s
 
 
June 2014 
CARE Haiti
Final Evaluation of the Food Voucher Program 
“Kore LavniNou – 2”
implemented in North-west& Upper Artibonite
Departments
Final Report
Contact
Diagnostic & Development Group
3, Rue VilleValeix, Peguy Ville, Pétion-Ville, Haïti
www.ddghaiti.com
info@ddghaiti.com
(509) 442-47176
2 
Acronyms and Abbreviations 
ACF Action Contre la Faim 
ASEC Assemblée de la section communale 
BMI Body Mass Index 
CASEC Conseil d’administration de la section communale 
CNSA Coordination nationale de la Sécurité Alimentaire 
HTG Haitian Gourde 
MARNDR Ministère de l’Agriculture, des Ressources Naturelles et du 
Développement Rural 
M&E Monitoring and Evaluation 
MOU Memorandum of Understanding 
PIN Personal Identification Number 
WFP World Food Program 
WHO World Health Organization 
3 
Table of content
Acronyms and Abbreviations.....................................................................................................2
List of figures.............................................................................................................................5
List of tables...............................................................................................................................6
Executive Summary ...................................................................................................................7
1. Introduction ......................................................................................................................9
1.1 Context.............................................................................................................................9
1.2 Scope of the evaluation ................................................................................................. 11
1.3 Summary of the Methodological approach ................................................................... 11
Data analysis................................................................................................................... 14
2. Key findings of the evaluation....................................................................................... 15
2.1 Analysis of the project’s relevance............................................................................ 15
2.2 Analysis of the project’s effectiveness and efficiency............................................... 16
2.2.1 Project’s outputs..................................................................................................... 16
2.2.2 Mechanism selection of beneficiary households and criteria................................. 18
2.2.3 Activities and indicators......................................................................................... 19
2.2.4 Project’s management and partnership .................................................................. 20
2.2.4.1 Overall management ....................................................................................... 20
2.2.4.2 Partnership mechanism ................................................................................... 21
2.2.5 Issues to achieve results and strategies used to overcome them ............................ 21
2.2.5.1 Issues with the community committees .......................................................... 21
2.2.5.2 Issues with the selection of the operating shops owners................................. 22
2.2.5.3 Issues with the electronic food voucher.......................................................... 23
2.2.5.4 Satisfaction with the project............................................................................ 24
2.3 Project’s effects......................................................................................................... 25
2.3.1 Effects on beneficiaries.......................................................................................... 25
2.3.1.1 General situation ............................................................................................. 25
2.3.1.2 Food availability .......................................................................................... 29
2.3.1.3 Children Nutrition........................................................................................ 33
2.3.1.4 Consumption and awareness of fruits and vegetable consumption................. 34
2.3.1.5 Effects on gender dynamic.............................................................................. 36
2.3.2 Effects on operating shops and petty traders.......................................................... 37
2.4 Project’s sustainability .................................................................................................. 38
3. Conclusions................................................................................................................... 40
3.1 Key lessons learnt...................................................................................................... 41
4 
3.2 Recommendations ..................................................................................................... 42
Annex:..................................................................................................................................... 43
5 
List of figures 
Figure 1. Kore Lavni Nou 2 electronic food voucher process 
Figure 2. Petty traders’ satisfaction level (%) 
Figure 3. Use of the savings due to food vouchers during the lifespan of the project (%) 
Figure 4. Fruit and vegetable consumption (%) 
Figure 5. Training on fruits and vegetable consumption awareness 
Figure 6. Profit allocation by petty traders 
6 
List of tables 
Tableau 1. Distribution of the samples by commune 
Table 2. Sample characteristics (% of households) 
Table 3. Number of interviews and focus groups held and number of participants 
Table 4. Beneficiary household composition 
Table 5. Large capacity operating shops distribution per section communal 
Table 6. Beneficiaries’ satisfaction level 
Table 7. Credit during the 4 weeks prior to the survey (%) 
Table 8. Food consumption during the day of the survey (% of yes) 
Table 9. Number of meals consumed per day 
Table 10. Food situation during the baseline and four weeks prior the survey 
Table 11. Coping strategies used by households 
Table 12. Percentage of beneficiary households sharing food 
Table 13. Households with children presenting malnutrition symptoms 
Table 14. Nutrition situation of 5-10 years aged children based on BMI 
Table 15. Frequency of fruit and vegetable consumption 
Table 16. Knowledge on nutriment element in fruits and vegetables 
Table 17. Distribution of women headed household by commune 
Table A1: Logical framework of Kore Lavni Nou – 2 
7 
Executive Summary 
This document reports findings from the final evaluation of Kore Lavni Nou – 2, a food voucher 
emergency relief project implemented by CARE in three communes of the Upper Artibonite and 
North-West Departments of Haiti. This intervention aimed at providing food to the vulnerable 
households affected by the extensive drought and recurring food insecurity to offset harmful coping 
strategies of the beneficiaries and support them in recovering assets. CARE and its partner pursued 
the following objectives: 
 providing food vouchers as a short-term safety net for 8,000 food insecure and extremely 
poor families to fill gaps exacerbated by the disasters and to reinforce local markets in 
Gonaives, Mole Saint Nicholas, and Terre Neuve and, 
 piloting healthy diet promotion with fresh fruit and vegetable paper vouchers approach in 
Terre Neuve commune to reach 1,000 food insecure and extremely poor families. 
CARE commissioned Diagnostic & Development Group to evaluate relevance, effectiveness, 
efficiency and effects of Kore Lavni Nou-2. We used both qualitative and quantitative 
methodology: a) focus groups and interviews with relevant stakeholders, and on other hand b) 
survey of representative samples of beneficiary households, non-beneficiary household, operating 
shops owners and petty traders. Data collection was held from May 13th to May 22nd in Môle 
Saint-Nicolas, Gonaives and Terre-Neuve. Using those data and the ones from baseline, the 
project’s effects were analyzed via first difference and double differences or difference in 
differences techniques.
Local authorities and beneficiaries acknowledged that Kore Lavni Nou – 2 was relevant to alleviate 
hunger in their community.Artibonite and North-West departments were severely affected by the 
shocks of hurricane Sandy. Households lost small farm plantations, livestock and assets. National 
Coordination for Food Security (CNSA) reported that prevalence of food insecurity reached 43% 
in the North-West department and the rate of prevalence of severe food insecurity reached 41 % in 
Artibonite department in end of 2012. From this study, local authorities and beneficiaries reported 
that natural disasters and severe food insecurity were the main issues of the communities. Some 
women reported that loss of their subsistence small farms lead to difficulty to feed newborns and 
children. In general, households reported that coping strategies were threefold: cutting trees for 
charcoal, increasing debts and consumption of immature mangoes with salt. 
Specific criteria and participation of community leaders eased the implementation process. Local 
authorities, registered vendors and beneficiaries also acknowledged that Kore Lavni Nou – 2 was 
innovative and safer compared to previous food distribution programs in the communities. 
Beneficiaries reported that they enjoyed the privacy of redemption process respecting their dignity. 
Following initial redemption, effective solutions were found to address crowd and pushing in 
Gonaives. In fact, CARE was challenged by low supply capacity of local shops and their informal 
status. Other challenge emerged from holdup in negotiating with Action Contre la Faim (ACF) a 
common strategy for promotion campaign of fresh food in Terre-Neuve.
Indeed, beneficiaries could hedge against hungry during project’s food supply. From the baseline 
to the final evaluation, in Terre-Neuve the percentage of household with children at risk of acute 
malnutrition dropped from 31,2 % to 28,1% (-3,1 points) while it increased for the control group at 
33,8%, which means a net decrease of 5,7 points. In Gonaives a net decrease of 2,1 points was 
observed. However, in Môle Saint-Nicolas, the situation worsened for 2 points. Coping strategies 
of households were improved from initial situation: borrow food dropped from 83,2% to 55,7%; 
begging from 48,1% to 17,4%, the households that decreased the number of meals they usually 
consumed daily decreased from 100% to 82,4%. Nonetheless, the contagion effect due to 
8 
beneficiaries sharing with non-beneficiaries limited the conclusion regarding the net effect. Kore 
Lavni Nou-2 contributed to recapitalization of households by cumulating savings from food 
vouchers. Those savings were mainly allocated to children education (34%), to health service care 
(20,9%) and to invest in livestock (9,5%) or in savings (9,5%). Similarly, vendors and petty traders 
reported that sales volume at least doubled and gross margins increased. 
This project was a short term emergency intervention. Interviewed beneficiaries and community 
leaders indicated resurgence of previous hunger situation. Petty traders in Terre-Neuve reported a 
decline in their number of clients, after end of Kore Lavni Nou-2, despite beneficiaries claimed a 
higher share of consumption of vegetable. Local authorities’ wondered why CARE did not invest 
or reinforce agriculture and irrigation systems for a sustainable intervention. 
Key lessons learnt are important for initiatives targeted or similar communities. Kore Lavni Nou￾2 revealed the prominence of community involvement and definition of specific, tangible criteria 
to select poor families among vulnerable ones. Second, assessment of signal coverage is necessary 
prior project design in order to choose either electronic or paper food voucher. Third, beneficiaries 
are sensitive and value self-respect in food distribution program. Fourth, nutrition awareness 
campaigns related to the fresh fruit and vegetable consumption are important to even it is hard to 
separate a consumption shift from necessity in such distress living conditions. Fifth, CARE and its 
partners should address leaders’ critics that those short sight interventions have no impact on the 
root causes of hunger and drought. Finally, implementers should be aware to mitigate unintended 
results of increasing dependence on food assistance by household. 
Following recommendations are set forth to inform future design of similar projects: 
Community members: A predefined quota of elected and non-elected community members 
should be used to counter balance political influence from deviating the selection process from 
purpose. Programs should budget trips and refreshment for committee participants. Elder without 
national ID should be prioritized as part of their exclusion in society. 
Electronic voucher system: Cost of electronic voucher program can be improved by using 
company who have pre-established money transfer system or technology. Therefore, price should 
not be increased with additional number of clients but opposite. Using beneficiary picture and a 
unique identification number should ease the process against fraud. 
Promotion of local fresh fruit and vegetable and local production: The fresh fruit program 
should be extended to other areas. Vegetables are short term crops, they are more likely to respond 
to demand within a short term project. 
Finally, short term intervention should be part of a more holistic and comprehensive program to 
address the root causes of the hunger and natural disasters. Partnerships should be built with 
government technical agencies and other non-governmental organization to leverage positive 
effects of emergency project such Kore Lavni Nou-2. 
Even, dependence behavior on food assistance can be alleviated by sustainable conditional 
community project such planting trees, irrigation system, watershed protection which may have a 
higher community pride component than only receiving food.
9 
1. Introduction 
This introduction section presents the general implementation context of the food voucher program, 
Kore Lavni Nou-2, in Upper Artibonite and Northwest in the beginning of year 2013. We provide 
a brief description of the situation leading to the conception and implementation of this project 
including the scope of this current evaluation. This section is closed by presentation of the 
methodology adopted to carry out the evaluation. 
1.1 Context 
Food insecurity and malnutrition affects a high proportion of Haitian population and those issues 
have been aggravated by natural disasters shocks such as hurricanes Isaac and Sandy1
 in October 
2012. Following hurricane Sandy, the National Commission for Food Security (CNSA)2
 estimated 
that 1.5 millions of peoples were in severe food insecurity conditions while six millions are 
vulnerable. Chronic food insecurity is directly related to insufficient food intake (quantity and 
quality) and precarious environmental and health conditions. Food insecurity is concentrated more 
in regions of the country with drought, poor road conditions and poor access to markets. As a result, 
food prices are increasing in those markets based on the lack of supply to local markets and mainly 
by low purchase power of habitants of those rural areas. In fact, Artibonite department, despite its 
larger size and its agricultural potential, had the highest proportion of habitants suffering from 
severe hunger which was 41,3 %, while this proportion was 21,7 % in the case of North-West 
department (CNSA, 2012)3
. For instance, about 5% of the 6-59 months children in the Artibonite 
department suffer from acute malnutrition4
. Emergency responses to this acute situation came 
mainly from the public sector from its social programs and non-governmental organization working 
in the region with the support of funders such USAID. 
Built on its experience in the region and a recent successful food security program, CARE proposed 
Kore Lavni Nou-2 to address hunger issue of extremely poor households. The project aims to 
provide food vouchers to protect the health of affected households, offsetting harmful coping 
strategies, and assisting them to recover from the loss of livelihood assets” by contributing to 
increase food security for 8,000 extremely poor and food insecure households in Upper Artibonite 
and North-West departments. CARE and its partners pursued the following objectives: 
 provide food vouchers as a short-term safety net for 8,000 food insecure and extremely 
poor families to fill gaps exacerbated by the disasters and to reinforce local markets in 
Gonaives, Mole Saint Nicholas, and Terre Neuve and, 
 pilot healthy diet promotion with fresh fruit and vegetable paper vouchers approach in 
Terre-Neuve commune to reach 1,000 food insecure and extremely poor families. 
Kore Lavni Nou-2 is a replication of Kore Lavni Nou, an innovative electronic food voucher 
program implemented by CARE in the Grand’Anse department. This program worked with local 
registered merchants that supplied dry foods to the beneficiary households. Being aware of the need 
 
1
 Haiti- Hurricane Sandy, Fact Sheet USAID. February, 2013. 
http://www.usaid.gov/sites/default/files/documents/1866/02.15.13%20-
%20Haiti%20Hurricane%20Sandy%20Fact%20Sheet.pdf 
21,5 million d’Haïtiens dans l’insécurité alimentaire élevée, Le Matin, 
http://www.lematinhaiti.com/contenu.php?idtexte=34399
3 Rapport d’évaluation d’urgence des impacts du cyclone Sandy sur la sécurité alimentaire.CNSA. PP 
36. http://documents.wfp.org/stellent/groups/public/documents/ena/wfp254368.pdf
4http://siteresources.worldbank.org/INTLACREGTOPNUT/Resources/HaitiNutritionAssessmentEngli
shFINAL.pdf
10
to promote diet diversity and intake of fresh fruits and vegetables, in partnership with Action Contre 
la Faim (ACF), CARE piloted a combined dry staple and fresh food voucher component of the 
program in Terre-Neuve. Beneficiaries in Môle Saint-Nicolas and Gonaives received only a 
voucher of 50 USD for dry food but those in Terre-Neuve benefited monthly 38 USD electronic 
voucher for dry foods such as beans, rice, corn and oil, and a 12 USD paper voucher for fresh fruits 
and vegetables5
.
In collaboration with community committees, the list of the beneficiaries who were eligible for the 
participation in the program was determined based on the project defined vulnerability criteria. 
Once the selection completed, the list of beneficiaries was provided to Digicel, a Haitian 
telecommunication company, to be uploaded on a centralized electronic platform. Digicel provided 
training to the participating merchants on the use of mobile transfer system, the reimbursement 
process at Fonkoze (cash-out point after the redemption of the vouchers). CARE field staffs were 
also trained to provide ongoing support. 
Operating local shops owners were approached and selected according to pre-established criteria 
to supply the beneficiaries with a basket of products composed of the above-mentioned items while 
petty traders supplied Terre-Neuve beneficiaries also with fresh fruits and vegetables. 
The beneficiaries were provided a voucher card and a pin scratch card. The PIN scratch card serial 
number was electronically linked with the voucher card number and computed in the electronic 
platform. Every month, the voucher was credited and beneficiaries were advised by mobilizers to 
visit a registered local merchant to redeem the food. Figure 1 below summaries this process. 
 
5Each household received a monthly transfer of 2000 HTG. 
Care Crediting account of 
Digicel
Digicel credit CARE 
account on Merchant Pro 
Platform on equivalent 
electronic money value 
CARE credits the voucher 
of beneficiaries
Beneficiaries redeem their 
voucher to merchants. 
Beneficiaries receive foods 
from merchants
( 3 weeks) 
DIGICEL send redemption 
report to FONKOZE in a 
3rd week interval.
Payment of merchants by 
FONKOZE
3rd week 
Figure 1. Kore Lavni Nou-2 electronic food voucher process
11
1.2 Scope of the evaluation 
As per the terms of reference, the objectives of this evaluation were to:
 Assess the outcomes of the project, intended and unintended, positive or negative, on the direct 
beneficiaries and the population living in the 3 target communes of the project versus non￾beneficiaries. 
 Determine the adequacy of the program design to meet the identified needs in the target area. 
 Compare the final conditions with the baseline findings and present persisting gaps in coping 
strategies used by the households to mitigate their vulnerability. 
 Relate lessons learned to appropriate stakeholders. 
 Identify key obstacles preventing the correct execution of the program. 
 Identify key successes and recommendations for future consideration. 
 Evaluate the partnership approach in the commune of Terre-Neuve and added value of the 
combined methodology for the food vouchers. 
To reach these objectives, first, we analyzed the project’s objectives and intended results in the 
lenses of the needs of the targeted population and their community. In order words, the evaluation 
looked at to what extent the activities and outputs of the program were consistent with the overall 
goal, the specific objectives, the intended effects, and how these specific objectives and expected 
effects met the identified needs in the targeted areas. 
Second, we sought to understand the extent to which the objectives were achieved or not. In the 
case the expected results were attained, we also analyzed the strategies that have been used to 
minimize the gap between planned and realized activities. 
Through another ax, we assessed to what extent the intended outcomes were achieved, and what 
net effect they produced in the daily living of the beneficiaries, the local markets and the 
community. We also sought to identify and evaluate the intended and unintended, positive or 
negative effects of the project on the beneficiary families and participating merchants. 
Lastly, the evaluation considered the extent to which the observed effects, if any, could be 
maintained beyond the project’s period. Therefore, we analyzed to what extent the program created 
long-term positive changes in the community, and how long those changes could be expected to be 
maintained after the project end. 
1.3 Summary of the Methodological approach 
This section presents a brief summary of the methodology designed to carry out the evaluation 
study. A detailed methodological note was submitted to CARE prior to data collection. 
To achieve the above-mentioned objectives, a mixed perspective was used: quantitative and 
qualitative approach. Data were collected from the following key informants: 
- CARE Staff 
- Mobile company responsible for the electronic food voucher service provision 
- Monitoring and evaluation staff 
- Local authorities 
- Local shops 
- Beneficiary households 
- High risk groups as under five children/handicapped and pregnant women 
- Control group (non-beneficiary households) 
12
We acquired also some documents generated during the lifespan of the projects that helped us in 
formulating a picture of the implementing course of the project, albeit a partial one. The documents 
surveyed included: 
- Project’s proposal 
- Narrative reports 
- Baseline study report 
- Data collection tools 
To complete our picture of the project, information was collected from all the stakeholders. 
Quantitative data were collected from a sample of beneficiary and non-beneficiary households and 
were compared to those of the baseline study.Non-proportional stratification was used to select the 
sample of beneficiaries, non-beneficiaries and shops in the three targeted communes (Gonaives, 
Môle St-Nicolas and Terre-Neuve). 
Considering the finite population of 8000 households, to guarantee a 95 % confidence level and a 
4.5% margin error, the minimal size of the sample of beneficiary households should be 448 but was 
rounded up to 500 beneficiaries to balance potential effects of non-answers. This margin of error 
was considered based on budget and timeframe constraints. 
For the control group (non-beneficiary households), the sample size was 60% of that of beneficiary 
group, say 300 non-beneficiary households. Table 1 records the distribution of the samples in the 
different communes. 
Tableau 1. Distribution of the samples by commune
Commune Section 
communale
Beneficiaries Non-beneficiaries Children
Targetted Realized Targetted Realized Realized
Labranle 125 129 75 76 7
Gonaïves Poteaux 125 131 75 75 4
Subtotal 1 250 260 150 151 11
Cote-de-Fer 62 69 38 35 0
Môles St-Nicolas Mare-Rouge 62 67 37 55 10
Subtotal 2 124 136 75 906 10
Bois Neuf 63 67 38 40 5
Terre-Neuve Lagon 63 63 37 40 8
Subtotal 3 126 130 75 80 13
Total 500 526 300 321 34
The control group was chosen from households having similar characteristics as the beneficiary 
group (table 2). They were selected from a different distance of beneficiaries and randomly by an 
interval of five households. Indeed, the proportion of households with one or more members 
suffering from chronic diseases was respectively 4,2% and 5% in the case of beneficiaries and non￾beneficiaries. In both groups, around 31% had at least one child fostered to other household. Other 
 
6High variances in the number realized compared to the number targeted are due to the fact some 
surveyors surveyed more respondents than what was requested. However, this does not affect the quality of 
data since the selection procedure was respected and because the size sample target was the minimal size 
required.
13
characteristics such as households with handicapped members or elderly people were quite similar 
in both cases, except for pregnancy cases. Indeed, the proportion of households with at least one 
pregnant woman was respectively 3,8% and 8,1% for beneficiary and non-beneficiary groups. 
 Table 2. Sample characteristics (% of households)
Characteristics Beneficiaries Non-beneficiaries
Households with one or more members suffering from chronic 
diseases 4,2 5
Households with children placed in domesticity 30,9 30,8
Lactating women 2,5 1,6
Pregnant 3,8 8,1
Households with physical handicapped members 15,4 12,8
Households with one or more members suffering from Mental 
diseases
2,5 1,8
Elderly people
12 15,6
Trained surveyors used a set of qualitative and quantitative tools to collect required data. For 
households, the questions were principally about their coping strategies, the types and frequency 
of food consumption, assets possession, etc. The goal was to analyze the conditions of the 
beneficiary group before and after the project as compared to the non-beneficiary group. 
Concerning local merchants, the questions addressed principally the value of sales, etc., before and 
after the project. A special questionnaire was also designed to assess the fruit and vegetable 
promotion program in Terre-Neuve. The questions were principally about respondents’ awareness 
of nutritional value of the fruits and vegetables, and the type and frequency of fruits and vegetables 
consumption. The questionnaire was administered to both beneficiary and non-beneficiary group. 
We also collected anthropometric measures (height, weight, waist/ hip circumference, arm 
circumference) on a random sample of 30 children aged between 5 and 10 years because they 
represented a very high risk group. These measures were used, among others, to compute Body 
Mass Index (BMI) and compared with Child Growth Standards fixed by the World Health 
Organization since they were not collected by the baseline study. 
To complete the quantitative data, interviews and focus groups were held with stakeholders of the 
project: local authorities (mayors, ASECs, CASECs), beneficiary households, participating local 
merchants, CARE staff, Monitoring and Evaluation (M&E) staff, etc. In addition, some focus 
groups were held with mixed groups (both males and females) and with only females in order to 
assess the effects of the projects on gender dynamics. 
Table 3 reports the number of interviews and focus groups realized and the number of participants 
of each. 
14
Table 3. Number of interviews and focus groups held and number of 
participants
Interviews Number of interviews conducted
1. CARE management staff 2
2. M&E staff 1
3. CREER/CECRET 1
4. ACF 0
5. Digicel 1
Total 5
Focus groups Number of focus 
groups
Number of 
participants per 
focus group
1. Beneficiaries (men and
women)
3 8
2. Beneficiary women 3 8
3. Local merchants 3 5
4. Petty traders 1 5
5. Local authorities 3 5
Total 13 31
Data collection was held from May 13th to May 22nd in the 3 targeted communes. 
Data analysis 
Triangulation was used to capture key findings from the qualitative data derived by concordance 
analysis. In this perspective, different sources of information were used to address each key 
question of the evaluation. However in some cases, we choose to report stand-alone but contextual 
comments (quotations) to illustrate specific general situations. 
Data related to the project’s effects were analyzed using first difference and double differences or 
difference in differences techniques. First difference allows comparing separately the situations of 
the beneficiaries and non-beneficiaries before and after the program (baseline data vs. final 
evaluation data). For instance, suppose the percentage of beneficiary households who used a 
specific coping strategy is A and C during the baseline and the final evaluation respectively, 𝑋 =
𝐶 − 𝐴 represents the first difference. That allows assessing the improvement of the situation during 
the project. However, to evaluate to what extend this improvement could be attributed to the project, 
we computed on one hand the same first difference for the control group, which suggests what 
would this improvement be were it not for the project. And, on other hand, the net effects of the 
project were assessed by computing the difference of these two first differences, the difference in 
differences. Suppose that the first difference is 𝑌 = 𝐵 − 𝐷 in the case of the control group. The net 
effect of the project on the aforementioned coping strategy is the double difference as measured by 
X − Y. Nonetheless, this technique supposes minimal or zero contagion between the two groups. 
15
2. Key findings of the evaluation 
This section presents the key findings of the evaluation. First we discuss the project’s relevance in 
the lenses of the acute food insecurity prevalence in Upper Artibonite and Northwest departments. 
This perspective of analysis is crucial for the interpretation of the results while reminding that Kore 
Lavni Nou-2 was an emergency response with purpose to mitigate the consequences of natural 
shocks. Second, the evaluation focuses on the project’s effectiveness and efficiency analyzing the 
outputs, the management structure and partnership and the strategies adopted to overcome 
difficulties. Afterwards, we present an analysis of the project’s effects on both direct and indirect 
beneficiaries; based on the short term scope of the project, we avoid using the term impacts. Lastly, 
the section is ended with general discussions on the project’s sustainability. 
2.1 Analysis of the project’s relevance 
The Artibonite and North-West departments were severely affected by the shocks of hurricane 
Sandy. People lost their cultures, livestock and assets. They are among the departments where food 
insecurity has been the highest. According CNSA7
, the prevalence rate of food insecurity has 
reached 43% of population from the North-West department. After the drought season prevailing 
in this department, this coordination observed that the households are out of food stock and then 
should be supplied from the market since the winter crops composed of maize and beans were 50% 
lower than its normal level. The report8 of CNSA following Sandy hurricane also confirmed that 
Artibonite was the department with highest rate of severe food insecurity.
Drought seasons and acute food insecurity were the main issues that came out from the focus groups 
with both local authorities and beneficiaries. Some beneficiaries reported having lost all (gardens, 
arable lands, livestock and even houses) after hurricane Sandy. Therefore, it was hard to provide 
food for household’s members and even newborns had to face the lack of food. A woman described 
her poignant living conditions following hurricane Sandy and affirmed: “I lost all during the 
hurricane; even my house was destroyed. If it wasn’t for friends to provide the clothing, I would 
have nothing.” The most common coping strategies reminded by the beneficiaries were cutting 
trees to make charcoal for sale on the market, contracting debts for basic needs or buying food on 
credit, and consuming immature mangoes with salt as vegetables. 
Another participant stressed the erosion and unproductive environment faced by thousands of 
peoples in the region of North-West department. As 
this woman, some peoples lost their relatives. She 
acknowledged: 
“In the aftermath of hurricane Sandy we lost our land, 
only the stones remained. Even before the hurricane we 
had to cut down a lot of trees because of hunger. We 
have no livestock and no land. All other persons that 
could help us are now dead.” Another woman reported 
that one night they had nothing to eat, and then they 
boiled premature mangoes with salt. 
 
7http://www.cnsa509.org/Web/Bulletin_Observatoires/OSANO/BULLETIN%20OSANO%2001.pdf
8Rapport d’évaluation d’urgence des impacts du cyclone Sandy sur la sécurité alimentaire. CNSA. PP 
36. http://documents.wfp.org/stellent/groups/public/documents/ena/wfp254368.pdf
Credit: DDG/Focus group 
with beneficiaries
16
Buying food on credit was also a very common coping strategy reported by the beneficiaries from 
almost all the “sections communales”9
. Some of them evoked they were sometimes outraged by 
merchants because they could not pay for food. The latter sometimes went to their home to claim 
their money. When that happened, they said they stayed inside and told the children to tell the 
merchants that they were absent. In addition, child malnutrition rate10 was12,7 % and 12,1 % 
respectively for North-West and Artibonite department and counted among highest rates while it 
was about 7% in the metropolitan area of Port-au-Prince. 
In such a context, the goal “Increased food security for 8,000 extremely poor and food insecure 
households in the Artibonite and North-West Departments of Haiti”, and the objective to “Provide 
food vouchers as a short-term safety net for 8,000 food insecure and extremely poor families to fill 
gaps exacerbated by the disasters and to reinforce local markets in Gonaives, Môle Saint-Nicolas, 
and Terre-Neuve”, at a first glance seem to be relevant. The needs were manifested. Therefore the 
goal and objective seem to be consistent with the needs of the targeted communities. 
Both local authorities and beneficiaries acknowledged that the project was of great support to 
mitigate the food crisis that prevailed in the sequel of hurricanes Sandy. During the six months of 
the lifespan of the project, the beneficiaries could hedge against hunger. However, both groups 
legitimately reported that at the end of the project the living conditions of the beneficiaries become 
the same as before the start of the project. Therefore, the project filled in the needs only temporarily 
since it was an emergency response and the severity of food insecurity is deeper than seasonal 
shocks or occasional natural disasters.
2.2 Analysis of the project’s effectiveness and efficiency 
This section focuses on the project’s effectiveness and efficiency. First, we analyse the outputs as 
understood by the stakeholders and how they were carried out through different activities. Second, 
we discuss the general management approach that was adopted from the design of the project to 
the delivery of services to the beneficiaries. Subsequently, we consider the main issues encountered 
to achieve those outcomes and the strategies developed to overcome them. The section closes with 
an analysis of the stakeholders’ satisfaction with the services provided through the project. 
2.2.1 Project’s outputs 
As mentioned previously, Kore Lavni Nou-2 was a short response intended to contribute to CARE’s 
primary humanitarian mission as well as USAID’s objective to support the people affected by 
natural disasters. The project also fell under Haitian Government objective to deliver emergency 
response to people in North-West department in the sequel of the last two years natural disasters. 
According to CARE staff, the project attained the 8,000 targeted beneficiaries in the all three 
communes except for the first two months where around 290 vouchers were not all redeemed but 
265 were authorized to have a last redemption at the end of the project due to the lack of capacity 
of program vendors for exchanging the coupons. This situation resulted from signal issues, and 
principally happened in Lagon. 
As mentioned in the beginning of the report, each household received a monthly transfer of 2000 
HTG for buying dry food except for the commune of Terre-Neuve where they have received 1600 
HTG for the dry food and 400 HTG for fresh food. Household size was respectively about 8,6 and 
7 in Gonaives, Terre-Neuve and Môles St-Nicolas as depicted in Table 4 below. Therefore, on 
 
9A section communale is the smallest territorial division in Haiti.
10 EMMUS V
17
average, each household member received food for an equivalent 262,58, 366,28 and 287,83 HTG 
in Gonaives, Terre-Neuve and Môle St-Nicolas respectively. This suggests that at an individual 
level, each household member in Terre-Neuve would benefit a greater consumption share from 
voucher than household member of the other two communes. 
In general, the monthly transfer amount was greater than the monthly food-consumption-related 
expenses reported by the households at the baseline moment. Indeed, at this time, about 28% 
estimated such expenses between 2004 and 4000 HTG, and about 51% between 404 and 2000 HTG. 
Thus, Kore Lavni Nou-2 contributed to considerably improve access to food for the beneficiary 
households. 
Table 4. Beneficiary household composition
Gonaïves Terre￾Neuve Môles St Nicolas Overall
Household average size 7,62 5,46 6,95 6,92
Average number of under 18 people 2,81 2,48 3,30 2,88
Average number of under 5 children 1,58 1,40 1,67 1,56
Ratio voucher value/Household size 262,58 366,28 287,83 287,83
All the interviewed stakeholders seemed to understand and discuss the project objective, which 
proves dissemination of information before the start of the project. The local merchants and the 
beneficiaries acknowledged training sessions were held periodically from the start to the lifespan. 
The local authorities participating to the focus 
groups acknowledged that the objectives of the 
project were discussed before the implantation but 
they criticized the short term approach that they 
considered insufficient to mitigate effects of chronic 
food insecurity. It seemed that they did not 
understand it was an emergency response project. 
They pointed out that at the end of the project the 
livelihood conditions of the beneficiaries go back to 
the starting point, since the project did not intervene 
on the causes of food insecurity but rather its effects. 
A local authority stated the following: “The conditions of the beneficiaries are now worse than at 
the start of the project; they were used to struggle with themselves to feed their family. The project 
brought hope to them for six months, and after this hope is broken.”
Local shop owners also shared the concerns of local authorities. All agreed that the project was 
relevant in the aftermath of hurricanes Sandy. However, they pointed out the number of 
beneficiaries selected was too small as compared to the scope of vulnerability in their areas. They 
also express the wish that those interventions tackled also structural challenge as drought and land 
productivity. One of them pointed out “There are so many vulnerable households in my area that 
the number of selected households was insufficient […]. The people really needed the food project. 
However, it would be better to strengthen the project with an irrigation component. This might 
better mitigate the vulnerability situation of the people.” 
Credit: DDG/Focus group 
with local authorities
18
2.2.2 Mechanism selection of beneficiary households and criteria 
According to World Food Program (WFP) to be considered food secure, a household should have 
all-time access to sufficient and nutritious food. This entails food availability in sufficient and 
consistent basis, food access through home production, purchase, etc. and food utilization in terms 
of nutritional impact on the households members11
.
Moreover, the World Bank defines vulnerability as “the probability or risk today of being in poverty 
or to fall into deeper poverty in the future”. In the lenses of this definition, vulnerability is then 
difficult to measure since evaluating the probability of being in poverty is hard to estimate. 
Therefore, this institution suggests income, consumption and variability as proxies for 
vulnerability12. Vulnerability analysis can be carried out considering variables such as health status, 
weigh, asset ownership, income level, etc.13
We also found the contribution of World Health Organization (WHO, 2002) complementary as it 
defines vulnerability as the degree to which a population, individual or organization is unable to 
anticipate, cope with, resist and recover from the impacts of disasters14. Therefore, children, 
pregnant women, elderly people, malnourished and people who are ill are at risk groups in disasters 
and emergency situations. 
Taking account of those considerations and the socio-economic context of the targeted regions, the 
selection process had to be specific, transparent and non-equivoque. From those lenses, the 
selection of beneficiaries based on the following criteria was relevant:
i. Women- headed families with multiple children 
ii. Families with adult members with chronic debilitative diseases, infirmity caused by old 
age or mental illness, or handicapped, not able to engage in sustained livelihoods and 
earn income, 
iii. Child- headed household 
iv. Pregnant and lactating women 
v. Orphans and internally displaced persons 
vi. Households with children who are “at-risk” or with acute malnutrition 
Since vulnerability is hard to measure empirically, it results that it is also a difficult task to rank 
households according to their degree of vulnerability. In socio-economic surveys it is common to 
construct socio-economic indices using sophisticated statistical models based on proxies such as 
households’ assets (both quantity and types), health status (including presence of at risk people), 
income, etc. Although this approach allows a more “objective” ranking of the households, however 
it may be both financially and time prohibitive since it requires census of the targeted populations. 
Therefore, such techniques cannot be applied in short term projects such as Kore Lavni Nou-2. 
To select the beneficiary households, CARE used a participatory approach involving community 
committees composed of mayors, ASECs, CASECs, religious authorities, etc., which, to our 
opinion, is a good alternative since those people can help differentiating between the most 
 
11 https://www.wfp.org/node/359289
12http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTPOVERTY/EXTPA/0,,contentMDK:
20238993~menuPK:492141~pagePK:148956~piPK:216618~theSitePK:430367,00.html
13Ibid
14Environmental health in emergencies and disasters: a practical guide.
19
vulnerable based on the proximity with the community. This approach also gives more credibility 
to the selection mechanism. 
Whether the selection criteria are well defined and the participatory approach adequate and 
credible, some local factors linked to the vulnerability seemed not to be considered which affected
the selection process. First, According to local authorities, to be selected, the beneficiaries also 
needed to have an identification document such as “Carte d’identification nationale”, birth 
certificate, etc. In some remote areas, some extremely poor elderly are living alone and are deprived 
of all even identification documents. Living alone and lack of national documents is evident sign 
of vulnerability. One CARE staff also reported the cases of people who were initially selected by 
the local authorities and who did not have any identification document. Thus, special measures 
should be taken to select those particular elderly. 
Secondly, some stakeholders acknowledged that some selected beneficiaries were not the most 
vulnerable. That was due to the fact that some remote areas are hard to access even by motorcycles, 
or because the community committee members did not have enough time to go to such areas. It 
emerged from the focus groups that some local authorities and other committee members did not 
go to some remote areas because they knew they would not be reimbursed their travelling fees. 
Though they recognized it was a good way to contribute to alleviate the hardship of the social 
conditions of the people, most of them reported that the project should at least reimburse such fees. 
One of them stated “We took part to all the meetings. Some of us live very far; we had to pay the 
travelling fees ourselves. CARE did not even give us water, though we were dying from thirst. So 
days we were looking for potential beneficiaries from 7am to 4 pm; some of them live very far from 
us.” Another one reported: “People from some remote areas are the most vulnerable; they are 
deprived of everything. For instance, 3e Terre-Neuve cannot be accessed by car neither 
motorcycles. When I got back from there, I could not do anything except going to bed.” 
The effectiveness of the beneficiary selection was also discussed in the focus groups with the local 
shops owners but in a more contrasted way than with the local authorities. Some operating shops 
owners accused the local authorities to have selected people that were not vulnerable at all; they 
selected people who voted on them. They also mentioned that at redemption days, some 
handicapped persons and elderly who were not selected came to their shops with the hope to receive 
food, while some beneficiaries were very young. They also reported that in some cases the holder 
of the voucher did not even live in the community, some of them dwelling in Port-au-Prince. This 
point of view was contrasted by other participants who clearly indicated that the most vulnerable 
people were actually selected; in some cases, they are either handicapped or very old, so they could 
not come for redemption and had to send someone else to get the food for them. 
2.2.3 Activities and indicators 
The following results were expected from the execution of Kore Lavni-2: 
(i) Increased access to staple foods at local markets. 
(ii) Increased awareness of the importance of fruits and vegetables in the diet. 
(iii)Increased access to these food products at local markets. 
These expected results as defined are verifiable in a reliable and credible fashion. However, we 
think that the expected results 1 and 3 are quite similar. Furthermore, a set of objectively verifiable 
indicators were defined in the project log frame as means of verification. Thus, the project could 
successful monitor activities and assess progress during implementation and achievements at
completion. However, since results 1 and 3 are similar, so are the corresponding indicators (See log 
frame in annex: Table A1). 
20
2.2.4 Project’s management and partnership 
Two key points are analyzed in this subsection. First the overall project’s management as seen by 
the different stakeholders is evaluated, and second the evaluation considers the partnership between 
CARE and ACF, and between ACF and CRECER/CECRET. 
2.2.4.1 Overall management 
From our analysis, overall, the management structure of the project was effective despite some 
issues encountered during the course of the project. The central coordination was based in Port-au￾Prince, while some managerial staffs were based in Gonaives, a strategic area for both Artibonite 
and North-West departments in terms of geographical considerations. To mitigate errors in 
exclusion and selection of beneficiaries, community committees participated to beneficiary 
selection since in such situation or quasi general poverty it is hard to differentiate the most 
vulnerable. 
A project manager coordinated the field activities and supervised a team of mobilizers based in 
different communes. Joint visits were organized with ACF team to contribute to the coordination 
in Terre-Neuve (where the two organizations were working with the same beneficiaries). 
The local authorities, the vendors and the beneficiaries acknowledged the good coordination of the 
project as compared to other previous food distribution initiatives. Before the start of the project, 
meetings were held with all stakeholders to explain the project’s objectives and its structure. 
Periodic trainings were held with beneficiaries and vendors based on the feedback received from 
M&E activities that proved to be very effective. Among others, M&E activities included post 
redemption survey of a random sample of beneficiaries to verify whether they actually received the 
food and price survey of samples of vendors to monitor prices for comparison with prices of the 
local markets. A vendor affirmed the following: “Every month CARE met with us to discuss the 
previous month outputs in order to analyze the gaps and how we should correct them. […] This 
partially explains the success of the project.” 
The local authorities pointed out that the effectiveness of the project’s coordination allows 
supplying the beneficiaries with dignity, which was highly promoted by CARE. In this perspective, 
one of them stated: “As compared to other programs, CARE respected the beneficiaries’ dignity. 
In other post cyclonic programs, the people were fighting to get food. Some of them even died. ” 
The beneficiaries also acknowledged an effective planning and respect of beneficiaries’ dignity. 
Generally, food distribution is complicated in areas where almost all the people are food insecure, 
which involves overcrowded distribution, physical insecurity, and often the use of force by the 
police or security staff. The use of the voucher system and the different points of redemption 
increased and facilitated the distribution of food in Kore Lavni Nou2. The fact that each participant 
has 30 days to redeem the voucher eliminated crowd and stigmatization compared to previous food 
distribution project in target communities. A woman from Labranle testified that: “It is the first 
project in the community where pregnant women did not breathe tear gas.” A vendor supported 
this observation by stating: “It is never easy to distribute food. It is for the very first time I observe 
food distribution activities where no one was beaten with a baton and no one was injured. This is 
the result of good planning […]”. 
Besides planning, the technology is one of the components that made the difference of Kore Lavni 
Nou-2 with others.
21
2.2.4.2 Partnership mechanism 
Based on lessons learnt from Kore Lavni Nou executed in the Grand’Anse department, CARE 
decided to reinforce Kore Lavni Nou-2 with health diet promotion of fresh fruit and vegetable 
consumption in Terre-Neuve. CARE partnered with ACF to implement this component. According 
to CARE, ACF demonstrated a good record of field mobilization capacity and experience of 
implementing fresh food vouchers in targeted communities. 
Interviews with CARE staff revealed that there were some difficulties to set up a common strategy 
for the coordination of the fresh fruit and vegetable consumption promotion campaign. It was then 
hard to have a planning accepted by both CARE and ACF teams. Efforts to share information and 
periodic meetings with both teams seemed to help solve this problem. Unfortunately, though our 
efforts, ACF did not complete the interview guide. The field manager hesitated to have a face to 
face interview without authorization of his supervisor, we sent them the interview guide and it was 
not returned properly. 
ACF also partnered with CREER a “cooperative” based in Gonaives with an extension in Terre￾Neuve (CECRET) with the purpose of reaching out to the petty traders at the markets and paying 
them back with no significant delays. The representative of CREER provided good feedback on the 
partnership with ACF. CREER did not participate in a voucher program in the past since its main 
activities are savings and microfinance. The representative expressed complete satisfaction with 
the experience and indicated that CREER would be prepared to enter into the same partnership in 
the future. The petty traders confirmed they were reimbursed by CREER with no delay. However, 
CARE notice that “a better organization on reporting format would have been needed.” 
2.2.5 Issues to achieve results and strategies used to overcome them 
The implementation of the project faced several types of issues mainly in the beginning. However, 
all the key respondents unanimously acknowledged that timely and appropriate actions were taken 
to mitigate them. In the following, we discuss the principal sources of issues encountered. 
2.2.5.1 Issues with the community committees 
One of CARE staff interviewed reported that members of the community committees complained 
on the fact neither themselves nor their relatives could benefit from the project. Finally they 
understood and accepted the rules. However some of those who participated in the focus groups 
still complained and did not demonstrate such understanding; they considered themselves as 
vulnerable as well as some of their relatives, which is reflected in the following declaration: “Some 
of the community members are vulnerable, really vulnerable; neither themselves nor their family 
could benefit from the project. It is hard to feed others while you or your family is hungry. That 
may generate conflict with your relatives since they actually are hungry.” 
As mentioned in the previous section, the local authorities and some other committee members also 
complained on the fact they were not provided with even travelling fees. Some of them had to travel 
up to five times a month to identify potential beneficiaries. For example, travelling from Terre￾Neuve to Lagon cost 350 HTG. 
However, according to this staff most of the local authorities were keen on continuing without paid. 
These complaints may be somewhat surprising since as aforementioned the community committees 
were composed of elected persons whose job is to represent and seek the interest of the population. 
Thus, it is clear that they could not benefit from the project and it is unfair that they considered 
themselves vulnerable as compared to other people. Furthermore, to avoid conflict of interest, their 
22
relatives could be selected neither. Nevertheless, taking account of the poor financial of the local 
authorities in such areas, it is worth budgeting their transport fees for long distance in order to 
increase efficiency and participation. 
2.2.5.2 Issues with the selection of the operating shops owners 
It was intended that local operating shops owners would be selected according to the following 
criteria: (i) fully registered businesses (with all permits and tax registration) or able to quickly 
become fully registered (merchants will be given a period of time within which to complete 
registration with the Government of Haiti); (ii) willing to sign an MOU with CARE Haiti; (iii) 
willing to undergo training and abide by the project’s guidelines; (iv) able to maintain adequate 
quality food stocks in a timely manner; and (iv) at least 1 year experience in food sales. 
In our opinion, these criteria seem to be relevant, although in rural areas some might be hard to 
meet. Indeed, not surprisingly, in some areas, no local operating shops satisfied such criteria and 
few of them were fully registered business; the informal sector dominating Haitian economy. For 
instance none of the local merchants operating in Lagon could be selected, and only 6 were selected 
in Gonaives to supply 75% of the beneficiaries, say 6,000. In Terre-Neuve, there was only one 
merchant but with very limited supplying capacity. Those issues resulted in complaints of 
beneficiaries for redemption during the first month of the project. In Gonaives they register some 
incidents like pushing during redemption because some vendors were not able to supply all the 
beneficiaries. Fortunately, a rapid solution was implemented by linking bigger suppliers with small 
vendors in some sections communales and asking other qualified vendors to open local shops in 
remote areas where none of the existing local merchants met the previously mentioned criteria (see 
Table 5). This we think was an effective strategy to counterbalance the low supplying capacity of 
the local operating shops. 
23
Table 5. Large capacity operating shops distribution per section communale
Department Communes Sections Communales Name of operating 
shops
.Artibonite
Gonaïves 2eme Section Bassin (Praville)
Mr & Mme Brice Delva 
Provisions Alimentaires 
Terre-Neuve Terre Neuve (Centre Bourg)
Gonaïves 4eme Section Poteaux (Carrefour
Gros Morne)
Gonaïves 3eme Section Bayonnais (Vieux 
Poste)
Gonaïves 4eme section Poteaux 
Gonaïves 1ere section Pont Tamarin
Artibonite
Gonaïves Gonaïves KGP Trading
Gonaïves Gonaïves 
Gonaïves Gonaïves 
Gonaïves Gonaïves 
Artibonite
Gonaïves Gonaïves Raco Produit 
Alimentaire Gonaïves Gonaïves 
Gonaïves Gonaïves 
Artibonite
Gonaïves Gonaïves Sel et Lumuiere Multi 
Service Gonaïves Gonaïves 
Gonaïves Gonaïves Martine Darius Oxima 
Provisions Alimentaires Gonaïves Gonaïves 
Gonaïves Gonaïves 
Môle Saint-Nicolas
Môle Saint-Nicolas 1ere Section- Cote de Fer 
(Rodolphe) Le Grand Dieu de 
Jehovah Môle Saint-Nicolas 2eme Section- Mare Rouge
Moreover, some merchants did not want to participate in the program. They did not trust such a 
payment system. They evoked their bad experiences in the past with the cooperative system. Local 
merchants from Grand’Anse who participated to Kore Lavni Nou were invited to share their 
experiences with potential participating local merchants and some were willing to participate only 
after the “success” of the first redemption. 
2.2.5.3 Issues with the electronic food voucher 
According to Digicel there were many issues with the technical component of the project. Some 
names were duplicated. Some people tried to register in the system more than once by moving from 
an area to another during the registration process. There also were some printing mistakes such 
identical number. Sometimes, the registration process took longer than expected and then increased 
stress in the process. One CARE staff interviewed confirmed the duplication issues and the cases 
of beneficiaries with different names but the same PIN number.
The voucher amount was only scheduled to be valid for 30 days, allowing redemption during this 
period. Upon the expiration of a 30-day period, unused amount was returned to CARE account. 
There also were cases of PIN loss and network coverage issue. Digicel was asked to issue another 
PIN to those beneficiaries. Some vendors reported issues with Digicel signal. Normally once the 
24
PIN number is entered, the vendor receives a success message. In some cases, the vendors received 
the message “requête en cours” (request is being proceed) or no message at all. In such cases, the 
vendors did not know whether the transaction was completed or not. According to these vendors, 
the beneficiary then had to go to CARE office for verification purposes. Afterwards, CARE staff 
called the vendors to ask them to deliver the food to the beneficiary. 
According CARE, “Digicel significantly improved their capacity to tweak all technical problems, 
however they are still somewhat disorganized as they are shifting the main platform and this affects 
their human capacity to cope with multiple projects”.
2.2.5.4 Satisfaction with the project 
In general, all the stakeholders seemed to be satisfied with the project. As previously mentioned, 
the local authorities, the beneficiaries and the local merchants expressed their satisfaction with the 
management structure of the project as compared to other food distribution programs. 
Globally, the beneficiaries did not report issues with the local merchants. They confirmed 
redemption happened in time and they received the pre-established basket of food. Thus, this 
confirms the achievement of the expected result 1. 
However, some beneficiaries and local authorities mentioned in the beginning some operating 
shops owners did not treat the beneficiaries with courtesy. However, they recognized that over the 
course of the project obvious changes were observed. The local authorities explained this change 
of behavior as consequences of the periodic trainings provided by CARE field staff. “[…] after 
that, those vendors supplied the beneficiaries with a big smile on their face” said a local authority. 
The beneficiaries also explained the courtesy of local vendors by the fact these latter knew that the 
beneficiaries could use their voucher at any of the local shops that participated to the project. One 
of them stated the following: “The voucher is mine as well as the money. Thus, if a vendor did not 
treat me with courtesy, I went to another one to redeem the food.” 
Those qualitative data are reinforced by the quantitative one. Indeed, from Table 6 below we may 
observe that all the surveyed beneficiaries were satisfied with the project and the quality of the 
products while only 0,7% reported being dissatisfied with the services from the vendors. 
Table 6. Beneficiaries’ satisfaction level 
Questions Satisfied Dissatisfied
Are you satisfied with the project? 100,0 0,0
Are you satisfied with the quality of the products? 100,0 0,0
Are you satisfied with services from the local 
shops? 99,3 0,7
All the local merchants seemed to be satisfied with the project. They expressed their satisfaction 
with the general management of the project. They particularly appreciated the monthly meetings 
with CARE which allowed discussing on the general issues and above all on solutions. They felt 
safe to use the voucher system though, as me mentioned earlier, at the beginning some did not trust 
it and did not want to integrate the project. Figure 2 shows that the participating pretty traders 
(Terre-Neuve) also were totally satisfied with the project. They felt safe using the paper voucher 
and the corresponding payment system. Few reported issues with some beneficiaries who wanted 
fruits and vegetables for only part of the voucher and requested the balance in cash. 
25
2.3 Project’s effects
This section of the report discusses the effects of the project on beneficiaries and local merchants 
(Operating shops owners and petty traders). 
2.3.1 Effects on beneficiaries 
This section discusses the effects of the project on the beneficiary households. It starts with a first 
general overview of the situations after the closure of the project as compared to their initial 
conditions. These discussions are principally based on information from the focus groups that are 
reinforced with some quantitative data. 
2.3.1.1 General situation 
The general context in which the project was implemented may be fairly described by this 
participant to a focus group with beneficiaries at Labranle who stated: “The project was 
implemented while we faced drought problems. It did not rain at all; all the trees died. The six￾month project helped us a lot and was very useful for our children.” 
Other beneficiaries interpreted the project implementation as a divine intervention, which is 
reflected in the following affirmation held by a woman “Our most urgent need was food and 
education for our children. [...] only God and mother and our father would do what USAID did for 
us.” Another woman participating to the same focus group supported: “I do not have a garden. I 
did not know where to go. When I was selected in the project, it was like God got down from heaven 
for me.” Thus, a local authority legitimately affirmed: “The people saw the project as windfall 
getting down from heaven [...]” 
All the stakeholders acknowledged the relevance of the project as we reported it in section 2.1 and 
then recognized that an emergency response was crucial to alleviate the burden of food insecurity. 
In such a context, long term positive effects are hard to achieve since the project did not intervene 
on the food insecurity factors. Thus, as reported at the beginning of this general section, the local 
authorities legitimately considered that the objective of the project was too much limited in time to 
produce long term positive changes. In this perspective, a beneficiary woman reported: “Now we 
are facing the same drought issues as before. It does not rain and life is very hard. Children have 
68.6
91.3
78.3 82.6
30.4
8.7
21.7 17.4
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Paper
voucher
system
Entire
project
Payment Security
of the
paper
voucher
Figure 2. Petty traders satisfaction level (%) 
Satisfied
Completely satisfied
26
to eat premature mangoes with salt as coping strategies to lack of food [...]”. Three other women 
from Terre-Neuve supported with the following: 
- “[...] somebody gave me those three mangoes; I am going to smoke and eat them.” 
- “Before the project I struggled with myself to eat; I cut off trees. Unfortunately, I have to 
do the same now.” 
- “Recently I sent someone to buy some flour on credit, but the merchant refused because I 
did not have money. We then went to bed without eating. We got back to the same point 
after the end of the project.”
For more potential long term effects, local authorities suggested to strengthen such food projects 
with interventions on agriculture and irrigation. Other suggested conditional aid such as requiring 
each beneficiary household planting a fruit tree during the lifespan of project. We found this 
suggestion interesting; however this would also require seedling distribution to the beneficiaries 
and therefore additional costs. Many local authorities and operating shops owners reported that 
during the project fewer trees were cut off than before. This also came out from the focus groups 
with beneficiaries. However, since the structural causes of food insecurity such as natural shocks 
threaten and the social conditions of the people are still the same, those effects were only very short 
as pointed out above. 
One could manifestly observe a general sense of despair from the beneficiaries participating to the 
focus groups. The general sentiment is that they are desperately waiting for another food project to 
come. Most of them emphasized they got back to the same point at the end of the project. The same 
coping strategies because of lack of financial means are used: cutting tree, borrow food on credit, 
eat premature food, etc. In such a context, the short term effect of the project may obviously be 
diluted. 
One of the main effects of the project is recapitalization of participants from savings derived from 
food costs. The savings from food vouchers were allocated principally to education, health, debt 
reimbursement and investment in livestock according to the surveyed beneficiaries (Figure 3). 
Overall, the main use of this saving was on children education (34%). This is a significant result 
since at the baseline, among the respondents with children who were not attending school, 81,3% 
reported their children were out of school because they could not afford the fees. The second 
allocation item of the savings due to food vouchers was health care services (20, 9%) (Physician 
fees, medicines, pills, etc.), followews by investment in livestock assets or saving (9, 5 % each 
one). In Môle Saint-Nicolas, despites education had the priority, debt reimbursement and savings 
were the two other most important components (23,5% and 19,9% respectively), while in Terre￾Neuve, the two most important components were livestock acquisition (16,4%) and health (14,8%). 
27
Table 7 below seems to confirm the resurgence of coping strategy pattern prior the project. For 
both beneficiary and non-beneficiary households, debt was contracted in the three target areas 
mainly to cover food needs during the 4 weeks before the survey. Indeed, in all three communes 
there were more than 57% of beneficiaries who have been liability for food. Nonetheless, 21,6% of 
beneficiaries clearly said they did not take credit against 18,1% of non-beneficiaries, say a 
difference of 3,5%. This difference is greater in Terre-Neuve where more than 27,7% of the 
beneficiaries did not take credit 4 weeks prior the final evaluation survey against 16,7% of non￾beneficiaries. Unfortunately, no comparison can be done with the baseline results because those 
data were not collected at that time. 
34.0
45.8
12.5
31.6
20.9
28.2
14.8 12.5
9.5
5.3
16.4
11
4.0
1.9
10.9
1.5
9.5
7.6 2.3
19.9
6.5 0 1.6
23.5
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Figure 3. Use of the savings due to food vouchers during the 
lifespan of the project (%)
Debt reimbursement
Saving
Repair the house
Invest in petty trade
Buy seed food/agriculture
Buy livestock
Health
Education
28
Table 7. Credit during the 4 weeks prior to the survey (%) 
Credit Overall Gonaives Terre-Neuve Môle Saint-Nicolas
Beneficiaries
No 21,6 21,7 27,7 15,4
Yes 57,4 67,2 36,6 58,1
I wanted but I had no access 21,0 11,1 35,7 26,5
If yes, why?
Food 53,3 56,3 33,3 58,2
Health 12,3 12,5 20,0 7,6
Education 24,0 21,0 31,1 26,6
Housing ,7 1,1 ,0 ,0
Marriage/funeral 1,3 3,4 ,0 3,8
Investments (goods) 1,3 0,6 4,4 1,3
Family 3,0 0,0 4,4 2,5
Other 4,0 5,1 6,7 0
Non beneficiaries
Credit Overall Gonaives Terre-Neuve Môle Saint-Nicolas
No 18,1 18,3 16,7 12,5
Yes 48,3 58,2 26,4 61,4
I wanted but I had no access 31,2 23,5 56,9 26,1
If yes, why?
Food 41,3 40,4 25 46,3
Health 18,7 24,7 8,3 11,1
Education 19,4 19,1 50 13
Housing 1,3 1,1 0 1,9
Marriage/funeral 7,1 7,9 8,3 5,6
Investments (goods) 7,1 1,1 0 18,5
Family 1,9 1,1 0 3,7
Other 3,2 4,5 8,3 0
Tables 7 also give insight on the allocation of credit by those who effectively took on. In all three 
communes, and for both beneficiaries and non-beneficiaries, food purchase was the main reason 
(53,3% and 41,3% respectively). In Gonaives and Môle Saint-Nicolas, such beneficiaries counted 
for 56,3% and 58,2% of the corresponding samples respectively. The two other most components 
of credit allocation were education and health weighting 24% and 12,3% respectively of the overall 
sample. In Terre-Neuve these proportions are 31,1% and 20% respectively. These results are yet 
more eloquent recalling that during the life span of the project, education and health were the two 
main allocations of the savings due the project as declared by the beneficiaries (Figure 2). 
29
2.3.1.2 Food availability 
Both beneficiaries and non-beneficiaries were asked whether they had breakfast the day of the 
survey and whether they would have lunch or dinner the same day. For both questions, overall the 
percentage of non-beneficiaries who answered yes was higher than that of beneficiaries who gave 
the same answer (Table 8). Indeed, 25, 6% of the beneficiaries said they had breakfast against 33, 
2% for the control group. About 36% of the surveyed beneficiaries declared they would have lunch 
or dinner the same day against approximately 42% of non-beneficiaries. However, for beneficiaries, 
concerning lunch or dinner, in Terre-Neuve and Môle Saint-Nicolas those percentages were 61% 
and 63% respectively against 22,2% and 55,7% respectively for non-beneficiaries. 
Table 8. Food consumption during the day of the survey (% of yes)
Overall Gonaives Terre-Neuve Môle Saint-Nicolas
Beneficiaries
Did you have breakfast today? 25,6 20,7 48,4 23,2
Will you have lunch or dinner today? 36,1 26,7 61,1 62,5
Non-Beneficiaires
Did you have breakfast today? 33,2 37,0 24,4 30,4
Will you have lunch or dinner today? 42,2 43,8 22,2 55,7
Overall, the general dynamic of the number of meals consumed per day did not dramatically change 
from the baseline to the final evaluation (Table 9). At the start of the project, 70,3% of the surveyed 
people reported consumption of at most one meal per day. At the final evaluation, this percentage 
was about 54,5% and 56,9% for beneficiaries and non-beneficiaries respectively. That suggests an 
increasing of the number of meals consumed per day for both beneficiaries and non-beneficiaries. 
Although, the difference between beneficiaries and non-beneficiaries is only 2,4 points, the sharing 
of food phenomenon observed during the project could be one the factors that explain this low 
difference. Thus, the food voucher program might indirectly contribute to improve the food 
situation of the non-beneficiaries. 
Terre-Neuve seemed to show better improvement than the two other communes. In this commune, 
at the baseline time, 98,9 % of the beneficiaries declared they had at most one meal daily. At the 
final evaluation, those beneficiaries counted only for 18,3% against 55,6% for the non￾beneficiaries, which represents an improvement of more than 80% for the beneficiaries and more 
than 43% for the non-beneficiaries. This positive difference between Terre-Neuve and the other 
communes might partially be explained by the fact that the situation in Terre-Neuve was worse 
than in the two other communes and the diversification of the vouchers (mix approach). 
30
Table 9. Number of meals consumed per day 
Frequency of 
food
consumption Baseline
Final evaluation
First 
Difference*
Double 
difference**
Beneficiaries
Non￾beneficiaries A B C
Overall
Sometimes 0,4 4,9 6,7 4,5 6,3 -1,8
Once a day 69,4 49,5 50,2 -19,9 -19,2 -0,7
Twice a day 28,5 38,9 32,9 10,4 4,4 6,0
Three times a day 0,9 6,6 8,6 5,7 7,7 -2,0
No answer 0,8 0,1 1,6
Gonaïves
Sometimes 0,4 6,1 7,8 5,7 7,4 -1,7
Once a day 60,2 52,3 63,4 -7,9 3,2 -11,1
Twice a day 36,9 38,2 24,2 1,3 -12,7 14,0
Three times a day 1,4 2,7 4,6 1,3 3,2 -1,9
No answer 1,1 0,8 0,0
Terre Neuve
Sometimes 1,1 0,8 1,4 -0,3 0,3 -0,6
Once a day 96,8 17,5 54,2 -79,3 -42,6 -36,7
Twice a day 1,1 60,3 16,7 59,2 15,6 43,7
Three times a day 0,0 21,4 25,0 21,4 25,0 -3,6
No answer 1,0 0,0 4,2
Môle Saint-Nicolas
Sometimes 0,0 5,9 9,1 5,9 9,1 -3,2
Once a day 69,1 73,5 23,9 4,4 -45,2 49,7
Twice a day 30,9 20,6 61,4 -10,3 30,5 -40,8
Three times a day 0,0 0 2,27 0 2,3 -2,3
No answer 0,0 0,0 3,4
* % of Beneficiaries-% in Baseline=A and % of non-Beneficiaries-
% in Baseline=B
** A-B=C
In general some changes had been observed in food insecurity situations as compared to the baseline 
(Table 10). Food insecurity situation being measured by the percentages of respondents being in 
the following situation 4 weeks prior to the corresponding survey: worried or concerned because 
of lack of food, obliged to eat food that he/she did not like because of lack of food and go bed 
without eating. At the baseline, in all the 3 communes, nearly 100%15 of the respondents reported 
being in such situation, while at the evaluation moment, in Gonaives, about 85% , 87% and 88% 
of the beneficiaries reported being concerned because of lack of food, obliged to eat food that they 
 
15 The declaration of mostly 100% in critical situation during the baseline suggests a potential 
anticipation effect by the population. 
31
did not like and got to bed respectively. This represents an improvement of about 15%, 13% and 
12% of the food situation, which is a critical change for a short term emergency project. By 
comparing these results with those of the non-beneficiaries, the difference in favour of the 
beneficiaries is not significant. However, as was previously mentioned, the effect of sharing the 
food with non-beneficiary neighbors might one possible factor to explain this low difference 
between the two groups. In Môle Saint-Nicolas, in the case of the beneficiary group, these figures 
(respectively 84,6%, 83,8% and 82,4%) were very similar to those of Gonaives; however, the non￾beneficiary group “exhibited a better performance than the beneficiary group”. For instance, only 
63,6% of the non-beneficiaries reported going to bed without eating against 82,4% for the 
beneficiaries, which accounts for a difference of 18,8% in favour in the non-beneficiaries. Again, 
the contagion bias observed may affect these results. 
In the case of Terre-Neuve, where the fresh fruit and vegetable voucher awareness campaign was 
piloted, it seems at a first glance that food insecurity situations were significantly mitigated as 
compared to the baseline situation. Indeed, the percentage of beneficiaries who reported being in 
the above-mentioned situations 55,6%, 55,6% and 49,2% compared to 100% for all the three cases 
at the baseline. This general “better improvement” of Terre-Neuve households might probably be 
attributed to the mix-approach used and the fact production of vegetables is more responsive to 
market demand. Vegetables are short term crops and they are planted in Sources Chaudes which is 
a commune close to Terre-Neuve even that some supply came also from Gonaives. 
Table 10. Food situation during the baseline and four weeks prior the survey
Food situation Baseline Final 
First 
Difference*
Double 
difference**
Beneficiaries
Non￾beneficiaries A B C
Gonaïves
Worried or concerned because 
of lack of food 99,3 85,5 85,6 -13,8 -13,7 -0,1
Obliged to eat food that you 
do not like
because of lack of resources
99,6 87,4 93,5 -12,2 -6,1 -6,1
Go to bed without eating 100 88,5 90,2 -11,5 -9,8 -1,7
Terre-Neuve
Worried or concerned because 
of lack of food 100 55,6 51,4 -44,4 -48,6 4,2
Obliged to eat food that you 
do not like
because of lack of resources
100 55,6 51,4 -44,4 -48,6 4,2
Go to bed without eating 100 49,2 56,9 -50,8 -43,1 -7,7
Môle Saint-Nicolas
Worried or concerned because 
of lack of food 100 84,6 76,1 -15,4 -23,9 8,5
Obliged to eat food that you 
do not like
because of lack of resources
100 83,8 72,7 -16,2 -27,3 11,1
Go to bed without eating 100 82,4 63,6 -17,6 -36,4 18,8
* % of Beneficiaries-% in Baseline=A and % of non-Beneficiaries-% in 
Baseline=B
** A-B=C
32
Table 11 depicts a list of coping strategies investigated. The percentage of beneficiaries who used 
such coping strategies systematically decreased for all the items from the baseline to the final 
evaluation. For instance, the proportion of households who borrowed food decreased from 83,2% 
to 55,7% (decreasing of 27,5%); at the baseline moment 48,1% of the respondents declared 
begging as coping strategy, at the final evaluation moment, this proportion decreased to 17,4% 
(decreasing of 30,7 points). The highest decreasing from baseline to the final evaluation (44,5%) 
was observed for the restriction of the consumption of adults so that children can eat. Thus, albeit 
the beneficiary complained their situation after the project goes back to the starting point, we can 
observe some important changes in the food insecurity situation. 
Although the non-beneficiary group once again exhibits better performances than the beneficiary 
group, these results should interpreted with attention. Indeed, it was observed from both focus 
groups and the quantitative survey a general solidarity situation between beneficiaries and non￾beneficiaries. Hence, uncontrolled contagions effects might potentially bias the results. For 
instance, from the focus groups a beneficiary declared: “The project intended to help us mitigating 
the hunger issues. Not all of us had vouchers. After redemptions, I shared the food with others who 
did not have vouchers. A lot of beneficiaries did the same.” 
Table 11. Coping strategies used by households
Strategies Baseline
Final evaluation
First 
Difference
Double 
difference
Beneficiaries
Non￾beneficiaries A B C
Borrow food 83,2 55,7 52,4 -27,5 -30,8 3,3
Restraint the consummation of 
adults so that children can eat 95,5 51 44,4 -44,5 -51,1 6,6
Buy food on credit 97,2 87,6 81,5 -9,6 -15,7 6,1
Consume seed stocks 71,7 37,8 41,5 -33,9 -30,2 -3,7
Begging 48,1 17,4 18,2 -30,7 -29,9 -0,8
Count on the support of a 
neighbor/ a family member to feed 
the household 84,7 59,5 49,8 -25,2 -34,9 9,7
Work for food 63,8 70,2 60,4 6,4 -3,4 9,8
Perform a risky job to acquire food 74,7 55,5 48,2 -19,2 -26,5 7,3
Consume inappropriate aliments 
(expired products) 86,2 35,4 33,9 -50,8 -52,3 1,5
Reduce the number of meals that 
you usually consume a day 100 82,4 77 -17,6 -23 5,4
* % of Beneficiaries-% in Baseline=A and % of non-Beneficiaries-% in 
Baseline=B
** A-B=C
Indeed, overall 96,3% of the surveyed beneficiaries declared having shared the basket of food they 
received from Kore Lavni Nou-2 with others. Generally, the food was principally shared with 
neighbours and family members respectively (Table 12). 
33
Table 12. Percentage of beneficiary households sharing food 
Share of food Overall Gonaives
Terre￾Neuve Mole Saint-Nicolas
% of yes 96,3 98,3 84,8 98,7
Whom with?
My children 1,4 0,0 7,5 1,3
Mother/Father 0,3 0,0 2,5 0,0
Other family 34,0 29,5 52,5 34,6
Neighbors 62,2 67,6 37,5 62,8
Other people 2,1 2,9 0,0 1,3
This positive solidarity situation was confirmed from the focus groups with the local authorities 
who considered that positive effect of the project. One of them reported that some beneficiaries 
wanted to share the food with him. Since he declined the offer, the beneficiary offered him other 
gifts such poultries. A participant to a focus group with the local merchants stated: “Sharing with 
others is part of rural communities culture. I have observed that after redemption some 
beneficiaries shared the food with others who did not have vouchers.” 
2.3.1.3 Children Nutrition 
The respondents were asked whether there were under-five children at their household with 
malnutrition symptoms (the symptoms were marasmus, oedemas, beriberi, etc.). In all three 
communes, the percentage of respondents who answered yes is reported in Table 13 where a 
decreasing tendency may be seen for the beneficiary group. This decreasing was respectively 0.3%, 
3.1% and 3.6% for Gonaives, Terre-Neuve (where the mix approach was used ) and Môle Saint￾Nicolas. A better result of improvement of children nutrition was observed in Terre-Neuve which 
might partially be the result of piloting the fresh fruit and vegetable consumption campaign.
Table13. Households with children presenting malnutrition symptoms
Commune Baseline
Final evaluation First Difference*
Double 
difference**
Beneficiaries
Non￾beneficiaries A B C
Gonaïves 10,6 10,3 12,4 -0,3 1,8 -2,1
Terre-Neuve 31,2 28,1 33,8 -3,1 2,6 -5,7
Môle Saint-Nicolas 12,4 8,8 6,8 -3,6 -5,6 2
* % of Beneficiaries-% in Baseline=A and % of non-Beneficiaries-% in 
Baseline=B
** A-B=C
34
Anthropometric data were also collected on a sample 
of 34 children. Eleven (11) were under five and 13 
aged between 5 and 10. Four of the under-five had 
unilateral oedemas and one bilateral oedemas. 
The Body Mass Index is a standard nutrition measure 
computed by comparing the weight of the children 
with their size; it allows to classify children based on 
their nutritional development. This measure was 
computed for 21 children aged between 5-10 years 
(only beneficiaries), and based on the World Health 
Organisation standards, four categories were 
considered: under nutrition, underweight, normal weight and overweight. For the 21 children under 
study, 14 were undernourished, 6 underweighted and the other overweighed (Table 14). That 
suggests despites the project, the nutritional development of those children remains critical. 
2.3.1.4 Consumption and awareness of fruits and vegetable consumption 
Based on experiences and lessons learnt from Grand’Anse, CARE introduced a fresh food 
component for ration delivered to beneficiaries in Terre-Neuve. Thus, all the data presented in this 
section were collected only in Terre-Neuve. 
Close to 78 % of the beneficiaries reported their households consumed fruits and vegetables before 
the project. However 97,7% reported increasing consumption of such items during the project 
lifespan and attributed this shift in consumption to healthy diet promotion. Moreover, 85,5% 
thought their households would keep this habit (Figure 4). It is worth noticing the result below that 
children in Terre-Neuve had a better nutrition result than other targeted communes. 
77.9
97.7
85.5
0.0 20.0 40.0 60.0 80.0 100.0 120.0
Fruit and vegetable consumption
before the project
Increased consumption of fruits
and vegetables during the project
Fruit and vegetable consumption
after the project
Figure 4. Fruit and vegetable consumption (%) 
Table 14. Nutritional development of 5-10 years aged children based on BMI*
Nutrition situation Number %
Under nutrition 14 66,7
Underweight 6 28,6
Normal weight
- -
Overweight 1 4,8
Total 21 100
*Body Mass Index = weight/(size*size)
Credit: DDG/Focus group with 
local authorities
35
Both beneficiaries and non-beneficiaries were asked how often their household consumed fruits 
and vegetables. Table 15 shows significant differences between the two groups. Most of the 
beneficiary household seemed to consume such items at daily basis or between 3 to 6 times a week 
(43,1% and 46,2% respectively), while 42,6% of the non-beneficiary households would consume 
fruits and vegetables rarely. 
Table 15. Frequency of fruit and vegetable consumption
Frequency Beneficiairies
Non￾Beneficiaries
First 
difference
Eachday 43,1 12,8 30,3
3-6 times 
aweek 46,2 25,5 20,6
1 to 2 
daysaweek 7,7 19,1 -11,5
Rarely 3,1 42,6 -39,5
About 93% of the beneficiaries acknowledged they received training on fruits and vegetables from 
Kore Lavni Nou-2. Figure 5 depicts the main topics they reported of such trainings as reported by 
the respondents. They mentioned nutrition contributions of fruits and vegetables, difference 
between fruits and vegetables, nutrition importance of fruits and vegetables, etc. 
The beneficiaries (95%) also showed awareness of potential health issues associated with unwashed 
fruits and vegetables such as microbes and bacteria. More than 90% of them (95,5%) also indicated 
they properly washed such items before consumption; 78,5% of such beneficiaries indicated it was 
change of habit following the fresh fruit and vegetable awareness promotion campaign of Kore 
Lavni Nou-2. 
Lastly, we assessed the awareness of the beneficiaries of nutrition contribution of fresh fruits and 
vegetables as compared to the control group. Table 16 shows that overall the beneficiary group 
demonstrated a better awareness than the control group. Indeed, Respectively 85,6%, 82,3% and 
83,2% of them were aware that such items are sources of vitamins, minerals and energy compared 
93.1
96.9
96.0
93.6
93.6
93.6
91.0 92.0 93.0 94.0 95.0 96.0 97.0 98.0
Training on fruit and vegetable awareness
Importance of fruits and vegetables
Difference between fruits and vegetables
Nutrion contribution of fruits and
vegetables
Potentional health issues of consumption of
unwashed fruits and vegetables
Potentional health issues of consumption of
unwashed fruits and vegetables
Figure 5. Training on fruits and vegetable consumption awareness
36
to 79,5%, 70,5% and 78,6% for the non-beneficiaries respectively. However, a less significant 
proportion of beneficiaries were aware that fresh fruits and vegetables are also rich in carbohydrates 
and lipids. 
Table 16. Knowledge on nutriment element in fruits and vegetables
Nutriment element Yes No Don't know
Beneficiaries
Vitamins 85,6 8,0 6,4
Minerals 82,3 8,9 8,9
Carbohydrates 48,4 18,0 33,6
Lipids 58,1 28,2 13,7
Energy (calories) 83,2 12,0 4,8
Non Beneficiaries
Vitamins 79,5 6,8 13,6
Mineralelement 70,5 13,6 15,9
Glucide 25,0 45,5 29,5
Lipide 62,8 20,9 16,3
Energy (calories) 78,6 14,3 7,1
Based on the previous results, we can obviously see that the expected result 2 was achieved. 
2.3.1.5 Effects on gender dynamic 
Most of the project beneficiaries were women, representing 75,63% of the group. Recall that two 
of the six criteria used for the selection of the beneficiaries were women- headed families with 
children and pregnant and lactating women. We also wished to examine whether this 
“overrepresentation” of the women affected gender dynamic within the households.
Table 17 shows the percentage of women headed household for both the baseline and the final 
evaluation. These results are surprising. In case of the beneficiaries, we would expect increasing 
percentages for the final evaluation as compared to the baseline based on the selection criteria 
mentioned above. Nevertheless, the women headed households decreased from 53,4% to 38,5% 
between the two surveys. In Gonaives, the percentage decreased from 50,5% to 22,2%. Only Terre￾Neuve exhibited a converse tendency with 44,7% of such households at the final evaluation against 
40,4% at the baseline. 
Table 17. Distribution of women headed household by commune
Commune Baseline
Final evaluation
First 
Difference
Double 
difference
Beneficiaries
Non￾beneficiaries A B C
Overall 53,4 38,5 28,8 -14,9 -24,6 9,8
Gonaïves 50,5 22,2 25,0 -28,3 -25,5 -2,8
Terre-Neuve 40,4 44,7 34,6 4,3 -5,8 10,0
Môle Saint￾Nicolas 74,2 49,1 21,6 -25,1 -52,61 27,51
37
Discussions from the focus groups revealed not much change within the household dynamic, except 
in rare cases where women considered themselves as the “chef” (chief or head). In general, the 
participating women reported good relationship with their husbands during the project as it was 
before. One stated: “Before the project we used to discuss on what we should do to feed the children. 
Why should I want to be the head just because I benefited from the project?” Though less spread 
out, some women felt their husband were respectful towards then because of their new role to feed 
the household. One reported the following: “He had a better respectful attitude towards me. Once 
I said to him: you spoke a lot when you fed the household. Now, it is my turn, you are quiet.” 
Rather rare situations of men who did not want to keep on working because of the project were also 
reported. In this perspective a women from Terre-Neuve stated: “Some men even liked the food 
more than their wives. Sometimes they did not even want to go to their gardens. One of them fought 
with his wife for the food.” 
Those results are not surprising at all. First, it was well stated at the start of the project that the food 
was granted to the entire household and not to the voucher holder who only was a representative of 
the household. Second, the women were aware of the short term scope of the project. Some of them 
used such argument to explain why there could not be much change in the household dynamic. 
Lastly, gender dynamic is more rooted in the cultural background. 
2.3.2 Effects on operating shops and petty traders 
During its lifespan, the project boosted the local economy. Both participating operating shops 
owners and petty traders reported significant increasing in the volume sales during the project. 
Some operating shops owners mentioned that it was the first time the volume of sales went up to 
200 thousands HTG monthly. Before the project, some of them only supplied people in their 
community, but because of the general hard social conditions, sales were very low. Since they 
needed to supply up 100 or 200 people at redemption, some increased the number of employers 
from 1 to 4. 
In focus group discussions in Môles Saint-Nicolas, vendors mentioned that at the start of the project 
some vendors supplied up to 150 beneficiaries at redemption while others supplied less. Meetings 
were organized with all the participating local merchants in the community to discuss on the 
“problem”. Finally they came to the resolution that no participating local merchant should supply 
more than 100 beneficiaries in order to “give a chance” to all participating vendors to significantly 
increase the volume of sales. 
Petty traders in Terre-Neuve also significantly increased their sales during the course of the project. 
From the focus groups, some reported 500 HTG as maximal daily sales before the project. 
Participating to the paper voucher program contributed to some extent to triple or quadruple that 
amount. Therefore, profits significantly increased. Eighteen of the 52 participating petty traders 
reported having owned a bank account or been member of a cooperative. This number went up to 
34 during the project. Figure 6 shows the allocation of profit during the life span of the project. 
Principally profit was reinvested (95,7%). Only 17,2% bought livestock and in 12,9% of the cases, 
profit also contributed to feed the household. 
38
However most of the petty traders reported after the project ended they went back to the same sales 
figures as before the project. Some reported that after the project volume of sales rarely attained 
250 HTG. 
2.4 Project’s sustainability
Kore Lavni Nou-2 was a short term emergency program as a response to natural shocks and 
therefore cannot be expected to impact vulnerability and food insecurity roots. In the current 
national food vulnerability mapping, Artibonite department, despite its larger size and its 
agricultural potential, and North-West department are among those where food insecurity is the 
highest. Moreover, the project targeted the most vulnerable households, though most of them are 
actually vulnerable.
The project proved to have short term effects on the direct beneficiaries. Changes in negative coping 
strategies have been reported by the beneficiaries: decreasing in cutting trees, none of them bought 
food on credit during the project, etc. During the lifespan of the project, most of them used the 
savings due the food voucher project to education, health care, debt reimbursement, assets 
acquisition, etc. However the structural causes of food insecurity legitimately seemed to reverse 
this positive trend only some months after the project ended. 
The threat of absorption of the short term effects by the structural conditions of the people were 
respectively paraphrased by a beneficiary and a CARE staff as follows: 
- “We bought some livestock during the project. For instance the money allocated to buy 
rice was used to buy a chicken or a goat. Now the project is ended, we need to send the 
children to school and care of them. Then we sold the assets we acquired during the project 
since we cannot leave them without schooling.” 
- “[...] during the project the beneficiaries might use the money they would allocate to food 
to buy some assets such as poultry and livestock. However other crisis and natural shocks 
may absorb the projects effects. ” 
The project also proved to have positive short term effects and local economy. Participating shop 
owners and petty traders reported significant increase in sales during the project though the 
structural causes of poverty reverse the trend after the project course. 
Therefore, lessons learnt from such program may be very useful to other NGOs and government 
structures in designing long term programs intended to address the structural causes of poverty. 
Therefore more sophisticated models might be used to set up vulnerability criteria and to target the 
95.7
17.2
12.9
4.3
1.3
.0 20.0 40.0 60.0 80.0 100.0 120.0
Reinvestment
Buy livestocks
Feed the household
Agriculture/Seeds
Children schooling
Figure 6. Profit allocation by petty traders
39
most vulnerable households. Such long term programs should also involve the households in their 
own recovery from poverty for instance by helping them developing income generating activities 
such agriculture and others. While local authorities pledged for conditioning the assistance for 
instance by imposing each household planting a tree, none of the beneficiary expressed the desire 
to be boosted in order to set up income generating activities. Their only hope was the replication of 
the project. This dependency mentality came out all the focus groups. We explicitly asked them 
what they could to keep the project’s short term effects; the common answer was “CARE should 
renew the project”. None of them expressed the need for a “boost”. This may be a very “dangerous” 
attitude. 
The electronic food voucher may also be a good example of social programs that respects the 
dignity of the people. Local authorities, participating merchants and beneficiaries pointed out the 
difference between Kore Lavni Nou-2 and other food distribution programs. The voucher system 
allowed a better management of the distribution and avoid situations such pushing and fighting that 
generally happen in food distribution. 
40
3. Conclusions 
This report presented the key findings of the end of project evaluation of Kore Lavni Nou-2. This 
short term program was implemented by CARE in Upper Artibonite and North-West department 
as a response to natural disasters shocks that worsened food insecurity in the Communes of 
Gonaives, Terre-Neuve and Môle Saint-Nicolas. 
This project was a replication of Kore LavniNou implemented in Grand’Anse department. Based 
on reflections with partners, CARE decided to reinforce the electronic food voucher project with 
healthy diet fresh fruit and vegetable consumption. CARE decided to partner with Action Contre 
la Faim that demonstrated good field mobilization and fresh fruit and vegetable promotion 
experiences. This component of the program was piloted in Terre-Neuve. 
To carry out the evaluation, moreover desk review of the project documents, interviews, focus 
groups and surveys were conducted with all the stakeholders. A mix methodology (qualitative and 
quantitative) was used as a complementary strategy to gather richer and more in-depth information 
on the project’s course. 
Eight thousands of households were targeted as the most vulnerable based on a set of prescribed 
criteria. In a context where food insecurity is quasi general, it is not an easy task to distinguish the 
most vulnerable. Thus, local authorities such mayors, CASECs and ASECs were part of the 
beneficiary selection and mobilization in the three targeted communes, which gave a more level of 
accountability to the project. However, most of them complained on the fact that neither themselves 
nor their relatives could not benefit from the project albeit they also are vulnerable. Selection and 
mobilization of beneficiaries required many trips in different areas among them some are hard to 
access. 
Local authorities, participating vendors and beneficiaries acknowledged the good management of 
the project as compared to other similar programs. There were no hurts, no deaths, no fighting 
during redemption except at the start of the project where pushing happened in Gonaives because 
of the lack of capacity of the local merchants to supply the beneficiaries. Prompt and effective 
solutions were found and no similar issues were reported afterwards. The electronic food voucher 
then helped to ensure respect of beneficiaries’ dignity. 
The project also faced other issues: technical problems with the electronic system: registration 
process, signal issues, etc. few local shops met the criteria, difficulty to set up a common strategy 
with ACF for the coordination of fresh fruit and vegetable consumption campaign at the start of the 
project, etc. However, appropriate actions were taken in a timely manner and those issues did not 
seem to have significant impacts on the project’s course. 
The project proved to have short term effects on both direct and indirect beneficiaries. Both 
operating shops owners and petty traders reported increasing in the volume of sales and gross 
margins. However, the petty traders reported this positive trend dramatically reversed at the end of 
the project. Indeed, both local authorities and beneficiaries reported that the general conditions of 
the households went back to the starting point after the end of the project. 
However, the project was not intended to mitigate the structural causes of food insecurity and 
poverty since it only was an emergency response to natural shocks worsening the social conditions 
of the targeted communities. We do believe that key lessons can be learnt from this experience to 
build more sustainable long term projects with a view to mitigate the structural causes of 
vulnerability. In this perspective, should a similar project be implemented, we suggest building on 
the following and propose the following recommendations. 
41
3.1 Key lessons learnt 
i. Involvement of community committees in the project: The participation of 
community members in the selection and mobilization of beneficiaries is a good 
strategy to effectively differentiating the most vulnerable households. However a good 
balance between elected and unelected members for more accountability and to 
minimize political influence. 
ii. Selection criteria and identification document: Both local authorities and local 
vendors reported the case of vulnerable people such elderly who were not selected 
because they did not have identification document. Special considerations should be 
adopted in such cases since absence of identification document may also be considered 
as a criteria of vulnerability.
iii. Electronic system: The electronic voucher may not be appropriate in remote areas 
where it is hard to receive signal. 
iv. Respect of the dignity of the beneficiaries: An effective organization of the voucher 
program helps to preserve the beneficiaries’ dignity; however local vendors should 
have the enough supplying capacity. In this perspective it might be a good strategy to 
a priori assign beneficiaries to local merchants based on their supplying capacities.
v. Success of the fresh fruit and vegetables consumption promotion campaign: 
During the life span of the project more households consumed fresh fruits and 
vegetables as compared to the start of the project. This reinforcement of such item 
consumption was attributed not only to 12USD voucher but also to the campaign 
promotion. In general, the beneficiaries showed more awareness of fresh fruit and 
vegetable nutrition contribution compared to non-beneficiaries. Vouchers exchange 
and repayment seemed to be realized in an effective manner.
vi. Setting stone to long term program design: The short term effects of the project are 
threaten by the structural causes of vulnerability and future shocks. Some months after 
the project ended, the beneficiaries seemed to go back to their stating conditions. 
During the project’s course, they cut off less trees, used the money normally allocated 
to food to afford children education and some cases to acquire assets such as poultry 
and assets. However since the social conditions are the same, these positive effects are 
temporary.
vii. Need to counterbalance dependency mentality: One side effect observed from the 
focus groups with the beneficiary was a dependency mentality which may increase 
with repetition of the same intervention. A general sense of despair emerged from the 
discussions. The only hope expressed was the renewal of the project and none of them 
mentioned the need to be “boosted” in order to set up individual initiatives such as 
income generating activities. Therefore, this emergency project should have other 
component of livelihood development. 
Thus, the following recommendations may help to improve future projects. 
42
3.2 Recommendations 
i. Community committee composition: A predefined quota of elected and elected 
members may be used in the composition of the community members to balance 
political influence. Furthermore, the members of these committees as well as their 
relatives should be less vulnerable as compared to the others to avoid family 
conflict. 
ii. Elderly selection: Vulnerable people such as elderly living alone should be given 
priority albeit they do not have identification document. 
iii. Mixed voucher system: A mixed electronic/paper voucher system should be used 
to avoid signal issues in remote areas. The electronic system may be used in more 
accessible areas while a paper voucher system may be used in the other areas. 
However CARE should be applied to guarantee anti-fraud mechanism. 
iv. Extension of the fresh fruit and vegetable consumption promotion campaign 
to other areas: This component of the project seemed to be successful. Hence the 
experience may be extended to other areas. Not only healthier diet promotion diet 
is carried out but also the local economy is reinforced. 
v. Promotion of locally grown products: In order to “boost” local economy a quota 
of locally grown products may be required in the food basket composition supplied 
to the beneficiaries, which could contribute to reinforce local markets dynamics. 
vi. Advocating for long term programs: More long term projects are needed to 
actually mitigate the causes and consequences of food insecurity. The food 
program may be reinforced with income generating activities to attain potential 
sustainable impacts. Strategies may be used to identify those who demonstrate 
ability and develop such activities and boost them. Specific trainings may also 
accompany such interventions to help them identify activities with potential 
income possibilities. Partnerships should be built with Government technical 
agencies and other non-governmental organization to leverage positive effects of 
emergency project such Kore Lavni Nou-2 
43
Annex: 
Table A1: Logical framework of Kore Lavni Nou – 2 
Kore Lavi Nou-2 Indicators Means of verification
Goal Increased food security for 8,000 extremely poor and food insecure households in the Artibonite 
and Northwest Departments, Haiti.
Objective One
Provide food vouchers as a short-term safety net for 8000 food insecure and extremely poor families 
to fill gaps exacerbated by the disasters and to reinforce local markets in Gonaives, Mole Saint Nicholas, 
and Terre Neuve.
Result 1: Increased 
access to staple foods at 
local markets. 
# of beneficiary /voucher 
months attained (electronic) Reports from Merchant Pro
% of recipients are 
female-headed households Project beneficiary lists
# of merchants 
participating in voucher 
redemption, disaggregated by 
sex
MOUs and registered merchant accounts to redeem the voucher, 
activity of merchants per Merchant Pro and receipts
% of beans and maize 
purchased are locally grown Review sample of receipts. Interview with merchants and recipients.
Percentage of households 
with moderate or severe 
hunger
Review results of Household Hunger Scale (HHS), baseline and end￾of project surveys
Objective Two Pilot healthy diet promotion with fresh fruit and vegetable paper vouchers approach in Terre 
Neuve commune to reach 1000 food insecure and extremely poor families. 
Result 2: Increased 
awareness of the 
importance of fruits and 
vegetables in the diet 
Result 3: Increased 
access to these food 
products at local 
markets. 
2.1 # of households 
reached with nutrition 
education, by sex 
Participant lists, partner activity reports
2.2 Level of awareness 
about importance of fresh 
fruits and vegetables in diets. 
Baseline and end-of project survey, focus group discussions
3.1 # of beneficiary 
/voucher months attained 
(paper vouchers for fresh fruit 
and vegetables)
Distribution lists
3.2 % of recipients are 
female-headed households Project beneficiary registration lists
3.3 # of merchants 
participating in voucher 
redemption, disaggregated by 
sex
List of participating merchants.