there is a 
difference 
p = 0.32>0.05, 
no difference 
Percentage of 
households where 
adults and children eat 
at least 2 meals a day 
(adults and children), 
age 0-59 months
63,6 ± 5.8% 40.0 ± 5.2% 72,9 ± 3.35% 49,6 ± 3.75%
p= 0.000< 
0.05, 
there is a 
difference 
p= 0.000< 0.05, 
there is a 
difference 
Prevalence of mothers 
who exclusively
breastfeed infants six 
months of age or 
younger
87,6 ± 9.65% 53,8 ± 15.45% 72,2 ± 6.9% 55,6 ± 9.0%
p= 
1>0.05, 
no 
difference 
p = 0.26>0.05, 
no difference 
Reduced Coping 
Strategies Index (rCSI)
Average 16,7 
± 0.9 16,7 ± 0.85 16,84 ± 0.95 12,6 ± 1.0
p= 
0.84>0.05, 
no 
difference 
 
p = 0.000< 0.05, 
there is a 
difference 
Standard 
deviation 10,5 10,9 12,84 13,13
Percentage of 
households reporting 
greater ability to cope 
with economic shocks.
0,9 ± 0.85% 0.0 ± 0.0% 4,7 ± 1.6% 7,4 ± 1.95%
p= 0.000< 
0.05, 
there is a 
difference 
p = 0.000< 0.05, 
there is a 
difference 
Prevalence of Global 
Acute Malnutrition 
(GAM)
10.5 ± 3.6% 20.4 ± 5.1% 3.3 ± 1.65% 9.3 ± 3.55%
p= 
1>0.05, 
no 
difference 
p= 1>0.05, no 
difference 
 
51
Beyond the figures shown for the indicators, local leaders recognize that the food insecurity 
and living conditions of the beneficiaries underwent positive changes since the food vouchers 
came with support in livelihoods such as agricultural production. In addition, nursing mothers 
are in a better position to support better nutrition for their offspring and apply better hygiene 
principles for the whole family. On the other hand, they believe that many unassisted families 
remain food insecure and their conditions have worsened following COVID-19. Since they 
were already living in precarious conditions, lockdown restrictions brought these families into 
an emergency phase of humanitarian assistance. 
The activities of the savings and credit groups helped families who never had access to financial 
services to save and access investment funds allowing them to diversify their economic 
activities. The technical support and agricultural training received helped beneficiaries better 
approach two windows of agricultural production. Unfortunately, the seeds were given out
late, outside the crop calendar, which reduced their productive potential. In addition, the 
trainings and maintenance of water points and pumps helped the whole communities have 
access to treated and drinking water, and prevent subsequent cases of cholera. 
Combined with training on the key aspects of family nutrition, pregnant and lactating women, 
the WV team believes the distribution of food vouchers helped beneficiaries and vulnerable 
families better feed themselves by considering the essential categories of a balanced diet. 
Beneficiaries emphasized that the voucher assistance they received helped their families access 
food quickly so that they did not need to resort to harmful coping strategies such as sending 
household members to beg, selling their livestock or productive assets (rCSI <19).
Beyond the mother-leaders strategy that encourages and anticipates the inclusive participation 
of women and girls, project activities had a clear strategy in their implementation to support 
and strengthen the role and participation of women and girls. For the agricultural training 
activities, there was no evidence that a gender analysis was conducted to understand the social 
dynamics and the role of women in the production chain and thus design training that takes 
these social subtleties into account.
As with the other criteria, the previous analysis of the project’s impact led us to consider a 
score of 4= Satisfied for this, on the satisfaction scale. 
5.6 Analysis of Project Sustainability 
The implementation approach, based on the search for ownership of the project by partners 
and local communities, is an important factor that can lead to sustainable outcomes. In this 
approach, two important elements to ensure sustainability are to be highlighted:
• The adoption of a clear communication strategy to facilitate understanding and strong 
involvement of beneficiaries, state and local community structures in both the planning 
and implementation of activities. 
• The adoption of a clear exit strategy to facilitate business continuity after the project.
Because they were carried out as part of a humanitarian action, the focus of project 
interventions on activities that can truly meet the needs of beneficiaries is a prerequisite to
improve community resilience. They must also be in line with the humanitarian-development 
nexus. On this point, the orientation of farming activities towards market value chains to 
ensure production, the construction/rehabilitation of important infrastructure for the 
52
community and the strengthening of beneficiaries’ capacities are actions that can ensure 
sustainability in the project. It can thus be admitted that the implementation strategy is quite 
correct and has a huge potential to produce sustainable and replicable outcomes in the 
medium and long term.
It is clear that the sanitation work and road section development are sustainable activities that 
will continue in the communities thanks to the commitments of the community leaders, 
CASEC and the local municipalities. However, the heavy sanitation work will require local 
investment funds through the cities (Ministry of Interior), perhaps, or the regional offices of 
the Ministries of Agriculture and Environment. Water point management committees continue 
to provide drinking water services in the communities, and mother leaders can join DINEPA 
structures and MSPP nutrition officers respectively to continue meeting the quality 
requirements for drinking water treatment and raise the awareness of lactating mothers on 
the importance of breastfeeding and a balanced diet for the whole family. In addition, due to 
the transfer of technical agricultural knowledge during training, farmers can continue applying 
good soil conservation practices to increase their resilience to recurring climate shocks, but 
they will need to better organize themselves, for example, into groups through the internal 
agricultural structures of farmers’ organizations to share technical information. The savings 
and credit groups will continue operating after the project. However, they believe the 
formation of other groups is a pressing need in the communities, especially in the communal 
sections. This is the part of the project where there is a higher community 
ownership! In the meantime, other groups have emerged without the help of the project. 
This ownership is reflected in the level of involvement of agents who worked without receiving 
project allowances or money from the savings groups. After receiving the necessary training, 
the members and agents – since they are also part of the groups – brought the solidarity 
funding to launch the activities.
The project also strengthened the visibility of local authorities and brought them closer to
their constituents, especially in the implementation of community work and the distribution 
of food vouchers, which represent concrete actions to inject funds into the different 
communities to alleviate the currently prevailing situation. 
Although food aid in cash or in kind is hardly a sustainable action, it is just a rapid intervention 
allowing disaster-affected populations to have quick access to food. Local authorities believe 
that working, at the same time, on reviving livelihood activities (agriculture, livestock, small 
businesses, etc.) is the best way to ensure better living for communities and, above all, to 
strengthen their capacity to adapt and transform themselves in the face of these recurring 
shocks (climatic, social, economic). For that purpose, the project was to support local 
mechanisms for adaptation to climate shocks through the decentralized DPC branches. 
In conclusion, the project set into motion dynamics that can continue after its closure, such 
as savings and credit groups and water point management committees insofar as these are 
linked to the TEPACs (Techniciens Eau Potable et Assainissement Communal) of the 
communes. However, some other components will have difficulty staying in the communities. 
Let’s consider, for example, mother-leaders. To what extent will they be able to be supervised 
by the ASCPs or other MSPP structures? We have seen in the efficiency section, their financial 
limitation to go into communities that are difficult to access. For all of this, a score of 
53
3=Moderately satisfied can be considered for the assessment of the sustainability criterion 
in the project’s dynamics. 
5.7 Analysis of Project Linkages and Exit Strategies 
A persistent challenge for projects is to ensure that benefits from interventions are sustained 
after the projects’ completion. Linkages, especially vertical linkages, such as those between 
community organizations or individuals and existing institutions in the public or private sector, 
are essential to the successful progressive transfer of responsibilities previously supported by 
projects. Project design should include indicators to measure not only impact but also the 
sustainability of change. Sustainability plans and exit strategies should include clear timelines 
and benchmarks for progress toward sustainability, separate from project impact indicators.
The team analyzed the capacity of WV during the project’s implementation to forge systemic 
links and partnerships with state structures or organized groups in the communities to support 
actions and interventions and to continue the dynamics initiated by the project’s many
components within the communities: 
Food aid: WV did not have direct collaboration with MAST to support the internal 
mechanism for strengthening social safety nets, protection and social promotion. Although 
WV has used SIMAST through WFP, the evaluation team believes this approach does not, in 
any way, ensure a transfer of lessons learned to the relevant state structure and especially 
allows MAST to appropriate data from WV in order to update SIMAST and strengthen their 
capacity to coordinate with other actors for better use of available resources and avoid 
duplication in future interventions. This brings the evaluation team to the following conclusion: 
the approaches included in the project proposal regarding links to social protection 
mechanisms to target the most vulnerable were not taken into account by the implementation 
team. While national governments should remain responsible for the design and management 
of social protection programs, according to the OECD Guidelines for Humanitarian 
Assistance17, humanitarian actors should play a role in strengthening existing policies and 
supporting the creation of new social protection systems in areas where such programs do 
not exist or are underdeveloped. There are opportunities to link humanitarian assistance and 
social assistance for two main reasons : 
- An existing social protection system can potentially help provide rapid response 
to a large number of people.
- Engaging with social assistance can help strengthen the State system.
Nutritional component: Project interventions are in line with the National Nutrition Policy 
(NNP) in its objectives of nutrition prevention, improvement of nutrition information systems 
and nutritional protection in emergency situations. WV has trained on key issues related to 
the promotion of proper nutrition during the life cycle of men, women and children in the 
communities, improved infant and young child feeding practices and strengthened the fight 
against micronutrient deficiencies. However, no vertical links or partnerships with the MSPP 
or horizontal links with the local agents of the nutrition section of the said Ministry were
created. Because of that, mother leaders themselves pointed out a lack of supervision and 
support, refresher training (continuous capacity building) and resources to continue 
supporting pregnant and breastfeeding women in their communities with clear objectives.
17 https://www.oecd.org/development/humanitarian-donors/docs/cashbasedresponse.pdf
54
Agricultural component: WV did not use a market system approach such as the M4P 
methodological framework to make market systems work for the most vulnerable. Farmers 
were not put into contact with potential sellers and buyers, which does not guarantee a secure 
market for surplus production (immediate results of input distributions), nor does it provide 
incentives (additional income accumulation) for farmers to continue applying improved
practices and market their crops. Furthermore, WV did not follow the MARNDR pre￾requisites that require any intervention in the seed sector to use the national seed service to 
ensure good quality and conduct seed adaptability tests before distribution to farmers. This 
would also help establish links between seed suppliers and the project’s localities and areas of 
intervention in order to ensure better bargaining power for future orders from the 
communities through organized groups of farmers (ensuring demand and supply). The internal 
structures of farmers’ organizations and associations were not intentionally involved in the 
technical agricultural training. According to community leaders, the agricultural committees
within farmers’ associations should be integrated through a Farmer Field Schools methodology 
and cascade training. These internal structures already exist and work voluntarily around the 
mission of farmers’ organizations to support communities to adapt to climatic and socio￾economic shocks. In addition, a reduced quota of women’s participation in the trainings was
observed (less than 20%), which also reduces the clear transfer of agricultural technical 
knowledge to ensure full sustainability in the application and subsequent implementation of 
these practices beyond the project cycle. CNSA, for example, through these observation 
centers, was not contacted or put in contact with farmers to provide information related to 
the monitoring of agricultural and food product prices in the main markets, in order to help
farmers better negotiate and value their products surplus on the markets.
Community activities: Excellent collaboration was observed between WV, local leaders 
and municipalities around community work. Respondents stressed their importance to 
communities and persisted on the need for local governments to dedicate future lines in 
municipal budgets to continue this work. It is important to remember that not all models for 
achieving sustainability are equally applicable to all technical sectors. On the other hand, the 
strengthening of community drinking water points is positively noted by the communities and 
project intervention areas as sustainable. Management committees are better able to continue 
serving the communities and replace defective pumps and equipment and especially train local 
technicians able to respond to subsequent technical problems. This will allow for continuity 
and better management of water points. 
Of all the above, the satisfaction score is quite low for this analysis criterion, 2=Not satisfied. 
Indeed, at this level, the links with system partners (the sectors linked to the project) are not 
very clear. 
The following table summarizes the satisfaction scores for each criterion:
Table 23: Level of Satisfaction with Criteria Achievement 
Criteria Score awarded*
Relevance 4
Effectiveness 4
Efficiency 3
Impact 4
Sustainability 3
Links and exit strategies 2
*1=Not satisfied at all 2=Not satisfied 3=Moderately satisfied 4=Satisfied 5=Very satisfied
55
5.8 Analysis of the Local Market in the Intervention Areas
An additional market study was also conducted to enrich the evaluation. Both a quantitative 
and qualitative approach were used, interviews were conducted with vendors, community 
leaders, the WV team, and there was a survey of beneficiaries. 
Market situation and availability of food products
The findings of the market assessment confirm that all major markets in the region were 
functioning adequately during project implementation and were able to absorb any changes in 
demand caused by cash injections. Markets were accessible to beneficiaries. Discussions with 
the WV team and vendors revealed that a variety of products were available in sufficient 
quantities. For the vendors, there was no marked period of shortages for voucher food 
products since most were imported products. Shortages could be recorded for local products 
such as certain vegetables and fruits that are marked by well-defined farming periods. For 
tubers and vegetables, in addition to their farming availability, their acquisition takes place in 
remote areas through the Madan sara. During the rainy episodes, it was particularly difficult 
to access these production points. Beneficiaries in both departments confirmed they were 
able to find all the food products listed on the vouchers at the project’s vendors who made 
efforts to ensure a constant flow of supplies, especially for local products mentioned earlier 
(nearly 97% of the surveys). Community leaders, in turn, said they observed no break in the 
supply of food, wholesalers were always accessible and, in addition, closer links were formed 
between vendors (especially those who were part of the program) and wholesalers, which 
opened up informal opportunities for loans on sales to retailers.
In order to reduce travel distances, the mobilization of traders was necessary to facilitate a 
market response in areas where vouchers were distributed in order to trigger a market 
response in areas where markets were absent (remote communal sections). The mobilization 
consisted in disseminating information on the potential increase in demand for commodities 
in order to stimulate commodity inventory planning. Indeed, vendors in both departments 
expressed a clear understanding of the selection process they went through to join the 
program: from the publication of the call for interest, surveys, audits, and visits to sales sites 
to the publication of the final list. Once selected, the vendors took part in meetings with the 
WV team to better understand the project and become aware of their role in the 
implementation of the food voucher component. They reported no difficulties during the 
selection process, and further reported that the WV team provided them with critical advice 
that would help them meet the defined selection criteria (e.g., guidelines on licensing their 
small businesses). There were no difficulties in meeting contractual requirements with WV, in 
payment mechanisms, product prices and voucher values, and the recipient verification system 
on the vouchers. 
Product prices
For the operational aspect of vouchers, the vendors emphasized how easy it was to identify 
project beneficiaries and the need to coordinate with WV on a clear distribution schedule. 
The value of vouchers was adjusted following complaints from beneficiaries about the 
depreciation of the Gourde, increased inflation and socio-political unrest that impacted local 
markets and increased the price of basic foodstuffs. According to community leaders, the 
project played no role in price fluctuation as local markets had sufficient food to 
meet the increased demand generated by the project. Moreover, the changes in food 
56
prices were not a response to increased demand but rather to factors external to the project 
such as rising inflation, devaluation of the local currency, and socio-political unrest. Indeed, in 
October 2019, CNSA reported a 14% monthly and 40% annual increase in the nominal cost 
of the basic food basket (rice, wheat flour, corn, beans, sugar and vegetable oil) with the highest 
increases recorded in Jérémie (62%), Fond des Nègres, Cayes, Hinche and Croix des Bossales 
(+41%). And imported products such as rice, vegetable oil, spaghetti, sugar, beans, flour and 
corn whose prices are strictly volatile compared to variations in the exchange rate, underwent
increases in annual rate (2018 to 2019) of +38% for rice, +49% for corn, +53% for wheat flour 
and 52% for sugar during project implementation (CNSA18). 59% of surveyed beneficiaries in 
the Center felt that the prices of products displayed on the vouchers were more expensive 
than on the local market, compared to 41% in the Northeast (Graph 21). As mentioned in 
the previous sections, WV made monthly adjustments to the price of products on the various 
markets in line with inflation and the depreciation of the Gourde, and consultations were made 
with vendors and beneficiaries about the prices set, but there were nevertheless some 
irregularities. The depreciation rate of the Gourde required a tighter adjustment, at least every 
15 days, and some vendors had to travel to areas that were difficult to reach for beneficiaries, 
which forced them to increase prices to absorb transportation costs and also use inflated
values to amortize any additional increase before the next adjustments. However, most local 
leaders felt that the purchase value of the vouchers should be adjusted to USD 100 to absorb 
fluctuations. 
Graph 20: Distribution of respondents according to their perception of products prices on the 
vouchers compared to prices for the same products at other sellers 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Redemption and use of vouchers 
According to the project team, WV made internal arrangements to combat voucher fraud. In 
the beginning of the project, WV used sealed paper coupons with logos and different colors, 
bar codes, graphics that could prevent their reproduction. As time went by, the paper 
vouchers were replaced by a digital card (NFC) for beneficiaries and telephones (MPOS) for
vendors with a link to the LMMS/EVS system. Information sharing, mobilization and awareness 
sessions were organized in the communities to reduce any risk of fraud, voucher theft, or 
exposure (security). Vendors were clearly identified with signs near their sales kiosk and 
practice sessions were conducted with beneficiaries on the required quality of food products 
for adequate consumption while maintaining their nutritional value. When redeeming vouchers 
18 https://fscluster.org/haiti/document/bulletin-du-panier-alimentaire-de
8%
59%
41% 41%
51%
Center Northeast
Lower coupon product
prices
Product prices on
higher coupon
Same price everywhere
57
at sales kiosks, nearly 98% of surveyed beneficiaries mentioned they had no difficulty 
redeeming vouchers with project vendors and 80% on average (81% in the Northeast versus 
79% in the Center) felt safe to complete such transactions. In addition, special attention was 
given to understanding the identification process, signature, balance on the vouchers, among 
others. On average, vendors received 20 voucher-holders per day and needed 10 minutes to 
complete the voucher redemption process for project beneficiaries, of which 3 minutes were 
dedicated to filling out the delivery form that had to be sent to WV for subsequent 
reimbursement. 
Within the families benefiting from the project, women (72% of respondents in the North East 
and 63% in the Center) made the decisions on redeeming the vouchers and the choice of 
products, followed by men. This confirms the program team’s targeting strategy that
prioritizes female-headed households and pregnant and lactating women. Not many
respondents said that these decisions were made in consultation (Graph 24). The same trend 
was noted where nearly 65% of respondents identified women as the household member who 
collected vouchers from distribution centers and kept them before they were redeemed with 
vendors. 
Graph 21. Distribution of respondents by who decided when to redeem and what products to get 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
89% of households in the Center used food products received from the program for internal 
family consumption, compared to 73% in the Northeast. They also helped other families and 
neighbors, as nearly 11% in the Center and 26% in the Northeast reported having consumed 
some of the food and shared the products received (Graph 25).
Graph 22: Distribution of respondents by use of products after voucher redemption 
63%
72%
17%
18%
18%
5%
1%
2%
Center
Northeast Other (nephew, cousin, ...)
Son/Daughter
Both (husband/wife)
Husband/Husband
Woman/wife
58
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Quality of products received from vendors
For vendors in the Northeast, no complaints were received from beneficiaries on the quality 
of food products received in exchange for vouchers. For the Center, beneficiaries reported 
to vendors the poor quality of beans and tubers, which was improved in future deliveries 
because the vendors changed suppliers. This trend is supported by surveyed beneficiaries, of 
whom 8% and 1% in the Center and Northeast respectively felt that products were not of 
good quality. Nearly 35% mentioned they were acceptable (normal quality), and nearly 60% 
felt that products were of good quality (Graph 26).
Graph 23: Distribution of respondents according to their satisfaction with the quality of food 
products after the voucher was redeemed 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
Impact on the supply chain
Community leaders highlighted complaints from non-program vendors who mentioned the 
loss of customers to registered vendors. Economic opportunities in monetary terms benefit
the communities, but the money flow and transactions benefit a smaller number of vendors. 
The WV team emphasized, however, that an inclusive and participatory approach was used to 
engage vendors in the program through clear selection criteria. It would be preferable, in these 
cases, to select as many vendors as possible and assign them a reasonable number of customers 
(splitting the additional flow of customers). 
Vendors ensured that the increase in demand and customer flow (traditional customers and 
project beneficiaries) was manageable and did not affect their capacity in any way, since they 
89%
73%
11%
26%
Center
Northeast
Sale of products
Storage of products
Consumed in part and shared with
neighbors
Consumed by the family
31%
43%
8%
1%
61%
56%
Center Northeast
Normal product quality
No, the products are not of
good quality!
The products were of good
quality
59
had the goods (agricultural products) to meet this increase in demand, although, in some cases 
(more than 30 clients per day), they were assisted by a helper whose role was to package the 
products for beneficiaries. For them, it was mainly difficult to explain to beneficiaries the 
decrease in the purchase value of vouchers due to the increase in the price of products on 
the local markets as a direct consequence of the rising exchange rate and inflation. 
Transportation costs increased and road conditions were not favorable during the rainy 
seasons to go and buy local foodstuffs such as tubers, fruits and vegetables. WV took more 
than two weeks to pay, following the presentation and validation of products delivery sheets, 
which created a challenge, for many of them, to replenish their stocks. This is mainly the case 
for vendors in Mombin-Crochu, Fort-Liberté and Cerca-Carvajal. In addition, many vendors 
mentioned delays from the WV team during open market days and the lack of coordination. 
In fact, they advise that each area should have an established market day for beneficiaries to 
redeem their vouchers. 
Beneficiary preference 
According to findings from the quantitative survey, beneficiaries prefer by far to receive 
humanitarian assistance in food vouchers, with nearly 60% for both departments, preferring 
receiving cash assistance. It should be noted that in-kind distribution is not much preferred 
and cash vouchers (i.e., from previously identified vendors) was not much appreciated (Graph
27).
Graph 24. Distribution of respondents by how they would like to receive assistance for a similar 
project 
Source: Final evaluation of the EFSP project in the Center and Northeast, World Vision Haiti, October 2020.
58%
62%
2% 4%
26% 27%
1%
6%
11%
1%
Center Northeast
Food Coupon
Cash coupon to be free in my
purchases
Cash
Food products through direct
distribution
Any modality
Other (agricultural project,...)
60
6) Conclusion and Recommendations 
The overall objective of this final evaluation is to analyze the level of achievement of the goal, 
objectives, and outcomes of the EFSP program and how these were done. For this purpose, a 
research combining quantitative and qualitative tools was conducted. 1,358 quantitative 
surveys were carried out among beneficiaries from the different project components. On the 
qualitative level, several dozen interviews were conducted not only with key persons in the 
project but also with beneficiaries, community leaders, local authorities and vendors.
Overall, the findings show that the majority of beneficiary households are headed by adult 
women and men (F&M) in both departments. Their profile is characterized by vulnerability 
points. For the most part, they have no education or were only able to get an incomplete 
primary education. They live mainly in tin roof houses which are not always of good quality. 
Agriculture (including animal husbandry) and trade are their main activities, with monthly 
incomes of around 5,500 HTG or less.
In terms of meeting OECD criteria, data analysis showed that the project was well positioned 
in terms of relevance, effectiveness and impact insofar as it responded to the priorities of the 
people in the communities affected by hurricane Irma by largely achieving its objectives and 
covering the targeted territories. The project is more or less well positioned, considering the 
aspects of efficiency and sustainability. This is because not all the components were able to get
adequate human resources to meet contractual deadlines and there were no total ownership
of actions by the communities in all the components. The project is rather weak in terms of 
linkages and exit strategies when considering the poor involvement of the regional structures 
of sectoral Ministries related to project activities, such as MARNDR for example.
The local market analysis confirmed that all major markets in the region were operating 
adequately over the life of the project. Vendors expressed a clear understanding of the 
selection process in which they participated. The project did not contribute to increasing 
prices in the local market. Unfortunately, one negative effect of the project observed is the 
market distortion created by draining a large part of the local clientele to contracted vendors. 
Other vendors outside the project complained about this to local authorities.
based on the findings of this study, recommendations were therefore expressed by project 
activity component:
Food Voucher Component
• Conduct analysis and monitoring of local markets to reflect fluctuations in food 
prices in order to anticipate and adjust the monetary values of food vouchers or simply 
define a lump sum that takes these fluctuations into account. Regularly check 
exchange rates to ensure that vouchers keep their value.
• When a voucher approach is used, ensure that as many merchants as possible 
participate. This ensures a competitive marketplace.
• Monitor food and non-food item prices in the nearby markets to ensure that 
your project area does not unfairly increase prices.
• Strengthen feedback mechanisms for beneficiaries (free numbers, suggestion 
boxes, volunteers, etc.).
• Ensure effective and continuous communication with beneficiaries and 
stakeholders throughout all phases of the project cycle (design, set-up, planning, 
61
implementation, evaluation, monitoring and transition at the end of project). WV staff 
noted that despite initial outreach efforts and messages, non-beneficiary households 
became angry at the lack of assistance and directed their anger and aggression at staff 
throughout program implementation. Although adequate communication and 
awareness raising can be considered a “best practice” in all humanitarian programs, 
and particularly in financial assistance programs, care must be taken to maintain 
consistent, constant, and accurate communication about selection criteria, resource 
availability, priority of assistance, and recognition of needs at the community level. The 
regular and repeated transmission of these messages will help reduce 
misunderstandings and resentment among potentially volatile and aggressive 
populations. 
• Reduce security risk by adjusting voucher distribution and redemption cycles 
according to the context. WV must avoid the risk of congestion and disruption or 
frustration in the vicinity of project offices or merchant sales areas by providing well￾communicated and established periods for voucher collection and redemption. For 
example, implement a ten-day cycle, allowing five days for voucher collection in a 
designated area and five days for voucher use at the vendors’ premises to reduce the 
risk of crowd formation. Crowding can lead to a lack of understanding of the program 
on a larger scale, community unrest, and the risk of theft or pressure on vendors. 
Vouchers distribution and redemption cycles should be determined according to the 
context and through a clearly defined participatory process. 
• Support the harmonization between beneficiaries and vendors and strengthen 
sensitization sessions with beneficiaries and vendors on WV’s internal 
accountability mechanism and thus help them channel their grievances for further 
processing, which will help in better planning and service delivery.
• Continuously and carefully raising awareness and informing non￾beneficiaries is essential to maintain stability in the implementation area.
• Extremely vulnerable populations require additional logistical support for
the process of collecting and redeeming food vouchers. Although the project’s 
beneficiary targeting criteria identified extremely vulnerable people for food assistance, 
community feedback indicated that the elderly, sick, and disabled should not have to 
wait in line for long periods of time or deal with the logistics of bringing their food 
basket to their homes. Future programs should consider special accommodations for 
these vulnerable groups to avoid excessive physical and mental stress or increased 
security risk during the voucher redemption process and transportation of purchased 
food.
• WV must anticipate community safety net and social solidarity systems in 
targeting and allocating vouchers. In all project sites, households frequently shared their 
food aid with family, friends and/or vulnerable community members. Community 
solidarity practices should be taken into account in the final determination of the 
amount and use of vouchers to ensure that food aid is sufficient for households that 
can engage in such practices. 
• The project was to support local mechanisms of adaptation to climate shocks 
through the decentralized branches of the DPC. Food aid in cash or in kind 
hardly represents sustainable actions, but just rapid interventions allowing disaster￾affected populations to have a quick access to food. Working in parallel to revive 
livelihood activities (agriculture, livestock, small businesses, etc.) is the best way to 
ensure better living for communities and, above all, to strengthen community capacity 
to adapt and transform themselves in the face of these recurring shocks (climatic, 
social, economic). 
62
• The project should take a more holistic approach to food security. It should also 
focus on issues such as access to food (including equitable access within households), 
absorption and maintenance. This would require a focused and logical approach aimed 
at specific changes in attitudes and behaviors.
• WV should play a role in strengthening the country’s existing policies and 
social protection system by using them to target the most vulnerable or by 
conducting assessments to enrich the MAST database.
• WV should integrate a community-based Disaster Risk Reduction (DRR) 
approach to help the community cope with the stressors and shocks associated with 
natural hazards.
• The project’s logical framework should be reviewed on the basis of a causal analysis. A 
multidimensional and comprehensive approach targeting the same cohort 
may be more effective in creating impacts in terms of resilience and food 
security. 
• It is recommended that field activities be monitored by dedicated and, to some 
extent, independent monitoring and evaluation staff. Field monitors will 
provide management with first-hand information on unilateral reporting of activities. 
This way, findings are verified and will contribute to transparency and accountability. 
Nutrition component
• Ensure that households not selected also benefit from sensitization sessions 
on good nutritional and hygiene practices in order to emancipate project impacts and 
thus improve the negative indicators of the nutritional situation in assisted areas. 
• The implementation of intervention activities to improve child nutrition is not only 
the responsibility of the health sector, but also requires the participation of local 
government, the agricultural sector, women’s and youth associations, farmers’ 
organizations and other local departments and mass organizations.
• As child rearing is not the sole responsibility and task of mothers, the project should 
have a clear framework to integrate and promote men’s participation in 
nutrition practices and awareness sessions.
• Adopt a decentralized approach for mother-leaders to reduce the need for 
transportation costs during awareness campaigns and thus ensure the sustainability of 
these activities beyond the project life cycle. 
• WV must ensure these implementation partners are able to meet quality 
requirements and project management standards. Provide better monitoring 
of nutritional activities under the lead of CEPAM to ensure objectives are achieved on 
time and with the desired performance. As an implementation partner, CEPAM has no 
internal monitoring and accountability system. 
Savings and Credit Community Group Component
• Participants in the savings groups would need more support, strengthening, and skill 
development to manage and support the groups independently.
Agricultural section
• The distribution of seeds to beneficiaries should be supported by a 
monitoring and evaluation system using beneficiary monitoring sheets to record 
the conditions of production, transfer and evolution of successive generations of crops 
63
and varieties. The members of the agricultural committees of local farmers’ 
associations can be trained to ensure the monitoring in this regard. 
• To ensure existing good conditions for the reception of the distributed seeds 
such as the existence of ecological fertilizers (manure and compost) whose absence 
compromised their performance.
• Allocate additional human resources to support the various stages of the 
agricultural component, from logistical planning for seed distribution to post-yield 
evaluations.
• There are no clearly defined exit strategies for the agricultural component beyond 
program support. WV needs to consider complementing agricultural 
interventions with broader actions to support local efforts to adapt to 
climate shocks, for example, providing training to communities on the Civil 
Protection Department’s early warning system, introduction of smart and climate￾smart seeds, promotion and introduction of eco-friendly agricultural practices, and 
establishment of firewood lots on steep slopes exceeding 40%.
• Integrate and ensure a good collaboration with MARNDR and the 
Communal Agricultural Offices during the whole project cycle. And above all, 
ensure that the National Seed Service is involved in the quality control of seeds to be 
distributed, and conduct germination tests through standard operating procedures 
(SOPs) that are clearly defined and approved by the WV team.
• Unlike other interventions, seed distribution must follow a strict schedule in 
accordance with the cropping calendar of the intervention areas. WV should 
strive in the future to respect the planting windows during seed distributions to ensure 
their full adaptation and the development of their productive potential. 
• WV should consider improving the productivity of local horticultural activities 
and also introduce other water-efficient crops such as the peanuts. This will help 
further diversify the agricultural economy and create a buffer against risks.
• Use a cascade training strategy (Training of Trainers) for agricultural 
training using already existing local structures such as the agricultural 
coordination of farmers’ organizations. This will ensure adequate transmission of 
knowledge and thus guarantee sustainability in the application of conservation and post￾harvest management techniques.
• Conduct analysis on gender and gender roles in farming and women’s 
working hours in order to produce adapted training modules that take into account 
women’s days and hours of availability and thus ensure their full participation and 
inclusion in training sessions. 
64
7) Bibliographical References 
1) Humanitarian Needs Overview (HNO), 2019, retrieved from: 
https://www.humanitarianresponse.info/en/programme-cycle/space/document/2020-
humanitarian-needs-overview-templates
2) Haiti, Supply and Market Outlook, Fews Net, September 2019 retrieved from the 
link: 
https://reliefweb.int/sites/reliefweb.int/files/resources/Haiti_Perspectives_de_lOffre_e
t_du_Marche_Septembre_2019_Juin_2020.pdf
3) UN, Designing Household Survey Samples: Practical Guidelines, 2005.
4) http://documents.worldbank.org/curated/en/673911467986337757/Haiti-Country￾partnership-framework-for-the-period-FY16-FY19
5) World Bank: Haiti, offering opportunities to all Haitians
6) FAO Haiti: Hurricane Matthew Situation Report, 2017
7) Haiti, Rural Livelihoods Profile, FEWS NET, March 2015.
8) An Investment Framework for Nutrition, World Bank 2017
65
8) Appendix I: FCS Indicator 
These groups include different types of foods such as:
Group No. Food Group Weighting
1 Staple food: Corn, rice, sorghum, other grain, roots and tubers 
(potatoes, yucca, yam, sweet potatoes, large breadfruit, small 
breadfruit) and plantain
2 
2 Legumes: White beans, black beans, red beans, pinto beans, green 
beans, nuts, peanuts (and other similar foods)
3
3 Vegetables/Leaves: Lyann panye, spinach, chives, cabbage, 
pumpkin, tomatoes, onions, broccoli, radishes (and all kinds of 
similar vegetables)
1 
4 Fruits: Mango, papaya, guava, apricot, cantaloupe, pineapple, 
orange, melon, watermelon, quince, cherries, lemon, grapefruit, 
avocado, banana, apple, plum, tamarind, strawberry, pear (and all 
kinds of fruits)
1
5 Meat, poultry and offal: goats, pigs, sheep, cows, horses, 
chickens, turkeys, guinea fowls, pigeons, liver, kidneys, hearts, 
intestines, offals, brains, (and all other types of meat)
Seafood: Fresh fish, salted fish, salted cod, crabs, shrimp, (and all 
kinds of seafood)
4
6 Milk and dairy: Cow’s milk, powdered milk, canned milk and 
batch milk, yogurt (and all other similar products)
4
7 Sugar and honey: White sugar, red sugar, honey (and all other 
similar products)
0.5
8 Oils and fatty products: Vegetable oil, olive oil, butter, 
shortening, fat (and all other similar products)
0.5
9 Spices/drinks: Coffee, tea, spices (parsley, thyme, garlic, clove), 
salt, fish powder, creamer
0
These scores were established based on information or experience from other surveys around 
the world. For the purposes of this survey, the following thresholds were considered: 
Poor food consumption: score between 0 and 21
Borderline food consumption: score between 21.5 and 35 
Acceptable food consumption: score above 35
66
9) Appendix II: HHS Indicator 
Households will be asked about their experiences of hunger in the last four (4) weeks/30 days 
preceding the survey. Their answers allow us to classify them into three (3) categories: mild 
or no hunger, moderate hunger and severe hunger. The first level of the scale is considered 
an acceptable or normal situation from the food access standpoint.
Three (3) frequency answers (Never=0, Seldom or Sometimes=1, Often=2) 
A score is calculated for each household (summing the three (3) responses), with a minimum 
possible score of 0 and a maximum possible score of 6. 
Three (3) categories of hunger are thus defined:
a. "No or mild hunger in households" (scores 0-1)
b. "Moderate hunger in households" (scores 2-3)
c. "Severe household hunger" (scores 4-6) 
67
10) Appendix III: Reduced Coping Strategy Index 
(rCSI) 
Survival strategies were categorized according to their severity :
Category Behavior Weighting
Stress Strategy
Buying or borrowing food on credit 2
Borrowing money 2
Spending savings 2
Use more casual work than usual 2
Crisis strategy
Selling productive assets 3
Removing children from school 3
Reducing health and education expenses 3
Emergency strategy
Sending household members to beg 4 
Selling the last breeding females 4 
Migration of the whole household 4 
In the so-called reduced strategy, only five (5) standard elements (standard strategies) are 
taken into account with their weighting, which tells their severity. 
No Strategy Weighting
1 Eat cheaper but less preferred food 1
2 Borrow food or money from friends or family 2
3 Reduce portion sizes in meals 1
4 Reduce adult consumption so that children eat more 3
5 Reduce number of meals a day 1
The maximum possible value of the score is 56 since a household uses all 5 strategies over all 
7 days. 
68
The situation was evaluated according to the index value:
Group Index Value
Less serious Less than 10
Moderate Between 10 and 19
Serious Between 20 and 29
Very serious More than 30
Source: CNSA, 2017
69
11) Appendix IV: Data Collection Tools (in separate 
files) 
1) M&E Team Interview Guide
2) Interview guide for the person in charge of access to drinking water and hygiene
3) Interview guide for the Agriculture Manager
4) Interview guide for the Food Voucher Distribution Manager
5) Interview guide for the S4T Manager
6) Project Manager Interview Guide
7) Interview guide for salespeople participating in the project
8) Interview Guide for the MARNDR/BAC Manager and Local Representative
9) Interview guide for the head of CEPAM
10)Interview guide for the MAST manager and representative
11)Community Leaders Interview Guide
12)Interview guide for the beneficiaries of each project component 
13)Interview guide for managers of other food security programs in the region and/or 
project target areas.
14) Quantitative questionnaire to evaluate project indicators