(2020) Emergency Food Security Program in La Gonâve, Haiti - Final Report

1 WORLD VISION INTERNATIONAL -HAITI Emergency Food Security Program in La Gonâve, Haiti USAID – EMERGENCY FOOD SECURITY PROGRAM (EFSP) Final Report Submitted by: Amisial LEDIX, Economist-Statistician/Monitoring & Evaluation Specialist In collaboration with: Amos MONTREUIL JEAN , Agronomist/Food Security and Rural Resilience Specialist Guibenson COLIN, Agroeconomist/Development and Society Specialist March 2020 2 Table of Contents Acknowledgements ............................................................................................................................. 3 List of Tables ......................................................................................................................................... 4 List of Figures ....................................................................................................................................... 4 List of Charts ........................................................................................................................................ 4 List of Acronyms .................................................................................................................................. 6 Executive Summary ............................................................................................................................ 7 1. Background ................................................................................................................................. 10 2. Baseline Objectives ................................................................................................................... 11 3. Methodology ............................................................................................................................... 11 3.1 Quantitative Survey of Project Beneficiaries ............................................................................... 13 3.1.2 Total Sample .......................................................................................................................... 17 3.2 Survey of Resellers (Retailers) and Wholesalers ......................................................................... 17 3.3 Field Data Collection ................................................................................................................... 18 3.4 Pre-Testing of Collection Tools................................................................................................... 18 4. Constraints and Limitations of the Study ............................................................................ 19 5. Reviewing the Findings of the Baseline Study..................................................................... 20 5.1 General Characteristics of Beneficiary Households .............................................................. 20 5.2 Regarding Beneficiaries’ Food Consumption ......................................................................... 28 5.3 Condition of Beneficiary Households in relation to Hunger................................................. 30 5.4 Information Related to Shock Management and the Agricultural Situation .......................... 36 5.5 Access to Drinking Water for Beneficiary Households ........................................................ 37 5.6 Regarding Exclusive Breastfeeding ......................................................................................... 40 6. Local Market Assessment in the Intervention Areas ........................................................ 41 6.1 General Characteristics of Vendors ...................................................................................... 41 6.2 Overview of the Local Market in Program Intervention Areas ............................................ 44 6.3 Current Status of Product Sales on the Local Market .......................................................... 50 7. Conclusion and Recommendations ....................................................................................... 55 8. Bibliographic References.......................................................................................................... 57 9. Appendix I: Calculating Indicators ........................................................................................ 58 10. Appendix II : Convergence Table ...................................................................................... 62 11. Appendix III: Baseline Questionnaire ............................................................................... 63 12. Appendix IV: Resellers and Wholesalers Questionnaire.............................................. 76 3 Acknowledgements The team of consultants in charge of the present study wishes to express their gratitude, in a special manner, to all those who contributed to the achievement of this work. Despite the many challenges and difficulties encountered, we were able to succeed because of your active and effective participation. First, we wish to thank the Monitoring and Evaluation team, in the World Vision Haiti Central Office, for their open and fruitful collaboration. Interacting with our interlocutors was essential to the successful completion of the consultation. The support provided for the use of tablets and the reference documentation made available to us made our work easier. Our gratitude, then, goes to the EFSP project manager with whom we also interacted during the scope of work meeting for this study. We also thank World Vision’s Regional Manager in La Gonâve and the project’s Monitoring and Evaluation key person for their respective support in logistics for the training of enumerators and for sharing the necessary information to facilitate the progress of field operations. We would also like to thank the enumerators and supervisors who went to the field to collect data from beneficiaries in the project’s various intervention communities. The consulting team remains convinced that the study’s findings and recommendations will effectively serve the EFSP project team in addressing the challenges and achieving expected results on behalf of the beneficiaries living in the organization’s areas of intervention. The Consulting Team 4 List of Tables Table 1: Key indicators to be measured ................................................................................................ 11 Table 2: Key indicators to be measured ................................................................................................ 12 Table 3: Number of Beneficiaries per Commune and Communal Section ........................................... 14 Table 4: Quantity of surveys to be carried out in the communal sections of the commune of Anse-à- Galets ..................................................................................................................................................... 16 Table 5: Quantity of surveys to be carried out in the communal sections of the commune of Pointe- à-Raquettes ............................................................................................................................................. 16 Table 6: Household composition in terms of dependents, orphans and pregnant women .................. 25 Table 7: Distribution of respondents by commune based on the head of household’s type of habitat ................................................................................................................................................................ 25 Table 8: Distribution of respondents by commune according to the head of household’s main activity ................................................................................................................................................................ 26 Table 9: Distribution of respondents by commune according to the head of household’s second activity .................................................................................................................................................... 26 Table 10: Dietary diversity score for the island of La Gonâve by commune ........................................ 29 Table 11: Food consumption score of project beneficiaries for the island of La Gonâve by commune ................................................................................................................................................................ 30 Table 12: Hunger score of project beneficiaries for the island of La Gonâve by commune ................ 31 Table 13: Reduced Coping Strategies Index (rCSI) ............................................................................... 32 Table 14: Expected values for the indicators under consideration ....................................................... 33 Table 15: IPC Classification for Anse-à-Galets ...................................................................................... 34 Table 16: IPC Classification for Pointe-à-Raquettes .............................................................................. 34 Table 17: IPC Classification for La Gonâve (both communes considered together) ........................... 35 Table 18: Percentage of households where adults and children have at least 2 meals a day (adults and children), according to their age ............................................................................................................ 35 Table 19: Percentage of food use by type for project beneficiaries on the island of La Gonâve and the communes .............................................................................................................................................. 36 Table 20: Percentage of households reporting having received information to cope with economic shocks for the island of La Gonâve and its communes ......................................................................... 36 Table 21: Major shocks leading to hunger issues in percentage ............................................................ 37 Table 22: Distribution of beneficiaries by commune based on their main source of drinking water supply ...................................................................................................................................................... 38 Table 23: Average supply cost by product and commune .................................................................... 42 Table 24: Distribution of vendors by commune based on food products on sale ............................... 44 Table 25: Average number of customers a day before the hard times and during crises .................... 50 List of Figures Figure 1: Age Pyramid for Anse-à-Galets............................................................................................... 21 Figure 2: Age Pyramid for Pointe-à-Raquettes  ...................................................................................... 21 List of Charts Graph 1: Distribution of respondents by commune based on gender (n=927) ................................... 22 Graph 2: Distribution of respondents by commune based on their relation to the head of household (n=926) ................................................................................................................................................... 22 Graph 3: Distribution of heads of households by commune and by level of education (n=927) ......... 23 Graph 4: Distribution of heads of household by commune and by marital status (n=919) ................. 23 Graph 5: Distribution of respondents by commune according to whether or not there is at least one orphan in the household ........................................................................................................................ 24 5 Graph 6: Distribution of respondents by commune according to whether or not there is a pregnant woman in the household ........................................................................................................................ 24 Graph 7: Distribution of respondents by commune according to whether or not there is at least one disabled person in the household .......................................................................................................... 24 Graph 8 : Distribution of respondents by commune according to monthly household income bracket ................................................................................................................................................................ 27 Graph 9: Distribution of respondents by commune according to household’s first annual expenditure item (n=914)........................................................................................................................................... 28 Graph 10: Distribution of respondents by commune according to household’s second annual expenditure item (n=887) ...................................................................................................................... 28 Graph 11: Range of reduced Coping Strategy Index (rCSI) .................................................................. 32 Graph 12: Distribution of beneficiaries by commune based on the amount of water available for use per day. ................................................................................................................................................... 39 Graph 13: Distribution of beneficiaries by commune based on the walking time to and from water collection ................................................................................................................................................ 40 Graph 14: Prevalence of mothers who exclusively breastfeed their infants six months of age or younger ................................................................................................................................................... 40 Graph 15: Distribution of vendors by commune based on whether or not they are on the project list ................................................................................................................................................................ 41 Graph 16: Distribution of vendors by commune based on the frequency of store opening................ 41 Graph 17: Distribution of vendors by commune based on gender ...................................................... 42 Graph 18: Distribution of vendors by commune based on age group .................................................. 42 Graph 19: Distribution of vendors by commune based on whether they are retailers or wholesalers ................................................................................................................................................................ 43 Graph 20: Distribution of vendors by commune based on whether or not they have storage space . 43 Graph 21: Distribution of vendors by commune based on the size of their vending space ................. 43 Graph 22: Distribution of vendors by commune based on the number of years since they have been using their space ..................................................................................................................................... 44 Graph 23: Distribution of vendors by commune based on who supplies rice ..................................... 45 Graph 24: Distribution of vendors by commune based on who supplies corn .................................... 45 Graph 25: Distribution of vendors by commune based on who supplies common beans ................... 45 Graph 26: Distribution of vendors by commune based on who supplies flour .................................... 45 Graph 27: Distribution of Pointe-à-Raquettes vendors by month based on their ability or not to get the products usually sold ....................................................................................................................... 46 Graph 28: Distribution of vendors by commune based on the reason given for buying more food products from their supplier .................................................................................................................. 47 Graph 29: Distribution of vendors by commune based on the goods’ delivery time........................... 47 Graph 30: Distribution of beneficiaries by commune based on whether they feel safe or not on the road to the area’s supplier or market ................................................................................................... 48 Graph 31: Distribution of beneficiaries by commune based on whether they usually have challenges reaching the market ............................................................................................................................... 48 Graph 32: Distribution of beneficiaries by commune based on the type of challenge met on the road to the area’s market (n=48) ................................................................................................................... 50 Graph 33: Distribution of vendors based on whether or not they have secondary suppliers ............. 51 Graph 34: Trend in the food basket’s minimum cost ............................................................... 52 Graph 35: Trend in the monthly variation of food prices ..................................................................... 53 6 List of Acronyms ASEC Assemblée de la Section Communale CASEC Conseil d’Administration de la Section Communale CNSA Coordination Nationale de la Sécurité Alimentaire EFSP Emergency Food Security Program FCS Food Consumption Score FEWSNET Famine Early Warning Systems Network FFP Food for Peace FVFA Food Vouchers for Asset GPS Global Positioning System IHSI Institut Haïtien de Statistique et d’Informatique HH Household HDDS Household Dietary Diversity Score HHS Household Hunger Scale HNO Humanitarian Needs Overview IPC Integrated Phase Classification rCSI Reduced Coping Strategies Index S4T Savings for Transformation SPSS Statistical Package for the Social Sciences SRS One-stage Simple Random Sample USAID U.S. Agency for International Development WVI World Vision International WVI-H World Vision International-Haiti 7 Executive Summary La Gonâve is one of the country’s regions with a high prevalence of food insecurity, with about 45% of households in the CPI 3 and 4 range (HNO, 2019). Such a situation could lead to a serious food crisis if no intervention is carried out to support the area’s population. The country has been facing an economic and political crisis with devastating consequences for some time now and the standard of living of households is increasingly deteriorating, forcing them to resort to various survival strategies further driving them into poverty. The La Gonâve area remains one of the worst affected areas. In order to provide assistance to this population, WVI-H decided to implement an emergency food security program that will last 15 months and that focuses on providing food baskets, supporting pregnant and lactating women and income-generating activities. To identify target households and assess the program’s final impacts, WVI-H decided to conduct this baseline study to rank households based on ICPs and establish initial values for baseline indicators. This study is based on a cross section of 927 households randomly selected from a population of 4,662 households previously selected by WVI-H. The field survey was extended to all the communal sections within the island’s two communes and also covered the peri-urban areas of Pointe-à-Raquettes and Anse-à-Galets. Data collection focused on the situation of households with regard to the items within the food security phase classification indicators and an analysis of the markets serving the households, taking into account the main food products consumed. The first facts resulting from the socio-economic data analysis show that: • The respondents are mostly young, with a marked predominance of age groups below 25 years. At the national level in Haiti, 23 is the median age of the population (IHSI, 2015). • Most respondents are women, 55% in Pointe-à-Raquettes and 65% in Anse-à-Galets. • Illiteracy affects a large proportion of the heads of households, as 40% of respondents say they have never been to school in their lives. • Heads of households have, on average, 5 dependents under their responsibility. • More than half of the houses (68%) are simple houses, with tin roofs, which is an indicator of the households’ level of vulnerability. In terms of livelihoods, agriculture and trade are the main economic activities of heads of household participating in the program in both communes, with percentages being respectively 41% and 20%. These figures reflect the reality of predominantly rural areas where there is little diversification of economic activities. Many heads of household only have one source of income. On this point, about 42% of them say they do not have a second economic activity. It has also been noticed that the level of income derived from their activities by households remains relatively low. Indeed, in both communes, we noted the predominance of the 1,500 – 3,500 HTG income bracket. As regards the main expenditure items within households, food ranks first for absorbing more than 85% of their income. School fees and health care expenditures rank respectively second and third. 8 The review of household dietary diversity scores shows that, overall, 6% of households have low dietary diversity with a significant difference between the two communes (3% versus 9%) and nearly 75% have a high dietary diversity. For food consumption scores, about 16% of households have a poor score and 35% are in the intermediate level. Therefore, nearly half of the households have an acceptable consumption score. Compared to the results on the household index, only 16% of households are in a situation of little or no hunger, while 21% are in a severe hunger situation. Thus, more than a third are moderately hungry. As for the calculation of the Reduced Coping Strategies Index (rCSI), the results found in this study are almost similar to those found in the last National Emergency Food and Nutrition Security Survey (Enquête Nationale d’Urgence sur la Sécurité Alimentaire et Nutritionnelle)(ENUSAN, MARNDR/CNSA, 2019). Indeed, 43% of households in La Gonâve have a moderate (4-18) rCSI score according to data collected in this baseline, virtually the same rCSI score at the national level. The most frequent strategies are: eating less valued/expensive foods and limiting portion sizes at mealtimes. Therefore, considering the last 4 indicators mentioned above, namely Household Food Diversity Score, Food Consumption Score, Household Hunger Index and Coping Strategies Index (CSI), it was established, by convergence, that beneficiaries in the communes of Anse-à-Galets and Pointe-à-Raquettes are classified as IPC3. This partly means that beneficiary households are marginally able to meet minimum food needs, but only by depleting essential livelihoods or through crisis management strategies. When it comes to the indicator related to the number of meals per day, in general, whether it is among children or adults, 72% of them, in Anse-à-Galets, have at least 2 meals per day against 62% for those in Pointe-à-Raquettes, i.e. 67% on the island of La Gonâve. As for the indicator related to the percentage of food use by type, in general, the agricultural production of households on the island of La Gonâve is used for household consumption (95%) and for sale (39%). As for the indicator related to information on managing economic shocks, only 16% of households in the commune of Anse-à-Galets and 20% of those in Pointe-à-Raquettes reported having received some information. Drought and plant disease were identified as the main shocks causing famine in the project intervention areas. Lastly, as for indicators, measuring the one related to the prevalence of mothers practicing exclusive breastfeeding, 86% of mothers with infants aged 6 months or less on the whole island of La Gonâve, exclusively used breast milk as food and drink. Access to water for various domestic uses remains a major concern for a large proportion of households. For drinking water, although almost half of households (48%) are supplied from the DINEPA network and public pumps, a significant number of them (25%) are supplied from unsafe water points (unprotected wells, uncollected springs). This exposes them to potential contamination by pathogens, which would worsen their nutritional status. This situation is more critical in the commune of Pointe-à-Raquettes, where they still have to travel a much longer distance to fetch water. In relation to the assessment of the local market in the intervention areas, concerning access to food products, the main supply points for the population of both communes of the island of La Gonâve are the markets of Anse-à-Galets and Pointe-à-Raquettes, which operate almost every day of the week. There are also other small markets serving as supply points in remote 9 areas. In these markets, the number of retailers far exceeds the number of wholesalers. These markets are supplied in food products from the port of Carriès, where boats travel between Anse-à-Galets/Carriès and the following supply centers: Port-au-Prince, Arcahaie, Saint-Marc and Gonaïves. The main products in the household food basket (oil, flour, corn, rice, wheat and common beans) are available on these markets. Overall, almost all the vendors surveyed offer rice (local and/or imported) as the main product, followed by corn (85% of vendors) and all of these products can be found with 57% of the vendors. It should be noted that commercial activities on the island are mainly carried out by women, who, according to the study’s findings, represent 67% of the vendors. Access to markets is not always easy for all households. The decline in the socio-political situation in the country creates conditions of insecurity and instances of road blockades, sometimes preventing certain households from getting regular supplies of food products. It should also be noted that the economic crisis is leading to a decrease in sales by traders as a result of the decline in household purchasing power following the increase in product prices, mainly over the last two years. 10 1. Background According to the latest Humanitarian Needs Overview (HNO) report in 2019, La Gonâve, one of Haiti’s most food-insecure regions, is currently experiencing extreme food insecurity in its rural areas. Poor spring and fall harvests in the 2017-2018 season have left 45 percent of households (about 40,000 people in the CPI 3 and 4 range) at risk of a severe food crisis if emergency assistance is not provided to this population, which is barely able to meet its basic food needs. Haiti’s dire economic situation, characterized by currency depreciation and inflation, continues leading to higher prices for basic foodstuffs, especially grain. Access to food is even worse for most people in La Gonâve. To cope with this situation, people are selling off their livestock and non-productive assets. Other coping strategies include migration, selling labor, prostitution, fishing, and some households have already consumed their seed reserves. In response to this humanitarian emergency, World Vision (WVI) is implementing an Emergency Food Security Program in La Gonâve to meet the immediate and basic food needs of 5,770 HH (approximately 28,842 people), currently in the IPC3 and IPC4 affected areas. This program will last approximately 15 months, from October 2019 to December 2020, and provide food vouchers to 100 percent of vulnerable households in IPC3 and 4 in rural areas, and to 58.5 percent of households in IPC3 and 4 in urban areas with funding from USAID. WVI builds on its 20 years of cumulative experience implementing USAID emergency and food security projects in Haiti. Indeed, over the past ten years, WVI has implemented, on the island of La Gonâve, two consecutive EFSP projects that have focused on child nutrition and building resilience. In this EFSP program in La Gonâve, the following fields of intervention to achieve the set goals and objectives are therefore: • Food Vouchers for Assets (Food Vouchers for Asset/FVFA) • Support to pregnant and breastfeeding women through training • Training for young people in IGA (Income-Generating Activity) • Savings for Transformation (S4T) In order to evaluate all the changes to be induced as a result of this program’s implementation, it is essential to outline the situation from the very beginning with precise data on the key indicators of the food security level. For that reason, this baseline study has been conducted among project beneficiaries, including an assessment of the local market in the intervention areas. 11 2. Baseline Objectives The purpose of this baseline is to establish benchmarks for project performance indicators against which project success will be measured. Specifically, this baseline is used for the following purposes: • Determine baseline values for key outcome indicators • Collect data comparable to that of the final evaluation to determine the level of change in outcome indicators. • Help establish/validate annual and end-of-project goals in the project M&E plan • Increase World Vision Haiti’s accountability with all its key stakeholders, including communities, partners, donors, supporters and the Government of Haiti. • Identify households affected by the drought shock and classify vulnerable households in IPC classifications. 3. Methodology The project’s baseline was based on the collection of primary data (surveys and field observations) and secondary data (documentary research). Data collection methods were established according to the indicators directly concerned by the baseline’s objectives, as indicated in the two tables below: Table 1: Key indicators to be measured Key Indicators Indicators Definition Data Collection Method Percentage of targeted households with an acceptable food consumption score (FCS) The frequency-weighted dietary diversity score is a score calculated from the consumption frequency for the different food groups consumed (Appendix 1) by a household over the 7 days preceding the survey. Quantitative survey of project beneficiary households Prevalence of households with little or no hunger (Household Hunger Scale – HHS) Is essentially a behavioral measure that tends to capture more serious behaviors (Appendix I). Quantitative survey of project beneficiary households Proportion of households consuming at least 6 food groups during the previous month Household Dietary Diversity Scale (HDDS): Dietary diversity is the number of different foods or food groups consumed during a given reference period. Quantitative survey of project beneficiary households Percentage of food use by type (household consumption, sale, exchange, feeding of livestock) This indicator considers food generally available at the household level, regardless of its origin and quantity. It analyses the different modes of use, namely: Self-consumption, sale, trade and feeding of livestock. The priority use of food can provide information on the household’s economic situation. Quantitative survey of project beneficiary households 12 Table 2: Key indicators to be measured Key Indicators Indicators Definition Data Collection Method Percentage of households where adults and children eat at least 2 meals a day (adults and children), by age The average number of meals consumed per day in a household is an indicator of their level of food security or insecurity. The number of meals in food insecure households is lower than in food secure households. In fact, this indicator records the number of households where adults and children eat at least 2 meals a day. Quantitative survey of project beneficiary households Prevalence of mothers exclusively breastfeeding infants six months of age or younger This indicator considers the proportion of mothers whose children aged six (6) months or less were exclusively breastfed. In the quantitative household survey, a subgroup of questions will be addressed to households with mothers with infants up to 6 months old. Reduced Coping Strategies Index (rCSI) rCSI measures behavior: what people do when they don’t have access to enough food by answering the question: What do you do when you don’t have enough food and you don’t have enough money to buy food? - Measures adjustments made by households in consumption and livelihoods. These may be changes in consumption, expenditure reductions or income growth. - rCSIs tend to measure less severe coping behaviors. - rCSIs use the five most common strategies with standardized weightings: 1- Relying on less valued and less expensive food? 2- Borrow food or rely on the help of a friend or relative? 3- Limit portion sizes at mealtimes? 4- Restrict consumption by adults so that young children can eat? 5- Reduce the number of meals consumed per day? (Appendix I) Quantitative household survey Percentage of households reporting greater ability to cope with economic shocks. This indicator considers households with information on shocks. The households most vulnerable to risks are those without information. Measuring this indicator therefore has to do with two (2) key elements: a. Shock identification (floods, wind, drought, plant disease, erosion, hurricanes and others) and the level of contribution from each type of shock in food insecurity. b. Information received by households to manage shocks. Quantitative survey of project beneficiary households 13 Rapid Market Assessment This baseline study also included a rapid analysis of local markets to assess their ability to meet additional demand without generating inflation risks or supply disruptions. USAID-IRC’s Market Information Framework was used and adapted to conduct this assessment around two main components: the capacity of markets and their relationship with beneficiaries (key information) and the key factors to consider for the design of voucher assistance. Since the program’s distribution mechanism had already been identified (vouchers), the market assessment proposed in this study sheds light on how markets work and then on the link between households and markets. 3.1 Quantitative Survey of Project Beneficiaries A quantitative sampling survey was conducted among project beneficiaries in the two intervention communes of the island of La Gonâve, namely Anse-à-Galets and Pointe-à- Raquettes. A probabilistic sampling method was considered to randomly select the various program beneficiaries to be surveyed. An initial internal work of the program had already written a list of beneficiaries by commune, by communal section, and by targeted localities with a contact telephone number, in most cases. This, therefore, constituted the sampling base from which a random sample of beneficiaries was extracted for interview. Then, we selected the One-stage Simple Random Sample (SRS) method. But all of that, each of the communes was assimilated to a stratum. The One-stage Simple Random Sample (SRS) method was conducted separately by stratum. It is true that access to the selected beneficiaries had caused a difficult logistical burden, but guides who were familiar with the localities concerned or who had participated with WVI Haiti in compiling the lists of beneficiaries were used to facilitate the work of the field interviewers. As it was planned in the study’s scope of work meetings with WVI-H counterparts to establish the values of indicators for each of the 2 communes, the number of beneficiaries to be surveyed was determined separately for each of them. The program targets 5,770 beneficiary households. The lists provided by the program provided contact information on nearly 4,662 beneficiaries. Hence the following table of the number of beneficiaries per commune and per communal section: 14 Table 3: Number of Beneficiaries per Commune and Communal Section Communes Communal Sections Number of Beneficiaries Anse-à-Galets 1ère Palma 719 2ème Petite Source 658 3ème Grande Source 462 4ème Grand Lagon 745 5ème Picmy 122 6ème Section Petite Anse 115 Total 2,821 Pointe-à-Raquettes 1ère Section la Source 47 2e Section Grand Vide 499 3e Section Trou Louis 464 4e Section Pointe-à-Raquettes 686 5e Section Gros Mangle 122 Town of Pointe-à-Raquettes 23 Total 1,841 Grand total 4,662 The overall sample is the sum of the 2 sub-samples n1 and n2 for the respective communes (or strata) of Anse-à-Galets and Pointe-à-Raquettes. The formula for calculating the sample size, with the chosen sampling method (One-stage Simple Random Sample (SRS)) per stratum is the following: ������ ��������������������������������=������ ��������[ � 1−�√2������(1−������) +������ 1−� √������ 1,�������� (1−������ 1,�������� )+������ 2,�������� (1−������ 2,�������� ) ������ ] 2 Where: ������ ��������������������������������∶ is the initial sample size required by the baseline and final study surveys for each of the two time points. ������ 1,�������� : represents a survey estimate of the true proportion ������1 of the population at baseline [If such an estimate is not available from previous surveys, use 0.5] ������ 2,�������� : represents a survey estimate of the true proportion ������2 of the population at the end of the project for the final study Since we are at the baseline, the value ������ 2,�������� does no longer apply in the previous formula. This is why we use the following equivalent formula in our situation (UN, Designing Household Survey Samples: Practical Guidelines, 2005a): ������=������ �������� ������ ������ (������ 1,��������∗(1−������ 1,��������))∗� ������ ������ 15 Where ������ �������� is the design effect, equal to 1 in a One-stage Simple Random Sample (SRS) design Z: Confidence level z-score (1.96 or 1.645 corresponding respectively to a 95% and 90% confidence level) �: Non-response rate (generally set at 1.1, corresponding to 10%, but may change depending on the context) ε: Margin of error set, in this work, at 0.06 N1 and N2 are less than 100,000 beneficiaries, so the n1 and n2 sought are multiplied by the sampling correction factor (N/(1+N)). Hence the n1 and n2 found from the formula become: N1 (Anse-à-Galets) = 2,821 beneficiaries, with a margin of error of 6%, n1 (Anse-à-Galets) ~268 beneficiaries. N2 (Pointe-à-Raquettes) = 1,841, with a margin of error of 6%, n2 (Pointe-à-Raquettes) ~256 beneficiaries. Given that, according to the selected sampling method, i.e. the One-stage Simple Random Sample (SRS) per commune, at least 339 beneficiaries are needed to calculate the FCS and HHS indicators, the sample sizes for the communes of Anse-à-Galets and Pointe-à-Raquettes were therefore increased from 268 to 339 and 256 to 339 beneficiaries respectively. We can see that the calculation formula gives a larger minimum size for Anse-à-Galets, which in fact has more participants in the program. It was in fact the adjustment to the minimum size required of 339 for the calculation of the indicators that removed the proportionality aspect of the respective sizes of the samples of the 2 communes compared to their population of program participants. In relation to one of the indicators, the prevalence of mothers exclusively breastfeeding infants six months of age or younger, investigators were clearly instructed to interview at least 5 breastfeeding mothers. The list of beneficiaries of the program which served as a sampling frame did not allow to identify all the breastfeeding mothers. Therefore, a separate sampling was not considered because we could not extract the subpopulation of the lactating women in the sampling frame. As in each commune there are communal sections in which listed beneficiaries live, the previously calculated sample size was broken down proportionally to the size (PPT) of each of these geographical units. 16 Table 4: Quantity of surveys to be carried out in the communal sections of the commune of Anse- à-Galets Communes Communal Sections Number of Beneficia ries Demograp hic Weight (pi=Ni/N) Quantity Allocated by Zone Minimu m Planned Quantity * Quantity of Surveys Conducted Anse-à-Galets 1ère Palma 719 0.25 86 86 101 2ème Petite Source 658 0.23 79 79 104 3ème Grande Source 462 0.16 56 56 49 4ème Grand Lagon 745 0.26 90 90 124 5ème Picmy 122 0.043 15 30 42 6ème Section Petite Anse 115 0.041 14 30 45 Total 2,821 1.00 339 371 465 * The number was adjusted for all communal sections with less than 30 households allocated. An adjustment is made for all the communal sections whose allocated quantity is less than 30 households because of the Central Limit Theorem which justifies the use of normal distribution if the sample size is large enough. ... Empirically, it is said to be enough if the sample size is at least 30. That is why we always consider 30 the minimum number or observations we need to conduct some statistical test, for instance if we need to compare two means. The quantity of beneficiary surveys to be carried out in the commune of Anse-à-Galets was therefore set at n1=371 beneficiaries. But 465 beneficiaries were actually surveyed in Anse-à-Galets. This therefore decreases the margin of error from 6% to 4% by deducting it in the previous formula for the random sampling plan. Table 5: Quantity of surveys to be carried out in the communal sections of the commune of Pointe- à-Raquettes Communes Communal Sections Number of Beneficiaries Demographic Weight (pi=Ni/N) Quantity Allocated by Zone Minimal Planned Quantity* Quantity of Surveys Conducted Town of Pointe-à- Raquettes 23 0.01 4 23 13 Pointe-à- Raquettes 1re Section la Source 47 0.03 9 30 30 2e Section Grand Vide 499 0.27 92 92 132 3e Section Trou Louis 464 0.25 85 85 92 4e Section Pointe-à- Raquettes 686 0.37 126 126 157 5e Section Gros Mangle 122 0.07 22 30 38 Total 1,841 1.00 339 381 462 * The number was adjusted for all communal sections with less than 30 households allocated. 17 The number of beneficiary surveys to be carried out in the commune of Pointe-à-Raquettes was therefore set at n2=381 beneficiaries. However, 462 beneficiaries were actually surveyed in Pointe-à-Raquettes. This therefore decreases the margin of error from 6% to 4% by deducting it in the previous formula for the random sampling plan. In the data analysis, the indicators are calculated by commune as mentioned above. The indicator values are not sample weighted. To analyze these indicators on the whole of La Gonâve Island, the data from representative samples of the 2 communes are aggregated in order to obtain a global sample of program participants for the whole. The values of the indicators for the whole Island of La Gonâve are thus calculated on this global sample also representative of program participants. 3.1.2 Total Sample The minimum size of the overall sample of beneficiaries to be surveyed in the 2 communes of La Gonâve was set to be: n = 747 beneficiaries In order to ensure this minimum number was surveyed, each investigator was required to interview between 5 and 10 beneficiaries in addition to the planned number. Considering this latter aspect, the final number of households surveyed in the 2 communes is then n’ = 927. 3.2 Survey of Resellers (Retailers) and Wholesalers To conduct a rapid assessment of the local market in project intervention areas, a quantitative survey of vendors was conducted. A data collection tool was developed for this category of study units to help identify the supply side while a section related to this market assessment was inserted in the baseline questionnaire for project beneficiaries to consider the demand side. A two-pronged approach was used to collect market assessment data by combining secondary data from observations and consultations with stakeholders and local authorities to identify and understand key aspects of market operations and size. Primary data (key information) was then collected from vendors/wholesalers in the project intervention communities. - Consultations Since no preliminary information was available, mainly with regard to the number of food vendors in the project areas, consultation sessions were organized with local authorities (mayors, CASEC and ASEC) to identify local markets in the 2 project communes. WVI was also asked to provide the list of traders in the food voucher distribution program to conduct the survey and analyze the inventory capacity and food prices the project intends to facilitate to food insecure beneficiaries. - Survey of vendors (retailers) and wholesalers 18 This survey was carried out, not just thoroughly among the vendors pre-selected by the project, but also in the mid-size and large markets identified in the communes on the basis of a convenience sample of merchants. Price comparisons could be made between the relevant and non-project merchants. This survey of vendors and wholesalers also made it possible to collect primary information on the inherent capacity of markets to respond to demand growth over the duration of the program. 3.3 Field Data Collection Primary data collection (beneficiary survey and vendor survey) was conducted electronically. The collection tools were designed using the digital survey platform called Ona. Each interviewer was therefore provided with a digital tablet on which the relevant survey form was installed. It was then possible to follow the entire collection process in real time from the Ona platform. The interviewers were also required to enter the GPS coordinates of the houses of the beneficiaries surveyed; this should reduce the risk of an interviewer filling in the form himself without actually visiting the specific survey respondent. For the survey of beneficiaries, each investigator received the list of beneficiaries to be surveyed. This list was previously prepared using MS EXCEL spreadsheet and the Random () function to randomly select the beneficiaries to be surveyed. With the list thus established, each interviewer was assisted by a guide with a great deal of knowledge of the different localities where the selected beneficiaries live. 3.4 Pre-Testing of Collection Tools The collection tools (baseline questionnaire and vendor questionnaire) were tested for applicability. This work, which was carried out by the technical team that had worked on the development of these collection tools, helped adjust them to field realities adapt the training modules for interviewers and supervisors. 3.5 Selection and Training of Interviewers and Supervisors Twenty (20) interviewers and two (2) supervisors were selected and mobilized in the 2 communes of La Gonâve to collect data for 5 to 6 days after receiving a one-day training on the survey forms and the study’s objectives. The data collected from the tablets and uploaded to the Ona digital platform were cleaned and processed. They were then exported to SPSS and MS Excel for further analysis. - Classification of Vulnerable Households in IPC Phases 19 In order to classify vulnerable households into the corresponding IPC phase(s), the convergence table in version 3.0 of the IPC manual 1 was used, considering the following indicators whose threshold values were established in this study: - Percentage of targeted households with an acceptable food consumption score (FCS) - Prevalence of households with little or no hunger (Household Hunger Scale – HHS) - Proportion of households who consumed at least 6 food groups in the previous month (HDDS) - Reduced Coping Strategies Index (rCSI) These indicators take into account the food consumption aspect in the ICP phases. The nutritional status is not considered since GAM (Global Acute Malnutrition) will not be measured according to the study’s terms of reference. The indicators considered are calculated for each of the 2 communes and for the 2 program intervention areas, as a whole. The percentages for each class of indicators are placed in the corresponding phases, as indicated in the convergence table (Appendix II). The accumulation of these figures from right to left (starting from phase 5 to phase 1) for each of the indicators allows to decide which IPC to use by applying the 20% rule. The phase in which aggregated percentages reach 20% is chosen for each of the indicator rows considered. In this way, the current population percentage ranges for each of the phases can be established, and then finally the consensus percentage of the population falling into each of the phases can be established. Where 20% is reached, it is chosen as the IPC phase for the commune or area under analysis. 4. Constraints and Limitations of the Study As indicated in the methodology, this study is being carried out on the whole island of La Gonâve. Thus, data collection was carried out on the basis of a random sample taken from a list of beneficiaries that had already been identified by WVI-H. In order to find the sample- selected beneficiaries in the field, the interviewers had to use the telephone numbers provided in the WVI-H database. This exercise sometimes proved difficult, despite the presence of the guides, because of communication issues related to the non-existent telephone signal in several localities on the island. Most of the initial beneficiaries selected could not be found, mainly in certain communal sections of Pointe-à-Raquettes, so the selection work had to be resumed for several localities and new lists had to be submitted to certain interviewers to facilitate their work. Another challenge faced in the field was due to the full names of beneficiaries. In some cases, the last or first names reported by beneficiaries or respondents differed somewhat from those found in the initial database. In such situations, interviewers had to ask additional questions to ensure the person they were interviewing was indeed the one they were 1 Technical Manual version 3.0, Evidence and Standards for Better Food Security & Nutrition Decisions, April 2019, pages 35-36 20 supposed to interview. This resulted in some rather marginal differences between the initial database and the one built from the baseline survey. As there already was a sampling frame that formed a clearly defined population from which the sample was drawn, this did not pose any methodological issues. However, given the size of the study area, which is made up of many small localities, it was not possible to take a sample that is equally spread over the entire area. However, all precautions were taken for the selected beneficiaries to cover as much of the study area as possible. This did not prevent some concentration of surveys in a few localities due to the corrections made to the first random selection where some beneficiaries could not be identified by the interviewer. Another limitation of this study is the fact that the survey was conducted on a sample of beneficiaries, a large part of which were from the 3,000 participants were part of in an emergency food distribution intervention in December 2019 after the political unrest causing the "peyi lok" (lockdown). This intervention has surely a positive effect on the initial values of the baseline indicators related to the food security in the households. There is also a limitation of this study related to the low sample size of lactating women who were questioned for the indicator titled: “Prevalence of mothers exclusively breastfeeding infants six months of age or younger”. A separate sample would have to be considered but we could not extract the subpopulation of the lactating women in the sampling frame. The list of beneficiaries did not have information the quantity of lactating women in the sampling frame. With this low sample size, we assume that the result for this indicator may not reflect the reality. 5. Reviewing the Findings of the Baseline Study In this section, the findings of the baseline are reviewed by highlighting the general characteristics of beneficiary households, establishing the values of the different indicators under study per commune and interpreting the food consumption indicators for the IPC classification by convergence. 5.1 General Characteristics of Beneficiary Households It was necessary to characterize the universe of the study in relation to the beneficiary households surveyed. In order to get an overview of the household structure in the project intervention communes, an age pyramid was designed for each of them (Figures 1 to 2). These graphs have the same configuration as the Haitian population’s age pyramid in general, i.e., very broad at the base and very slender at the top. This means that the population is very young. The most represented age groups are those under 25 years of age. We note that in the 2 communes, this similarity exists, i.e. a peak in the 5-9 and 10-14 age groups for both sexes. In both communes, there also tends to be more women in the population. This reflects the reality of the Haitian population, which has more women than men. [... middle sections omitted for long document ...] 77 Produce Type of suppliers (farmers, middleman, wholesalers etc) Where this supplier is located? How many of these products do you usually buy? (kilos, bags etc) provide accuracy How often do you buy these products from them? is there a special day of the week that this purchase is made? transaction amount (including transportation costs) Corn Wheat flour Wheat (grains) Cooking oil Pwa enkoni (local beans variety) Rice Are there other suppliers that you can contact for supplies if your normal suppliers cannot supply you?? o Yes, who ________________________________________ ________________ o No. 3.2 On what basis do you source food products? o When all my stock runs out o When my stock is less than 20% o When I have between 20 to 50% products stored o Other (specify) ________________ 3.3 When buying goods, how long can it take before these goods come to you? ___ As soon as the order is placed ____ Less than 3 days ____ in 1 week ____More than 2 weeks 3. 4 Do you sometimes not find these products from your suppliers? How often? o Corn ____Often _____Sometimes ___ It Never Happens o Wheat (grains) ____ Often _____Sometimes ____ It never happens o Cooking oil ____ Often _____Sometimes ____ It never happens o Rice ____Often _____Sometimes ___ It Never Happens o Wheat Flour ____Often _____Sometimes ___ It Never Happens o Pwa enkoni (local beans variety) ____Often _____Sometimes ___ It Never Happens If this happens you don't get the products, tell us in which month these products were not available to you ? o Corn _______/_______/________/________/_________/_________/_________/_______ o Wheat (grains) /_________/________/_________/_________/_________/_______ o Cooking oil _______/_______/________/________/_________/_________/_________/_______ o Rice _______/_______/________/________/_________/_________/_________/_______ o Wheat Flour _______/_______/________/________/_________/_________/_________/_______ o Pwa enkoni (local beans variety)______/________/________/_________/_________/_________/_______ 3.4 Why are these products not usually available for purchase? o Corn ___ There was no harvest 78 ___ Roads are not good, transportation becomes more expensive ___ Products are too expensive, I can't buy them ___ other ___________________________________________________________ o Wheat (grains) ___ There was no harvest ___ Roads are not good, transportation becomes more expensive ___ Products are too expensive, I can't buy them ___ other___________________________________________________________ o Cooking Oil ___ There was no harvest ___ Roads are not good, transportation becomes more expensive ___ Products are too expensive, I can't buy them ___ other ___________________________________________________________ o Wheat flour ___ There was no harvest ___ Roads are not good, transportation becomes more expensive ___ Products are too expensive, I can't buy them ___ other___________________________________________________________ Pwa enkoni (local beans variety) ___ There was no harvest ___ Roads are not good, transportation becomes more expensive ___ Products are too expensive, I can't buy them ___ Lot ___________________________________________________________ o Rice ___ There was no harvest ___ Roads are not good, transportation becomes more expensive ___ Products are too expensive, I can't buy them ___ other___________________________________________________________ 3.5 What often makes it difficult for goods (foodstuffs) to enter or exit the market? ___ all products come in and out normally ___ insecurity ____ roads are not good ___ we can't find the products to buy ___ products are too expensive, taxes and transportation take all the money ___ other ___________________________________ Section 4 : Sale 4.1 Can you tell us at least the quantity and price you sell each day for these products ? : Products how much do you sell in each of these products per day ? kilo, to be precise Cost of a unit ? gourd Total Corn Wheat flour Wheat grains Cooking oil 79 Pwa enkoni (local beans variety) Rice 4.2 Can you tell us at what price you sell these products for each month of the year? see reference 3.4 which gives the month that the products are hard to find, the trend should be in the months the products are hard to find, the prices should increase Produce Unit Jan. / gourd Feb. Mar. Av May June July August Sept. Oct. Nov. Dec. Corn Wheat flour Wheat grains Cooking oil Pwa enkoni (local beans variety) Rice 4. 3 After selling normally each day, How much extra can you sell in each of these products? that is, the amount you have in stock if you ever need to sell still. Products the additional quantity you can still sell (kilos, bags, provide accuracy) Corn Wheat flour Wheat grains Cooking oil Pwa enkoni (local beans variety) rice What would you have to do if you wanted to sell more products every day?: o 20% extra per day ________________________________________________________ o 30% to 40% extra per day ___________________________________________________________ o More than 40% extra every day ___________________________________________________________ 4. 4 Where do your customers come from? specify communes and communal sections 80 4. 5 Can you tell me how many such customers come to buy from you when things were good before this crisis? ____________ 4. 6 Can you tell me how many such customers come to buy from you at this time during this crisis hits the community ______________ 4. 7 What causes changes in the number of people who come to buy from you? ____ Products become more expensive ____The price of the product does not change but it is the people who don’t have enough money to buy ____ I don’t have the products so customers don’t come ____ other Section 5: Access to Credit 5.1 Do you have a bank account or a credit union savings account? o Yes o No. 5.2 Do you take credit for doing this business? o Yes o No. 5.3 (If yes) where do you get this credit? ____ Bank ___ credit union ____ Community Bank ____ individual ___other ______________ 5.4 Do you sell on credit to your customers? o Yes o No. Section 6: Markets in general 6.1 Since the crisis began last year, have you seen a change in the number of retailers and wholesalers in the market? o Yes ( If yes, how could you define this change ? _ __ There is 2 times less vendors than before ___ there is 2 to 4 times less vendors ____ The number of vendors decreased more than 4 times __ There is more than 2 times vendors since last year ___ Between 2 and 4 more vendors ____ More than 4 times vendors o No, there is always the same quantity of vendors selling in the market 6.2 What do you think causes these changes? ( If the person's response was to decrease the number of vendors ) ___ Products become more expensive 81 ___ the cost of transportation becomes higher ___ It is difficult to renew our stocks because the products are not available ____ roads are not even good ____ where we are used to buy are no longer given credit ____ is a 2 to 3 places where the products are sold, if they do not have the product, then we can not find it ____ other ____________________________ ( If the person's response was to increase the number of vendors ) ___ gets more opportunities to trade ___ we get credit ___ products are available, we just buy them ____ other 6.3 Are you accustomed to participating in programs where customers come up with a cash voucher to buy from you? o Yes o No. (If Yes ) Well tell us how the experience was for you, what would you like to change and how ? ______________________________________________________________________________