1 Quantitative impact evaluation of the SHOUHARDO II Project in Bangladesh May 2015 TANGO, International Inc. 406 S. 4th Avenue Tucson, Arizona 85701 USA www.tangointernational.com 2 Table of Contents Acknowledgements ....................................................................................................................................... 1 Acronyms ...................................................................................................................................................... 2 List of Tables ................................................................................................................................................. 3 List of Figures ................................................................................................................................................ 4 Executive summary ....................................................................................................................................... 5 1. Introduction ............................................................................................................................................. 8 2. The SHOUHARDO II project: Beneficiary selection process and interventions ....................................... 9 2.1 Beneficiary selection process ...........................................................................................................10 2.1.1 Identification of project geographical areas ................................................................................10 2.1.2 Household selection .....................................................................................................................10 2.2 Project interventions ........................................................................................................................12 2.3 Participation in project interventions ..............................................................................................15 3. Conceptual framework and measurement of stunting and its determinants .......................................16 3.1 UNICEF conceptual framework .........................................................................................................16 3.2 Measures of stunting and its determinants .....................................................................................17 3.2.1 Stunting .....................................................................................................................................18 3.2.2 Determinants of stunting ...........................................................................................................18 4. Data ........................................................................................................................................................19 5. Impact evaluation methodologies .........................................................................................................20 5.1 Descriptive methods .........................................................................................................................21 5.2 Instrumental variables regression analysis ......................................................................................22 5.3 Propensity score matching ...............................................................................................................25 6. Results: Descriptive evidence of project impacts .................................................................................28 6.1 Trends in stunting among project households compared to nationally .....................................28 6.2 Shift in the age trajectory of stunting among project households ..................................................30 6.3 Difference-in-difference analysis .....................................................................................................31 7. Results: Instrumental Variables evidence of project impacts .............................................................33 8. Results: Propensity Score Matching evidence on the impact of specific interventions .......................38 8.1 Mother and child health and nutrition interventions ......................................................................38 8.2 Women’s empowerment interventions ...........................................................................................42 3 8.3 Livelihoods promotion interventions ...............................................................................................44 8.4 Water and sanitation interventions ................................................................................................46 9. Summary and conclusions......................................................................................................................47 Appendix 1. Determinants of stunting employed as dependent variables: Variable descriptions ..........51 Appendix 2. Household, village and district-level characteristics used for Propensity Score Matching analysis: Variable descriptions ...................................................................................................................53 Appendix 3. Instruments employed for Instrumental Variables tests and 2SLS regressions ....................55 Appendix 4. Probit propensity score models for participation in SHOUHARDO II project intervention ...56 References ..................................................................................................................................................64 1 Acknowledgements I would like to extend my gratitude to the CARE Bangladesh staff who made this impact evaluation study possible and provided timely information regarding the SHOUHARDO II project under difficult working conditions. I would especially like to thank Monzu Morshed, Deputy Chief of Party of SHOUHARDO II, AKM Abdul Wadud, Monitoring and Evaluation Coordinator, and Khan Tawhid Parvez, Senior Technical Manager for Health, Hygiene and Nutrition. The survey team from Mitra and Associates that collected the data employed here, in addition to Lloyd Banwart and Towfique Aziz of TANGO, International are gratefully acknowledged for their hard work in the field. Tom Bower provided excellent research assistance in support of the data analysis and report write-up. Finally, the individuals and families who gave freely of their time to be interviewed during the data collection are thanked in hope that the findings of this study will enable future programs to contribute to further improvements in food security and children’s nutritional status throughout Bangladesh. Lisa C. Smith, Ph.D. (Principal author) Senior Economist TANGO, International 2 Acronyms 2SLS Two-stage Least Squares ATT Average treatment effect on the treated CHD Comprehensive Homestead Development CHV Community health volunteer COG Core Occupational Group(s) DHS Demographic and Health Survey ECCD Early Child Care for Development EKATA Empowerment, Knowledge, and Transformative Action HAZ Height-for-age Z-score IGA Income Generating Activities IV Instrumental Variable(s) MCHN/PEP Maternal and Child Health and Nutrition/PEP MT Metric ton NGO Non-governmental organization(s) OLS Ordinary Least Squares PEP Poor and Extreme Poor PCA Principal Components Analysis PM2A Preventing Malnutrition in Children Under Two PPS Probability proportionate to size PSM Propensity Score Matching RCT Randomized Controlled Trial SHOUHARDO Strengthening Household Ability to Respond to Development Opportunities WHZ Weight-for-height Z-score 3 List of Tables Table 1:Design of the randomized controlled trial to evaluate the relative effectiveness of the MCHN/PEP versus the PM2A approach .........................................................................................12 Table 2: Participation in SHOUHARDO II project interventions, by region .................................................15 Table 3: Planned versus actual treatment status, by type of intervention ................................................24 Table 4: Change in the prevalence of stunting between the SHOUHARDO II ............................................28 Table 5: Age trajectory of stunting among 0-5 year olds: Comparison of SHOUHARDO II participant children with Bangladeshi children ................................................................................................. 31 Table 6: Difference-in-difference analysis: Changes in child undernutrition and its determinants from baseline to endline for eligible versus non-eligible households .....................................................32 Table 7: Instrumental variables estimates of the impact of participation in the SHOUHARDO II project on children's height-for-age z-scores .................................................................................................. 34 Table 8: Instrumental variables estimates of the impact of participation in the SHOUHARDO II project on household food security ..................................................................................................................36 Table 9: Instrumental variables/OLS estimates of the impact of participation in the SHOUHARDO II project on determinants of children’s nutritional status................................................................37 Table 10: Propensity Score Matching estimates of the average treatment effect on the treated: Full participation in Mother and Child Health and Nutrition interventions ..........................................40 Table 11: Propensity Score Matching estimates of the average treatment effect on the treated (ATT): Mother’s participation in women’s empowerment interventions .................................................43 Table 12: Propensity Score Matching estimates of the average treatment effect on the treated (ATT): Participation in livelihood promotion interventions .........................................................................45 Table 13: Propensity Score Matching estimates of the average treatment effect on the treated: Participation in water and sanitation interventions........................................................................46 Table 14: Summary of findings from single difference, difference-in-difference, instrumental variables, and propensity score matching analyses ........................................................................................48 Table 15: Instrumental variables used for endogeneity testing and 2SLS estimates of the impact of participation in the SHOUHARDO II project ....................................................................................55 Table 16 Probit propensity score model estimation for full participation in SHOUHARDO II MCHN interventions .....................................................................................................................................56 Table 17 Probit propensity score model estimation for full participation in SHOUHARDO II women’s empowerment interventions ..........................................................................................................58 Table 18: Probit propensity score model estimation for participation in livelihoods promotion interventions .....................................................................................................................................60 Table 19: Probit propensity score model estimation for participation in water and sanitation interventions ...................................................................................................................................62 4 List of Figures Figure 1: Map of SHOUHARDO II program area..........................................................................................11 Figure 2: UNICEF conceptual framework for the causes of maternal and child undernutrition ................16 Figure 3: Change in the prevalence of stunting between the SHOUHARDO II baseline and endline surveys ....................................................................................................................................................................28 Figure 4: Change in stunting prevalence among children under five: SHOUHARDO I and II ....................29 Figure 5: Age trajectory of stunting among 0-59 month olds in project area at baseline .........................30 Figure 6: Instrumental variables estimates of the impact of participation in the SHOUHARDO II project on children's height-for-age z-scores (HAZ) ......................................................................................35 Figure 7: Common support: Propensity scores of participant and non-participant households for full participation in MCHN interventions .............................................................................................41 5 Executive summary The “Strengthening Household Ability to Respond to Development Opportunities II” (SHOUHARDO II) project was implemented by CARE Bangladesh from June 2010 through September 2015 in 1,573 villages located in the poorest and most marginalized districts in the country. The overall goal of SHOUHARDO II was to reduce households’ vulnerability to food insecurity. However, a key project outcome indicator was the prevalence of stunting—or chronic, long-term undernutrition—among preschool children. It employed an integrated approach to reducing food insecurity and child undernutrition, combining nutrition-specific interventions with those that address underlying causes, such as poverty, economic and gender inequality, and poor sanitation. The objective of this impact evaluation was to determine whether the observed reductions in the prevalence of stunting that took place over the project’s implementation period, from 61.7 to 48.8 percent for children under five, were caused specifically by the project’s interventions. It further aimed to understand how the reductions were brought about by examining whether the project had an impact on a set of underlying and immediate determinants of stunting, as defined in the UNICEF Conceptual Framework for the Causes of Maternal and Child Undernutrition. It also did so by examining which of the project’s interventions had an impact, focusing on four sets of interventions: maternal and child health and nutrition (MCHN), women’s empowerment, livelihoods promotion, and water and sanitation. The evaluation employed a variety of methods, including temporal comparisons of changes in indicators for project households compared to Bangladeshi households nation-wide, difference-in-difference (DID) analysis, Instrumental Variables (IV) testing and regression, and Propensity Score Matching (PSM). The results from these analyses were triangulated to draw conclusions regarding the project’s overall impact and how it was brought about. The data employed are from cross-sectional, population-based surveys of project villages conducted near the project’s inception (December 2010) and near its end (December 2014). Given the nature of the data employed, this impact evaluation was not able to evaluate the influence of the project’s activities designed to empower the poor and assist households and communities in preparing for, responding to and mitigating the impacts of disasters and climate change. Overall, the evidence presented in the report indicates that the SHOUHARDO II project was very successful in reducing child stunting. While it is not possible to pinpoint the exact amount of stunting reduction caused with accuracy, it seems likely that a large portion, if not all, of the 12.9 percentage– point reduction in the prevalence of stunting among under-fives observed between the baseline and endline surveys can be attributable to the project. Combined, the following findings support this conclusion:  The average annual decline in the stunting prevalence among eligible project households (3.2 percentage points per year) was far higher than that of rural Bangladeshi households in recent years (0.6 percentage points);  The normal large increase in stunting prevalence seen for children as they age from the 6-18 to the 48-60 month age group was not found for the group of children whose households participated in SHOUHARDO II interventions;  The DID analysis comparing the changes over time for eligible project households compared to non-eligible project households indicates that the stunting prevalence fell more for eligible households; 6  IV estimates of the impact of participation in the project confirm that it had a substantial, positive impact on children’s height-for-age z-scores;  The DID, IV and PSM analyses all indicate that the project’s interventions led to improvements in a broad array of determinants of stunting, improvements which are necessary for reducing stunting. The findings regarding project impacts on the determinants of stunting reveal that the stunting reductions were brought about by improvements in all three underlying determinants—household food security, the quality of caring practices for mothers and children, and household health environments— and, additionally, in mother’s and children’s food consumption. With respect to household food security, the project’s interventions increased the amount of food households have access to, increased household dietary diversity (an indicator of the dietary quality), and reduced household hunger. With respect to caring practices for mothers during pregnancy, all methods point to project impacts on antenatal care, including whether that care is received in a medical facility. They suggest that the project led to women consuming more food and getting more day-time rest during their pregnancies. Finally, because of the project more women are receiving Vitamin A supplementation within six weeks of their delivery and iron/folic acid supplementation during pregnancy. For caring practices for children, project interventions led to greater knowledge among mothers of the appropriate times for hand washing and an increase in the practice of safely disposing of children’s feces. It also increased Vitamin supplementation for children, including vitamin A and multivitamin supplementation. No clear evidence was found of an impact on child immunization. Some of the improvement in household health environments seen between the baseline and endline surveys among eligible households, including improvement in access to safe water and access to sanitary toilet facilities, can likely be attributed to the project. Note, however, that the results from the different analyses are incongruent on this important determinant of child stunting. Finally, with regard to food consumption, the results suggest that dietary diversity was enhanced for households as a whole and for mothers and children living in them. They confirm that the large increase in the percent of children 6-23 months who have a minimum acceptable diet, from 10 to 46 percent, was at least partially caused by the project’s interventions. Taking into account the results for all analyses, the evidence on the impact of the project is ambiguous for diarrhea incidence among children under five, the only indicator of children’s health measured as part of the project surveys. While the evidence of an impact on mother’s nutritional status is not straightforward, it appears likely that the project’s interventions did lead to some improvement in mothers’ Body Mass Index, which is an important step towards preventing low birth weight. The PSM results give insight into the question of which of the four sets of intervention examined brought about the reductions in stunting and improvements in its determinants. While none of the interventions were found to improve children’s nutritional status (which is likely due to the weakness of the PSM method in controlling for the targeting of undernourished children that took place), they were each found to have contributed in some way. In sum: 7  The MCHN interventions had a broad influence, improving household, mother’s and children’s dietary diversity; a wide variety of the caring practices for mothers during pregnancy; a wide variety of the caring practices for children; and access to sanitary toilet facilities.  The women’s empowerment interventions also facilitated improvements in many important determinants of stunting, including household and mother’s dietary diversity, household hunger, antenatal care during pregnancy, taking more food during pregnancy, post-delivery Vitamin A supplementation of mothers, and indicators of the knowledge and use of hygiene practices.  The livelihoods promotion activities increased household, mother and children’s dietary diversity, reduced household hunger, and improved mothers’ nutritional status.  The project’s water and sanitation interventions are found to have increased access to sanitary toilet facilities. In conclusion, this report finds that the SHOUHARDO II project was successful in reducing child stunting. Two factors that contributed to its success were: 1) it addressed a broad range of underlying and immediate causes of chronic undernutrition; and 2) it brought to bear not only nutrition-specific MCHN interventions to address the problem, but also interventions designed to empower women, to promote households’ livelihoods, and to improve households’ health environments. 8 1. Introduction The “Strengthening Household Ability to Respond to Development Opportunities II” (SHOUHARDO II) project, funded by the United States Agency for International Development and the Government of Bangladesh, was implemented by CARE Bangladesh from June 2010 through September 2015. Carried out in 1,573 villages located within eleven of the poorest and most marginalized districts in Bangladesh, it is one of the largest non-emergency food security development programs in the world. The project follows on the experience of its predecessor, the SHOUHARDO I program implemented from 2005-2009, which piloted an integrated approach to reducing child undernutrition, combining nutrition-specific interventions with those that address key underlying determinants of stunting using a rights-based, livelihoods programming approach. Some of these underlying determinants are poverty and food insecurity, economic and gender inequality, poor sanitation and vulnerability to natural disasters. As shown by Smith et al. (2013), SHOUHARDO I was exceptionally successful in applying this approach to reducing child undernutrition. While the overall goal of SHOUHARDO II was to reduce households’ vulnerability to food insecurity, a key project outcome indicator was the prevalence of stunting—or chronic, long-term undernutrition— among preschool children. In addition to child mortality, stunting is associated with poor school and work performance and an increased likelihood of overweight, chronic disease and mental health issues among adults. Such personally damaging effects for young children and their families, along with its intergenerational transmission, have severe consequences for entire communities and countries, dampening their wider development (Smith and Haddad 2015)—and certainly compromising long-term food security. As documented in this report, the prevalence of stunting among children under five dropped from 61.7 at the time of the project’s inception to 48.8 four years later, a total reduction of 12.9 percentage points. This reduction of 3.2 percentage points per year is impressive when compared to the annual decline for rural Bangladeshi households as a group, which was 0.6 of a percentage point between 2007 and 2013.1 The reduction for children under two was equally impressive. The current momentum within developing countries and internationally to address the problem of child undernutrition has never been higher. The rise of the Scaling Up Nutrition (SUN) movement starting in 2010, and the publication of the Lancet Maternal and Child Nutrition Series in 2008 have both served to raise awareness of its extent and consequences. The development community is increasingly recognizing that slower-than-expected progress towards reaching the Millennium Development Goals (MDGs) by 2015—including those for poverty, secondary education, child mortality and maternal health—is due, in large part, to lack of investment in children’s nutrition (World Bank 2013). Nutrition has consequently been greatly elevated on the development agenda, and global commitment to reducing undernutrition is stronger than ever (Gillespie and Haddad et. al. 2013). In turn, answers to the question of how to accelerate reductions in undernutrition in the coming decades are in great demand. To address this increased demand, a wide evidence base is building regarding the roles of nutrition￾specific interventions, such as micronutrient supplementation and nutrition education, as well as those promoting more fundamental, underlying and basic determinants of nutritional status such as safe water access, sanitation, women’s education and empowerment, food security, national income growth and governance (Bhutta et al. 2013; Ruel et. al. 2013; Haddad 2012; Ruel and Alderman 2013; Smith and 1 See Section 6 below. 9 Haddad 2015). The experience of the SHOUHARDO II project in reducing child stunting in Bangladesh— a country with one of the highest prevalences in the world, at 41 percent of all children under five (Niport et. al. 2013)—provides a unique opportunity to gain insight into how integrated, participatory development projects implemented at the local level can contribute to accelerating reductions in child undernutrition. The objective of this impact evaluation is to determine whether the observed reductions in stunting that took place over the SHOUHARDO II project’s implementation period were caused specifically by the project’s interventions. It further aims to understand how the reductions were brought about by examining whether the project had an impact on a set of determinants of stunting—including household food security, caring practices for mothers and children, household health environments, mother’s nutritional status, and children’s health. The reason the study includes analysis of the determinants of stunting in addition to stunting itself is because they give insight into the pathways through which stunting was influenced by the project and, being necessary conditions for children’s nutritional health, alternative evidence regarding the impact of the project on stunting. The evaluation looks at the impacts of the following subsets of the project’s interventions that were implemented at the household level: 1) maternal and child health and nutrition; 2) women’s empowerment; 3) livelihoods promotion; and 4) water and sanitation. The evaluation employs a variety of methods, including temporal comparisons of changes in indicators among project households compared to Bangladeshi households nation-wide, difference-in-difference analysis, Instrumental Variables (IV) testing and regression, and Propensity Score Matching (PSM). The results from these analyses are triangulated to draw conclusions regarding the project’s overall impact and how it was brought about. The data employed are from cross-sectional, population-based surveys of project villages conducted near the project’s inception (December 2010) and near its end (December 2014). The next section of the report describes the beneficiary selection process and project interventions. Section 3 lays out the conceptual framework and outcome indicators employed as dependent variables. Section 4 describes the data collection process and Section 5 the impact evaluation methods used. Sections 6, 7 and 8 present the main empirical results. Finally, Section 9 provides a summary of the results and conclusions. 2. The SHOUHARDO II project: Beneficiary selection process and interventions The SHOUHARDO II project was implemented within the context of CARE Bangladesh’s long-term program goals, which are to eradicate poverty and promote social justice through improving social equity, livelihood security and governance in the areas in which it works. The project partnered with a variety of institutions including 16 local NGOs who are responsible for 90 percent of overall implementation coverage, and technical partners such as the International Union for the Conservation of Nature, WorldFish, the International Rice Research Institute. It was implemented with the active participation of 13 ministries within the Government of Bangladesh. The project was funded at US$130,000,000, including 287,420 MT worth of commodities for both direct distribution and monetization. This section first describes the project’s beneficiary selection process and interventions 10 in detail. It then provides data on the percent of households in project villages participating in each intervention. 2.1 Beneficiary selection process 2.1.1 Identification of project geographical areas National databases were used to identify the remote areas most vulnerable to shocks and food insecurity within Bangladesh. The following criteria for area selection were used:  Degree of food insecurity and child undernutrition  Susceptibility of the area to natural disasters and shocks  Remoteness, illiteracy and poverty rates  Avoiding duplication and overlap with other projects Figure 1 locates the resulting four SHOUHARDO II project areas—Coast, Haor, Mid Char and North Char—within Bangladesh. The northcentral Chars are riverine islands surrounded by water most of the year. They are prone to dramatic erosion and floods, which results in crop loss, isolation, and poor access to markets and services. Also highly flood-prone and with similar food insecurity issues to the Chars is the northeastern Haor area, characterized by vast expanses of depressed wetlands with scattered, elevated mounds that become largely inhabitable islands during the wet season. The delta￾like Coast region is in the deep southeast of the country where food security is threatened by regular storm surges and slow-onset disasters such as water-logging and land salinization, and the impacts of climate change. Within these four regions, 11 of the most marginalized and poor districts were chosen, followed by 30 Upazilas and 171 unions within them. Project villages were selected through Focus Group Discussions with local and national government representatives and NGOs. 2.1.2 Household selection Household beneficiary selection was guided by both socio-economic targeting and randomization required by a research project incorporated into the project’s design. Socio-economic targeting: selection of PEP households Following the SHOUHARDO I project’s approach, household targeting within each village began with the use of Participatory Rural Appraisal tools to identify the poorest households. The tools included social and resource mapping and a “well-being analysis”. Community members representing the broad range of interest groups and classes grouped households into five economic categories: extreme poor, poor, lower middle, middle, and rich. The classification criteria used included land ownership, housing condition, income level, income sources, occupation and food insecurity. Following, the “Poor and Extreme Poor” (PEP) households were selected as the project’s key targeted beneficiaries. 11 Figure 1: Map of SHOUHARDO II program area 12 Targeting associated with the RCT of the PM2A programming approach Embedded within the SHOUHARDO II project design was a Randomized Controlled Trial (RCT) implemented in order to evaluate the relative effectiveness of two approaches to targeting Maternal and Child Health and Nutrition interventions. These are: the Maternal and Child Health and Nutrition/PEP (MCHN/PEP) approach, established with SHOUHARDO I, and the Preventing Malnutrition in Children Under Two (PM2A) approach (FANTA-2, 2010). As summarized in Table 1, the MCHN/PEP approach includes only PEP households as participants in MCHN activities, including educational activities, child growth monitoring and food ration receipts (described below). By contrast, the PM2A approach includes as participants all eligible2 women and children in project villages regardless of socio-economic status. Table 1:Design of the randomized controlled trial to evaluate the relative effectiveness of the MCHN/PEP versus the PM2A approach RCT intervention arm 1: MCHN/PEP RCT intervention arm 2: PM2A PEP Eligible to participate in MCHN and all other project interventions Eligible to participate in MCHN and all other project interventions Non￾PEP Not eligible to participate in any project interventions. Eligible to participate in MCHN interventions only SHOUHARDO II project villages were randomly selected (using a computer program) into the MCHN/PEP and PM2A intervention arms, with roughly 17 percent of villages chosen to follow the PM2A approach in order to facilitate the RCT research design (see map in Figure 1). The addition of the RCT to the project design means that some non-PEP households are included as project beneficiaries. 2.2 Project interventions This section focusses on the SHOUHARDO II project’s interventions that were implemented at the household level. Many project interventions were implemented at the community level and thus could not be directly evaluated using the household level data employed for this study. These include empowerment of the poor through the establishment of Village Development Committees, efforts to increase the accountability of local elected bodies and government service providers to the PEP, and the project’s disaster preparation, response and mitigation and climate change adaptation activities. Maternal and child health and nutrition (MCHN) The SHOUHARDO II package of MCHN interventions was expected to most directly address the problem of chronic undernutrition in the project area. In line with global best practices of targeting the first 1,000 days of life, including the time in-utero and ending with the child’s second birthday, to achieve sustained impact on nutritional status, the package prioritizes children under age 2 and pregnant and 2 The eligibility criteria are based on children’s ages and the pregnancy status of women (see Section 2.2.). 13 lactating women. During this time the child has increased nutritional needs to support rapid growth and development, is more susceptible to infections, and is completely dependent on others for nutrition, care and social interactions. Growth faltering typically begins during pregnancy and continues to about 24 months of age. The loss in linear growth is not recovered, and catch-up growth later on in childhood is minimal (UNICEF 2013). A key component of the MCHN package was promotion of health and nutrition behavior change through nutrition education in two forms. The first was monthly “courtyard sessions” led by trained community health volunteers (CHVs)3 with topics including optimal breastfeeding, complementary feeding and weaning practices, care for mothers during pregnancy and delivery, and hygiene practices. The second was cooking and feeding demonstration sessions. A third component of the package was monthly Growth Monitoring and Promotion sessions in which children’s height and weight was monitored. Children whose growth was faltering received follow-up care from CHVs. A fourth MCHN intervention was the provision of monthly food rations to pregnant women, women with children under two, and children under two living in eligible households. The ration was provided both to fill gaps in nutritional intake and to provide an incentive to participate in behavior change activities. It contained wheat, vegetable oil and yellow split peas. Integrated into these MCHN interventions were efforts to establish linkages with preventive and curative health and nutrition services, build capacity for community-based integrated management of childhood illnesses, and facilitate linkages of mothers with the Ministry of Health and Family Welfare program to provide micro-nutrient supplementation for pregnant and lactating mothers. Women’s empowerment Although there have been gains in women’s empowerment in Bangladesh in recent years, discrimination against women remains strong and pervasive in Bangladesh (Nosback, Champion and Mutahara 2014). At the start of the SHOUHARDO II project, very few women could make basic economic decisions on their own, their freedom of movement was restricted, only five percent earned cash income, and over a quarter had experienced some form of domestic violence in the previous year (Caldwell, Ravesloot and Smith 2011). CARE’s commitment to women’s empowerment as a means of addressing underlying causes of child undernutrition is a distinguishing feature of the SHOUHARDO II design. The central intervention designed to do so was Empowerment, Knowledge and Transformative Action (EKATA) groups for promoting life-skills education, empowerment and social change. Made up of 20 women and 15 adolescent girls recruited from among interested community members, and facilitated bi-weekly by a paid volunteer, the groups provided a platform for empowering women and adolescent girls through education, solidarity, group planning, and rights advocacy. The EKATA intervention had a broad range of goals: increasing women’s decision making power at household and community levels, reducing gender-based violence, raising awareness of educational entitlements for women and girls, building women’s leadership, advocacy, and literacy skills, and consciousness-building around women’s’ rights in 3 The CHVs were married women with children who were at least 20 years old with secondary education, previous experience in health related work and socially accepted by their community. They received a four-day training program as well as counseling and facilitation skills. For continued education and support, they gathered for a one-day meeting once per quarter and received technical support from CARE and partner NGO technical staff. 14 existing legislation and important social issues, including dowry, early marriage, divorce, and violence against women. More directly focused on promoting the economic empowerment of women was the establishment of self-help savings groups. While men could be members of these groups, they were directed at women, and the majority of members were women. The groups provided a means for women to save for investment purposes, pool their incomes in times of need, and avoid taking loans from money lenders. Note that another project intervention directed at women’s empowerment in the long run was the establishment of Early Child Care for Development (ECCD) centers, preschools that introduce a learning process, flow of information, and preparation for entering formal schooling that has been traditionally denied to girls. An equal number of girls and boys are enrolled. In addition, parenting sessions are held for mothers and fathers with the aim of improving parenting skills, especially related to girls’ enrollment in school. ECCD is not evaluated in this study as it is not expected to directly empower the current generation of women. Livelihoods promotion: Core Occupational Groups This set of interventions was designed to directly address food insecurity and poverty in the project area by increasing food production and incomes. Project beneficiary households were divided into four distinct Core Occupational Groups (COGs) based on asset holdings (availability of land, access to water bodies, and labor availability) for the receipt of packages of input support and training. The sets of interventions are: (1) Crop production Provision of seeds/seedlings, organic fertilizer and training in irrigation, field preparation and crop management to support the production of key field crops (e.g., rice, wheat and maize). (2) Fisheries Provision of fingerlings, lime, fish meal and fertilizer for fish culture, in addition to fish nets, boats, and aluminum patil/pots for fish capture. (3) Comprehensive homestead development (CHD) Provision of saplings, seeds, organic fertilizer for homestead gardens and animals (chickens, ducks and goats) and fencing for animal rearing (4) Income generating activities (IGA) Entrepreneur development and business management training; skill training based on selected trade. Water and sanitation Diarrheal disease is a key cause of child undernutrition in Bangladesh, with lack of access to safe water and sanitary latrines being its main structural cause (United Nations Integrated Regional Information Network (IRIN), 2010). At the start of the project, while 61 percent of households had access to safe water, only 26 percent had access to a sanitary latrine. This problem was addressed by assisting households in obtaining safe, arsenic-free drinking water through the installation of tube wells and arsenic testing, as well as access to sanitary latrines. 15 2.3 Participation in project interventions Table 2 presents data on the percent of households in project villages with children under five that participated in each intervention by region. Table 2: Participation in SHOUHARDO II project interventions, by region Coast Haor Mid Char North Char All (Percent of households) Mother and Child Health and Nutrition (MCHN) Courtyard sessions 63.9 66.7 64.3 67.2 66.3 Cooking/feeding sessions 53.2 56.7 63.0 67.5 61.1 Child growth monitoring 54.9 63.6 60.7 66.6 63.6 Food ration 62.0 61.9 57.3 55.2 58.9 All MCHN interventions (Full participation) 41.0 43.9 45.1 45.8 44.6 Any MCHN intervention 70.4 72.9 72.2 77.3 74.1 Women's empowerment Mother is EKATA group member 4.9 2.8 8.2 9.6 6.1 Mother is savings group member 10.8 10.3 8.7 9.5 9.8 Any empowerment intervention 14.3 11.8 14.0 15.6 13.6 Livelihoods promotion Crop production 11.3 12.1 16.1 10.2 12.1 Comprehensive Homestead Development (CHD) 29.5 28.6 17.5 20.8 24.2 Fisheries 6.4 6.5 2.1 5.6 5.5 Income generating activities (IGA) 20.9 23.7 28.0 26.9 25.3 Any livelihoods promotion intervention 66.2 67.4 58.2 58.5 62.9 Water and sanitation Any water and sanitation intervention 24.6 26.0 6.4 13.0 18.4 Any SHOUHARDO II intervention 77.8 78.9 80.2 81.0 79.8 Three-quarters of all households participated in at least one MCHN intervention over the life of the project, with roughly equal participation in the educational, growth monitoring, and food ration interventions. Near 45 percent of households participated in all four MCHN interventions, hereafter termed “full participation” in MCHN. The next most commonly participated in intervention is livelihoods promotion, with a prevalence of 63 percent. Roughly a quarter of households participated in CHD and IGA, the most popular of the livelihoods promotion interventions. Twelve percent participated in crop production and only five percent in fisheries. Participation of mothers living in households with children under five in EKATA was quite low, at six percent, perhaps due to the child care time constraints felt by these mothers. Participation in savings groups was somewhat higher, at 10 percent, giving a total overall participation prevalence in the two women’s empowerment interventions considered here of 14 percent. Finally, 20 percent of households participated in the project’s water and sanitation interventions. 16 Overall, 80 percent of households living SHOUHARDO II project villages participated in the project’s interventions in some form. Overall participation prevalences vary little across the four regions, although there are some substantial regional differences for membership in EKATA groups (higher in Mid and North Char) and participation in water and sanitation interventions (higher in Coast and Haor). 3. Conceptual framework and measurement of stunting and its determinants 3.1 UNICEF conceptual framework The conceptual framework guiding this report’s analysis is the UNICEF conceptual framework for the causes of maternal and child undernutrition (see Figure 2). The framework lays out the hierarchical relationship between the immediate, underlying, and basic causes of undernutrition. Figure 2: UNICEF conceptual framework for the causes of maternal and child undernutrition Source: UNICEF (2013). 17 The immediate causes, which manifest themselves at the level of the individual child, are inadequate dietary intake (energy, protein, fat, and micronutrients) and disease. These factors themselves are interdependent. A child with inadequate dietary intake is more susceptible to disease; disease in turn depresses appetite, inhibits the absorption of nutrients in food, and competes for a child’s energy. The underlying causes, which impact child nutritional status through the immediate causes, manifest themselves at the household level. The first is household food insecurity, or the inability of a household to access enough food of adequate quality for all of its members to live an active, healthy life. The second is inadequate quality of caring practices for children and their mothers. Examples of caring practices for children are child feeding, health-seeking behaviors, and cognitive stimulation. The most obvious aspect of care for women that affects children’s nutritional wellbeing is care and support during pregnancy and lactation. Women are typically the main caretakers of children after birth, and in order to provide quality care they need continued adequate food consumption and health care, rest and measures to protect their mental health, such as protection from abuse. The third underlying cause is an unhealthy household environment and inadequate health services, which condition children’s exposure to pathogens and the use of preventative and curative health care. Elements of a health environment include access to safe water, to sanitary facilities for disposing of human waste and to health services. Physiologically, a mother’s nutritional status is closely tied that that of her child. Adequate maternal nutrition and health are crucial to prevent child undernutrition. Pregnancy increases nutrient needs and is a time when illness and environmental and psychosocial stress can contribute to undernutrition of an unborn child through impaired fetal development and low birthweight. Undernourished girls have a greater likelihood of becoming undernourished mothers, who then have a greater likelihood of giving birth to a low birthweight baby, leading to an intergenerational cycle of undernutrition. The issue of maternal undernutrition is particularly important to take into account in Bangladesh, which has both a high prevalence of maternal undernutrition (24 in 2011) and low birthweight (22 in 2006) (UNICEF 2013). Finally, the basic causes, which in turn impact nutritional status through the underlying causes, manifest themselves at broader geographical levels, such as national, regional or global. They form the economic, political, environmental, social and cultural context in which children’s nutritional status is determined. While the SHOUHARDO II project addresses some of the basic causes of child undernutrition, such as poverty and the disempowerment of women, this impact assessment focusses only on the underlying and immediate causes (in addition to stunting itself). 3.2 Measures of stunting and its determinants In this section the measures of stunting and its determinants employed as dependent variables in this study are described. As noted in the introduction, one of the reasons the study includes analysis of the determinants of stunting in addition to stunting itself is because it helps understand the pathways through which stunting may have been influenced by the project. A second reason is that improvements in the determinants are necessary for bringing about improvements in children’s nutritional health. Evidence that the project brought about such improvements thus give alternative 18 evidence regarding the impact of the project on stunting. In the case of the evaluation of SHOUHARDO II, the need for such alternative evidence is heightened: as discussed below (Section 4.2), the selection of households into the project’s MCHN interventions was partially based on whether or not the household had a child under five who was undernourished. Statistically, this makes it more difficult to draw out the impact of the project on stunting itself. 3.2.1 Stunting Stunting is a result of inadequate growth of the fetus and child and results in a failure to achieve expected height compared to a healthy, well-nourished child of the same age. It is a cumulative indicator of growth failure and a marker of chronic insufficient protein and energy intake, frequent infection, sustained inappropriate feeding practices, and impaired brain development (Black et al 2013; UNICEF 2013). The rationale for employing stunting as an indicator of undernutrition for this impact evaluation is four￾fold. First, it is a key SHOUHARDO II project outcome indicator against which progress towards project goals was assessed. Second, replacing underweight, it has become the consensus measure among the international community to mark the damage that is done from the interaction of poor diet and repeated infections (Black et. al. 2013; UNICEF 2013). Third, it is a measure of long-term, chronic undernutrition rather than undernutrition as a result of short-term fluctuations in dietary intake and/or health. It is thus particularly well suited to the evaluation of this project, which took place over more than four years. Fourth, stunting was more prevalent than either wasting (measuring acute undernutrition) or underweight (a composite measure of both chronic and acute undernutrition) at the start of the project and thus represented a more widespread problem. The specific indicator employed as a dependent variable for this analysis is children’s height-for- age z￾score (HAZ) measured using data collected on height or length and months of age. A child is considered stunted if her or his HAZ is less than -2 standard deviations below the median of a global reference population of children who are well nourished and received key recommended caring practices. The current reference is the World Health Organization 2006 Child Growth Standards (de Onis et al. 2004). 3.2.2 Determinants of stunting The selection of the determinants of stunting included in this analysis is guided by the conceptual framework presented above. It is also influenced by the project outcome and impact indicators included in its Indicator Performance Tracking Table (see TANGO 2015) as well as the statistical methods employed and data availabilities. The determinants include indicators of all three underlying causes of child undernutrition (food insecurity, inadequate caring practices, and an unhealthy household environment) and both immediate causes (inadequate dietary intake and disease). The variables are listed in Box 1 and described in detail in Appendix 1. 19 4. Data The data from two cross-sectional, population-based surveys of all households in SHOUHARDO II villages, whether eligible to participate in project interventions or not, are employed for this impact evaluation. The first is the project baseline survey, conducted between December 8, 2010 and January 2, 2011. The second is the project endline survey, conducted between November 17 and December 12, 2014. To ensure comparability, the data collection methodology was identical for the two surveys. The survey questionnaires, which can be found in Caldwell, Ravesloot and Smith (2011) and TANGO, International (2015), were designed by TANGO, International in collaboration with CARE Bangladesh and Food and Nutrition Technical Assistance-II. The data collection was conducted by Mitra and Associates and TANGO, International. A two-stage, stratified sampling design was employed, with two levels of stratification. The first was a division of the SHOUHARDO II operational area into its four geographical regions – Coast, Haor, Mid Box 1. Determinants of stunting employed as dependent variables Household food security  Number of months of adequate household food provisioning  Household dietary diversity score  Household hunger score Caring practices for mothers during pregnancy  Antenatal care during pregnancy  Antenatal care in a medical facility during pregnancy  More food during pregnancy  More rest during pregnancy  Vitamin A six weeks from delivery  Iron/folic acid during pregnancy Caring practices for children  Mother’s knowledge of hand washing at five critical times  Safe disposal of feces of children 0-35 months  Number of vaccinations received (0-23m)  Vitamin A capsule in the last six months (6-23m)  Monomix multivitamin supplement (6-23m) Household health environment  Access to safe water  Access to sanitation Mother’s and children’s food consumption  Mother’s dietary diversity  Child minimum dietary diversity (6-23m)  Child minimum meal frequency (6-23m)  Child minimum acceptable diet (6-23m) Children’s health  Child diarrhea (0-59m) Mother’s nutritional status  Mother’s Body Mass Index Note: Variable definitions are given in Appendix 1. 20 Char, and North Char – reflecting the distinct geographic areas where the project was implemented. The second level of stratification was into the two intervention arms defining the RCT embedded into the project’s design: MCHN/PEP and PM2A, as described above. An equal number of villages and households were sampled in the resulting eight strata. Following stratification, sampling took place in two stages. In the first, 25 villages were randomly chosen within each stratum using probability proportionate to size (PPS) sampling. In the second, 45 households were randomly selected in each village, for a total of 9,000 households. Sample size calculations were based on ensuring the ability to detect a 10 percentage-point change in stunting prevalence between the project’s baseline and endline surveys from a projected initial prevalence of 50%. Assumptions of a 95 percent confidence level, 80 percent power, and a design effect of 2.0 yielded a minimum sample size of 666 households per stratum. To keep the sample size reasonable, a single sample of households was selected to collect both socio-economic data (from all households) and health and nutrition data (needed only from households with children under five). To do so, the sample size factored in the proportion of the population in Bangladesh that is aged 6-59 months and the average household size. Applying the required sample size above to these factors plus adding in a 10 percent cushion to account for non-response yielded a final sample size of 1,119 households per strata, or a total of 8,952 households. The sampling of 45 households within 200 villages met this sample size requirement. Only the data collected from households with children under five with valid anthropometric data were employed for this study. In these households, an index child was randomly chosen for collection of data on children’s and mother’s health and nutrition.4 After cleaning of the anthropometric data, the analytic sample size for the study is 2,471 children under five (6-59 months) and 871 children under two (6-23 months). For household-level variables, such as the food security indicators, data are employed for households with children under six months as well, increasing the sample size to 2,844. 5. Impact evaluation methodologies An impact evaluation is a study conducted in order to determine the extent to which changes in outcomes can be attributed to a project or intervention. Evaluating such attribution requires comparing what happened to the outcome with an intervention (the factual) to what would have happened to the outcome without it, the latter referred as the “counterfactual”. The counterfactual is never known with certainty because the exact same participants in an intervention cannot not participate in it at the same time. Given this issue, two necessary conditions for an impact evaluation to be conducted in a rigorous manner are that (1) a non-participant control group be available so that a counterfactual can be identified; and (2) that the problem of selection bias be addressed (Waddington et. al. 2012). This latter 4 For the baseline survey child anthropometric data were collected only for one index child in each household having a child under five. For the endline, following Food-for-Peace guidance, anthropometric data were collected for all children under five in each household, with data for other child-level variables being collected only for an index child. In this report’s analysis of the endline data, only that for the selected index child is employed for two reasons. First, doing so allows valid comparisons of stunting prevalences over time (endline households with multiple children and thus greater child care burdens have greater representation than they do in the baseline). Second, for the impact analysis methods involving regression, it is not possible to properly control for intra-household correlations statistically when only some households are represented more than once. 21 problem arises because, in most cases, either purposeful targeting of project interventions to specific populations (e.g., the most poor) and/or self-selection of participants into interventions takes place. This renders the control group and the participant group fundamentally different from one another prior to the commencement of project activities (Waddington et. al. 2012; Khandker, Koolwal & Samad, 2010). The SHOUHARDO II surveys are population based. Ample data for households that did not participate in the project’s interventions at all or only in its MCHN activities— whether due to the PM2A RCT allocation mechanism or by choice—are available, thus providing a pool of potential control group households. Further, as outlined below, the endline survey was extended to allow collection of data necessary for addressing the problem of selection bias, that is, data on factors affecting households’ participation in various interventions. A special effort was made to collect data on factors that are typically “unobservable”, the exclusion of which can lead to bias in estimates of the impacts of interventions. As described here, the data allow use of a variety of impact evaluation methodologies, including descriptive and regression-based methods, to determine whether and how the SHOUHARDO II project led to the 12.9 percentage-point reduction in the prevalence of stunting among children under five observed since the project’s inception. 5.1 Descriptive methods The first descriptive method is a comparison of the change in stunting in the project area with the change nationally over the same time period. This analysis is important for ruling out the possibility that the change in stunting in the project area was due to forces external to the project. Specifically, we examine the change in stunting among children under five living in eligible project households between the baseline and endline surveys compared to the change that took place for this age group in rural areas countrywide in recent years. The data used are from three nationally-representative Demographic and Health Surveys (DHS), those undertaken in 2004, 2007 and 2011 (NIPORT et al. 2005, 2009 and 2013) and a survey conducted in 2013 administered using the same methods as those of the DHS (Shahin et al. 2014).5 Note that the project’s population of 370,000 people is very small relative to that of Bangladesh as a whole (roughly 160 million), such that changes in the project area had negligible influence on the stunting prevalence country wide. The second method is a comparison of the actual age trajectory of the stunting among a specific age cohort of children living in eligible project households compared to the projected age trajectory of that cohort at the time of the baseline, that is, before the project’s activities commenced. The cohort is children who were 6-18 months old at baseline (in December 2010) and 48-60 months at endline (December 2014). Stunting typically shows a large increase over these age groups. Depending on whether their household actually participated in them, this cohort of children was exposed to the project’s MCHN interventions for an average of 12 months and to the rest of its interventions for an 5 This survey was conducted by the National Institute of Population Research and Training (NIPORT), the same organization that conducts Bangladesh’s Demographic and Health Surveys. The sampling scheme was similar to that of the DHS to ensure comparability of data across the surveys. 22 average of three and a half years. We explore whether the change for these children shows an altered pattern from that projected at baseline. The third method is a comparison of the changes in stunting and its determinants between the baseline and endline surveys for the group of households who were eligible to participate and the group who were not eligible to participate. This intent-to-treat (effect of treatment as assigned), difference-in￾difference analysis allows determination of whether the eligible households did better than non-eligible households while taking into account any initial differences between the groups at baseline. By doing so it controls for any changes that took place in the project area that are not related to project interventions or that are only indirectly related to them through spillover effects. Spillover effects occur when an intervention has an impact on households that do not participate in it. Examples of how this could have occurred in the SHOUHARDO II project were MCHN behavior change messages and technical skills gained through COG groups being disseminated to non-participants by participants. 5.2 Instrumental variables regression analysis Instrumental variables analysis is a regression technique that allows us to rigorously estimate the impact of participation in the SHOUHARDO II project using the endline survey data by correcting for systematic differences between the households that actually participated in it and those that did not. It does so by controlling for selection bias due to both observable factors affecting participation and outcomes and unobservables. Examples of such unobservable factors that are typically not measured are ability, entrepreneurship, attitudes towards risk, weather shocks, social capital, and pre-project outcome levels (Habicht et al. 2009; Gilligan and Hoddinott 2007; Linnemayr and Alderman 2011). By controlling for these factors we are ensuring that in our estimations only the causal effect is being identified, and that only the causal portion of the observed relationship is represented by regression coefficient estimates. Of particular importance in the case of SHOUHARDO II is to control for the fact that the project’s MCHN interventions were purposefully targeted towards households whose children were identified to be undernourished. These children’s mothers were given greater priority by CHVs during the implementation of MCHN activities and special encouragement to participate in courtyard sessions, cooking/feeding sessions, and growth monitoring of their child (Wadud 2015). Given that food rations were used as an incentive to participate in the other interventions, these children’s mothers were also probably more likely to receive a food ration than mothers whose children were not undernourished. The IV method corrects for this type of reverse causality, where the treatment variable itself is influenced by the outcome. The basic regression model used to evaluate the impact of the SHOUHARDO II project on children’s long￾term nutritional status and other dependent variables is: , where Ti is a dummy variable equal to 1 if the household participated in the project, γ is the treatment effect, and the Xi are child, mother and household characteristics believed to influence outcomes. The term εi is the unobserved error term. If the decision to participate, T, is correlated with the error term, Ordinary Least Squares (OLS) regression will yield biased estimates of project impact. Two-stage least squares (2SLS) is used to correct for this bias. In the first stage, a set of instruments, Z, along with the 23 child, mother and household characteristics, X, are used to predict the treatment status of each household: In the second stage, the predicted value of T is used to estimate project impact: The term γ measures project impact. The child, mother and household characteristics used as independent variables in the IV regression analysis (the Xi) are listed in Box 2. These variables are the typical observables found in reduced-form analyses of child undernutrition (e.g., Smith et al. 2003). Box 2. Child, mother and household characteristics used as independent variables for Instrumental Variables/OLS regressions Child characteristics  Age in months, age-squared  Whether child is a girl Mother characteristics  Mother’s age  Mother’s education level a/ None Primary Secondary Household characteristics  Age of household head  Whether household is headed by a female  Education of household head a/ None Primary Secondary  Occupation of household head: Farming Agricultural laborer Non-agricultural laborer Salaried employment Self employment Unpaid household work Other  Household size  Household age-sex composition Percent females 0-16, 16-30, and 30+ Percent males 0-16, 16-30, and 30+  Well-being category at baseline Extreme poor Poor Middle Middle-rich Rich  Region of residence Coast Haor Mid Char North Char a/ For models with a child or mother-level dependent variable, mother’s education is employed. For models with household-level dependent variables, the education of the household head is employed. 24 With respect to the instrumental variables employed (the Zi), a valid instrument must satisfy two conditions. First, the “relevance” condition specifies that the instrument must be sufficiently correlated with participation in the intervention. Second, the “overidentification” condition specifies that the instrument must only be correlated with the outcome of interest through T. That is, it must only affect the outcome through its effect on participation in the project and not through any other means (Bazzi and Clemens 2013; Baum, Schaffer and Stillman 2007). The random allocation of villages into PM2A and MCHN/PEP groups discussed in Section 2.1.2 is an exogenous allocation mechanism underlying the planned treatment status of households that satisfies both conditions. As will be seen, planned treatment status has a strong correlation with actual participation and, for most of the dependent variables of this study, has no influence on outcomes except through influencing participation. Linnemayr and Alderman (2011) successfully used such planned treatment status as an instrument for actual treatment status in the case of an impact evaluation where significant deviation from the (randomized) planned treatment status occurred (see also Ten Have et al. 2008). As shown in Table 3, deviation from planned treatment status also occurred in the case of SHOUHARDO II. A substantial proportion of non-eligible households in MCHN/PEP villages participated in project interventions (35 percent of these households), and non-eligible households in PM2A villages participated in women’s empowerment, livelihoods promotion and water and sanitation interventions. Table 3: Planned versus actual treatment status, by type of intervention (Percent of households participating in interventions) RCT intervention arm 1: MCHN/PEP RCT intervention arm 2: PM2A PEP Non-PEP PEP Non-PEP Eligible Actual Eligible Actual Eligible Actual Eligible Actual MCHN Any MCHN intervention 100 87.7 0 30.9 100 90.4 100 84.1 Received a food ration 100 75.1 0 6.5 100 79.1 100 73.8 Women's empowerment 100 17.5 0 2.5 100 19.4 0 7.9 Livelihoods promotion 100 86.3 0 7.8 100 85.4 0 6.2 Water and sanitation 100 21.6 0 6.8 100 25.3 0 19.4 Any SHOUHARDO II intervention 100 94.6 0 35.0 100 95.4 85.7 Note: Highlighted areas of table represent deviations from planned treatment. In some cases the actual instrument employed here for IV testing and estimation is planned treatment status, while in others it is the PM2A status of the village of residence. These instruments are complemented by several others (specified in the IV results section below), as the use of multiple instruments is required for testing the overidentification condition. While the main instruments employed are arguably correlated with participation and theoretically exogenous, statistical tests of the relevance and overidentification conditions are undertaken for formal verification. With regard to instrument relevance, a test of whether the instruments are strong enough to remove a substantial portion of the OLS bias if it exists is undertaken. The Kleibergen-Paap rk Wald first-stage F statistic is reported and compared to critical values developed by Stock and Yogo (2005) for weak-instruments hypothesis tests. The null hypothesis that the maximum bias in the coefficient 25 estimate for each potentially endogenous variable is greater than 5, 10 or 20 percent of the OLS bias is assessed. This test identifies cases of weak instruments, which can arise even when the correlations between the endogenous regressors and instruments are significant at conventional levels (5% and 1%). Next, Hansen’s J test for overidentification of all instruments, which is robust to heteroskedasticity and within-group correlation, is conducted. If the J-statistic p-value is <0.1, the instruments are considered to not be valid. Given relevant and valid instruments, the test for endogeneity employed and implementation of 2SLS is undertaken using the STATA command xtivreg2 developed by Schaffer (2010). Where testing indicates that the treatment variable is not endogenous, OLS is used for estimation rather than 2SLS. 5.3 Propensity score matching The IV analysis can only be undertaken for participation in the SHOUHARDO II project as a whole and not for its individual interventions.6 Yet knowledge of the impacts of individual interventions is needed to understand how the SHOUHARDO II project brought about a reduction in stunting if it in fact did. Was the reduction due solely to the MCHN “direct nutrition” interventions, including the monthly distributions of food aid, or did the interventions that addressed deeper causes and were likely to set in motion sustainable impacts contribute as well? Using the endline survey data, this question is investigated using PSM to create comparable-on￾observables control groups for each intervention from among households that did not participate in them to serve as the counterfactual. The impact of interventions is estimated using the difference in HAZ (and its determinants) between the control group and intervention group. To isolate the independent impact of each intervention, the fact that there may be differences in participation in the other project interventions across the participant and control groups is accounted for in the analysis. Note that for MCHN, the analysis relies on the “full participation” variable, defined as participation in all four MCHN interventions. Doing so allows construction of an adequately-sized control group. The matching process in PSM takes place using measured indicators of characteristics that are believed to influence participation in an intervention as well as those influencing the outcome of interest. If these observed characteristics are the only ones influencing participation, the estimates are deemed unbiased and the important “conditional independence” condition is met. However, if unobserved characteristics also influence participation, then the estimates will be biased (Khandker, et al., 2010). The challenge then, is to collect data on the entire universe of such characteristics so that none can be deemed unobserved. In addition to planned participation established at baseline (see Section 4.2), households’ participation in SHOUHARDO II interventions was influenced by two broad factors: (1) targeting conducted by project administrators; and (2) households’ and individuals’ own decisions on whether to participate. As part of the SHOUHARDO II endline survey, data were collected for many of the determinants of participation and outcomes typically employed in conventional PSM impact analyses. These are the same variables used in the IV analysis (listed in Box 1). 6 This is because the participation decision for individual interventions, as we will see in this report, was highly dependent on participation in other interventions. It is not possible to address the endogeneity of multiple treatment variables simultaneously in the context of this research project. 26 Others factors affecting participation are not typically measured, and are thus relegated to the “unobservables” category. For the SHOUHARDO II project these were identified to be the following: aspirations and confidence to adapt, peoples’ time constraints, social capital, women’s decision making power in their households, personal familiarity with project staff, and household shock exposure. To capture these factors and thus render them “observables”, a module on factors affecting participation in the project was added to the endline questionnaire so that they could be explicitly included in the PSM analysis. Also important for a valid assessment of impact using PSM is that the characteristics affecting participation used for matching not be affected by project activities themselves. Given that panel data were not collected (that is, the baseline and endline surveys were not administered to the same households), retrospective recall was used to collect data on the characteristics that may have been affected by the project. That is, households were asked to answer questions regarding their current situation and then give information that allows estimation of the values of the variables as they were before the inception of the project (specifically, in 2009 or “five years ago”). Since each households’ ”well-being category” (extreme poor, poor, middle, middle-rich, and rich) was assigned before the baseline survey was administered, retrospective data are not needed for measuring initial socio-economic status. The variables used for matching are listed in Box 3 and described in detail in Appendix 2. Box 3. PSM: Child, mother, household and village characteristics used for matching Child and mother characteristics and household socio-demographic characteristics: See Box 2 Other Household characteristics  Current shock exposure/relative shock exposure in 2009  Bonding social capital/relative bonding social capital in 2009  Exposure to alternatives (outside of village)  Absence of fatalism  Number of SHOUHARDO II project staff known in 2009  Leisure time in 2009  Women’s decision making score in 2009  Participation in other interventions (than the one being evaluated)  Receipt of a food ration from another project Village characteristics  Classified as extremely vulnerable at baseline  Total number of households  Whether CARE is implementing NGO  Whether nearest town is greater than one walking hour away  PM2A village Baseline district mean child nutritional status  Mean height-for-age z-score  Mean weight-for-height z-score 27 For any intervention, PSM estimates of impact are generated in three steps. The first is to estimate a probit participation model using data on both participants and non-participants to compute a probability of participation, or “propensity score”, for each household conditional on the observed characteristics. In the second step, participant households are matched with non-participant households based on similarity of propensity scores. An important condition for the success of this step is “common support”. Participant households must be similar enough to non-participant households in the observed characteristics so that there are sufficient non-participant households close by in the propensity score distribution with which to make matches (Khandker, et al., 2010). Participant propensity scores that are higher than the maximum or lower than the minimum of the non-participant distribution are dropped. In the third step of PSM, the average value of the outcome variable of the matched participant and non￾participant groups of households are compared to calculate an estimate of the impact of the intervention, or the “average treatment effect on the treated” (ATT). Of the many techniques available, PSM is conducted here using kernel matching, for which each treated household is matched to a group of non-treated households with propensity scores within a certain radius.7 The control group outcome is computed as a weighted average, with a lower weight given the greater is the propensity score difference from the treated household. The analysis is conducted using PSMATCH2 in STATA along with PSTEST to test for matching effectiveness (Leuven & Sianesi, 2003). Matching effectiveness is evaluated by conducting t-tests for equality in the mean values of the characteristics on which matching is based across the participant and matched non-participant groups of households. An overall summary measure is given by the p-value from a likelihood ratio test for the joint insignificance of the characteristics after matching (that is, using the matched sample only). If the characteristics are no longer jointly significant (p>0.10), then matching has succeeded. 7 The radius depends on the bandwidth of the kernel. After finding that variations between 0.01 and 0.10 make little difference to the ATT estimates, a bandwidth of 0.05 is used for all estimates. 28 6. Results: Descriptive evidence of project impacts 6.1 Trends in stunting among project households compared to nationally Table 4 (also illustrated in Figure 3), reports the change in the prevalence of stunting between the SHOUHARDO II baseline and endline surveys. The prevalences for both under-fives and under-twos dropped by 12.9 percentage points. Because the under-two prevalence was lower at baseline than that for under-fives, the percentage change in stunting for under twos was somewhat higher. Note that in both age groups the prevalence was far higher for boys than for girls at baseline. However, the drop over the four years was also comparatively greater for boys, and thus the gap between boys and girls was narrowed considerably by baseline. Table 4: Change in the prevalence of stunting between the SHOUHARDO II Baseline (December 2010) Endline (December 2014) Difference Percent difference Under fives (6-59 m) All 61.7 48.8 -12.9 -20.9 Girls 56.5 47.8 -8.7 -15.4 Boys 66.1 49.7 -16.4 -24.8 Under twos (6-23 m) All 55.8 42.9 -12.9 -23.1 Girls 50.4 41.2 -9.2 -18.3 Boys 61.0 45.0 -16.0 -26.2 Figure 3: Change in the prevalence of stunting between the SHOUHARDO II baseline and endline surveys 61.7 55.8 48.8 42.9 0 10 20 30 40 50 60 70 Under-fives (6-59m) Under-twos (6-23m) Baseline Endline 29 A comparison of the change in the prevalence of stunting for under-fives among SHOUHARDO II’s participant population with trends in rural Bangladesh is given in Figure 4. Although less than that of the SHOUHARDO I project,8 compared to the national trend, the SHOUHARDO II population saw a rapid reduction over the period. The average annual decline was 3.2 percentage points while the trend in rural Bangladesh whole was 0.6 percentage-points per year. This comparative evidence rules out the possibility that the decline in stunting seen among the SHOUHARDO II project population was brought about by positive forces emanating from wider favorable economic, climatic or policy-related trends in the country. Figure 4: Change in stunting prevalence among children under five: SHOUHARDO I and II Sources: SHOUHARDO I data: Smith et al. (2012). SHOUHARDO II data: TANGO, International (2015). National (rural) prevalences: NIPORT et al. (2005, 2009, 2013) and Shahin et al.(2014). 8 The total reduction for the SHOUHARDO I project was of 15.7 percentage points over 3.5 years, or 4.5 percentage points per year. 56.1 40.4 61.7 52.7 48.8 44.3 45.0 42.7 41.3 30 35 40 45 50 55 60 65 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 SHOUHARDO I SHOUHARDO II National (rural) 30 6.2 Shift in the age trajectory of stunting among project households Following the typical pattern for children from poor households in developing countries, in Bangladesh there is normally a steep increase in stunting as children age over the six month to 2 year￾old range. This increase is associated with poor weaning practices and exposure to infectious disease. Continued high prevalences for older age groups are due to the initial growth failure at younger ages as well as poor household food access (Beaton et al. 1990). The SHOUHARDO II baseline data exhibit this pattern, as can be seen in Figure 5. Figure 5: Age trajectory of stunting among 0-59 month olds in project area at baseline Source: SHOUHARDO II baseline survey. Table 5 shows the pattern for Bangladeshi children in 2011, giving stunting prevalences for the age cohort of interest, the group of children who were 6-18 months old at the time of the baseline and 48-60 by the time of the endline. The prevalence was 30.5 among 6-18 month olds, rising to 41.9 for 48- 60 month olds. By contrast, there was no increase in stunting prevalence among the children that had been exposed to SHOUHARDO II project interventions (the change was -0.6 percentage points). This finding is even more notable given that not all children in the 6-18 month group at baseline were exposed to the project’s MCHN interventions for the full 18 month eligibility period (6-24 months), simply because they were not in the eligible age range for that long. For example, the 18 month olds were only exposed to project interventions for six months. 10 20 30 40 50 60 70 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 Percent stunted Age (months) 6-18 month olds 48-60 month olds 31 Table 5: Age trajectory of stunting among 0-5 year olds: Comparison of SHOUHARDO II participant children with Bangladeshi children Stunting among 6-18 month olds Stunting among 48-60 month olds Increase (percentage points) Bangladeshi children (2011) 30.5 41.9 11.4 Project participant children 49.3 48.7 -0.6 (baseline) (endline) Source: Data for Bangladeshi children are from NIPORT et al. (2013). We can deduce from this evidence that something happened to the children living in project households that prevented many of them from becoming stunted as they aged, an indication that the project’s interventions plausibly led to a reduction in stunting. 6.3 Difference-in-difference analysis As noted in Section 2.1, the SHOUHARDO II project was designed such that all households in project villages randomly assigned to the PM2A programming approach were eligible to participate in project interventions. By contrast, in MCHN/PEP villages, only the PEP were eligible to participate, leaving a group of non-eligible surveyed households that can serve as a control group for intent-to-treat comparison purposes. The group is non-PEP households in MCHN/PEP villages. As mentioned, this design was adhered to for the most part. We can thus compare the change over time between baseline and endline for the evaluation outcome indicators listed in Box 1 across the two groups while taking into account the baseline differences between them. Doing so allows us to gain some insight into whether the SHOUHARDO II project’s interventions themselves led to any changes in the outcomes. This difference-in-difference analysis is presented in Table 6. Statistically significant differences between the baseline and endline at the 5% or lower level are indicated with a star (*). Note first that, as would be expected given its higher economic status as a group, the non-eligible group started out at baseline with more favorable outcomes than the participant group. The only exceptions are for two indicators: safe disposal of feces and the percent of children 6-23 months with minimum meal frequency. By contrast, by the time of the endline survey, the eligible group was doing better than the non-eligible group for 16 of the 24 indicators despite starting out poorer than them. It is also important to note that while eligible households’ actual participation rate, at 94 percent, was quite high compared to non-eligible households, the participation rate for the latter was not negligible. Thirty-five percent of non-eligible households participated in the project. This means that we can expect to see some improvement for these groups associated with the project’s interventions if the interventions are having a positive impact overall. It is also possible and likely that these households experienced the positive benefits of the project through spillover effects (see Section 5.2). For almost every indicator, the absolute change over time was more favorable for eligible households than non-eligible households. That is, in the case of indicators for which an increase indicates better well-being, the increase was greater for eligible households. In the case of indicators for which a decrease indicates better well-being, the decrease was greater for eligible households. The indicators that improved the most for eligible versus non-eligible households are: 32  The percent of mothers who received Vitamin A within six weeks of delivery;  The percent of children 6-23 months who received Vitamin A in the last six months; and  The percent of children 6-23 months with minimum dietary diversity. There are two exceptions to these more favorable trends for eligible households: The percent of children with minimum meal frequency increased slightly more for non-eligible households, and the decline in the prevalence of diarrhea was greater for non-eligible households. Table 6: Difference-in-difference analysis: Changes in child undernutrition and its determinants from baseline to endline for eligible versus non-eligible households Eligible households Non-eligible households (Comparison group) Difference in Baseline Endline Change Baseline Endline Change difference Household food security Number of months of adequate food 6.3 11.1 4.8 * 8.6 11.4 2.8 * 2.0 Household dietary diversity 4.8 9.0 4.2 * 6.1 8.7 2.6 * 1.6 Household hunger score 2.00 0.33 -1.67 * 0.85 0.27 -0.58 * -1.1 Caring practices for mothers during pregnancy (%) Antenatal care during pregnancy 48.0 86.7 38.7 * 60.8 80.6 19.8 * 18.9 Antenatal care in a medical facility 33.6 64.1 30.5 * 49.0 65.9 16.9 * 13.6 More food during pregnancy 13.1 58.5 45.4 * 18.1 51.6 33.5 * 11.9 More rest during pregnancy 23.6 66.2 42.6 * 27.0 57.3 30.3 * 12.3 Vitamin A 6 weeks from delivery 34.6 83.4 48.8 * 41.1 64.8 23.7 * 25.1 Iron/folic acid during pregnancy 45.4 86.6 41.2 * 49.3 74.4 25.1 * 16.1 Caring practices for children Hand washing at five critical times (%) 9.5 31.9 22.4 * 10.3 22.2 11.9 * 10.5 Safe disposal of feces (0-35m) (%) 47.2 69.3 22.1 * 46.5 60.7 14.2 * 7.9 No. of vaccinations received (0-23m) 5.8 7.0 1.2 * 6.4 6.9 0.5 * 0.7 Vitamin A capsule last 6m (6-23m) (%) 58.1 85.5 27.4 * 62.4 64.8 2.4 25.0 Child receiving Monomix (6-23m) (%) 2.4 31.7 29.3 * 2.6 14.4 11.8 * 17.5 Household health environment Access to safe water (%) 58.0 76.1 18.1 * 62.9 68.3 5.4 12.7 Access to improved toilet facility (%) 20.8 52.9 32.1 * 40.5 59.6 19.1 * 13.0 Mother's and children's food consumption (6-23m) Mother's dietary diversity 4.6 8.4 3.8 * 5.7 7.85 2.2 * 1.7 Child: minimum dietary diversity (%) 13.9 59.9 46.0 * 21.9 44.2 22.3 * 23.7 Child: minimum meal frequency (%) 47.4 63.2 15.8 * 36.1 52.3 16.2 * -0.4 Child: minimum acceptable diet (%) 9.7 46.4 36.7 * 12.3 35.3 23.0 * 13.7 Children's health (%) Diarrhea in last two weeks (0-59m) 12.6 5.5 -7.1 * 17.4 6.1 -11.3 * 4.2 Mother's nutritional status Mother's Body Mass Index 19.5 20.4 0.9 * 20.0 21.4 1.4 * -0.5 Child stunting (%) Under fives 61.7 48.8 -12.9 * 52.0 44.8 -7.2 -5.7 Under twos 55.8 42.9 -12.9 * 46.5 43.1 -3.4 -9.5 Notes: Stars represent statistical significance of the difference at the 5% or lower level. 33 The prevalence of stunting, our main indicator of interest, for children under five of eligible households declined by 12.9 percentage points, while that for children under five of ineligible households declined by only 7.2 percentage points. This yields a difference-in-difference of -5.7 . That for under-twos is even greater, at -9.5. These difference-in-difference results, along with those associated with the outcome variables that are determinants of stunting, are evidence that the SHOUHARDO II project interventions caused reductions in stunting among project participants. Because of the lack of a true randomized control group for this comparison, it is not possible to estimate the actual amount of the stunting reduction that was brought about. However, given the high participation in some project interventions by ineligible households and spillover effects, we can safely say that the difference-in-difference estimates are lower bounds on the amount of the stunting reduction caused by the project. 7. Results: Instrumental Variables evidence of project impacts IV estimates of the impact of participation in SHOUHARDO II on height-for-age z-scores of children under five and under two are reported in Table 7. The instruments employed are: a dummy variable representing the planned treatment status of households and a dummy variable indicating whether or not the household is more than a one-hour walk to the nearest town, which was collected at the household level. Note that the instruments for all regressions reported in this section are listed in Appendix 3 and described in Appendix 2. The regressions in Table 7 satisfy the relevance condition (see Kleinbergen-Paap Walk F-statistic) and pass the overidentification test (chi-sq p-value>0.1), indicating they are valid for this analysis. The endogeneity test further indicates that participation is indeed endogenous (chi-sq p-value<0.1), and that 2SLS is thus the appropriate estimation technique. The 2SLS regression coefficient for the specification using HAZ of under-fives as the dependent variable is 0.49 z-scores; that for under-twos is 0.71, 26 percent higher.9 Both are statistically significant at the 5% level and provide further evidence that the project had a positive and substantial impact on HAZ for both age groups. Figure 6 illustrates the results and shows those for boys and girls. The girl-boy difference is particularly stark for under-twos and indicates that the project had a much greater impact on boys’ long-term nutritional status than girls, explaining why the decline in stunting prevalence over the project’s operational period was so much higher for boys (see Table 4). 9 For reference, the total increase in HAZ between the baseline and endline surveys was 0.41 z-scores for under-fives and 0.42 z-scores for under-twos. 34 Table 7: Instrumental variables estimates of the impact of participation in the SHOUHARDO II project on children's height-for-age z-scores Under fives (6-59 months) Under twos (6-23 months) Coefficient (2SLS) z￾statistic Coefficient (2SLS) z￾statistic Participation in SHOUHARDO II 0.488 2.37 ** 0.706 2.20 ** Child's age -0.055 -5.55 *** -0.133 -1.93 * Child's age-squared 0.001 4.22 *** 0.003 1.11 Girl child 0.094 1.31 0.305 2.24 ** Mother's age 0.015 2.44 ** -0.002 -0.18 Mother's education: None a/ Primary 0.090 1.18 0.074 0.57 Secondary 0.247 2.73 *** 0.347 2.51 ** Age of household head 0.001 0.27 0.003 0.56 Female household head -0.117 -0.71 -0.096 -0.38 Occupation of head: Farming a/ Agricultural laborer 0.087 0.91 -0.013 -0.08 Non-agricultural laborer 0.118 1.03 0.337 1.37 Salaried employment 0.096 0.68 0.005 0.02 Self employment 0.032 0.33 0.040 0.29 Unpaid household work 0.311 1.62 0.375 1.34 Other 0.092 0.84 0.159 0.87 Household size 0.002 0.09 0.018 0.56 Age-sex composition: % females 0-16 a/ Percent females 16-30 0.010 2.31 ** 0.006 0.66 Percent females 30+ 0.002 0.40 -0.007 -0.99 Percent males 0-16 0.005 2.25 ** 0.007 1.72 * Percent males 16-30 0.006 1.96 ** 0.005 0.78 Percent males 30+ 0.007 1.63 -0.003 -0.37 Well-being category: Extreme poor a/ Poor -0.057 -0.56 -0.212 -1.07 Lower middle 0.344 2.36 ** 0.323 1.33 Middle 0.309 1.96 * 0.127 0.46 Rich 0.425 2.37 ** 0.239 0.79 Region: Coast a/ Haor -0.441 -5.16 *** -0.505 -3.64 *** Mid Char 0.008 0.09 -0.065 -0.41 North Char 0.068 0.70 -0.066 -0.38 Number of observations 2,475 871 Weak instrument test Kleibergen-Paap rk Wald F-stat 110.3 43.5 Maximal IV relative bias b/ b/ Overidentification test (chi-sq p-value) 0.899 0.317 Endogeneity test (chi-sq p-value) 0.016 0.021 a/ Reference category. b/ Maximal IV relative bias statistics not reported by STATA because the estimation is not sufficiently overidentified, rendering the test not well defined (Shaeffer 2012). Notes: z-statistics are robust to heteroskedasticity and clustering by village. Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 35 Figure 6: Instrumental variables estimates of the impact of participation in the SHOUHARDO II project on children's height-for-age z-scores (HAZ) Turning to the determinants of stunting, Table 8 reports the regression results for the measures of household food security. They indicate that the SHOUHARDO II project’s interventions served to increase the number of months in which households had adequate food, to increase the diversity of households’ diets, an indicator of dietary quality, and to reduce household hunger. Note that the regressions for household-level variables employ the education of the household head as a dependent variable while those for child and mother –level variables employ mother’s education. Table 9 reports results for the remaining determinants of stunting. In this table the dependent variables are listed in the far-left column, and the coefficient estimates are only reported for the impact of participation in the project. The next column to the right gives the estimation technique employed, which depends on the endogeneity test statistic. The relevance, overidentification, and endogeneity test statistics are given in the four far-right columns. Among the underlying determinants, in addition to food security, the results indicate that the project’s interventions led to improvements in the quality of caring practices for mothers and children and in household health environments. With regard to caring practices for mothers during pregnancy, they led to increases in antenatal care, increased the likelihood that mother will receive more food and rest during pregnancy, and increased Vitamin A and iron/folic acid supplementation among pregnant mothers. With regard to caring practices for children, they increased the use of hygiene practices by mothers and vitamin supplementation for children. The estimates suggest that they did not, however, serve to increase the number of vaccinations received by children. Finally, the results indicate that the increases in access to safe water among project households (see Table 6) were brought about by the project’s interventions 0.488 0.562 0.484 0.706 0.562 0.934 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 All Girls Boys All Girls Boys Under-fives (6-59m) Under-twos (6-23m) H A Z 36 Table 8: Instrumental variables estimates of the impact of participation in the SHOUHARDO II project on household food security Months of adequate food provisioning Household dietary diversity score Hunger score Coefficient (2SLS) z￾statistic Coefficient (2SLS) z￾statistic Coefficient (2SLS) z￾statistic Participation in SHOUHARDO II 1.175 3.06 *** 9.365 5.29 *** -0.444 -2.06 ** Age of household head 0.004 1.68 * -0.007 -0.84 -0.003 -1.8 * Female household head -0.294 -1.63 -0.411 -1.04 0.331 1.88 * Education of household head: None a/ Primary 0.027 0.41 0.780 4.44 *** -0.018 -0.51 Secondary 0.214 2.96 *** 0.869 3.36 *** -0.050 -1.1 Occupation of head: Farming a/ Agricultural laborer -0.495 -5.54 *** -0.566 -2.42 ** 0.213 3.83 *** Non-agricultural laborer -0.429 -3.59 *** -0.277 -0.97 0.173 2.64 *** Salaried employment 0.126 1.30 0.126 0.35 -0.118 -2.7 *** Self employment 0.045 0.62 0.052 0.22 0.033 0.63 Unpaid household work 0.275 1.44 0.570 1.3 -0.261 -1.46 Other -0.240 -2.21 ** -0.610 -2.31 ** 0.130 2.62 *** Household size -0.041 -2.36 ** 0.054 1.1 0.010 1.05 Age-sex composition: % females 0-16 a/ Percent females 16-30 0.010 3.02 *** 0.032 3.55 *** -0.003 -1.25 Percent females 30+ 0.005 1.43 0.029 2.93 *** 0.000 -0.08 Percent males 0-16 0.001 0.78 0.005 1.08 0.000 -0.41 Percent males 16-30 0.012 3.70 *** 0.034 3.25 *** -0.003 -1.49 Percent males 30+ 0.010 2.59 *** 0.043 3.32 *** -0.004 -1.75 * Well-being category: Extreme poor a/ Poor 0.345 3.56 *** -0.561 -2.25 ** -0.115 -1.89 * Lower middle 0.795 5.21 *** 1.819 3.23 *** -0.247 -2.7 *** Middle 1.058 6.40 *** 2.828 4.25 *** -0.393 -3.86 *** Rich 1.338 6.80 *** 4.150 5.02 *** -0.463 -3.95 *** Region: Coast a/ Haor 0.116 1.07 -0.085 -0.24 0.009 0.16 Mid Char -0.053 -0.40 -0.900 -2.7 0.028 0.43 North Char -0.079 -0.57 -0.872 -2.15 ** 0.127 1.87 * Number of observations 2,844 2,844 2,844 Weak instrument test Kleibergen-Paap rk Wald F-stat 59.6 15.7 59.6 Maximal IV relative bias 5% 10% 5% Overidentification test (chi-sq p-value) 0.335 0.851 0.463 Endogeneity test (chi-sq p-value) 0.001 0.000 0.016 Notes: z-statistics are robust to heteroskedasticity and clustering by village. Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 37 Table 9: Instrumental variables/OLS estimates of the impact of participation in the SHOUHARDO II project on determinants of children’s nutritional status Estim￾ation method Coeff￾icient z￾statistic N Weak instrument test Overident￾ification test (chi-sq p￾value) Endogen￾eity test (chi-sq p￾value) Kleibergen￾Paap rk Wald F￾stat Maximal IV relative bias Household food security Number of months of adequate food 2SLS 1.18 3.06 *** 2,844 59.6 5% 0.335 0.001 Household dietary diversity 2SLS 9.37 5.29 *** 2,844 15.7 10% 0.851 0.000 Household hunger score 2SLS -0.44 -2.06 ** 2,844 59.6 5% 0.463 0.016 Caring practices for mothers during pregnancy Antenatal care during pregnancy 2SLS 4.960 4.66 *** 2,840 61.6 5% 0.311 0.000 Antenatal care in a medical facility OLS 0.092 3.37 *** 2,840 90.2 10% 0.754 0.710 More food during pregnancy 2SLS 1.010 3.55 *** 2,829 24.9 10% 0.325 0.001 More daytime rest during pregnancy 2SLS 0.608 4.33 *** 2,824 45.9 5% 0.190 0.000 Vitamin A within 6 weeks of delivery OLS 0.307 4.71 *** 2,730 110.7 a/ 0.592 0.015 Iron/folic acid during pregnancy 2SLS 0.417 3.68 *** 2,831 32.3 5% 0.161 0.013 Caring practices for children Hand washing at five critical times 2SLS 0.423 2.92 *** 2,844 59.8 5% 0.129 0.004 Safe disposal of feces (0-35m) 2SLS 0.375 2.29 ** 1,845 40.6 5% 0.220 0.078 No. of vaccinations received (0-23m) OLS -0.008 -0.07 918 16.7 5% 0.239 0.158 Vitamin A capsule last 6m (6-23m) OLS 0.156 3.47 *** 873 22.1 5% 0.700 0.612 Child receiving multivitamin (6-23m) 2SLS 0.537 3.30 *** 871 21.4 5% 0.512 0.002 Household health environment Access to safe water 2SLS 0.276 3.70 *** 2,844 92.4 5% 0.584 0.000 Access to an improved toilet facility OLS -0.019 -0.69 2,844 91.4 5% 0.204 0.147 Mother's and children's food consumption Mother's dietary diversity 2SLS 7.950 4.24 *** 2,734 14.3 10% 0.925 0.000 Minimum dietary diversity (6-23m) OLS 0.215 4.78 *** 845 20.0 5% 0.140 0.649 Minimum meal frequency (6-23m) 2SLS 0.438 2.17 ** 766 20.5 5% 0.747 0.065 Minimum acceptable diet (6-23m) OLS 0.203 4.35 *** 740 19.0 5% 0.338 0.317 Mother's nutritional status and food consumption Mother's Body Mass Index IV set 1 2SLS 1.870 1.75 * 2,522 39.0 5% 0.360 0.041 Mother's Body Mass Index IV set 2 OLS -0.253 -0.10 2,522 93.3 a/ 0.690 0.129 Children's health Diarrhea in last two weeks (0-59m) 2SLS -0.002 -0.12 2,834 73 5% 0.791 0.528 a/ Maximal IV relative bias test statistics not reported by STATA because the estimation is not sufficiently overidentified, rendering the test not well defined (Shaeffer 2012). Notes: z-statistics are robust to heteroskedasticity and clustering by village. Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 38 Turning to the immediate determinants of children’ nutritional status, the IV evidence indicates that all four indicators of mother’s and children’s food consumption were positively impacted by the SHOUHARDO II project, including mother’s dietary diversity and, for children, minimum dietary diversity, minimum meal frequency and minimum acceptable diet. The data give ambivalent results for mothers’ nutritional status, with one set of instruments indicating a positive impact on mother’s Body Mass Index and another indicating no impact. Lastly, the results indicate no impact of the project’s interventions on diarrhea among children under five. Overall these results suggest that the project had a positive impact on children’s nutritional status and that this was brought about by:  Increases in household food security;  Improvements in the quality of caring practices for mothers during pregnancy;  Increased use of hygiene practices by mothers;  Increased vitamin supplementation for children;  Improvements in access to safe water;  Improved food consumption for mothers and children; and  Possibly, improved nutritional status of mothers. 8. Results: Propensity Score Matching evidence on the impact of specific interventions In this section, the PSM estimates of the average treatment effect on the treated (ATT) of the four intervention sets of focus—MCHN, women’s empowerment, livelihoods promotion, and water and sanitation—are presented. As discussed in the methods section, underlying these estimates are probit￾regression predictions of each household’s propensity score for participating in the intervention of interest. The full participation regression results are presented in Appendix 4, but will be briefly summarized in each section here. Note that for each intervention the assessment is only undertaken for outcomes they would be expected to influence. 8.1 Mother and child health and nutrition interventions Recall that the participation variable for the MCHN interventions indicates whether the household participates in all four MCHN interventions: courtyard sessions, cooking/feeding sessions, child growth monitoring, and receipt of a food ration. Forty-five percent of households in project villages did so, leaving an ample potential pool of households for matching. It was not possible to undertake analysis for each of the four interventions individually because of their high participation rates, which meant that a large enough pool of households for matching was not available. 39 The probit propensity score model for full participation in MCHN interventions given in Appendix 4, Error! Reference source not found. reveals that the following factors influenced households’ and others’ participation in MCHN interventions:  Participation in the other three interventions  Whether the household received a food ration from another project  Household demographic characteristics: age and sex of the child, mother’s and household head’s age, education of household head, age-sex composition, region of residence  Whether the household resides in a PM2A village  Relative shock exposure in 2009  Relative bonding social capital in 2009  Number of SHOUHARDO II project staff known in 2009  Mother’s leisure time in 2009  Baseline district-level HAZ and weight-for-height z-scores. It is important to keep in mind that the PSM estimates are only able to account for observable, measured determinants of households’ participation in interventions as opposed to the IV estimates, which are designed to account for observable and unobservable determinants. As such, the selection bias caused by targeting of MCHN interventions to mothers in households with children that are undernourished is not corrected for in the estimates presented here. This bias is likely to lead to underestimation of the impacts of the MCHN interventions on HAZ and other variables closely related to it in the hierarchy of causality, in particular, mothers’ nutritional status. Baseline district-level anthropometric z-scores were included to help control for this selection bias. However, doing so is not likely to adequately control for household-level selection bias. Table 10 presents the PSM results. The far-right column reports on the key statistic that allows one to assess the degree of matching quality. Matching is of adequate quality for all of the dependent variables of interest (chi-squared p-value>0.1). The percent of sample households falling in the common support is also very high. As illustrated in Figure 7 for the example of the number of months of adequate food provisioning, the common support condition is strongly satisfied. This figure shows the propensity score distribution of participating versus non-participating households, and that there are ample non-participating households with propensity scores close by in the distribution with which to be matched (with the exception of a few households having very high propensity scores). Note that matching quality and common support statistics, although not reported, are of adequate quality for all PSM results presented in the rest of Section 8. The results point to a positive impact of MCHN participation on at least some aspect of all three underlying determinants of children’s nutritional status. They indicate that they served to increase household dietary diversity, to improve all six caring practices for mothers and all five caring practices for children, and to increase access to sanitary toilet facilities. Among the immediate determinants, the results indicate that the MCHN interventions increased mothers’ dietary diversity and the likelihood that a child has minimum dietary diversity. The PSM results indicate no impact of the MCHN interventions on children’s HAZ and in fact a negative impact on mothers’ body mass index. This is likely related to the negative selection bias discussed above. 40 Table 10: Propensity Score Matching estimates of the average treatment effect on the treated: Full participation in Mother and Child Health and Nutrition interventions Average treatment effect on the treated (ATT) z￾statistic Number of observations Percent of households on common support Chi-squared p-value for matching quality Partici￾pants Controls Household food security Number of months of adequate food 0.047 0.73 1,331 1,494 99.7 1.00 Household dietary diversity 0.526 4.39 *** 1,331 1,494 99.7 1.00 Household hunger score 0.015 0.46 1,331 1,494 99.7 1.00 Caring practices for mothers during pregnancy Antenatal care during pregnancy 0.083 5.78 *** 1,332 1,490 99.7 99.9 Antenatal care in a medical facility 0.059 2.92 *** 1,332 1,490 99.7 1.00 More food during pregnancy 0.085 3.58 *** 1,331 1,480 99.8 1.00 More daytime rest during pregnancy 0.049 2.10 ** 1,332 1,480 99.8 1.00 Vitamin A within 6 weeks of delivery 0.116 6.53 *** 1,301 1,412 99.7 99.9 Iron/folic acid during pregnancy 0.095 5.95 *** 1,332 1,481 99.8 1.00 Caring practices for children Hand washing at five critical times 0.105 5.09 *** 1,331 1,494 99.7 1.00 Safe disposal of feces (0-35m) 0.140 5.29 *** 967 866 99.8 0.99 No. of vaccinations received (0-23m) 0.471 3.07 *** 483 421 99.0 0.62 Vitamin A capsule last 6m (6-23m) 0.122 2.42 ** 475 382 98.4 0.94 Child receiving multivitamin (6-23m) 0.097 2.36 ** 470 383 98.4 0.96 Household health environment Access to safe water -0.023 -1.22 1,331 1,494 99.7 1.00 Access to an improved toilet facility 0.068 2.82 *** 1,331 1,494 99.7 1.00 Mother's and children's food consumption Mother's dietary diversity 0.576 5.20 *** 1,306 1,413 99.9 1.00 Minimum dietary diversity (6-23m) 0.095 1.76 * 451 374 97.7 0.96 Minimum meal frequency (6-23m) 0.005 0.11 415 322 96.2 0.98 Minimum acceptable diet (6-23m) 0.061 1.07 396 314 96.2 0.99 Children's health Diarrhea in last two weeks (0-59m) 0.012 1.52 1,331 1,491 99.9 0.98 Mother's nutritional status Mother's Body Mass Index -0.475 -2.45 *** 1,218 1,320 99.7 1.00 Children's height-for-age z-scores Under fives -0.134 -1.44 1,278 1,414 99.9 0.96 Under twos -0.033 -0.32 562 533 99.6 0.70 Notes: Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 41 Figure 7: Common support: Propensity scores of participant and non-participant households for full participation in MCHN interventions Note: The dependent variable used for this example is number of months of adequate household food provisioning. 0 .2 .4 .6 .8 1 Propensity Score Untreated Treated: On support Treated: Off support 42 8.2 Women’s empowerment interventions The probit propensity score model for participation in the empowerment interventions is given in Appendix 4, Table 17. The following factors influenced households’ and mothers’ participation in these interventions:  Participation in the other three interventions  Whether the household received a food ration from another project  Household demographic characteristics: child’s age, gender and occupation of household head, age-sex composition, region of residence  Household well-being category  Whether the village of residence is classified as extremely vulnerable  Total number of households in the village  Whether CARE is the implementing NGO in the village  Whether the household resides in a PM2A village  Relative shock exposure in 2009  Bonding social capital  Number of SHOUHARDO II project staff known in 2009. Table 11 presents the PSM results. Overall, they suggest that the empowerment interventions led to some important improvements in the areas of household food security, caring practices for mothers and children, mothers’ food consumption, and children’s health. Membership in an EKATA group is associated with increased antenatal care during pregnancy, increased likelihood that a mother will receive vitamin A within six weeks of delivery, and knows the five critical times for hand washing. Note that the small sample of mothers participating in EKATA limits our ability to detect statistically significant results for this intervention, especially for the outcomes applying to children under two. Membership in a savings group is positively associated with increased household and mother’s dietary diversity and with reduced household hunger. With respect to caring practices, it increases post￾delivery Vitamin A supplementation for mothers and the safe disposal of children’s feces. Together, the PSM results suggest that the women’s empowerment interventions increase household food security, increase women’s dietary diversity, lead to women consuming more food during their pregnancies and increase the likelihood that they will receive vitamin A supplementation. Women participating in the interventions are more likely to know about or practice hygienic behaviors, and perhaps this is why their children are less likely to have diarrhea. It is important to keep in mind that some health and nutrition behavior change messages were reinforced in the EKATA groups. Thus it is not clear that the impacts seen here are due to this factor or to improvement in women’s empowerment itself, an important subject for future research. Here, again, no positive influence on HAZ can be detected. 43 Table 11: Propensity Score Matching estimates of the average treatment effect on the treated (ATT): Mother’s participation in women’s empowerment interventions Mother is a member of an EKATA group Mother is a member of a savings group Mother is a member of an EKATA group or a savings group ATT z-statistic ATT z-statistic ATT z-statistic Household food security Number of months of adequate food 0.136 1.25 0.133 1.29 0.065 0.83 Household dietary diversity 0.373 1.50 0.514 2.92 *** 0.547 3.28 *** Household hunger score 0.001 0.02 -0.089 -1.89 * -0.080 -1.91 * Caring practices for mothers during pregnancy Antenatal care during pregnancy 0.046 1.76 * 0.006 0.3 0.025 1.40 Antenatal care in a medical facility -0.01 -0.35 0.014 0.44 0.024 0.90 More food during pregnancy 0.045 1.19 0.053 1.43 0.048 1.80 * More daytime rest during pregnancy 0.057 1.35 0.036 1.27 0.023 0.91 Vitamin A within 6 weeks of delivery 0.048 1.72 * 0.068 3.27 *** 0.064 2.97 *** Iron/folic acid during pregnancy 0.004 0.14 0.015 0.67 0.013 0.76 Caring practices for children Hand washing at five critical times 0.094 2.19 ** 0.008 0.28 0.034 1.62 Safe disposal of feces (0-35m) 0.026 0.67 0.077 2.08 ** 0.082 2.51 ** No. of vaccinations received (0-23m) 0.247 0.99 0.093 0.41 0.245 1.33 Vitamin A capsule last 6m (6-23m) 0.007 0.12 0.008 0.16 0.028 0.66 Child receiving multivitamin (6-23m) -0.05 -0.55 -0.019 -0.33 -0.014 -0.30 Mother's and children's food consumption Mother's dietary diversity 0.280 1.23 0.610 3.24 *** 0.550 4.58 *** Minimum dietary diversity (6-23m) 0.056 0.58 -0.028 -0.32 0.011 0.18 Minimum meal frequency (6-23m) 0.038 0.37 0.012 0.15 0.038 0.57 Minimum acceptable diet (6-23m) 0.105 0.96 0.046 0.68 0.103 1.57 Children's health Diarrhea in last two weeks (0-59m) -0.01 -0.93 -0.021 -1.63 -0.026 -2.80 *** Mother's nutritional status Mother's Body Mass Index -0.21 -0.07 -0.134 -0.53 -0.236 -0.98 Children's height-for-age z-score Under fives -0.06 -0.43 -0.109 -0.96 -0.065 -0.73 Under twos -0.21 -0.73 0.020 0.09 0.043 0.25 Notes: Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 44 8.3 Livelihoods promotion interventions The following factors influenced households’ and mothers’ participation in the four livelihoods promotion interventions, that is, all or one of: Crop production, CHD, Fisheries and IGA (see Appendix 4, Table 18):  Participation in the other three interventions  Whether the household received a food ration from another project  Household demographic characteristics: mother’s and household head’s age and education, occupation of household head, age-sex composition, region of residence  Household well-being category  Whether the household resides in a PM2A village  Current bonding social capital and relative bonding social capital in 2009  Number of SHOUHARDO II project staff known in 2009  Index of leisure time10  Women’s decision making score in 2009  Baseline district-level HAZ. The PSM results (Table 12) suggest that the livelihoods promotion interventions had an impact on household food security, mother’s and children’s food consumption, and mother’s nutritional status. All four of the interventions had a positive impact on either household dietary diversity or mother’s dietary diversity. Two of the interventions had a positive impact on both: CHD and Fisheries. Fisheries additionally served to reduce household hunger and increase the likelihood of a child having minimum dietary diversity. The analysis suggests that two of the interventions lead to improvements in mothers’ body mass index: CHD and Fisheries. 10 This index refers to the leisure time of the respondent for Part I of the questionnaire, which was typically either the household head or the spouse of the household head. 45 Table 12: Propensity Score Matching estimates of the average treatment effect on the treated (ATT): Participation in livelihood promotion interventions Crop production Comprehensive Homestead Development Fisheries Income Generating Activities Any agriculture/ income generation intervention ATT z￾statistic ATT z￾statistic ATT z￾statistic ATT z￾statistic ATT z￾statistic Household food security Months of adequate food 0.045 0.35 0.054 0.41 -0.09 -0.47 -0.09 -0.69 -0.012 -0.12 Household dietary diversity 0.646 2.30 ** 0.612 2.15 ** 0.566 1.37 0.64 2.74 *** 0.756 3.46 *** Household hunger score 0.015 0.23 0.036 0.50 -0.28 -1.81 * 0.064 0.85 0.014 0.18 Mother's and children's food consumption Mother's dietary diversity 0.354 1.3 0.448 2.19 ** 0.728 2.26 ** 0.253 1.04 0.464 2.27 ** Minimum dietary diversity (6-23m) 0.067 0.67 0.104 1.11 0.205 1.68 * 0.07 0.70 0.069 0.74 Minimum acceptable diet (6-23m) 0.036 0.28 0.055 0.45 -0.04 -0.26 -0.14 -1.14 -0.032 -0.29 Mother's nutritional status Mother's Body Mass Index -0.23 -0.64 0.590 1.99 ** 0.807 1.76 * 0.42 1.38 0.430 1.55 Children's height-for-age z-score Under fives -0.04 -0.23 -0.15 -1.19 0.040 0.22 -0.06 -0.43 -0.059 -0.41 Under twos -0.15 -0.46 -0.15 -0.61 0.300 0.68 -0.20 -0.85 -0.117 -0.49 Notes: Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 46 8.4 Water and sanitation interventions Participation in water and sanitation interventions was influenced by (see Appendix 4, Table 19):  Participation in the other three interventions  Whether the household received a food ration from another project  Household demographic characteristics: mother’s education, occupation of household head, age-sex composition, region of residence  Household well-being category  Village characteristics: total number of households, CARE is the implementing NGO, walking distance to nearest town is greater than 1 hour, and village assigned to the PM2A intervention arm  Relative shock exposure in 2009  Number of SHOUHARDO II project staff known in 2009  Index of leisure time11  Women’s decision making score in 2009  Baseline district-level HAZ and weight-for-height z-score (WHZ). Regarding the baseline district-level HAZ and WHA, interestingly, they are highly significant. The coefficient on HAZ is strongly negative and that on WHZ strongly positive. According to the PSM results, the only outcome that the water and sanitation interventions had an impact on was access to sanitary toilet facilities, on which it had a positive influence. Table 13: Propensity Score Matching estimates of the average treatment effect on the treated: Participation in water and sanitation interventions Average treatment effect on the treated (ATT) z-statistic Number of observations Percent of households on common support Chi￾squared p-value for matching quality Partici￾pants Controls Caring practices for children Hand washing at five critical times 0.031 1.07 564 2,269 100 0.96 Safe disposal of feces (0-35m) 0.005 0.13 374 1,459 99.8 0.99 Household health environment Access to safe water 0.005 0.20 564 2,269 99.0 0.96 Access to an improved toilet facility 0.066 2.15 ** 564 2,269 99.0 0.96 Children's health Diarrhea in last two weeks (0-59m) 0.004 0.34 562 2,259 99.9 1.00 Children's nutritional status Under fives -0.009 -0.01 528 2,161 99.7 1.00 Under twos -0.041 -0.23 207 879 98.8 1.00 Notes: Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. 11 This index refers to the leisure time of the respondent for Part I of the questionnaire, typically either the household head or the spouse of the household head. 47 9. Summary and conclusions Overall, the evidence presented in this report indicates that the SHOUHARDO II project was very successful in reducing stunting among children under five. While it is not possible to pinpoint the exact amount of stunting reduction caused with accuracy, it seems likely that a large portion, if not all, of the 12.9 percentage–point reduction in the prevalence of stunting observed between the baseline and endline surveys can be attributable to the project. Combined, the following findings support this conclusion:  The average annual decline in the stunting prevalence among eligible project households was 3.2 percentage points while the trend in rural Bangladeshi households in recent years has been a lower 0.6 percentage points per year. This comparative evidence rules out the possibility that the decline among project children was due to positive forces emanating from wider favorable economic, climatic, or policy-related trends in the country.  The normal large increase in stunting prevalence seen for children as they age from the 6-18 to the 48-60 month age group was not found for the group of children whose households participated in SHOUHARDO II interventions. Something happened that prevented many children from becoming stunted as they aged.  A difference-in-difference (DID) analysis comparing the changes over time for eligible project households compared to non-eligible project households indicates that the stunting prevalence fell more for eligible households. The difference is particularly strong for children under two.  Instrumental variables (IV) estimates of the impact of participation in the project confirm that it had a substantial, positive impact on children’s height-age-z-scores, particularly for children under two and for boys.  Propensity score matching (PSM) analysis indicates no impact of the project on child stunting. This can be attributed to the inability to control for a known, yet unobservable, factor affecting participation in the project’s MCHN activities: the purposeful targeting of children who were already undernourished.  The DID, IV and PSM analyses all indicate that the project’s interventions led to improvements in a broad array of determinants of stunting, improvements which are necessary for reducing stunting. The findings regarding project impacts on the determinants of stunting give insight into how the stunting reductions were brought about. Table 14 summarizes these findings from the various analyses. The left-hand panel focusses on the DID and IV analyses as well as the “single-difference” results, that is, the change from baseline and endline for eligible project households (reported in Table 6). The right￾hand panel focuses on the PSM results for individual project interventions. Positive impacts revealed by a particular analysis are indicated by purple shading. Negative impacts are indicated by red shading. Table 14’s summary reveals that the stunting reductions were brought about by improvements in household food security, in the quality of caring practices for mothers during pregnancy, in the quality of caring practices for children, in household health environments, in mother’s and children’s food consumption and, most likely, in mother’s nutritional status. With respect to household food security, DID, IV and PSM evidence all indicate that the project’s interventions increased household dietary diversity, an indicator of dietary quality, and reduced 48 Table 14: Summary of findings from single difference, difference-in-difference, instrumental variables, and propensity score matching analyses Evidence of impact of participation in the SHOUHARDO II project Evidence of the impact of participation in project interventions (PSM) Single difference Single difference Difference -in￾difference IV/OLS MCHN Women's empower -ment Livelihoods promotion Water and sanitation Household food security Number of months of adequate food Household dietary diversity Household hunger score Caring practices for mothers during pregnancy Antenatal care during pregnancy Antenatal care in a medical facility More food during pregnancy More daytime rest during pregnancy Vitamin A within 6 weeks of delivery Iron/folic acid during pregnancy Caring practices for children Hand washing at five critical times Safe disposal of feces (0-35m) No. of vaccinations received (0-23m) Vitamin A capsule last 6m (6-23m) Child receiving multivitamin (6-23m) Household health environment Access to safe water Access to an improved toilet facility Mother's and children's food consumption Mother's dietary diversity Minimum dietary diversity (6-23m) Minimum meal frequency (6-23m) Minimum acceptable diet (6-23m) Children's health Diarrhea in last two weeks (0-59m) Mother's nutritional status and food consumption Mother's Body Mass Index IV set 1 Mother's Body Mass Index IV set 2 Child height-for-age z-score Under fives Under twos Note: Purple shading indicates evidence of a positive impact for any of the interventions in an intervention set. Red shading indicates evidence of a negative impact. Single difference and difference-in-difference results are presented in Table 6. IV/OLS estimates are presented in Tables 7 through 9. PSM estimates are presented in Tables 10 through 13. 49 household hunger. The DID and IV analyses, but not PSM, indicate that they also led to increases in the amount of food households have access to, as measured by reported number of months of adequate food provisioning. With respect to caring practices for mothers during pregnancy, all methods point to project impacts on antenatal care, including whether that care is received in a medical facility. They suggest that the project led to women consuming more food and getting more day-time rest during their pregnancies. Finally, because of the project more women are receiving vitamin A supplementation within six weeks of their delivery and iron/folic acid supplementation during their pregnancy. The DID, IV and PSM results indicate that project interventions led to greater knowledge among mothers of the appropriate times for hand washing and an increase in the practice of safely disposing of children’s feces. They also increased vitamin supplementation for children, including Vitamin A and multivitamin supplementation. While the DID and PSM results suggest that they increased child immunization, the IV evidence does not support this conclusion. Different methods imply different project impacts in the area of household health environments. The DID and IV analyses suggest that the project brought about increased access to safe water, but not to sanitary toilet facilities. The DID and PSM analyses suggest that it brought about increased access to sanitary toilet facilities, but not to safe water. In all, while these results are incongruous, we can safely say that at least some of the improvement in household health environments seen between the baseline and endline surveys among eligible households (an increase in prevalences of 18.1 and 32.1 percentage points, respectively, for safe water and sanitary toilet facilities) can be attributed to the project. With regard to food consumption, dietary diversity was enhanced not only for households as a whole but specifically for mothers and children as well according to all three types of analysis. The DID and IV analyses confirm that the large increase in the percent of children 6-23 months who have a minimum acceptable diet, from 10 to 46 percent, was at least partially caused by the project’s interventions. The evidence is ambiguous for diarrhea incidence among children under five. The DID analysis showed that the reduction in diarrhea prevalence seen for eligible project households from baseline to endline (from 12.6 to 5.5 percent) was less than for non-eligible households. While the PSM results indicate a favorable impact, the IV analysis indicates none. Mother’s BMI changed little over the project’s implementation period. Yet the IV results give some evidence of a positive impact, and the PSM results that one of the project interventions (livelihoods promotion) had a positive impact. The negative PSM results seen for MCHN are likely due to a combination of the close relationship between mother’s and children’s nutritional status (see Section 3.1) and, again, the negative selection bias associated with the targeting of undernourished children. While the evidence presented in this report is not strong, it appears that the project likely led to some improvement in mothers’ nutritional status. The PSM results give insight into the question of which interventions brought about the reductions in stunting and improvements in its determinants. None of the interventions were found to improve children’s nutritional status. Again, this is likely due to the inability of the PSM method to control for the (unobservable) targeting of undernourished children. 50 The PSM results suggest that the MCHN interventions had a broad influence, improving household, mother’s and children’s dietary diversity; all of the caring practices for mothers during pregnancy; all of the caring practices for children; and access to sanitary toilet facilities. The women’s empowerment interventions also facilitated improvements in a variety of determinants of stunting, including household and mother’s dietary diversity, antenatal care during pregnancy, taking more food during pregnancy, post-delivery Vitamin A supplementation of mothers, and indicators of the knowledge and use of hygiene practices. The PSM results additionally suggest that they served to reduce hunger and prevent of diarrhea among children. The livelihoods promotion activities increased household, mother and children’s dietary diversity and reduced household hunger. According to the PSM results, that they also improved mothers’ nutritional status. Finally, the project’s water and sanitation interventions are found to have increased access to sanitary toilet facilities. In conclusion, this report finds that the SHOUHARDO II project was successful in reducing child stunting. Two factors that contributed to its success were: 1) it addressed a broad range of underlying and immediate causes of chronic undernutrition; and 2) the integrated approach that brought to bear not only nutrition-specific MCHN interventions to address the problem, but also interventions designed to empower women, to promote households’ livelihoods, and to improve households’ health environments. Suggested areas for future research are whether the project’s women’s empowerment interventions actually empowered women and whether the SHOUHARDO II project was able to shield households and children from the impacts of climate shocks. 51 Appendix 1. Determinants of stunting employed as dependent variables: Variable descriptions Variable Description Household food security Number of months of adequate food The number of months of months in the previous year that the household had adequate food, as reported by the member responsible for preparing food (Bilinsky and Swindale 2010). Household dietary diversity Number of food groups, out of 15, from which food was consumed by any household member in the previous 24 hours. Household hunger score A score assigned based on the frequency of occurrence of three situations in the last four weeks (no food to eat of any kind in household, going to sleep at night hungry, going a whole day and night without eating), with frequency scores of no=0, rarely or sometimes=1, and often=2. The scale ranges from 0 to 6 (Ballard et al. 2011). Caring practices for mothers during pregnancy Antenatal care during pregnancy Mother received antenatal care during current or recent pregnancy. Antenatal care in a medical facility Mother received antenatal care during current or recent pregnancy in a medical facility (government hospital, other government health facility, private hospital/clinic, or community clinic). More food during pregnancy Mother indicated she took more food than she usually takes during current or recent pregnancy. More rest during pregnancy Mother indicated she took more daytime rest than she usually takes during current or recent pregnancy. Vitamin A 6 weeks from delivery Mother received Vitamin A within one and a half months of delivery of child. Iron/folic acid during pregnancy Mother took iron and folic acid during the last pregnancy. Caring practices for children Hand washing at five critical times Mother indicated it is important to wash hands at all five of the following critical times: before eating, before breastfeeding/feeding child, before cooking/preparing food, after defecation/urination, after cleaning child that defecated/changing child diaper. Safe disposal of feces (0-35m) Mother of child 0-35 months indicated that the last time her child defecated it was in toilet or the feces was disposed of in toilet. Number of vaccinations received (0-23m) Total number of vaccinations received, out of a total of eight (BCG, Polio 1, 2 and 3, DPT/Penta 1, 2 and 3, and measles) by child 0-23 months. Vitamin A capsule last 6m (6-23m) Mother of child 6-23 months indicated they gave child a Vitamin A capsule in the last six months. Child receiving Monomix (6-23m) Mother of child 6-23 months indicated they are giving child Monimix or other sprinkles packets in food. Household health environment Access to safe water Household has access to water from one of the following sources: hand tube well, tara pump, deep tube well, shallow tube well, ring well/indara, piped water, pond sand filter, or rainwater harvesting system. The water must be normally available from the source and it must not have been unavailable for a day or longer in the two weeks preceding the survey. Access to an improved toilet facility Household has access to a ring-slab/offset latrine with a water seal, a pit latrine that is covered, a septic latrine, or a local adopted hygienic latrine. 52 Mother's and children's food consumption (6-23m) Mother's dietary diversity Number of food groups, out of 15, from which food was consumed by mother in the previous 24 hours. Child: minimum dietary diversity Child consumed food from at least four out of seven food groups in the previous 24 hours. The seven food groups are: grains; roots and tubers; legumes, nuts and pulses; milk and dairy products; eggs; flesh foods (meat, fish, poultry and liver/organ meats) (WHO 2008). Child: minimum meal frequency Breast fed children 6-23 months who either consumed two meals a day (6-8 month olds) or three meals a day (9-23 month olds) (WHO 2008). Child: minimum acceptable diet Breastfed children with both minimum dietary diversity and minimum mea frequency (WHO 2008). Children's health Diarrhea in last two weeks (0-59m) Child 0-59 months had diarrhea (3 or more loose stools in 24 hours) in the last two weeks. Mother's nutritional status Mother's Body Mass Index Mother’s weight divided by height-squared. 53 Appendix 2. Household, village and district-level characteristics used for Propensity Score Matching analysis: Variable descriptions Basic child, mother and household demographic characteristics included in all PSM analyses are the same as those used in the Instrumental Variables analysis and are described in Box 2. This appendix describes additional characteristics used for PSM matching. Variable Description Household characteristics Current shock exposure Total number of shocks experience in the past 12 months from among four types of climate shocks, five types of family event shocks, and four types of economic shocks. Relative shock exposure in 2009 For each of the three types of shock, endline survey respondents were asked to answer the question “Do you feel that the situation was better, the same or worse five years ago (before SHOUHARDO II began) than it is now with regards to these kinds of shocks?”. Answers to the three questions were then combined into an additive index ranging from 3 to 9, with higher numbers indicating an increasingly better shock exposure situation in 2009 than in the year prior to the endline survey. Exposure to alternatives An index calculated using Principal Components Analysis (PCA) based on answers to the following questions: (1) Does anyone in your household communicate regularly with at least one person outside this village?; (2) During the past week, has anyone in your household engaged in economic activities with members of other village? For example, farming, trading, employment, borrowing or lending money?; (3) How many times in the past month has anyone in your household got together with people to have food, either in their home or in a public place?; (4) How many days in the past month has anyone in your household attended a mosque or other religious service?; (5) In the last year, how many times has anyone in your household stayed more than two days outside this village? The first principal component was used for calculating the index. Absence of fatalism An index calculated using PCA based on a set of three dummy variables equal to 1 if the respondent indicated they agree with these statements: (1) When I get what I want, it is usually because I worked hard for it; (2) Some things that happen to me are God’s will and some things are because of my own actions; (3) To be successful, above all one needs to be lucky. The second principal component (for which the variables correlated with the expected sign) was used for the index. Number of SHOUHARDO II project staff known in 2009 The number of SHOUHARDO II staff members known by either the respondent or another household member before the project started. Index of leisure time in 2009 After reporting on the hours spent in the “typical day in the last month” in a variety of non-leisure activities, respondents were asked “During the daytime do you spend time doing other things like socializing, watching TV, taking naps or reading?” and, if yes, to specify “On a typical day in the last month, how many hours did you spend in these kinds of “leisure” activities where you were not working? With this information as context, they were then asked to “Imagine ten steps, where on the bottom, the first step is a person who spends no time in the day doing these leisure activities, and on the highest step, the 10th, is a person who spends the whole day doing them. On 54 Variable Description a typical day in the last month, which step were you on?”. The respondent was shown a picture of a ladder with ten steps to point to. Finally, they were asked “Which step were you on five years ago (before SHOUHARDO II began?”. The answer to this question, ranging from 1 to 10, was used as an index of leisure time in 2009. Index of leisure time of child’s mother in 2009 Calculated as above, but for the index child’s mother. Decision making score of a women in the household in 2009 An adult women in the household was asked questions (without men present) about her ability to take part in 12 types of decisions, ranging from buying small food items, to moving to a shelter during a time of disaster. The possible responses were: Can decide alone, can decide with husband or other adult male family member, husband makes decision after discussion with wife, not involved in decision. The responses were used to create an index based on the mean across the various types of decisions, with only those included that the woman felt was applicable to her situation. Scores were calculated only for women reporting that at least five types of decisions were applicable to her situation. Decision making score of child’s mother in 2009 Calculated as above, but for the index child’s mother. Participation in other interventions For each type of the four types of intervention, participation in other interventions is measured using three dummy variables equal to one if the household participated in the intervention and zero otherwise. Receipt of a food ration from another project A dummy variable equal to one if the household received a food ration from a project other than the SHOUHARDO II project. Village characteristics Classified as extremely vulnerable at baseline Dummy variable equal to one if the village was classified as extremely vulnerable by project administrators at baseline. Total number of households Total number of households in the village. Whether CARE is the implementing NGO Dummy variable equal to one if CARE is the implementing agency for the village (as opposed to another of the 16 Bangladeshi implementing NGOs). Whether the nearest town is greater than one walking hour away Average across households in the village of a dummy variable equal to one if household reports that the nearest town is greater than one hour away by walking. PM2A village Dummy variable equal to one if the village was randomly assigned to the PM2A intervention arm. District mean child nutritional status at baseline District mean HAZ at baseline Mean HAZ across households in each district calculated using the baseline survey data. District mean WHZ at baseline Mean WHZ across households in each district calculated using the baseline survey data. 55 Appendix 3. Instruments employed for Instrumental Variables tests and 2SLS regressions Table 15: Instrumental variables used for endogeneity testing and 2SLS estimates of the impact of participation in the SHOUHARDO II project Planned treatment status of household Household located in a PM2A village Number of project staff knew in 2009 More than a 1 hour walk to nearest town More than a 2 hour walk to Upazila head￾quarters Friend/ relative of Upazila elected leader Absence of fatalism index Shock exposure & relative shock exposure in 2009 Household food security Number of months of adequate food, household hunger score x x x Household dietary diversity x x x Caring practices for mothers during pregnancy Antenatal care during pregnancy x x X Antenatal care in a medical facility x x More food during pregnancy x x More daytime rest during pregnancy x x x x Vitamin A within 6 weeks of delivery x x Iron/folic acid during pregnancy x x x Caring practices for children Hand washing at five critical times, safe disposal of feces x x x No. of vaccinations received, Vitamin A in last 6 m, Multivitamin x x x Household health environment Access to safe water x x x Access to an improved toilet facility x x x Mother's and children's food consumption Mother's dietary diversity x x x Minimum dietary diversity, acceptable diet x x x Minimum meal frequency (6-23m) x x x Children's health Diarrhea in last two weeks (0-59m) x x x Mother's nutritional status and food consumption Mother's Body Mass Index IV set 1 x x x x Mother's Body Mass Index IV set 2 x x Child height-for-age z-score x x 56 Appendix 4. Probit propensity score models for participation in SHOUHARDO II project intervention Table 16 Probit propensity score model estimation for full participation in SHOUHARDO II MCHN interventions Models with child￾level outcomes (6-59 months) (1) Models with child￾level outcomes (6-23 months) (2) Models with household-level outcomes (3) Coefficient z￾statistic Coefficient z￾statistic Coefficient z￾statistic Participation in other interventions Women's empowerment 0.429 5.50 *** 0.426 2.79 *** 0.429 5.49 *** Livelihoods promotion 0.964 11.80 *** 1.212 8.31 *** 0.963 11.79 *** Water and sanitation 0.249 3.41 *** 0.458 3.31 *** 0.250 3.4 *** Food ration from other project 0.142 1.98 ** 0.245 1.75 * 0.142 2.0 * Child characteristics Child's age 0.047 7.14 *** 0.153 2.44 ** Child's age-squared -0.001 -8.77 *** -0.005 -2.14 ** Girl child 0.092 1.34 0.224 1.78 * Mother characteristics Mother's age -0.010 -2.00 ** -0.010 -0.90 Mother's education: None a/ Primary 0.029 0.42 0.178 1.36 Secondary 0.071 0.85 0.148 0.96 Household characteristics Age of household head -0.006 -2.13 ** -0.007 -1.54 -0.005 -1.75 * Female household head -0.112 -0.65 0.200 0.66 -0.027 -0.16 Education of household head: None a/ Primary 0.214 3.42 *** Secondary 0.135 1.71 * Occupation of head: Farming a/ Agricultural laborer -0.013 -0.16 -0.177 -1.07 -0.008 -0.10 Non-agricultural laborer -0.057 -0.56 -0.103 -0.51 -0.023 -0.23 Salaried employment -0.076 -0.59 -0.120 -0.51 -0.092 -0.72 Self employment 0.035 0.41 -0.020 -0.13 0.056 0.68 Unpaid household work -0.117 -0.63 -0.277 -0.84 -0.102 -0.57 Other 0.048 0.48 0.052 0.29 0.064 0.68 Household size 0.019 1.13 0.010 0.36 0.023 1.43 Age-sex composition: % females 0-16 a/ Percent females 16-30 -0.012 -3.21 *** -0.022 -2.74 *** -0.007 -2.25 ** Percent females 30+ -0.011 -2.96 *** -0.009 -1.29 -0.012 -3.26 *** Percent males 0-16 0.002 0.76 0.006 1.50 0.000 0.25 Percent males 16-30 -0.006 -1.69 * 0.010 1.65 * -0.003 -0.92 Percent males 30+ -0.006 -1.54 0.008 1.06 -0.006 -1.45 Well-being category: Extreme poor a/ Poor -0.082 -0.96 0.010 0.06 -0.039 -0.48 57 Lower middle -0.129 -1.01 0.128 0.54 -0.113 -0.93 Middle -0.208 -1.51 -0.196 -0.79 -0.151 -1.15 Rich -0.081 -0.49 -0.248 -0.80 -0.141 -0.89 Region: Coast a/ Haor 0.106 1.13 -0.036 -0.21 0.179 2.02 ** Mid Char 0.524 5.35 *** 0.530 2.83 *** 0.547 5.88 *** North Char 0.585 6.28 *** 0.679 3.84 *** 0.535 6.09 *** Village characteristics Classified as extremely vulnerable 0.054 0.89 0.070 0.61 0.084 1.45 Total number of households 0.000 -1.19 0.000 -0.37 0.000 -1.35 CARE is implementing NGO -0.058 -0.46 -0.065 -0.25 -0.034 -0.28 Nearest town > 1 hour away -0.001 -0.49 -0.001 -0.28 0.000 -0.50 PM2A village 0.394 7.08 *** 0.513 4.95 *** 0.360 6.78 *** Other potential participation determinants Current shock exposure -0.025 -1.14 -0.001 -0.02 -0.032 -1.53 Relative shock exposure in 2009 -0.048 -3.35 *** -0.082 -2.96 *** -0.047 -3.49 *** Bonding social capital (index) -0.019 -0.67 -0.054 -0.99 -0.023 -0.85 Relative bonding social capital in 2009 0.136 3.82 *** 0.188 2.79 *** 0.127 3.72 *** Exposure to alternatives (index) -0.063 -0.27 -0.202 -0.46 -0.015 -0.07 Absence of fatalism (index) 0.017 0.56 0.003 0.05 0.030 1.08 Number of project staff known (2009) 0.032 2.29 ** 0.002 0.08 0.026 1.97 ** Mother's leisure time index (2009) 0.014 1.05 -0.044 -1.72 * 0.012 0.93 Decision making score of mother (2009) 0.004 0.28 -0.002 -0.06 0.002 0.16 District mean child nutritional status (2009) Height-for-age z-score -0.356 -1.96 * -0.590 -1.48 -0.242 -1.40 Weight-for-height z-score -0.290 -1.67 * -0.445 -1.35 -0.234 -1.41 Number of observations 2,696 871 2,834 Pseudo R-squared 0.207 0.276 0.172 Notes: Stare represent statistical significance at the 1(*), 5%(**) and 1%(***) levels. Dependent variables employed for example models: (1) height-for-age z-score; (2) dummy variable for whether child received a vitamin A capsule in the last 6 months; (3) months of adequate household food provisioning. 58 Table 17 Probit propensity score model estimation for full participation in SHOUHARDO II women’s empowerment interventions Models with child-level outcomes (6-59 months) (1) Models with child-level outcomes (6-23 months) (2) Models with household-level outcomes (3) Coefficient z-statistic Coefficient z-statistic Coefficient z-statistic Participation in other interventions MCHN 0.905 6.72 *** 1.067 4.01 *** 0.903 6.92 *** Livelihoods promotion 0.467 4.30 *** 0.409 2.14 ** 0.454 4.29 *** Water and sanitation 0.139 1.69 * 0.105 0.68 0.152 1.91 * Food ration from other project -0.156 -1.87 * -0.298 -1.82 * -0.140 -1.72 * Child characteristics Child's age 0.014 1.73 * -0.070 -0.95 Child's age-squared 0.000 -1.82 * 0.002 0.78 Girl child -0.067 -0.83 -0.061 -0.41 Mother characteristics Mother's age -0.002 -0.25 0.014 1.01 Mother's education: None a/ Primary 0.109 1.37 0.208 1.39 Secondary 0.050 0.50 0.173 0.95 Household characteristics Age of household head 0.001 0.38 -0.004 -0.56 0.002 0.48 Female household head 0.297 1.45 0.904 2.44 ** 0.319 1.64 Education of household head: None a/ Primary 0.071 0.95 Secondary -0.032 -0.32 Occupation of head: Farming a/ Agricultural laborer 0.160 1.60 0.485 2.60 *** 0.126 1.30 Non-agricultural laborer 0.081 0.68 0.255 1.11 0.073 0.63 Salaried employment 0.024 0.15 -0.570 -1.68 * 0.090 0.57 Self employment 0.032 0.31 0.100 0.53 0.026 0.26 Unpaid household work -0.206 -0.92 -0.615 -1.51 -0.202 -0.95 Other -0.129 -1.03 -0.240 -0.98 -0.130 -1.06 Household size -0.003 -0.12 0.005 0.13 0.000 -0.02 Age-sex composition: % females 0-16 a/ Percent females 16-30 -0.011 -2.18 ** -0.012 -1.21 -0.009 -2.19 ** Percent females 30+ -0.010 -2.10 ** -0.019 -2.04 ** -0.010 -2.21 ** Percent males 0-16 -0.001 -0.40 0.002 0.54 0.000 0.11 Percent males 16-30 0.002 0.48 0.004 0.55 0.003 0.74 Percent males 30+ 0.008 1.51 0.023 2.43 ** 0.009 1.84 * Well-being category: Extreme poor a/ Poor -0.091 -0.92 0.076 0.38 -0.047 -0.49 Lower middle -0.298 -1.75 * 0.040 0.13 -0.271 -1.63 Middle -0.094 -0.53 0.108 0.34 -0.054 -0.31 Rich -0.616 -2.24 ** -0.370 -0.69 -0.493 -1.92 * 59 Region: Coast a/ Haor -0.221 -1.96 * -0.531 -2.33 ** -0.156 -1.43 Mid Char -0.088 -0.74 -0.132 -0.60 -0.075 -0.65 North Char -0.014 -0.13 -0.042 -0.21 0.002 0.02 Village characteristics Classified as extremely vulnerable -0.169 -2.34 ** -0.043 -0.31 -0.173 -2.45 ** Total number of households 0.000 -3.75 *** 0.000 -1.39 0.000 -3.55 *** CARE is implementing agency 0.286 2.00 ** 0.908 3.28 *** 0.313 2.25 ** Nearest town > 1 hour away -0.001 -0.66 -0.002 -0.94 -0.001 -0.85 PM2A village 0.109 1.64 0.201 1.60 0.120 1.83 * Other potential participation determinants Current shock exposure 0.036 1.40 0.019 0.37 0.037 1.46 Relative shock exposure in 2009 0.054 3.26 *** 0.099 3.13 *** 0.055 3.47 *** Bonding social capital (index) 0.090 2.56 ** 0.109 1.62 0.081 2.37 ** Relative bonding social capital in 2009 -0.035 -0.83 -0.011 -0.13 -0.022 -0.54 Exposure to alternatives (index) 0.207 0.73 0.585 1.13 0.188 0.68 Absence of fatalism (index) -0.004 -0.11 -0.084 -1.29 -0.011 -0.31 Number of project staff known (2009) 0.027 1.73 * -0.011 -0.36 0.027 1.76 * Mother's leisure time index (2009) 0.009 0.58 0.040 1.29 0.004 0.26 Decision making score of mother (2009) 0.002 0.12 -0.024 -0.75 0.003 0.21 District mean child nutritional status (2009) Height-for-age z-score 0.081 0.36 -0.040 -0.08 0.172 0.77 Weight-for-height z-score -0.112 -0.53 0.034 0.08 -0.014 -0.07 Number of observations 2,696 871 2,834 Pseudo R-squared 0.124 0.186 0.120 Notes: Stare represent statistical significance at the 105**), 5%(**) and 1%(***) levels. Dependent variables employed for example models: (1) height-for-age z-score; (2) dummy variable for whether child received a vitamin A capsule in the last 6 months; (3) months of adequate household food provisioning. 60 Table 18: Probit propensity score model estimation for participation in livelihoods promotion interventions Models with child￾level outcomes (6-59 months) (1) Models with child￾level outcomes (6-23 months) (2) Models with household-level outcomes (3) Coefficient z￾statistic Coefficient z￾statistic Coefficient z￾statistic Participation in other interventions MCHN 1.140 12.62 *** 1.420 8.03 *** 1.143 13.21 *** Women's empowerment 0.452 4.14 *** 0.435 2.19 ** 0.450 4.23 *** Water and sanitation 0.519 5.07 *** 0.301 1.65 * 0.481 4.92 *** Food ration from other project 0.237 2.44 ** 0.142 0.75 0.228 2.44 ** Child characteristics Child's age -0.002 -0.27 -0.088 -1.07 Child's age-squared 0.000 0.71 0.003 1.23 Girl child -0.050 -0.55 -0.121 -0.70 Mother characteristics Mother's age 0.012 1.81 * 0.011 0.79 Mother's education: None a/ Primary -0.029 -0.33 -0.163 -0.98 Secondary -0.192 -1.83 * -0.238 -1.21 Household characteristics Age of household head 0.017 4.48 *** 0.017 2.57 *** 0.013 3.68 *** Female household head 0.189 0.90 0.161 0.39 0.163 0.80 Education of household head: None a/ Primary -0.242 -3.09 *** Secondary -0.306 -3.01 *** Occupation of head: Farming a/ Agricultural laborer 0.215 1.93 * 0.457 2.12 ** 0.203 1.90 * Non-agricultural laborer 0.053 0.42 0.160 0.63 0.080 0.64 Salaried employment -0.010 -0.06 -0.106 -0.35 0.064 0.38 Self employment 0.024 0.22 0.231 1.12 0.046 0.43 Unpaid household work 0.056 0.24 0.026 0.06 -0.010 -0.04 Other 0.126 0.97 -0.060 -0.25 0.156 1.24 Household size 0.017 0.77 0.002 0.07 0.009 0.44 Age-sex composition: % females 0-16 a/ Percent females 16-30 -0.008 -1.65 * -0.014 -1.32 -0.013 -2.99 *** Percent females 30+ -0.003 -0.56 -0.009 -0.89 -0.002 -0.41 Percent males 0-16 -0.003 -1.03 -0.008 -1.54 -0.001 -0.70 Percent males 16-30 -0.012 -2.83 *** -0.010 -1.24 -0.013 -3.27 *** Percent males 30+ -0.002 -0.48 0.007 0.69 -0.003 -0.53 Well-being category: Extreme poor a/ Poor 0.172 1.66 * 0.369 1.91 * 0.200 2.00 ** Lower middle -2.236 -15.18 *** -2.386 -8.28 *** -2.147 -15.17 *** Middle -2.278 -13.67 *** -1.991 -6.69 *** -2.119 -13.41 *** Rich -2.975 -10.38 *** -3.138 -5.35 *** -2.765 -10.56 *** Region: Coast a/ Haor 0.430 3.38 *** 0.375 1.50 0.320 2.64 *** Mid Char -0.162 -1.32 -0.426 -1.75 * -0.179 -1.51 North Char -0.077 -0.66 -0.196 -0.86 -0.111 -1.00 61 Table 18: Probit propensity score model estimation for participation in livelihoods promotion interventions Models with child￾level outcomes (6-59 months) (1) Models with child￾level outcomes (6-23 months) (2) Models with household-level outcomes (3) Coefficient z￾statistic Coefficient z￾statistic Coefficient z￾statistic Village characteristics Classified as extremely vulnerable -0.081 -1.05 0.054 0.35 -0.054 -0.73 Total number of households 0.000 0.05 0.000 1.05 0.000 0.09 CARE is implementing agency 0.066 0.39 -0.289 -0.86 0.026 0.16 Nearest town > 1 hour away 0.001 1.04 0.002 0.84 0.001 1.19 PM2A village -0.148 -2.03 ** -0.142 -1.02 -0.165 -2.35 ** Other potential participation determinants Current shock exposure 0.027 1.00 -0.009 -0.16 0.035 1.30 Relative shock exposure in 2009 0.007 0.37 -0.043 -1.17 0.010 0.53 Bonding social capital (index) -0.051 -1.43 0.011 0.16 -0.061 -1.76 * Relative bonding social capital in 2009 0.181 4.12 *** 0.213 2.52 ** 0.190 4.40 *** Exposure to alternatives (index) 0.441 1.43 0.415 0.72 0.364 1.23 Absence of fatalism (index) -0.042 -1.15 -0.049 -0.68 -0.039 -1.12 Number of project staff known (2009) 0.031 1.70 0.024 0.73 0.036 2.08 ** Leisure time index (2009) 0.023 1.32 0.061 1.85 * 0.015 0.90 Women's decision making score (2009) 0.056 3.02 *** 0.096 2.65 *** 0.064 3.56 *** District mean child nutritional status (2009) Height-for-age z-score 0.465 1.94 ** 0.840 1.47 0.294 1.29 Weight-for-height z-score -0.039 -0.17 -0.013 -0.03 -0.064 -0.28 Number of observations 2,696 843 2,834 Pseudo R-squared 0.529 0.565 0.521 Notes: Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. Dependent variables employed for example models: (1) height-for-age z-score; (2) dummy variable for whether child has minimum dietary diversity; (3) months of adequate household food provisioning. 62 Table 19: Probit propensity score model estimation for participation in water and sanitation interventions Models with child￾level outcomes (6-59 months) (1) Models with child￾level outcomes (6-23 months) (2) Models with household-level outcomes (3) Coefficient z￾statistic Coefficient z￾statistic Coefficient z￾statistic Participation in other interventions MCHN 0.568 5.30 *** 0.541 5.27 *** 0.539 5.28 *** Women's empowerment 0.144 1.70 * 0.153 1.85 * 0.149 1.81 * Livelihoods promotion 0.522 4.72 *** 0.482 4.57 *** 0.484 4.61 *** Food ration from other project 0.244 3.14 *** 0.248 3.28 *** 0.248 3.29 *** Child characteristics Child's age -0.001 -0.18 -0.001 -0.12 Child's age-squared 0.000 0.14 0.000 0.19 Girl child 0.096 1.21 0.100 1.30 Mother characteristics Mother's age 0.005 0.81 0.005 0.87 Mother's education: None a/ Primary -0.193 -2.50 -0.174 -2.33 ** Secondary -0.157 -1.63 -0.141 -1.50 Household characteristics Age of household head 0.003 0.95 0.002 0.72 0.002 0.70 Female household head -0.004 -0.02 0.074 0.39 0.075 0.39 Education of household head: None a/ Primary -0.085 -1.16 Secondary -0.051 -0.53 Occupation of head: Farming a/ Agricultural laborer -0.287 -2.86 *** -0.265 -2.74 *** -0.250 -2.60 *** Non-agricultural laborer -0.135 -1.14 -0.137 -1.19 -0.132 -1.16 Salaried employment -0.227 -1.39 -0.284 -1.77 * -0.288 -1.75 * Self employment -0.076 -0.77 -0.075 -0.78 -0.082 -0.86 Unpaid household work 0.010 0.04 -0.073 -0.36 -0.077 -0.37 Other -0.241 -2.04 ** -0.245 -2.12 ** -0.260 -2.26 ** Household size 0.010 0.52 0.017 0.90 0.013 0.73 Age-sex composition: % females 0-16 a/ Percent females 16-30 0.005 1.08 0.004 0.88 0.000 -0.04 Percent females 30+ -0.006 -1.23 -0.005 -1.12 -0.006 -1.42 Percent males 0-16 -0.002 -0.91 -0.002 -0.93 -0.004 -2.18 ** Percent males 16-30 -0.003 -0.65 -0.001 -0.18 -0.003 -0.86 Percent males 30+ -0.006 -1.15 -0.003 -0.66 -0.005 -1.11 Well-being category: Extreme poor a/ Poor 0.275 2.59 *** 0.241 2.37 ** 0.239 2.36 ** Lower middle 0.321 1.94 * 0.229 1.44 0.227 1.43 Middle 0.380 2.13 ** 0.288 1.69 * 0.280 1.65 * Rich 0.263 1.16 0.128 0.58 0.139 0.64 Region: Coast a/ Haor -0.004 -0.04 -0.013 -0.13 -0.027 -0.28 Mid Char -0.776 -6.23 *** -0.825 -6.78 *** -0.850 -7.08 *** 63 Table 19: Probit propensity score model estimation for participation in water and sanitation interventions Models with child￾level outcomes (6-59 months) (1) Models with child￾level outcomes (6-23 months) (2) Models with household-level outcomes (3) Coefficient z￾statistic Coefficient z￾statistic Coefficient z￾statistic North Char -0.491 -4.44 *** -0.558 -5.17 ** -0.579 -5.45 *** Village characteristics Classified as extremely vulnerable 0.045 0.63 0.032 0.47 0.030 0.43 Total number of households 0.000 2.91 *** 0.000 2.89 *** 0.000 2.94 *** CARE is implementing agency -1.000 -5.16 *** -1.005 -5.25 *** -1.013 -5.28 *** Nearest town > 1 hour away -0.003 -2.64 *** -0.003 -2.55 ** -0.003 -2.59 *** PM2A village 0.207 3.15 *** 0.199 3.11 *** 0.193 3.02 *** Other potential participation determinants Current shock exposure 0.024 0.88 0.027 1.04 0.030 1.14 Relative shock exposure in 2009 -0.076 -4.23 *** -0.070 -4.08 *** -0.069 -4.01 *** Bonding social capital (index) 0.000 -0.01 -0.006 -0.17 -0.008 -0.24 Relative bonding social capital in 2009 -0.004 -0.11 0.003 0.08 0.004 0.11 Exposure to alternatives (index) 0.003 0.01 -0.049 -0.18 -0.038 -0.14 Absence of fatalism (index) -0.033 -0.89 -0.022 -0.62 -0.024 -0.69 Number of project staff known (2009) 0.069 4.29 *** 0.068 4.32 *** 0.068 4.35 *** Leisure time index (2009) 0.016 1.00 0.020 1.30 0.018 1.14 Women's decision making score (2009) -0.022 -1.31 -0.028 -1.69 * -0.028 -1.68 * District mean child nutritional status (2009) Height-for-age z-score -0.824 -4.08 *** -0.753 -3.90 *** -0.740 -3.87 *** Weight-for-height z-score 0.928 4.86 *** 0.909 4.90 *** 0.931 5.07 *** Number of observations 2,696 2824 2,834 Pseudo R-squared 0.223 0.218 0.215 Notes: Stars represent statistical significance at the 10%(*), 5%(**) and 1%(***) levels. Dependent variables employed for example models: (1) height-for-age z-score; (2) whether child under five had diarrhea in the last two weeks; (3) whether mother has knowledge of five critical times for hand washing. 64 References Ballard, Terri, Coates, Jennifer; Swindale, Anne, and Deitchler, Megan. 2011. Household Hunger Scale: Indicator Definition and Measurement Guide. Washington, DC: Food and Nutrition Technical Assistance II Project, FHI 360. Baum, C. F., Schaffer, M. E., & Stillman, S. 2007. Enhanced routines for instrumental variables/generalized method of moments estimation and testing. The Stata Journal, 7(4), 464–506. Bazzi, S., & Clemens, M. A. 2013. Blunt instruments: Avoiding common pitfalls in identifying the causes of economic growth. American Economic Journal: Macroeconomics, 5(2), 152–186. Bhutta, Z. A., Das, J. K., Rizvi, A., Gaffey, M. F., Walker, N., Horton, S., et al. 2013. Evidence-based interventions for improvement of maternal and child nutrition: What can be done and at what cost?. The Lancet. Bilinsky, Paula and Anne Swindale. 2010. Months of Adequate Household Food Provisioning (MAHFP) for Measurement of Household Food Access: Indicator Guide (v.4). Washington, D.C.: FHI 360/FANTA Black, R. E., Victora, C. G., Walker, S. P., Bhutta, Z. A., Christian, P., de Onis, M., et al. 2013. Maternal and child undernutrition and overweight in low-income and middle-income countries. The Lancet. Caldwell, Richard, Bruce Ravesloot and Lisa Smith. 2011. CARE Bangladesh SHOUHARDO II program: Baseline study report. TANGO, International, Tucson, AZ. de Onis, M., Garza, C., Victora, C. G., Bhan, M. K., & Norum, K. R.Guest editors. 2004. The WHO multicentre growth reference study (MGRS): Rationale, planning, and implementation. Food and Nutrition Bulletin, 25(1), S3–S84. FANTA-2. 2010. Title II Technical Reference Materials. TRM-01: Preventing Malnutrition in Children Under 2 Approach (PM2A): A Food-Assisted Approach. Revised November 2010. Washington, D.C.: Food and Nutrition Technical Assistance II Project (FANTA-2). Gillespie, S., Haddad, L., Mannar, V., Menon, P., & Nisbett, N. The Maternal and Child Nutrition Study Group. 2013. The politics of reducing malnutrition: Building commitment and accelerating progress. The Lancet, 382(9891), 552–569. Gilligan, D. O., & Hoddinott, J. 2007. Is there persistence in the impact of emergency food aid? Evidence on consumption, food security, and assets in rural Ethiopia. American Journal of Agricultural Economics, 89(2), 225-242. Habicht, J. P., Gretel, H., & Lapp, J. 2009. Methodologies to evaluate the impact of large scale nutrition programs. Doing Impact Evaluation, 13. Haddad, L. 2012. How can we build an enabling political environment to fight undernutrition? European Journal of Development Research, 25(1), 13–20. 65 Integrated Regional Information Network. 2010. Bangladesh: Towards “sanitation for all by 2010”. IRIN Asia. IRIN humanitarian news and analysis. Retrieved from: . Khandker, S. R., Koolwal, G. B., & Samad, H. A. 2010. Handbook on impact evaluation: Quantitative methods and practices. Washington, DC: The World Bank. Leuven, E., & Sianesi, B. 2003. PSMATCH2: Stata module to perform full Mahalanobis and propensity score matching, common support graphing, and covariate imbalance testing. Version 3.1.5 2 May 2009. Linnemayr, Sebastian and Harold Alderman. 2011. Almost random: Evaluating a large-scale randomized nutrition program in the presence of crossover. Journal of Development Economics 96(1): 106-114. National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ICF International. 2013. Bangladesh Demographic and Health Survey 2011. Dhaka, Bangladesh and Calverton, Maryland, USA: NIPORT, Mitra and Associates, and ICF International. National Institute of Population Research and Training (NIPORT), Mitra and Associates, and Macro International. 2009. Bangladesh Demographic and Health Survey 2007. Dhaka, Bangladesh and Calverton, Maryland, USA: National Institute of Population Research and Training, Mitra and Associates, and Macro International. National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ORC Macro. 2005. Bangladesh Demographic and Health Survey 2004. Dhaka, Bangladesh and Calverton, Maryland [USA]: National Institute of Population Research and Training, Mitra and Associates, and ORC Macro. Ruel, M. T., & Alderman, H. The Maternal and Child Nutrition Study Group. 2013. Nutrition-sensitive interventions and programmes: How can they help to accelerate progress in improving maternal and child nutrition? The Lancet. Schaffer, M.E. 2010. xtivreg2: Stata module to perform extended IV/ 2SLS, GMM and AC/HAC, LIML and k￾class regression for panel data models. . Schaeffer, Mark E. 2012. Statalist discussion “ivreg questions”. March 19, 2012. Shahin S, Bhadra SK and Ahsanul Alam M. Utilization of Essential Service Delivery (UESD) Survey 2013. 2014. Dhaka: National Institute of Population Research and Training (NIPORT). http://www.nfpcsp.org/agridrupal/sites/default/files/1385218571- UESD%20Survey%202013%20provisional%20report.pdf Smith, Lisa and Haddad, Lawrence. 2015. Reducing Child Undernutrition: Past Drivers and Priorities for the Post-MDG Era. World Development. (68): 180-224. Smith, L.C., Khan, F, Frankenberger, T.R., Wadud, A.K.M. Abdul. 2013. Admissible Evidence in the Court of Development Evaluation? The Impact of CARE’s SHOUHARDO Project on Child Stunting in Bangladesh. World Development (41): 196-216. 66 Smith, L. C., Ramakrishnan, U., Ndiaye, A., Haddad, L., & Martorell, R. 2003. The importance of women’s status for child nutrition in developing countries. IFPRI Research Report #131. Washington, DC: International Food Policy Research Institute. Stock, J. H., & Yogo, M. 2005. Testing for weak instruments in Linear IV regression. In J. H. Stock, & D. W. K. Andrews (Eds.), Identification and inference for econometric models: Essays in Honor of Thomas J. Rothenberg. New York: Cambridge University Press. Ten Have, Thomas R., Sharon-Lise T. Normand, Sue M. Marcus, C. Hendricks Brown, Philip Lavori, and Haihua Duan. 2008. Intent-to-treat vs. non-intent-to-treat analyses under treatment non-adherence in mental health randomized trials. Psychiatr Ann. 38(12): 772-783. United Nations Children’s Fund (UNICEF). 2013. Improving child nutrition: The achievable imperative for global progress. UNICEF: New York. Waddington, Hugh, Howard White, Birte Snilstveit, Jorge Garcia Hombrados, Martina Vojtkova, Philip Davies, Ami Bhavsar, John Eyers, Tracey Perez Koehlmoos, Mark Petticrew, Jeffrey C. Valentine and Peter Tugwell. 2012. How to do a good systematic review of effects in international development: a tool kit. Journal of Development Effectiveness 4(3): 359-387. Wadud, Abdul. 2015. Personal email communication. February 8, 2015 World Bank. 2013. Improving nutrition through multisectoral approaches. Washington, DC: World Bank. World Health Organization (WHO). 2008. Indicators for assessing infant and young childfeeding practices. World Health Organization, Geneva. Annex 1: References Ballard, T., J. Coates, A. Swindale and M. Deitchler. 2011. Household Hunger Scale: Indicator Definition and Measurement Guide. FANTA-2 Bridge, FHI 360. August, Washington, D.C. Bilinsky, P. and A. Swindale. 2010. Months of Adequate Household Food Provisioning (MAHFP) for Measurement of Household Food Access: Indicator Guide, Version 4. FANTA III/USAID. June, Washington D.C. British Broadcasting Corporation (BBC). 2014. Bangladesh Profile. News Asia. 3 November. Accessed at: http://www.bbc.com/news/world-south-asia-12650940 FAO, WFP and IFAD. 2013. The State of Food Insecurity in the World 2013. The multiple dimensions of food security. Rome, FAO. Government of the People’s Republic of Bangladesh (Ministry of Planning, General Economics Division) and UNDP Bangladesh. 2014. Millennium Development Goals: Bangladesh Progress Report 2013. August, Dhaka. Government of the People’s Republic of Bangladesh (Ministry of Planning, General Economics Division). 2012. Sixth Five Year Plan Fy2011-Fy2015. Part‐1: Strategic Directions and Policy Framework, Dhaka. IFPRI, Concern and Welthungerhilfe. 2014. Global Hunger Index: The challenge of hidden hunger. Washington, D.C., IFPRI. Karmalkar, A., C. McSweeney, M. New and G. Lizcano. 2010. UNDP Climate Change Country Profiles: Bangladesh. Accessed at: http://country-profiles.geog.ox.ac.uk/ National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ICF International. 2013. Bangladesh Demographic and Health Survey 2011. Dhaka, Bangladesh and Calverton, Maryland, USA. January. United Nations Development Programme (UNDP). 2014. Human Development Report 2014: Sustaining human progress: Reducing vulnerabilities and building resilience. New York, UNDP. Swindale, A., and Bilinsky, P. 2006. Household Dietary Diversity Score (HDDS) for Measurement of Household Food Access: Indicator Guide (v.2). Washington, D.C.: FHI 360/FANTA. World Bank. 2014. Bangladesh Data/ ASPIRE Key Indicators. Accessed at: http://data.worldbank.org/country/bangladesh#cp_wdi and http://datatopics.worldbank.org/aspire/country/bangladesh Annex 2: Scope of Work Quantitative Impact Evaluation of the SHOUHARDO II Project in Bangladesh on Household Food and Livelihood Outcomes Technical Proposal submitted by TANGO, International Inc. 406 S. 4th Avenue Tucson, Arizona 85701 USA Telephone: (1-520) 617-0977 Fax: (1-520) 617-0980 www.tangointernational.com to Monzu Morshed Acting Chief of Party SHOUHARDO II Project CARE, Bangladesh December 5, 2014 1. Introduction As is now well-documented, child malnutrition represents a fundamental squandering of human potential. In addition to child mortality, it is associated with poor school and work performance and an increased likelihood of overweight, chronic disease and mental health issues among adults. As one might expect, such personally damaging impacts of malnutrition for the world’s youngest citizens and their families, along with its intergenerational transmission, have severe consequences for entire economies, dampening economic growth and poverty reduction (Smith and Haddad 2014). The development community is increasingly recognizing that slower-than-expected progress towards reaching the Millennium Development Goals (MDGs) by 2015—including those for poverty, secondary education, child mortality and maternal health—is due, in large part, to lack of investment in children’s nutrition (World Bank 2013). The current momentum within developing countries and internationally to address the problem of child malnutrition has never been higher. The rise of the Scaling Up Nutrition (SUN) movement starting in 2010, and the publication of the Lancet Maternal and Child Nutrition Series in 2008 have both served to raise awareness of its extent and consequences. Nutrition has consequently been greatly elevated on the development agenda, and global commitment to reducing malnutrition is stronger than ever (Gillespie and Haddad et. al. 2013). A case in point: “Food Security and Good Nutrition” is one of 12 Development Goals proposed in the UN’s High Level Panel on Development After 2015. Given the above momentum, answers to the question of how to accelerate reductions in child malnutrition in the coming decades are in great demand. A wide evidence base is building regarding the key roles of nutrition-specific interventions, such as micronutrient supplementation and nutrition education, as well as more fundamental, underlying and basic determinants of malnutrition (Bhutta et al. 2013; Ruel et. al. 2013; Haddad 2012; Ruel and Alderman 2013). With regard to the latter, a recent study by Smith and Haddad (2014), for example, demonstrates that improvements in safe water access, sanitation, women’s education, gender equality, and the quantity and quality of food available in countries have been key drivers of stunting reductions since the 1970s. Poverty-reducing income growth and advances made in the quality of governance played essential facilitating roles. The experience of the SHOUHARDO II project in Bangladesh—a country with one of the highest stunting prevalences in the world, at 41 percent of all children under five (Niport et. al. 2013)—provides a unique opportunity to gain insight into how development projects implemented at the local level can contribute to accelerating reductions in child malnutrition. Funded by the United States Agency for International Development and the Government of Bangladesh, SHOUHARDO II (Strengthening Household Ability to Respond to Development Opportunities II) is being implemented by CARE and its Bangladeshi NGO partners in eleven of the poorest and most marginalized districts in Bangladesh. The project takes a novel approach to reducing malnutrition, integrating nutrition-specific interventions with those that address key underlying determinants of stunting using a right-based, livelihoods programming approach (Frankenberger, Drinkwater and Maxwell, 2000). The underlying determinants the project is addressing include the dis-empowerment of women and the poor, low food production and income, poor hygiene conditions, recurrent disasters and environmental (climate) change. Smith et al. (2012) and TANGO, International (2009) provide evidence that the project’s predecessor, the “SHOUHARDO I” project, was exceptionally successful in applying this approach to reduce child stunting. Early indications from the mid-term evaluation of the SHOUHARDO II project are that reductions in stunting in the project’s operational area are likely to be similarly strong: the stunting rate among children under five at baseline (December 2010) was 61.7. It had fallen to 52.7 percent (a total of 9 percentage points) by the midterm (December 2012). Beyond monitoring progress towards strategic objectives and intermediate results, the rich data collected in the SHOUHARDO II baseline, midterm and endline surveys can be drawn on to determine whether and why the project actually brought about these recorded reductions in child stunting. As detailed below, they present the opportunity to conduct a more rigorous evaluation than those collected in conjunction with the SHOUHARDO I project. The data allow methods to be used that can more definitively determine which interventions caused reductions in child malnutrition and which have synergistic impacts with the nutrition-specific interventions implemented, thereby providing valuable lessons for future projects striving to reduce child malnutrition in Bangladesh and other developing countries. 2. Research questions The overall objective of this research project is to determine whether and by how much the SHOUHARDO II project as a whole, and four sets of interventions implemented at the household level, served to reduce stunting among children under five in the project area. The four sets of interventions are: (1) Nutrition-specific interventions -Provision of supplementary food rations to pregnant women and women with children under 2 -Nutrition education through Mother’s Groups -Growth monitoring and promotion (2) Interventions aimed at enhancing women’s empowerment -Empowerment, Knowledge and Transformative Action (EKATA) women’s groups (3) Interventions aimed at increasing food production and incomes -Training and provision of inputs to promote field crop and fisheries production, homestead gardening, livestock rearing, and cash income generating activities -Food-for-work -Cash-for-work -Savings groups (4) Interventions aimed at increasing access to safe water and sanitation -Provision of sanitary latrines -Provision of tube wells -Provision of sanitation drains. An additional objective is to determine whether there are synergistic impacts of the nutrition-specific interventions with those addressing the underlying determinants of malnutrition (2, 3 and 4 above). The objective will help improve understanding of whether the project’s integrated Rights-Based Livelihoods approach made a difference. The specific questions that will be investigated are: Overall impact of the project  Did the SHOUHARDO II project’s activities cause the reduction in stunting that took place among project beneficiaries over the life of the project? Impact of specific types of interventions  Did the project’s nutrition-specific interventions serve to reduce child stunting?  Did the project’s women’s empowerment intervention serve to reduce child stunting?  Did the project’s interventions to increase food production and incomes serve to reduce child stunting?  Did the project’s interventions to improve access to safe water and sanitation serve to reduce child stunting? Synergistic impacts  Were greater impacts achieved when the project’s nutrition-specific interventions were combined with those aimed at enhancing women’s empowerment?  Were greater impacts achieved when the project’s nutrition-specific interventions were combined with those aimed at increasing food production and incomes?  Were greater impacts achieved when the project’s nutrition-specific interventions were combined with those aimed at increasing access to safe water and sanition? 3. Methods Impact evaluation is essentially about determining the extent to which changes in outcomes can be attributed to a project or intervention. The two key necessary conditions for an impact evaluation to be conducted in a rigorous manner are that (1) a control group be available so that a counter-factual can be identified; and (2) that the problem of selection bias be addressed (Waddington et. al. 2012). This latter problem arises because, in most cases, either purposeful targeting of project interventions to specific populations (e.g., the most poor) and/or self-selection of participants into interventions takes place. This renders the control group and the “intervention” group fundamentally different from one another even prior to the commencement of project activities. Survey data collected as part of the SHOUHARDO I project were used to answer similar questions to those being posed here for SHOUHARDO II (see Smith et al. 2012). However, the fact that the data were only collected for project beneficiaries and the questionnaire was designed specifically for performance monitoring and not to conduct an impact evaluation constrained the rigor with which the analysis could be undertaken. Such a situation necessitated that a “mixed-methods” approach, triangulating information from the project data sets with that from various other sources, be used in order to establish evidence of impact. Further, while the evidence was strong, it was necessary to classify it as “suggestive” rather than definitive. The SHOUHARDO II data sets, by contrast, contain ample data for households that did not participate in the project’s interventions at all or only in its MCHN activities, thus providing a pool of potential control group households.1 Further, as outlined below, the endline survey can be extended to allow the collection of the data necessary for addressing the problem of selection bias, that is, data on factors affecting households’ participation in various interventions. For the SHOUHARDO I analysis only such factors for which data had already been collected were included, leaving out important “unobservable” factors, the exclusion of which can lead to bias in estimates of the impacts of interventions. The three types of analysis proposed to answer the research questions laid out above are outlined below. The first, a set of descriptive analyses, provides key contextual information about the project’s impact on child stunting. The second two—propensity score matching and instrumental variables estimation—allow for a more rigorous evaluation design than was possible using SHOUHARDO I data, and thus a stronger evidence base for the lessons to be learned from the SHOUHARDO approach. 3.1 Descriptive analysis of trends in stunting and the age trajectory of stunting Following the SHOUHARDO I evaluation protocol, we will use the following descriptive analyses to provide evidence on the overall question: Did the SHOUHARDO II project’s activities cause the reduction in stunting that took place among project beneficiaries between the baseline and endline surveys (December 2010-December 2014)? 1. Comparison of trends in stunting in the project area, and among project participants, compared to those of rural children 0-5 years old nationally. The national data sets that will employed are Demographic and Health Surveys (DHS) conducted in 2011 and 2014 (forthcoming) and Hellen Keller International (HKI) Food Security and Nutrition Surveillance surveys conducted in 2010 (Round 3), 2012 (Round 9) and 2014.2 The SHOUHARDO II baseline, mid-term and endline survey data will be used for this analysis. 2. Comparison of the age trajectory of stunting over the 0-5 year old range in the project area, and among project beneficiaries, compared to nationally. The national data employed will be from the 2014 DHS, pending availability. This analysis will help determine whether the typical pattern for children in Bangladesh of a sharp increase in stunting as they age over the six month to two year old range and continually high prevalences thereafter is exhibited among SHOUHARDO II project children. If not, then it can be surmised that the project’s activities as a whole had a preventative effect. The endline survey data will be used for this analysis. 3.2 Propensity Score Matching and Multiple Treatment PSM Propensity Score Matching (PSM) will be used to construct a valid control group to compare with the group of households participating in each set of project interventions being evaluated (the “intervention group”). This control group mimics that which would be used if the project’s interventions had been 1 See the SHOUHARDO I baseline and mid-term reports for a full explanation of the survey sampling designs. 2 It has yet to be determined whether and when data were collected by HKI in 2014. randomly allocated across households in the project area. The impact of interventions is estimated using the difference in the stunting prevalence between the control group and intervention groups. Multi-treatment PSM (MT-PSM) is an extension of PSM that will allow determination of whether there are positive synergies between the nutrition-specific and three other sets of interventions. We will do so following the analysis strategy laid out in Smith et al. (2012). The PSM and MT-PSM analyses will be conducted using the endline survey data. In PSM, identification of a control group takes place through a matching process using measured indicators of characteristics that are believed to influence participation in the interventions being evaluated as well as those influencing the outcome of interest, in this case stunting. If these observed characteristics are the only factors influencing participation, then impact estimates can be deemed unbiased and serve to give valid, causal evidence of impact. However, if unobserved characteristics also influence participation, then the estimates will be biased (Khandker et al. 2010). The challenge then, is to collect data on the entire universe of such characteristics so that none can be deemed unobserved. As noted above, households’ and individuals’ participation in SHOUHARDO II interventions was influenced by two factors: (1) targeting conducted by project administrators and (2) households’ and individuals’ own decisions on whether to participate. The SHOUHARDO II endline questionnaire is already set up to capture many of the indicators of participation determinants typically employed in conventional PSM impact analyses (see, for example Raza et. al.’s 2012 evaluation of BRAC). These are: Households’ socio-economic status Household size and age-sex structure Education Employment status Occupation Gender of household head Region of residence. Others factors are not typically measured and thus relegated to the “unobservables” category. For the SHOUHARDO II project these fall into the following categories: Aspirations and confidence to adapt3 Time constraints Strength of social connections Women’s decision making power in their households Indicators of personal familiarity with project staff Shock exposure. To capture these factors and thus render them “observables”, an additional module on “Factors affecting participation in the SHOUHARDO II project” will be added to the endline questionnaire. It is also important for a valid assessment of impact that the characteristics affecting participation used for matching not be affected by project activities themselves. Given that panel data were not collected 3 This factor will pick up on any Hawthorn effects (Waddinton et. al. 2012), whereby differences in outcomes across the control and intervention groups are influenced by participant motivation as a result of the program itself. (that is, the baseline and endline surveys were not administered to the same households), for the characteristics listed above that may have been affected by the project retrospective recall data will be collected. That is, households will be asked to answer questions regarding their current situation and then whether (and in some cases, by how much) they feel it has changed over the last five years, since the inception of the project. Note that as part of the SHOUHARDO II targeting process, each household in the project’s operational area, including those not participating in the project, was assigned a ”well-being category” (extreme poor, poor, middle, middle-rich, and rich) using Participatory Rural Appraisal techniques. The assignment took place before the baseline survey was administered. Thus retrospective data are not needed for measuring socio-economic status. 3.3 Instrumental Variables Estimation Instrumental Variables (IV) estimation is an alternative technique that can be used to control for selection bias in the estimation of the impacts of project interventions. It involves identifying a variable (termed an “instrument”) that is highly correlated with participation in an intervention but that does not influence the outcome of interest through any other pathway and thus can be deemed “exogenous”. Following, two-stage least squares regression is used to estimate the impact of a project or intervention (Khandker et al. 2010). A number of potentially valid instruments will be used for this impact evaluation. The first is an indicator of a participation assignment mechanism that took place as part of a Randomized Controlled Trial (RCT) embedded into the sampling design for the SHOUHARDO II surveys. Specifically, villages in the project’s operational area were randomly assigned to one of two groups: (1) those subject to the traditional SHOUHARDO mechanism for the allocation of food assistance and other MCHN (Maternal and Child Health and Nutrition) interventions, titled “PEP/MCHN”, in which only extreme poor or poor households receive assistance; or (2) those subject to an alternative allocation mechanism, titled “PM2A” (Preventing malnutrition under two (see Ruel et. al. 2008), whereby all households, regardless of socio-economic status, receive the assistance. Such an exogenous rule for determining eligibility for the program is likely to have influenced child malnutrition only through its influence on who and who did not participate in the MCHN interventions. It possibly also affected participation in the other project interventions. Other potential instruments are reported travel time to project activities (for example, EKATA group meetings) and indicators of personal familiarity with project staff, both of which are likely to have influenced child stunting only through influencing project participation decisions. Current techniques to ensure that the instruments used for IV estimation are relevant or “strong” (correlate sufficiently with participation) and valid (do not affect stunting through channels other than participation) will be employed (see Bazzi and Clemens 2013). 4. Research team, time requirements, budget and schedule 4.1 Research team The analysis will be conducted by Lisa Smith and Tim Frankenberger of TANGO, International, who have extensive experience conducting impact evaluations and working on Bangladesh development issues in various capacities. Lisa Smith holds a doctoral degree in Agricultural and Applied Economics from the University of Wisconsin, Madison. Currently a Senior Economist at TANGO International, she specializes in the areas of development economics, food and nutrition security, gender, and project impact evaluation. As an American Association for the Advancement of Science Post-Doctoral Fellow she gained experience providing technical assistance in the area of food security to USAID. While a Research Fellow at the International Food Policy Research Institute, she conducted various studies of the causes of child malnutrition and food insecurity in developing countries using both Demographic and Health Survey data and national-level data to conduct cross-country analyses. In her current position at TANGO she has conducted numerous food and livelihood security assessments as well as impact evaluations of large scale NGO-implemented development projects in Zimbabwe and Bangladesh. Timothy Frankenberger is the President and co-founder of TANGO International (14 years) and has over 30 years of experience in international development activities. He previously was the Senior Food Security Advisor and Livelihood Security Coordinator at CARE for six years. Prior to this he was a farming systems research specialist at the University of Arizona for eight years. He has published numerous articles on household food security and is considered an expert on livelihood approaches and resilience. He received a B.S. in Anthropology from Iowa State University and an M.A. in Anthropology with a minor in Agricultural Economics from the University of Kentucky. Tim also carried out his doctoral studies at the University of Kentucky in Anthropology with a minor in Agricultural Economics. 4.2 Time requirements Days Lisa Smith Tim Frankenberger Total Development of endline questionnaire modules on participation and factors affecting participation 3.5 1 Variable creation 4 1 Descriptive analysis 4 PSM analysis 4 Multi-treatment PSM analysis 4 IV analysis 5 Write first draft of report 8 1 Write final draft of report 2.5 1 35 4 39 4.3 Schedule The study will be conducted between December 15 and 28 of February, 2015. 4.4 Budget Questionnaire preparation, data analysis, and report write-up LSmith 35 days x 680/day + VAT(15%) 27,370.00 TFrankenberger 4 days x 690/day + VAT(15%) 3,174.00 Field work: Data collection by Mitra & Associates 9,000.00 TOTAL 39,544.00USD References Bazzi, Samuel and Michael A. Clemens. 2013. Blunt Instruments: Avoiding Common Pitfalls in Identifying the Causes of Economic Growth. American Economic Journal: Macroeconomics 5(2): 152-86. Bhutta , Zulfiqar A, Jai K Das, Arjumand Rizvi, Michelle F Gaffey, Neff Walker, Susan Horton, Patrick Webb, Anna Lartey, Robert E Black, The Lancet Nutrition Interventions Review Group, and the Maternal and Child Nutrition Study Group. 2013. Evidence-based interventions for improvement of maternal and child nutrition: what can be done and at what cost? The Lancet, June 6, 2013. Frankenberger, Tim, M. Drinkwater and D. Maxwell. 2000. Operationalizing household livelihood security: A holistic approach for addressing poverty and vulnerability. Atlanta, GA: CARE International. Gillespie, Stuart , Lawrence Haddad, Venkatesh Mannar, Purnima Menon, Nicholas Nisbett, and the Maternal and Child Nutrition Study Group. 2013. The politics of reducing malnutrition: building commitment and accelerating progress. The Lancet 382(9891):552-569. Haddad, L. 2012. How can we build an enabling political environment to fight undernutriton? European Journal of Development Research, 25(1), 13-20. Khandker, Shahidur R., Gayatri B. Koolwal and Hussain A. Samad. 2010. Handbook on impact evaluation: quantitative methods and practices. The World Bank, Washington, D.C. National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ICF International. 2013. Bangladesh Demographic and Health Survey 2011. Dhaka, Bangladesh and Calverton, Maryland, USA: NIPORT, Mitra and Associates, and ICF International. Raza, Wameq A., Narayan C. Das and Farzana A. Misha. 2012. Can ultra-poverty be sustainably improved? Evidence from BRAC in Bangladesh. Jounral of Development Effectiveness 4(2): 257-276. Ruel , Marie T, Harold Alderman, and the Maternal and Child Nutrition Study Group. 2013. Nutrition-sensitive interventions and programmes: How can they help to accelerate progress in improving maternal and child nutrition? The Lancet, June 6, 2013. Ruel, Marie, Purnima Menon, Jean-Pierre Habicht, Cornelia Loechl, Gilles Bergeron, Gretel Pelto, Mary Arimond, John Maluccio, Lesly Michaud and Bekele Hankebo (2008). Age-based preventive targeting of food assistance and behaviour change and communication for reduction of childhood undernutrition in Haiti: a cluster randomized trial. Lancet, 371, pp. 588-595. Smith, Lisa C. and Lawrence Haddad. 2014. Reducing child undernutrition: Past drivers and priorities for the post-MDG era. IDS Working Paper No. 441. Institute of Development Studies, Brighton. (in review World Development). Smith, Lisa C., Faheem Khan, Timothy R. Frankenberger and AKM Abdul Wadud. 2012. Admissible evidence in the court of development evaluation? The impact of CARE’s SHOUHARDO project on child stunting in Bangladesh. World Development (41): 196-216. Waddington, Hugh, Howard White, Birte Snilstveit, Jorge Garcia Hombrados, Martina Vojtkova, Philip Davies, Ami Bhavsar, John Eyers, Tracey Perez Koehlmoos, Mark Petticrew, Jeffrey C. Valentine and Peter Tugwell. 2012. How to do a good systematic review of effects in international development: a tool kit. Journal of Development Effectiveness 4(3): 359-387. World Bank. 2013. Improving nutrition through multisectoral approaches. World Bank, Washington D.C. CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 14 of 142 Annex 3: Endline Survey Instrument Quantitative Performance Evaluation Population Based Sample Survey 2014 for USAID Title II CARE SHOUHARDO II Program Quantitative Questionnaire CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 15 of 142 CARE Bangladesh SHOUHARDO II Program and TANGO International 406 S. 4th Avenue Tucson, Arizona 85701 USA Mitra and Associates (Centre for Research and Consultancy) 2/17 Iqbal Road, Mohammadpur Dhaka-1207, Tel: +88-02-8118065, +88-02-9115503 Note for Enumerators:  Insert all data very clearly in the appropriate boxes/cells.  Open ended responses must be entered in Bengali  Use permanent, non-washable and black ink.  Do not use whiteout/fluid.  Do not overwrite. CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 16 of 142 INFORMED CONSENT Salutation (Assalamualaikum/Namashkar/hello): My name is ____________________ and I am representing Mitra and Associates, a private consultancy firm that conducts social, livelihood, health and demographic surveys in Bangladesh. With technical support from TANGO International, Mitra Associates will conduct a quantitative performance evaluation population based sample survey for CARE Bangladesh, USAID/FFP, LGRD and Cooperative Ministry funded project Strengthening Household Ability to Respond to Development Opportunities II (SHOUHARDO II). Your household/you have been selected by chance for this interview. Your participation in the survey is voluntary and your participation does not guarantee that you will be provided any benefits from the project. I would like to collect information on socio-economic condition and mother and child health and anthropometric measures for children U5 of your household. The survey is voluntary and you/your family can choose not to take part. The information (e.g. age, date of birth, socio-economic status etc.) that you/your family give will be used to prepare reports. The report will be used for the program and will be used as a reference. Please let me know if you are agreed with it ornot. Do you agree to provide me with birthdates and other information of yourself and other household members, with the understanding that the information will be strictly confidential and not released to any organization or individual? Yes No ……..if no, thank the respondent and finish the interview It will need 1-2 hours to complete the interview. Could you please spare some time for the interview? At this time, do you want to ask me anything about the survey? If you have any question as a participant of this survey, you may contact here: Address: Director, Mitra and Associates, 2/17 Iqbal Road, Mohammadpur, Dhaka-1207, Bangladesh Phone: 8118065, 9115503. Name: Respondent 1________________________ Yes No End Name: Respondent 2________________________ Yes No End Name: Respondent 3________________________ Yes No End Signature of Interviewer: _______________________Date____________________________ CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 17 of 142 Question Serial HOUR MINUTE Cluster Number RECORD TIME THE INTERVIEW STARTED HH WBA ID HH Unique ID** Survey Team’s ID **From household MBL/sample list MODULE A: INFORMATION ON INTERVIEW AND AREA IDENTIFICATION Interview information A1 Date of interview dd mm yy ID Signature A2 Name of Interviewer 1 A3 Name of Interviewer 2 A4 Reviewed by (Supervisor/ Name & Code) A5 Reviewed by (Team Leader/Name & Code) A6 Reviewed by others (Name & Code) A7 Data Entry by (Name/Code) A8 Data Entry Date A9 Data entry checked by (Name/Code) A10 Data Entry Checked Date Area Identification (follow MBL/ sample list) CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 18 of 142 Area A11 Region Coast ................................................................................ 1 Haor ................................................................................. 2 Mid Char .......................................................................... 3 North Char ....................................................................... 4 A12 District (Use Geo-code) A13 Upazila(Use Geo-code) A14 Union (Use Geo-code) A15 Village (Use Code provided) A16 Village Type MCHN............................................................................... 1 PM2A ............................................................................... 2 A17 ECCD Yes.................................................................................... 1 No..................................................................................... 2 A18 Savings Group Yes.................................................................................... 1 No..................................................................................... 2 A19 Implementing Organization A20 Interview result Complete.....................................................................1 Incomplete..................................................................2 Write reason: PART I: HOUSEHOLD INFORMATION [Instruction: Respondent should be head of household, knowledgeable adult household members, preferably the head of sampled household and spouse] MODULE B: RESPONDENT IDENTIFICATION FOR PART- I QUESTION CODE/RESPONSE SKIP B1 CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 19 of 142 Name of household head: _______________________________________ [INSTRUCTION: Use name from the sample list] B2 Part I Respondent’s name: __________________________(Household head if possible) B2.1 Part I Respondent Line number (Module C) [INSTRUCTION: Fill in after interfiew is complete] B3 Relationship to household head (see codes below) B4 Part II Respondents Name: __________________________________ B5 Relationship to household head (see codes below) B6 Do you have cell/home phone for your personal/adult household members/household use? Yes............................................................. 1 No ............................................................. 2 2→B7 B6a Cell or home phone number of household head or other adult household member B6b Ownership type of the cell phone / home phone? Self ............................................................ 1 Family member......................................... 2 On request................................................ 3 Others (SPECIFY) ........................................ 4 B6c Since when are you using this phone? __ __ / __ __ __ __ m m y y y y CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 20 of 142 B7 Marital status of household head Married ..................................................... 1 Single ........................................................ 2 Divorced/separated .................................. 3 Widowed................................................... 4 B8 Disability status of household head List responses and circle code number of response Not disabled.............................................. 1 Physically disabled (temporary)................ 2 Physically disabled (permanent)............... 3 Mentally disabled ..................................... 4 B9a Household well-being category [INSTRUCTION: Extract from HH sample sheet] Extreme Poor............................................ 1 Poor .......................................................... 2 Lower Middle............................................ 3 Middle....................................................... 4 Rich ........................................................... 5 B9b Household WBA ID [INSTRUCTION: Extract from HH sample sheet] B9c Household Unique ID [INSTRUCTION: Extract from HH sample sheet] Codes for B3 and B5: Relationship to household head 1= Household head 2= Wife of household head 3= Husband of household head 4 =Caregiver of children (in absence of mother) 5= Son 6= Daughter 7= Father 8= Mother 9= Daughter in law/son in law 10= Brother 11= Sister 12= Father/mother in law 13= Nephew/niece 14= Grandfather/Grandmother 15= Grandson/Granddaughter 16= Sister-in-law/Brother-in-law 17= Brother’s wife 18= Others (e.g. servant) CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 21 of 142 MODULE C: BASIC INFORMATION ON HOUSEHOLD MEMBERS Circle line number of member Line number Name of member (Start with household head) Sex 1= Male 2= Female Age in years (write “0” for less than 1 year) Education (for ages 6 and up) Literacy 1= Can read 2 = Can write 3 = Can read and write 4 = Neither Primary occupation (see codes below) Secondary occupation (see codes below) Eligible for Part II: women’s empowerment Woman 18 years or older? Part III: Information on children under 6 years old Child under 6? Enter line # of caregiver Only for adults (18 or older) C1 C2 C3 C4 C5 C6 C7 C8 C9 1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 CARE Bangladesh SHOUHARDO II Performance Evaluation Survey 2014 Page 22 of 142 9 9 9 10 10 10 11 11 11 12 12 12 13 13 13 14 14 14 15 15 15 Codes for C3:0 = No class, 1 = Class 1, 2 = Class 2, 3 = Class 3, 4 = Class 4, 5 = Class 5, 6 = Class 6, 7 = Class 7, 8= Class 8, 9 = Class 9, 10 = SSC pass, 11 = HSC pass, 12 = Graduate, 13 = Masters Codes for C5 and C6: 1=Farming, 2 =Agricultural day labor/contract labor, 3 =Fishing, 4 =Poultry and livestock rearing, 5 =Non-agricultural day labor/contract labor, 6 =Casual labor, 7 =Regular salaried employment, 8 =Self-employed in business/petty business, 9 =Paid “volunteers”, 10 = House work (child care, home care), 11 =Servant/ Maid, 12 =Student, 13 =Beggar, 14 =Old/ Disabled, 15 =Unemployed, 16 =Other. CARE Bangladesh SHOUHARDO II Baseline Survey Page 23 of 142 MODULE D: HOUSEHOLD ECONOMIC SECURITY [INSTRUCTION: Head of the household (or knowledgeable adult) is the respondent for Section D] Indicators of economic distress QUESTION CODE/RESPONSE SKIP D1 Did any resident household member migrate out of the village for part of the last 12 months to find employment? Yes.................................................. 1 No................................................... 22→D2 D1.1 If yes, how many members migrated? Members D1.2 What was the longest time any one person was gone (in days) Days D2 Did any resident household member sell labor in advance for part of the last 12 months? Yes.................................................. 1 No................................................... 2 D3 Did any resident household member take out an interest bearing loan from non-formal sources in the last 12 months? Yes.................................................. 1 No................................................... 2 Housing Characteristics QUESTION CODE/RESPONSE SKIP D4 What is the main construction material of the walls of your main house? [INSTRUCTION: Observe and circle code number of response] Brick ................................................1 C.I. Sheet / wood.............................2 Mud wall .........................................3 Bamboo...........................................4 Straw/jute stick/leaves ...................5 Thatched bamboo/polythene .........6 Other.....................................................7 CARE Bangladesh SHOUHARDO II Baseline Survey Page 24 of 142 QUESTION CODE/RESPONSE SKIP D5 What is the main construction material of the roof of your main house? [INSTRUCTION: Observe and circle code number of response] Concrete..........................................1 C.I. Sheet/wood...............................2 Tiles.................................................3 Bamboo...........................................4 Straw/jute stick/leave .....................5 Thatched bamboo/polythene .........6 Other...............................................7 D6 How many rooms do you have for your family to live in your house? Rooms Ownership and sales of assets Asset D7 D8 Number currently owned Sales in last year Now I’m going to ask you about some of the items you own in your house. How many (_____) do you own? [INSTRUCTION:If do not own, write “0”] D8_1. Did you sell any of these kinds of items in the last year? Domestic assets 1A Chairs Yes .................................................1 No ..................................................2 DNK................................................8 1B Khat 1C Cupboard 1D Tables 1E Show case 1F Dressing table 1G Watch 1H Clock 1I Lantern 1J Radio CARE Bangladesh SHOUHARDO II Baseline Survey Page 25 of 142 1K TV 1L Cassette player 1M Electric fan 1N Mobile Phone 1O Gold ornaments/jewelry (ana) 1P Silver ornaments/jewelry (ana) Now I’m going to ask you about some of the transportation and agricultural items you own. How many (_____) do you own? [INSTRUCTION:If do not own, write “0”] D8_2. Did you sell any of these kinds of items in the last year? Transport/Agricultural Assets 2A Boat Yes ..................................... 1 No ...................................... 2 DNK.................................... 8 2B Motorcycle 2C Rickshaw/van 2D Bicycle 2E Shallow / hand tube well 2F Power tiller 2G Paddle thresher 2H Spray machine 2I Plough 2J Fishing net 2K Pump 2L Hoe 2M Axe 2N Shovel/spade 2O Weedier 2P Power thresher Now I’m going to ask you about the farm animals you own. How many (_____)do you own? [INSTRUCTION: If do not own, write “0”] D8_3. Did you sell any of animals in the last year? Animal Assets CARE Bangladesh SHOUHARDO II Baseline Survey Page 26 of 142 3A Cow Yes ..................................... 1 No ...................................... 2 DNK.................................... 8 3B Buffalo 3C Goat 3D Sheep 3E Chicken 3F Duck 3G Pigs Now I’m going to ask you about the trees and plants you own. How many(_____) do you own? [INSTRUCTION: If do not own, write “0”] D8_4. Did you sell any of these trees or plants in the last year? Trees and Plants 4A Timber tree Yes ..................................... 1 No ...................................... 2 DNK.................................... 8 4B Fruit tree 4C Bamboo tree 4D Medicinal plants Land ownership How much of these types of land do you own (in decimals)? QUESTION CODE/RESPONSE SKIP Amount owned (Decimals) D9A Own homestead land . D9B Own agricultural land . D9C Land lease-IN . D9D Land lease-OUT . D9E Mortgage-IN . D9F Mortgage-OUT . CARE Bangladesh SHOUHARDO II Baseline Survey Page 27 of 142 QUESTION CODE/RESPONSE SKIP D9G Haor (extended marsh) . D9H Pond/ditch . D9I Other type of land . D10 Did anyone in your household sell any assets in the last year in order to be able to purchase food, pay for medicine, pay school fees, or meet any other urgent household need? Yes ..................................................1 No ...................................................2 DNK.................................................8 2,8→D12 D11 How much money did you get from selling assets for these things? (Taka) Taka ................ DNK...............................................98 CARE Bangladesh SHOUHARDO II Baseline Survey Page 28 of 142 Household income/Earnings D12 Did any resident household member bring cash income into the household in the last year? If yes, then write their name and line no. in following column. If necessary, see module C columns C5,C6. Yes..........................................................................1 No...........................................................................2 DNK.........................................................................8 NR...........................................................................9 2,8,9→D18A CARE Bangladesh SHOUHARDO II Baseline Survey Page 29 of 142 What activities did you make money from in last year? For each activity, record number of months and monthly income. Prompt for more activities till respondent indicates no more. (if less than 1 month, enter 1) Number of months in different activities and (net) income for last 12 months Person 1 Name ______________ Line number from Module C Person 2 Name ______________ Line number from Module C Person 3 Name ______________ Line number from Module C Person 4 Name ______________ Line number from Module C Person 5 Name ______________ Line number from Module C # of months Monthly Income (Taka) # of months Monthly Income (Taka) # of months Monthly Income (Taka) # of months Monthly Income (Taka) # of months Monthly Income (Taka) D13.1 D13.2 D14.1 D14.2 D15.1 D15.2 D16.1 D16.2 D17.1 D17.2 a Agricultural day labor b Agricultural contract labor c Non-agricultural day labor d Non-agricultural contract labor e Casual labor f Regular salaried employment g Self-employment in business/service provision h Petty business i Business, using hired labor j Paid “volunteer” CARE Bangladesh SHOUHARDO II Baseline Survey Page 30 of 142 k Rickshaw/rickshaw van pulling l Boatman m Working as servant/maid n Begging o Cash-for-work p Student stipend (including cash value of food received) q Others (specify) 31 Other source Income (Taka) D18a How much income did your household receive from remittances in the last year? D18b from gifts in the last year? D18c from pensions/retirement fund? D18d from leases? D18e sales of agricultural products? If “0”, skip to D18_g D18f How much did you spend on agricultural inputs (e.g., seeds, fertilizer, etc) in the last year? D18g How much income did your HH receive from sales of animals or animal products (including cattle, poultry and fish) in the last year If “0”, skip to D19_a D18h How much did you spend on inputs needed to raise the animals (e.g., feed, veterinary services) in the last year? Remoteness and access to markets QUESTION CODE/RESPONSE SKIP D19a How long would it take to walk to the nearest town? Less than 30 minutes............................1 30 minutes to 1 hour............................2 1 to 2 hours...........................................3 More than 2 hours................................4 DNK.......................................................8 32 QUESTION CODE/RESPONSE SKIP D19b How long would it take to walk to ______________ (Upazilla/thana headquarter)? Less than 30 minutes............................1 30 minutes to 1 hour............................2 1 to 2 hours...........................................3 More than 2 hours................................4 DNK.......................................................8 D19c Did anyone in your household buy any food in the last year? Yes .......................................................1 No .......................................................2 DNK.......................................................8 2,8→ D19e D19d How long does it take to walk to a place to buy food? Less than 30 minutes............................1 30 minutes to 1 hour............................2 1 to 2 hours...........................................3 More than 2 hours................................4 DNK.......................................................8 D19e Some people have their own businesses making things to sell like baskets, rugs or furniture. Does anyone in your household do this? Yes ........................................................1 No .........................................................2 2 → D19g D19f How long does it take to walk to the place to sell these things? Less than 30 minutes............................1 30 minutes to 1 hour............................2 1 to 2 hours...........................................3 More than 2 hours................................4 Sell at home (0 minutes).......................5 DNK.......................................................8 D19g Do you ever sell food that you grow? Yes ........................................................1 No .......................................................2 Don’t grow food .................................3 Refused to respond ..............................8 2,3,8→D19i D19h How long does it take to walk to the place to sell the food, for example to a market or to a buyer pick-up location? Less than 30 minutes............................1 30 minutes to 1 hour............................2 1 to 2 hours...........................................3 More than 2 hours................................4 DNK.......................................................8 D19i Do you ever buy inputs for crop production like seeds and fertilizer? Yes ........................................................1 No .........................................................2 DNK.......................................................8 2,8→D20 33 QUESTION CODE/RESPONSE SKIP D19j How long does it take to walk to the nearest place to buy inputs such as seeds and fertilizer? Less than 30 minutes............................1 30 minutes to 1 hour............................2 1 to 2 hours...........................................3 More than 2 hours................................4 DNK.......................................................8 34 Household loans QUESTION CODE/RESPONSE SKIP D20 How many loans does your household currently have? [INSTRUCTION: Write total loans] 00 = No loans 99 = Refuse to respond 00, 99 → D27 Details by loan Loan # Sex of the household member who took out the loan 1 = Male 2 = Female What was the source of the loan? What was the main reason for taking out the loan? Total amount borrowed (Taka) Amount of loan still outstanding (Taka) Rate of interest paid/agreed upon (%) D21 D22 D23 D24 D25 D26 1 2 3 4 5 35 QUESTION CODE/RESPONSE SKIP Codes for D22: Source of loan Money lender/pawnshop.................. 1 Bank/formal lending institution ........ 2 Informal savings group...................... 3 Neighbor/friend/relative................... 4 NGO/CBO .......................................... 5 Trader/grocer.................................... 6 Other................................................. 7 DNK ................................................... 8 Refused to respond ........................... 9 Codes for D23: Main reason of taking out the loan Purchase agricultural tools................................. 1 Purchase agricultural inputs............................... 2 Land purchase .................................................... 3 Livestock purchase ............................................. 4 Purchase of other productive assets.................. 5 Purchase of non-productive assets .................... 6 Consumption (food, clothes, etc.) ...................... 7 Pay for treatment/medicine............................... 8 Education expenses............................................ 9 Housing/repairing (including housing tax)...................................................... 10 Wedding ........................................... 11 Bride price/Dowry ............................ 12 Funeral ............................................. 13 Religious event ................................. 14 Loan repayment ............................... 15 Legal dispute/expenses.................... 16 Migration.......................................... 17 Rental of house / shop ..................... 18 Starting small business..................... 19 Land/Home Mortgage ...................... 20 Other ................................................ 21 DNK................................................... 88 Household Savings QUESTION CODE/RESPONSE SKIP D27 Does any member of your household have any cash savings (money put aside for some future use)? Yes.......................................................... 1 No........................................................... 2 DNK 8 Refused to respond................................9 2,8,9→E1 Savings #1 D28.1 Sex of person saving Name:_______________________________ _ Roster Line Number:_______________ Male.................................................. 1 Female ............................................. 2 All household.................................... 3 D29.1 Main method of saving used Bank .................................................1 Savings Scheme/ Coops..................2 Savings or credit group ....................3 Post Office........................................4 36 QUESTION CODE/RESPONSE SKIP Home................................................5 Insurance company..........................6 NGO .......................................................7 Other (specify) .......................................8 DNK 9 D30.1 Total amount of Taka in savings Taka……….... D31.1 Reasons for saving [MULTIPLE RESPONSES] To buy household goods..................1 To start/help busines........................2 To buy land/house............................3 For education/training ......................4 For marriage ....................................5 To build/repair house .......................6 To pay interest from lending.............7 Difficult times....................................8 To meet medical expenses ..............9 To replace lost assets ....................10 To purchase large asset.................11 To meet children’s needs ...............12 To meet all members needs...........13 Other ..............................................14 DNK................................................98 Savings #2 [INSTRUCTION: If household has only one form of savings skip additional savings and skip to E1] D28.2 Sex of person saving Name:_______________________________ _ Roster Line Number:_______________ Male..................................................1 Female .............................................2 All household....................................3 37 QUESTION CODE/RESPONSE SKIP D29.2 Main method of saving used Bank .................................................1 Savings Scheme/ Coops..................2 Savings or credit group ....................3 Post Offices......................................4 Home................................................5 Insurance company..........................6 NGO .......................................................7 Other (specify) .......................................8 DNK ........................................................9 D30.2 Total amount of Taka in savings Taka……….... D31.2 Reasons for saving [MULTIPLE RESPONSES] To buy household goods..................1 To start/help busines........................2 To buy land/house............................3 For education/training ......................4 For marriage ....................................5 To build/repair house .......................6 To pay interest from lending.............7 Difficult times....................................8 To meet medical expenses ..............9 To replace lost assets ....................10 To purchase large asset.................11 To meet children’s needs ...............12 To meet all members needs...........13 Other ..............................................14 DNK................................................98 Savings #3 [INSTRUCTION: If household has only two forms of savings skip to E1] D28.3 Sex of person saving Name:_______________________________ _ Male.................................................. 1 Female ............................................. 2 All household.................................... 3 38 QUESTION CODE/RESPONSE SKIP Roster Line Number:_______________ D29.3 Main method of saving used Bank .................................................1 Savings Scheme/ Coops..................2 Savings or credit group ....................3 Post Offices......................................4 Home................................................5 Insurance company..........................6 NGO .......................................................7 Other (specify) .......................................8 DNK.................................................. 9 D30.3 Total amount of Taka in savings Taka……………… D31.3 Reasons for saving? [MULTIPLE RESPONSES] To buy household goods.................. 1 To start/help busines........................ 2 To buy land/house............................ 3 For education/training ...................... 4 For marriage .................................... 5 To build/repair house ....................... 6 To pay interest from lending............. 7 Difficult times.................................... 8 To meet medical expenses .............. 9 To replace lost assets .................... 10 To purchase large asset................. 11 To meet children’s needs ............... 12 To meet all members needs........... 13 Other .............................................. 14 DNK................................................ 98 39 MODULE E. ACCESS TO SOCIAL SERVICES AND COMMON PROPERTY RESOURCES Access to and use of social services Which of the following services are available in your village/union? [INSTRUCTION: List each service one-by-one and record answers. If answer is “yes” for either a or b, ask question on utilization (c) and circle appropriate code. Respondent for this module is Head of Household or Knowledgeable Adult] Type of Service Available in village Available in union Utilization Yes................................1 No ................................2 DNK ..............................3 Yes ............................... 1 No ................................ 2 DNK.............................. 3 Frequently .....................1 Sometimes.....................2 Never.............................3 40 [If 2,3 in both A and B, SKIP C and go to Next] DNK ...............................8 NR/NO...........................9 A B C E1 Primary health care services E2 Family planning services E3 Primary school E4 Pre-school E5 Union Parishad E6 GrammoShalish Services from the Government provided by the … E7 Department of Social Services E8 Department of Women’s Affairs E9 Department of Agriculture Extension (DAE) E10 Department of Fisheries (DOF) E11 Department of Livestock (DOL) E12 Government Land Office 41 E13 BADC seed department E14 Department of Youth Development E15 Department of Cooperatives E16 Government Family Planning E17 Government immunization services Community clinic E17X Community Clinic IF E17X_A and E17X_B both 2 or 3 SKIP to E18 QUESTION CODE/RESPONSE SKIP E17X.1 How many days per week does the community clinic in your locality provide services? DNK = 9 E17X.2 Did you (any household member) receive any service from the community clinic in last 12 months? Yes..................................................................... 1 No...................................................................... 2 DNK.................................................................... 8 2,3→E18 E17X3 What type of services did you/your household members receive in last 12 months? Antenatal Care services (ANC)........................... 1 Post-natal Care services (PNC) .......................... 2 Newborn care services...................................... 3 Measuring child weights/GMP ......................... 4 42 QUESTION CODE/RESPONSE SKIP [MULTIPLE RESPONSE] For diarrhea/cough/cold ................................... 5 For immunization .............................................. 6 Family planning services.................................... 7 For newly married couple registration.............. 8 For pregnant women registration ..................... 9 For birth/death registration ............................ 10 Nutrition education......................................... 11 Iron table, vitamin-A, deworming tablet......... 12 Health and family planning education and counseling ....................................................... 13 Identification of infectious disease (TB, Malaria etc.) 14 Referrals.......................................................... 15 Others (Specify)............................................... 16 DNK 98 E17X4 How do you rate the services that you or another household member received from the community clinic? Highly satisfied .................................................. 1 Somewhat satisfied ........................................... 2 Neutral............................................................... 3 Not very satisfied............................................... 4 Not at all satisfied.............................................. 5 DNK 9 Participation in social safety nets QUESTION CODE/RESPONSE SKIP 43 E18 In the last 12 months did your household receive assistance from any of the following programs? [INSTRUCTION: Read all programs] Yes=1, No=2 1. Government VGD ................................. 2. Government VGF.................................. 3. Govt. cash-for-work.............................. 4. “100 days work..................................... 5. “40 days work....................................... 6. Aged allowance .................................... 7. Widow allowance ................................. 8. Disability allowance.............................. 9. Maternal allowance.............................. 10. Non-Govt cash-for- work...................... 11. Non-Govt food-for-work ...................... 12. Community based savingsgroup .......... 13. Self-help group ..................................... 14. SHOUHARDO II ..................................... 15. Other .................................................... 44 Access to and use of common property resources [INSTRUCTION: Read each item one-by-one and fill in response codes in column a. Next, for all items with response “yes”, fill in response code for column b. For all items with response “yes” in column b, fill in codes in column c. If “no” then ask next item.] E19. Which of the following common properties are available and used by members of your household? 45 Available Yes ................... 1 No .................... 2 DNK.................. 8 Utilized Yes..................... 1 No...................... 2 DNK ................... 8 Activities (see codes) A B C E19.1 Roadside sloping 2,9→ E19.2 2,9→ E19.2 E19.2 Embankments 2,9→ E19.3 2,9→ E19.3 E19.3 Railway grounds 2,9→ E19.4 2,9→ E19.4 E19.4 Beel/Haor 2,9→ E19.5 2,9→ E19.5 E19.5 River/Canal 2,9→ E19.6 2,9→ E19.6 E19.6 CBO water body 2,9→ E19.7 2,9→ E19.7 E19.7 Grazing land 2,9→ E19.8 2,9→ E19.8 E19.8 Forest land 2,9→ E19.9 2,9→ E19.9 E19.9 Hills 2,9→ E19.10 2,9→ E19.10 E19.10 Khas pond 2,9→ E19.11 2,9→ E19.11 E19.11 Khas land 2,9→ E19.12 2,9→ E19.12 46 E19.12 Other____________________ 2,9→ E19.13 2,9→ E19.13 E19.13 Other ___________________ 2,9→ E19.14 2,9→ E19.14 Codes for column C Fishing ..................................................................... 1 Collecting aquatic animals...................................... 2 Collecting aquatic foods.......................................... 3 Irrigation ................................................................. 4 Grazing .................................................................... 5 Collecting fruit ........................................................ 6 Collecting firewood................................................. 7 Collecting soil ......................................................... 8 Collecting water..................................................... 9 Crop cultivation.................................................... 10 Fish culture........................................................... 11 Tree plantation .................................................... 12 Other.................................................................... 13 MODULE F. DISASTER RISK MANAGEMENT AND CLIMATE CHANGE (Instruction: Respondent is the Head of Household or Knowledgeable Adult) QUESTION CODE/RESPONSE SKIP Disaster risk management F1 In the last 12 months, what type(s) of natural disasters were experienced by your household? [MULTIPLE RESPONSE] A. Heavy rains.............................................01 B. Wildfire ..................................................02 C. Hurricane................................................03 D. Wind storms...........................................04 E. Erosion (river, wind)...............................05 47 QUESTION CODE/RESPONSE SKIP [DO NOT READ THE ANSWERS] F. Earthquake .............................................06 G. Cyclone...................................................07 H. Floods.....................................................08 I. Cold wave ...............................................09 J. Hail .........................................................10 K. Draught...................................................11 L. Other ......................................................12 M. None.......................................................13 N. DNK.........................................................98 F2 What was the most recent natural disaster your household experienced? [SINGLE RESPONSE] [DO NOT READ THE ANSWERS] Heavy rains.............................................01 Wildfire ..................................................02 Hurricane................................................03 Wind storms...........................................04 Erosion (river, wind)...............................05 Earthquake .............................................06 Cyclone...................................................07 Floods.....................................................08 Cold wave ...............................................09 Hail .........................................................10 Draught...................................................11 Other ......................................................12 None.......................................................13 DNK.........................................................98 13,98→F15 F3 In what year did you experience this disaster? Year................................ F4 How did the most recent disaster affect your household? [MULTIPLE RESPONSE] [DO NOT READ THE ANSWERS] A. Loss of family member...........................01 B. Loss of livelihood....................................02 C. Loss of home ..........................................03 D. Physical disability/injury.........................04 E. Loss of assets..........................................05 F. Loss of Latrine/Drainage ........................06 G. Loss of water supply...............................07 H. Disease ...................................................08 I. Crop Loss................................................09 J. Having to care for others........................10 48 QUESTION CODE/RESPONSE SKIP K. Additional household members.............11 L. Stress/anxiety/fear.................................12 M. Other ......................................................13 N. No effect.................................................14 O. DNK.........................................................98 14,98→F6 F5 How did your household cope with the most recent disaster? [MULTIPLE RESPONSE] [DO NOT READ THE ANSWERS] A. Loan from neighbors/relatives ................01 B. Loan from money lender .........................02 C. Loan from NGO........................................03 D. Loan form bank........................................04 E. Reduced # or quantity of meals...............05 F. Mortgaged farmland out.........................06 G. Leased farmland out................................07 H. Sold HH productive assets (tools, livestock, vehicles, etc.)...........................................08 I. Sold other household assets (furniture, radios, jewelry, etc.) ................................09 J. Sold agricultural products in advance or low price................................10 K. Sold advance male labor..........................11 L. Sold advance female labor ......................12 M. Sold farmland...........................................13 N. Sold homestead land ...............................14 O. Ate famine foods .....................................15 P. Accepted aid ............................................16 Q. Accepted help from others......................17 R. Migrated ..................................................18 S. Used savings ............................................19 T. Purchased goods on credit ......................20 U. Postpone medical treatment...................21 V. Sent child to work....................................22 W.Other........................................................23 X. No Coping ................................................24 Y. DNK..........................................................98 49 QUESTION CODE/RESPONSE SKIP F6 What could have been done differently to reduce the impact of future disasters in your community? [MULTIPLE RESPONSE] [DO NOT READ THE ANSWERS] A. Structural improvement to home .........01 B. Improvement to infrastructure (shelters, roads, bridges) ......................................02 C. Community disaster response plan.......03 D. Food stocks ...........................................04 E. Water stocks .........................................05 F. Medical supplies stocks.........................06 G. First aid training ....................................07 H. Increased collaboration/ coordination w/ neighbors ...................08 I. Increased collaboration/ coordination with communities............09 J. Better forecasting .................................10 K. Earlier/better warning .........................11 L. Increased collaboration/ coordination w/ local govt....................12 50 QUESTION CODE/RESPONSE SKIP M. Income alternatives/more diversified income.................................13 N. Evacuation routes/plans .......................14 O. Improved alternative modes of communication (i.e. shortwave radio, etc.).............................................15 P. Other.....................................................16 Q. Nothing .................................................17 R. DNK .......................................................98 F7 Did you receive any early warning signal/message before the last natural disaster (you had in your area)? Yes..................................................................1 No...................................................................2 2F10 F8 How long before the disaster did you receive the warning signal message? minutes hours [Don’t know “000”] F9 Who gave the early/signal message? [MULTIPLE RESPONSE] A. CPP volunteers......................................01 B. Radio .....................................................02 C. Television ..............................................03 D. Union parishad......................................04 E. VDC........................................................05 F. Disaster management committee ........06 G. Disaster volunteers...............................07 H. NGOs.....................................................08 I. Mosque miking......................................09 J. Neighbor/relatives................................10 K. Other (Specify) ......................................11 F10 Did you/your household members move to Yes..................................................................1 1F12 51 QUESTION CODE/RESPONSE SKIP another place to take shelter before/after the last natural disaster? No...................................................................2 F11 If no, why not? [MULTIPLE RESPONSE] A. No shelter..............................................01 B. No space available in the shelter..........02 C. Shelter not functional ...........................03 D. Did not receive messages .....................04 E. No transport..........................................05 F. Did not want .........................................06 G. To protect home/assets........................07 H. Live in protected house.........................08 I. Not required..........................................09 J. Others ...................................................10 K. DNK .......................................................98 AnyF15 F12 Where did you move to take shelter before/after the last natural disaster? [MULTIPLE RESPONSE] A. ‘Pacca’ House (cement).......................01 B. ‘Kacha’ house ........................................02 C. Cyclone or flood shelter ......................03 D. Union parishad building........................04 E. School/institution building....................05 F. Boat.......................................................06 G. Highways/ Embankment.......................07 H. Raised hillock ........................................08 I. Mosque/Temple/Church.......................09 J. Market place .........................................10 K. Other (SPECIFY).....................................11 F13 Did anybody help you to take shelter? Yes..................................................................1 No...................................................................2 DNK ................................................................8 2,8 →F15 F14 Who did help you out to take shelter? A. CPP volunteers........................................1 B. Union parishad........................................2 C. VDC..........................................................3 D. Disaster management committee ..........4 E. Disaster volunteers.................................5 52 QUESTION CODE/RESPONSE SKIP [MULTIPLE RESPONSE] F. NGOs.......................................................6 G. Neighbor/relatives..................................7 H. Other (Specify) ........................................8 I. DNK .......................................................98 F15 Are you aware of any members of the community trained to help you during disaster? Yes..................................................................1 No...................................................................2 2F17 F16 Who are they? [MULTIPLE RESPONSE] A. CPP volunteers......................................01 B. Union parishad chairman/member.......02 C. NGOs.....................................................03 D. Teacher .................................................04 E. Students................................................05 F. Village leaders.......................................06 G. Village Development Committee..........07 H. Disaster volunteers...............................08 I. Other (specify) ......................................09 F17 Have you or any member of your HH received any disaster preparedness training? Yes..................................................................1 No...................................................................2 2F19 F18 Who provided the training? [MULTIPLE RESPONSE] A. CPP volunteers...................................... 01 B. Union parishad chairman/member...... 02 C. NGOs..................................................... 03 D. Teacher................................................. 04 E. Students................................................ 05 F. Village leaders ...................................... 06 G. Village Development Committee ......... 07 H. Disaster volunteers............................... 08 I. Other (specify)...................................... 09 53 F19 What do you plan to with your household members in the event of a disaster (cyclone/flood/other natural)? [MULTIPLE RESPONSE] A. Evacuation of vulnerable members ........ 01 B. Visit shelter centers in normal time ........ 02 C. Identify safe shelter center ..................... 03 D. Plan for dry food ..................................... 04 E. Plan to protect HH valuables/assets ....... 05 F. Identify safe shelter for livestock ............ 06 G. Other (Specify) ........................................ 07 H. No plan ................................................... 08 I. Don’t know.............................................. 98 F20 Do you consider disaster risk as an issue of your daily activities? Yes.................................................................. 1 No .................................................................. 2 DNK................................................................ 8 2,8 →F22 F21 If yes, in which activities? [MULTIPLE RESPONSE] A. Cultivation - time, seed selection ... 1 B. House construction/repair............. 2 C. food preservation ......................... 3 D. Other(Specify) .............................. 4 Climate change F22 Do you think the climate is changing in your area? Yes.................................................................. 1 No .................................................................. 2 No opinion/DNK............................................. 8 2,8 →G1 F23 If yes, in what ways do you think it is changing? A. It is becoming warmer............................ 1 B. It is becoming colder .............................. 2 C. It is becoming dryer................................ 3 D. It is becoming wetter.............................. 4 E. Rains are more unpredictable ................ 5 F. Rains are coming earlier......................... 6 54 [MULTIPLE RESPONSE] [DO NOT READ THE ANSWERS] G. Rains are beginning later........................ 7 H. Rains are stopping earlier....................... 8 I. Rains are stopping later.......................... 9 J. Temperatures are unpredictable.......... 10 K. Strong winds are more common .......... 11 L. Other (Specify)...................................... 12 MODULE G. HOUSEHOLD FOOD SECURITY Food consumption [INSTRUCTION: The respondent should be an adult female knowledgeable about food in the household. If currently administering survey to male head of household, continue with module H, and return to complete this module once module J2 is completed] [INTRODUCTION: Now I would like to ask you about the types of foods that you or anyone else in your household ate yesterday during the day or at night. Please include all foods, including the foods eaten here at your house or somewhere else (e.g., other homes, street stalls, given by employer.] Read the list of foods one-by-one and record coded response. QUESTION CODE/RESPONSE SKIP G1 Any cereals (e.g. rice, bread, wheat, wheat bread, rice flakes, puffed rice, barley, wheat grain, popcorn)? Yes.............................1 No..............................2 G2 Any pumpkin, carrots, squash, or sweet potatoes or vegetables that are yellow or orange inside? Yes.............................1 No..............................2 G3 Any white potatoes, white yams or other foods made from roots and tubers? Yes.............................1 No..............................2 55 QUESTION CODE/RESPONSE SKIP G4 Any dark green, leafy vegetables, (e.g., ipomoea, amaranth, spinach, parwar sag, and drumstick leaves)? Yes.............................1 No..............................2 G5 Any other vegetables, (e.g. cucumber, radish, pepper, string beans, cabbage, cauliflower, radish, onion)? Yes.............................1 No..............................2 G6 Any ripe papaya, mangoes or other fruits that are yellow or orange inside? Yes.............................1 No..............................2 G7 Any other fruits, (e.g. banana, papaya, sithphal, grapefruit, apple, orange, jackfruit, jambu fruit, plums, melon, tomato, date, lemon, etc.)? Yes.............................1 No..............................2 G8 Any meat, such as, liver, beef, poultry, lamb, pork, etc.? Yes.............................1 No..............................2 G9 Any Egg? Yes.............................1 No..............................2 G10 Any fresh or dried fish or shellfish? Yes.............................1 No..............................2 G11 Any legumes/pulses, (e.g. Bengal gram, black gram dal, lentil, Khesarl)? Yes.............................1 No..............................2 G12 Any Milk or Milk products, (e.g. cow milk, buffalo milk, goat milk, yogurt, curd, cheese)? Yes.............................1 No..............................2 G13 Any foods prepared using fat,, e.g., oil, butter, dalda or ghee? Yes.............................1 No..............................2 56 QUESTION CODE/RESPONSE SKIP G14 Any sweet foods? e.g. sugar, gur, honey, Chocolate? Yes.............................1 No..............................2 G15 Any other foods such as condiments, coffee, tea? Yes.............................1 No..............................2 Months of Insufficient Food [INSTRUCTION: the respondent should be an adult female if possible.] QUESTION CODE/RESPONSE SKIP G15.1 Now I would like to ask you about your household’s food supply during different months of the year. When answering these questions, please think back over the last 12 months, from now to the same time last year.Were there months, in the past 12 months, in which you did not have enough food to meet your family’s needs? Yes.......................................................1 No........................................................2 2→G17.1 Which were the months in the past 12 months in which you did not have enough food to meet your family’s needs?This includes any kind of food, such as food you produced yourself, food purchased, food given to you by others, food aid, or food you borrowed. INSTRUCTION:  Do not read the list of months  Place a “1” in the box if the respondent mentions the month. If the respondent does not mention the month, place a “2” in the box. 57 QUESTION CODE/RESPONSE SKIP G16.1 January G16.2 February G16.3 March G16.4 April G16.5 May G16.6 June G16.7 July G16.8 August G16.9 September G16.10 October G16.11 November G16.12 December Household Hunger QUESTION CODE/RESPONSE SKIP G17.1 In the last 4 weeks was there a time when there was no food to eat of any kind in the house, Yes..........................................................1 58 because of lack of resources to get food? No...........................................................2 2→G17.3 G17.2 How often did this happen? Rarelyor sometimes (1-10 times)...........1 Often (more than 10 times) ...................2 G17.3 In the last 4 weeks, was there a time when you or any household member went to sleep at night hungry without eating anything at all because there was not enough food? Yes..........................................................1 No...........................................................2 2→G17.5 G17.4 How often did this happen? Rarely or sometimes (1-10 times)..........1 Often (more than 10 times) ...................2 G17.5 In the last 4 weeks was there a time when you or any household member went a whole day and night without eating anything at all because there was not enough food? Yes..........................................................1 No...........................................................2 2→G18.1 G17.6 How often did this happen? Rarely or sometimes (1 – 10 times) .......1 Often (more than 10 times) ...................2 59 Household Food Access: Food insecurity coping strategies [INSTRUCTION: read each question and then ask how often the event happened in the last year.] QUESTION CODE/RESPONSE SKIP G18.1 In the last year, how often did you or any of your family have to eat potato, wheat, or another grain although you wanted to eat rice (not including when you were sick)? Never......................................................1 Rarely (1-6 times in last year).................2 Sometimes (7-12 times in last year).......3 Often (a few times each month) ............4 Regularly (almost or every day) .............5 G18.2 In the last 12 months how often did you yourself skip entire meals due to scarcity of food? Never......................................................1 Rarely (1-6 times in last year).................2 Sometimes (7-12 times in last year).......3 Often (a few times each month) ............4 Regularly (almost or every day) .............5 G18.3 In the last 12 months how often did you personally eat less food in a meal due to scarcity of food? Never......................................................1 Rarely (1-6 times in last year).................2 Sometimes (7-12 times in last year).......3 Often (a few times each month) ............4 Regularly (almost or every day) .............5 G18.4 In the past 12 months how often did your family purchase food (rice, lentils etc.) on credit (or loan) Never......................................................1 60 from a local shop? Rarely (1-6 times in last year).................2 Sometimes (7-12 times in last year).......3 Often (a few times each month) ............4 Regularly (almost or every day) .............5 G18.5 In the past 12 months how often did your family have to borrow /take food from relatives or neighbors to make a meal? Never......................................................1 Rarely (1-6 times in last year).................2 Sometimes (7-12 times in last year).......3 Often (a few times each month) ............4 Regularly (almost or every day) .............5 61 MODULE H. AGRICULTURAL PRODUCTION, FISHERIESAND LIVESTOCK REARING [Instructions: Respondent should be Head of Household or Knowledgeable Adult] Field crop production QUESTION CODE/RESPONSE SKIP H1 Did you cultivate any field crops like cereals, ground nuts, jute, or fruits and vegetables for selling to others? Yes.............................................................1 No ..............................................................2 DNK ...........................................................8 2,8→ H8 H2 In the last year did anyone in your household cultivate any of these crops? Example: [READ THE CROPS THAT ARE LISTED] Yes=1No=2 H.2.1. Rice (HYV)................................... H.2.2. Rice (LIV)..................................... H.2.3. Rice (Local) ................................. H.2.4. Rice (Hybrid) ............................... H.2.5. Vegetables (commercial) ............ H.2.6. Fruits (commercial)...................... H.2.7. Wheat .......................................... H.2.8. Groundnut.................................... H.2.9. Maize ........................................... H.2.10.Pulses .......................................... H.2.11.Oilseeds....................................... H.2.12.Spices .......................................... H.2.13. Jute.............................................. H.2.14.Tobacco....................................... H.2.15.Other (Specify) ............................ H.2.16.Other (Specify) ............................ 62 H3 Which improved cropping practices did you apply to any of your crops mentioned in H2 in the last year? Yes=1 No=2 H.3.1. Use high quality seed ................. H.3.2. Use 2-3 seedling per hill for rice. H.3.3. Maintain proper spacing ............. H.3.4. Intercrop ..................................... H.3.5. Use IPM...................................... H.3.6. Use organic fertilizers ................. H.3.7. Use recommended seed storage Methods ..................................... H.3.8. Balanced fertilizer use ................ H.3.9. Green manure ............................ H.3.10.Other (Specify) ........................... H4 Which agricultural inputs did you purchase and/or receive before or during the last cropping season? Yes=1 No=2 H.4.1. Improved seed H.4.2. Seedlings........ H.4.3. Saplings.......... H.4.4. Irrigation water H.4.5. Fertilizer.......... H.4.6. Ploughing........ H.4.7. Use of pesticides ........................ H.4.8. Use of weedicides...... H.4.9. Other (Specify) H.4.10.Other (Specify) 63 H5 Which agricultural financial services did you or your household use in the previous cropping season? Yes=1No=2 H.5.1.Agricultural loan................................. H.5.2.A company provided advance inputs H.5.3.Government subsidy.......................... H.5.4.Other (Specify)................................... H6 Have you or any member of your household participated in any training programs on improved food crop production technologies? Yes................................................................1 No .................................................................2 DNK ..............................................................8 2,8→H8 H7 What kind of agricultural techniques were you trained in? [MULTIPLE RESPONSES] A. Use of improved seeds..................................... 1 B. Use of new food crops...................................... 2 C. Proper use of fertilizer...................................... 3 D. Weed control (herbicides, weeding) ................ 4 E. Conservation agriculture (zero /minimal tillage, composting) ................. 5 F. Pest management practices (pesticides).......... 6 G. Improved post-harvest techniques .................. 7 H. Improved water management ......................... 8 I. Others (Specify)................................................ 9 Vegetable Production/Gardening 64 [INSTRUCTION: Ask to the person who normally does gardening if possible.] QUESTION CODE/RESPONSE SKIP H9 In the previous year did any member of your household grow any vegetables in a garden? Yes................................................................1 No .................................................................2 DNK ..............................................................8 2,8 →H12 H10 Which of the following vegetables did you grow? Yes=1 No=2 H.10.1. Green Gourd ................................... H.10.2. Radish.............................................. H.10.3. Birinjal ............................................. H.10.4. Lalshak (red amaranth)................... H.10.5. Pumpkin (yellow) ............................ H.10.6. Corriandor leaf /Kalijira/Ginger ...... H.10.7. Potato/Keshur (white potato)......... H.10.8. Data Shak (Green amaranth) .......... H.10.9. Potol/Shajina (drum stick) .............. H.10.10. Chichinga/Jhinga ............................. H.10.11. Bean ................................................ H.10.12. Pui Shak (Indian spinach) ................ H.10.13. Kacha Kola (green banana) ............. H.10.14. Ladies Finger ................................... H.10.15. Green Chili....................................... H.10.16. Onion .............................................. H.10.17. Garlic ............................................... H.10.18. Sweet potato................................... H.10.19. Tomato............................................ H.10.20. Korolla (bitter gourd) ...................... H.10.21. Papaya............................................. 65 H.10.22. Other green leafy vegetables..... H.10.23. Other (Specify)........................... H11 Which improved practices did you apply to any of your vegetable crops in the last year? Yes=1 No=2 H.11.1. Improved bed system........................ H.11.2. Improved pit/heap system ................ H.11.3. Quality seed....................................... H.11.4. Organic fertilizer................................ H.11.5. Compost preparation......................... H.11.6. Multi storied cropping ....................... H.11.7. Relay cropping ................................... H.11.8. Multiple cropping .............................. H.11.9. Thinning ............................................. H.11.10. Pruning............................................... H.11.11. Mulching ............................................ H.11.12. Bagging .............................................. H.11.13. Stalking/sticking/trellis...................... H.11.14. Non-chemical pesticides.................... H.11.15. Artificial pollination ........................... H.11.16. Weed management........................... H.11.17. Water management........................... H.11.18. Other (Specify)................................... 66 Fish Production/Rearing [INSTRUCTION: Ask person normally engaged in fish production if possible.] QUESTION CODE/RESPONSE SKIP H12 In the last year did you or your household raise/rear any fish? Yes................................................................. 1 No .................................................................. 2 DNK ............................................................... 8 2,8→H14 H13 Which improved fish production practices did your household use in the last year? Yes=1 No=2 H.13.1. Testing water color to determine if food is adequate................................ H.13.2. Maintaining stocking density............. H.13.3. Species selection................................ H.13.4. Pond cleaning .................................... H.13.5. Liming ................................................ H.13.6. Providing supplementary feed .......... H.13.7. Employing fish disease management ..................................... H.13.8. Using polyculture............................... H.13.9. Providing fish seed............................. H.13.10. Others (Specify) ................................. Livestock Production/Rearing [INSTRUCTION: Ask person normally engaged in livestock production if possible.] 67 QUESTION CODE/RESPONSE SKIP H14 During the last 12 months, did you raise any livestock or poultry? Yes raised livestock....................................... 1 Yes raised poultry.......................................... 2 Yes raised both.............................................. 3 No .................................................................. 4 DNK ............................................................... 8 4,8→H16 H15 What are the following improved practices do you apply in the last year to raising poultry and rearing livestock? Yes=1 No=2 H.15.1. Improved breeding ............................ H.15.2. Vaccination ........................................ H.15.3. Supplementary poultry feed.............. H.15.4. Fattening............................................ H.15.5. Artificial insemination.................... H.15.6. Supplementary poultry feed.............. H.15.7. Other (Specify) ............................. Technical support QUESTION CODE/RESPONSE SKIP H16 Do you know where to get technical guidance for agriculture, livestock rearing, gardening, or pond/fish management? Yes................................................................. 1 No .................................................................. 2 DNK ............................................................... 8 2,8  I1 H17 In the past year did you or any member of your household receive any type of assistance (technical, materials, financial) from any of the following sources? Yes=1 No=2 H.17.1. Neighbors/relatives/other farmers ..... H.17.2. Dept. of Agriculture ............................. H.17.3. Dept. of Fisheries................................. 68 H.17.4. Department of livestock ...................... H.17.5. NGO ..................................................... H.17.6. Seed/pesticide companies................... H.17.7. Fish/poultry/livestock feed and pharmaceutical companies.................. H.17.8. PDF (Positive Deviant Farmer)............. H.17.9. Paravet................................................. H.17.10. Community Agriculture Volunteers..... H.17.11. Other (Specify) .............................. MODULE I. WATER AND SANITATION [Instructions: Respondent should be Head of Household or Knowledgeable Adult] QUESTION CODE/RESPONSE SKIP 69 QUESTION CODE/RESPONSE SKIP I1 What is the main source of drinking water for members of your household? [PROMPT IF NECESSARY] Hand tube well ......................................................1 Tara pump .............................................................2 Deep tube well ......................................................3 Shallow tube well ..................................................4 Ring well/ indara....................................................5 Pond ......................................................................6 River/canal.............................................................7 Piped water ...........................................................8 Pond sand filter .....................................................9 Rainwater harvesting system ..............................10 Other (specify) ....................................................11 I2 Is water normally available from this source? Yes .........................................................................1 No ..........................................................................2 I3 In the last two weeks was water unavailable from this source for a day or longer? Yes .........................................................................1 No ..........................................................................2 70 QUESTION CODE/RESPONSE SKIP I4 How much time does it usually take to go to the drinking water source, get water, and come back? 0 (in or at house) ...................................................1 30 minutes or less .................................................2 30 minutes to 1 hour.............................................3 1 hour to 2 hours...................................................4 2 to 3 hours ...........................................................5 More than 3 hours.................................................6 I5 If source is a tubewell (hand, deep, shallow or Tara pump, has the well been tested for arsenic? Yes ........................................................................1 No ..........................................................................2 DNK........................................................................8 N/A ........................................................................9 2,8,9→I8 I6 If tested, does the tubewell/Tara pump have arsenic? Yes .........................................................................1 No ..........................................................................2 DNK........................................................................8 2,8→I8 I7 If yes, is it marked red or green? Red.........................................................................1 Green.....................................................................2 Neither...................................................................3 I8 Does the household have access to a toilet facility? Yes .........................................................................1 No ..........................................................................2 2→J1 71 QUESTION CODE/RESPONSE SKIP I9 What kind of toilet facility do members of your households usually use? [PROMPT IF NECESSARY] Ring-slab/offset latrine (water seal)......................1 Ring-slab/offset latrine (water seal broken) .........2 Pit latrine (covered)...............................................3 Pit latrine (uncovered)...........................................4 Septic latrine..........................................................5 Hanging/open latrine ............................................6 Local adopted hygienic latrine ..............................7 I10 Which members of your household use this toilet? [SELECT MUPLTIPLE] A. Male adults.............................................. 1 B. Female adults.......................................... 2 C. Male children .......................................... 3 D. Female children....................................... 4 E. All ...........................................................5 I10a May I see your latrine? Yes ......................................................................... 1 No .......................................................................... 2 2→J1 [INSTRUCTION: Observe the latrine directly and record condition] I11 Is the latrine functioning? Yes .........................................................................1 No ..........................................................................2 I12 Does the latrine show signs of use? Yes .........................................................................1 No ..........................................................................2 I13 Is the latrine itself clean?For example, is the pan and slab (or place to sit while defecating) clean? Yes .........................................................................1 No ..........................................................................2 72 QUESTION CODE/RESPONSE SKIP I14 Is the surrounding area of the latrine clean? Yes .........................................................................1 No ..........................................................................2 I15 Does the latrine have an unbroken water seal? Good water seal ....................................................1 Broken water seal..................................................2 No water seal.........................................................3 I16 Is there a hand washing station inside the latrine or within 10 paces of the latrine? Yes .........................................................................1 No ..........................................................................2 2→J1 I17 Is there a cleansing agent at this hand washing station? (soap, detergent, ash or clay) Yes .........................................................................1 No ..........................................................................2 MODULE J1.PARTICIPATION IN THE CARE SHOUHARDO II PROGRAM FROM 2010-2014 [INSTRUCTIONS: The respondent to module J1 should be the person in the household most knowledgeable about household’s engagement with SHOUHARDO II activities.] QUESTION CODE/RESPONSE SKIP Agriculture, Comprehensive Homestead Development, Fisheries, and Income Generation Activities Could you please tell me which of the following supports (e.g., training, inputs, technical back-stop) that the 73 QUESTION CODE/RESPONSE SKIP household has received from the SHOUHARDO II Program? Did you receive….. J1.1 a. Agricultural support? Yes...................................................................1 No....................................................................2 2→J1.2 b. Agriculture – field crop (Cereals) Yes...................................................................1 No....................................................................2 DNK .................................................................8 c. Agriculture – field crop (Vegetable) Yes...................................................................1 No....................................................................2 DNK .................................................................8 J1.2 a. Comprehensive Homestead Development (CHD) Yes...................................................................1 No....................................................................2 2→J1.3 b. CHD – Vegetable garden Yes...................................................................1 No....................................................................2 DNK .................................................................8 c. CHD – Fruit production Yes...................................................................1 No....................................................................2 DNK .................................................................8 d. CHD – Goat Yes...................................................................1 No....................................................................2 74 QUESTION CODE/RESPONSE SKIP DNK .................................................................8 e. CHD – Poultry Yes...................................................................1 No....................................................................2 DNK .................................................................8 f. CHD – Other livestock Yes...................................................................1 No....................................................................2 DNK .................................................................8 J1.3 a. Fisheries Yes...................................................................1 No....................................................................2 2→J1.4 b. Fisheries: fish culture Yes...................................................................1 No....................................................................2 DNK .................................................................8 c. Fish – capture/ fishing Yes...................................................................1 No....................................................................2 DNK .................................................................8 J1.4 a. IGA (On-farm/Off-farm) Yes...................................................................1 No....................................................................2 2→J1.5 b. Non agriculture Yes...................................................................1 75 No....................................................................2 DNK .................................................................8 c. Agriculture - seed, technical assistance Yes...................................................................1 No....................................................................2 DNK .................................................................8 d. Goat rearing Yes...................................................................1 No....................................................................2 DNK .................................................................8 e. Cow rearing Yes...................................................................1 No....................................................................2 DNK .................................................................8 f. Poultry/Bird rearing Yes...................................................................1 No....................................................................2 DNK .................................................................8 g. Other livestock rearing Yes...................................................................1 No....................................................................2 DNK .................................................................8 h. EDBM/ ME Value chain Yes...................................................................1 No....................................................................2 DNK .................................................................8 i. Technical Training Yes...................................................................1 76 No....................................................................2 DNK .................................................................8 j. Others(specify if not mentioned above/below) Yes...................................................................1 No....................................................................2 DNK .................................................................8 J1.5 Training/Technical support a. Did anyone in your household receive any training/technical support from the SHOUHARDO II Program? Yes...................................................................1 No....................................................................2 DNK .................................................................8 2,8→J1.6 b. If yes, please list down all those specifically below (max major three) 1. Day ............................................................... DNK/Can’t remember....................................98 2. Day ............................................................... DNK/Can’t remember....................................98 3. Day ............................................................... DNK/Can’t remember....................................98 J1.6 Food ration beneficiary a. Did anyone in your household receive a supplementary food ration from the SHOUHARDO II Program? Yes...................................................................1 No....................................................................2 DNK .................................................................8 2,8→J1.7 77 b. If yes, did she receive it as a pregnant woman or a lactating mother or for both? Pregnant woman.............................................1 Lactating mother.............................................2 Both.................................................................3 DNK .................................................................8 c. Number of months received ration (Check Ration Card if Available) Months.................................................... J1.7 Participation in SHOUHARDO II groups/membership in community group Is anyone in your household a member of the following groups? [INTERVIEWER: Read list one-by-one and record responses] a. Agriculture group Yes...................................................................1 No....................................................................2 DNK .................................................................8 b. Comprehensive Homestead Development (CHD) group Yes...................................................................1 No....................................................................2 DNK .................................................................8 c. Income Generating Activities (IGA) group Yes...................................................................1 No....................................................................2 DNK .................................................................8 d. Fisheries group Yes...................................................................1 78 No....................................................................2 DNK .................................................................8 e. Village Development Committees (VDC) Yes...................................................................1 No....................................................................2 DNK .................................................................8 f. Savings Group Management Committee Yes...................................................................1 No....................................................................2 DNK .................................................................8 g. ECCD Management Committee Yes...................................................................1 No....................................................................2 DNK .................................................................8 h. EKATA Management Committee Yes...................................................................1 No....................................................................2 DNK .................................................................8 i. PIC/ Food Commodity Distribution Management Committee Yes...................................................................1 No....................................................................2 DNK .................................................................8 79 j. Mother’s Group Yes...................................................................1 No....................................................................2 DNK .................................................................8 k. Other (Specify) Yes...................................................................1 No....................................................................2 DNK .................................................................8 J1.8 EKATA a. Are there any women or adolescent girls living in your household? Yes ...................................................................1 No ....................................................................2 2→ J1.9 b. If yes, are they enrolled in or do they participate in a SHOUHARDO II EKATA? Yes ...................................................................1 No ....................................................................2 DNK..................................................................8 2,8,→J1.9 c. If yes, is she is an adult woman or adolescent girl? Adult woman ...................................................1 Adolescent.......................................................2 Both .................................................................3 J1.9 EARLY CHILDHOOD CARE DEVELOPMENT (ECCD) 80 a. Are there any children under 7 years old living in your household? Yes ...................................................................1 No ....................................................................2 DNK..................................................................8 2,8,→J1.10 b. Does any child attend preschool at a SHOUHARDO II Early Childhood Development Center (ECCD) (SBK, preschool)? Yes ...................................................................1 No ....................................................................2 2→J1.10 c. Is the child a boy or girl? Boy=1, Girl=2 Child-1 .......................................................... Child-2 .......................................................... Child-3 .......................................................... J1.10 WATER AND SANITATION a. Did SHOUHARDO II Program help your household get access to a latrine? Yes ...................................................................1 No ....................................................................2 DNK..................................................................8 b. Did SHOUHARDO II Program help your household get access to a tube well? (Do not count tube well platform) Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 c. Did the SHOUHARDO II Program help construct/ upgrade/ repair a sanitation drain near your home? Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 J1.11 OTHERS 81 a. Did your household receive support for house rising from the SHOUHARDO II Program? Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 b. Did your household receive support for mound protection or extension from the SHOUHARDO II Program? Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 c. Did anyone in your household participate in the cash￾for-work program of the SHOUHARDO II Program? Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 d. Did anyone in your household participate in thefood￾for-work program of the SHOUHARDO II Program? Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 e. During the past two years, did anyone in your household directly receive any aid or food from programs other than SHOUHARDO II? Yes ...................................................................1 No ....................................................................2 DNK................................................................. 8 2,8,→J2.1 f. If “yes”, what was the program? [ASK TO SEE ANY RATION CARDS AND LIST SOURCE OF AID] __________________________ MODULE J2.FACTORS AFFECTING PARTICIPATION IN SHOUHARDO II 82 [Instructions: Respondent should be Head of Household or Knowledgeable Adult] Shocks experienced in the last year Now I’m going to ask you some questions about the shocks, or bad surprises, that hurt the household financially in the last year. The first type of shock is climate shocks, like floods. J2.1During the past 12 months did your household experience any of the following events? a. Floods Yes…………………….1 No……………………..2 b. Cyclone Yes…………………….1 No……………………..2 c. Drought Yes…………………….1 No……………………..2 d. Had to move due to river erosion Yes…………………….1 No……………………..2 e.Do you feel that the situation was better, the same or worse 5 years ago (before SHOUHARDO II began) than it is now with regards to these kinds of climate shocks? Better……………….1 Same………………..2 83 Worse………………3 Next I’m going to ask you about family events. During the past 12 months did your household experience any of the following events? f. Loss of income due to illness or injury of a family member Yes…………………….1 No……………………..2 g. Divorce or abandonment Yes…………………….1 No……………………..2 h. Dowry and wedding expenses Yes…………………….1 No……………………..2 i. Court case/legal problems Yes…………………….1 No……………………..2 j. Death of family member Yes…………………….1 No……………………..2 k. Do you feel that the situation was better, the same or worse 5 years ago (before SHOUHARDO II began) than it is now with regards to these kinds of family events? Better……………….1 Same………………..2 Worse………………3 Next I’m going to ask you about economic shocks. During the past 12 months did your household experience any of the following events? l. Sharp food price increase Yes…………………….1 No……………………..2 84 m. Very poor harvest from fields or fish ponds Yes…………………….1 No……………………..2 n Deaths of livestock, including fish Yes…………………….1 No……………………..2 o. Massive loss of household income Yes…………………….1 No……………………..2 p.Do you feel that the situation was better, the same or worse 5 years ago (before SHOUHARDO II began) than it is now with regards to these kinds of economic events? Better……………….1 Same………………..2 Worse………………3 Social capital J2.2 If your household had a problem and needed TK 1000 or food urgently, would you be able to borrow it from relatives living in this community? Yes...................................................................1 No....................................................................2 DNK..................................................................8 J2.3 If your household had a problem and needed TK 1000 or food urgently, would you be able to borrow it from people in your community who are not your relatives? Yes...................................................................1 No....................................................................2 DNK..................................................................8 J2.4 Compared to five years ago (before SHOUHARDO II began) has the number of people you think you could ask for help if your household needed money or food urgently: Increased ........................................................1 Stayed the same ..............................................2 Decreased........................................................8 85 J2.5 If a relative in this community had a problem and needed TK 1000 or food urgently, would you be able to lend it to them? Yes...................................................................1 No....................................................................2 DNK..................................................................8 J2.6 If someone who is not your relative, but lives in this community had a problem and needed Tk 1000 or food urgently, would you be able to lend it to them? Yes...................................................................1 No....................................................................2 DNK..................................................................8 J2.7 Compared to five years ago (before SHOUHARDO II began) has your ability to give this type of assistance: Increased ........................................................1 Stayed the same ..............................................2 Decreased........................................................3 J2.8 Is anyone in your household a friend or relative of a government official in this area? Yes...................................................................1 No....................................................................2 DNK..................................................................8 J2.9 Is anyone in your household a Union Parishad Elected Leader? Yes...................................................................1 No....................................................................2 DNK..................................................................8 J2.10 Is anyone in your household a friend or relative of a Union Parishad elected leader? Yes...................................................................1 No....................................................................2 DNK..................................................................8 Aspirations and confidence to adapt 86 J2.11 Does anyone in your household communicate regularly with at least one person outside this village? Yes .............................. 1 No ............................... 2 DNK............................. 8 J2.12 During the past week, has anyone in your household engaged in economic activities with members of other villages? For example, farming, trading, employment, borrowing or lending money. Yes .............................. 1 No ............................... 2 DNK............................. 8 J2.13 How many times in the past month has anyone in your household got together with people to have food, either in their home or in a public place? [____] J2.14 How many DAYS in the past month has anyone in your household attended a mosque or other religious service? [____] J2.15 In the last year, how many times has anyone in your household stayed more than 2 days outside this village? [____] J2.16: Below is a series of statements that you may agree or disagree with. I’m going to ask you to tell me whether you agree with each one. Do you Agree? a. When I get what I want, it is usually because I worked hard for it. Yes...............................1 No................................2 87 DNK .............................8 b. Some things that happen to me are God’s will and some things are because of my own actions. Yes...............................1 No................................2 DNK .............................8 c. To be successful, above all one needs to be lucky. Yes...............................1 No................................2 DNK .............................8 88 Personal familiarity with SHOUHARDO II project staff QUESTION CODE/RESPONSE SKIP J2.17 For how many years has the SHOUHARDO II project had activities in your community? [____] J2.18 Did you or any member of your household know a staff member of the SHOUHARDO II project before it started having activities in your community? Yes .............................. 1 No ………………….. 2 2→ J2.21 J2.19 How many staff members did you or another household member know before the SHOUHARDO II project started here? [____] J2.20 How did you, or household members, know the staff members at that time? Were they…. [INSTRUCTION: Read all choices and circle answers) [MULTOPLE RESPONSE] 1. A family member or relative 2. A friend 3. A neighbor 4. An acquaintance Time constraints QUESTION CODE/RESPONSE SKIP J2.21 Do you do any agricultural production, gardening, or animal raising? Yes...............................1 No................................2 2→ J2.23 J2.22 On a typical day in the last month, how many hours did you spend on these activities? [____] J2.23 Do you do any post-harvest activities like threshing? Yes...............................1 No…………………… 2 2→ J2.25 J2.24 On a typical day in the last month, how many hours did you spend on post-harvest activities? [____] J2.25 Do you earn any cash income? Yes...............................1 No…………………… 2 ......... 2→ J2.27 J2.26 On a typical day in the last month, how many hours did you spend earning cash income? [____] J2.27 Do you do any housework, like cooking, washing dishes and clothes, fetching wood and water? Yes...............................1 No…………………… 2 2→ J2.29 89 J2.28 On a typical day in the last month, how many hours did you spend on housework? [____] J2.29 Do you take care of any children? Yes...............................1 No................................2 2→ J2.31 J2.30 On a typical day in the last month, how many hours did you spend taking care of children? [____] J2.31 During the daytime do you spend time doing other things like socializing, watching TV, taking naps, or reading? Yes...............................1 No................................2 2→ J2.33 J2.32 On a typical day in the last month, how many hours did you spend in these kinds of “leisure” activities where you were not working? [____] J2.33 Imagine ten steps, where on the bottom, the first step, is a person who spends no time in the day doing these leisure activities, and on the highest step, the 10th, is a person who spends the whole day doing them [INSTRUCTION: Show picture of steps]. On a typical day in the last month, which step were you on? [____] J2.34 Which step were you on five years ago (before SHOUHARDO II began)? [____] 90 PART II. INFORMATION ON WOMEN’S EMPOWERMENT [INSTRUCTION]: The questions in Part II should be asked of an adult woman member of the household without men present. To help find a woman, see circled line numbers from column c7 of household roster. The preferred respondent is the female head of household or spouse of the male head of household.] MODULE K1. RESPONDENT IDENTIFICATION FOR PART II K1.1 Name of respondent: _______________________________________ K1.2 Line number of respondent __________________________________ [INSTRUCTION: Record line number from Module C] K1.3 Relationship to household head [SEE CODES BELOW] Codes for K1.3: 1= Household head, 2= Wife of household head, 3= Daughter, 4=Granddaughter, 5=Niece, 6=Mother, 7= Daughter in law, 8=Sister, 9=Sister-in-law, 10=Brother’s wife. MODULE K2. INFORMATION ON WOMEN’S EMPOWERMENT Decision making in household QUESTION CODE/RESPONSE SKIP 91 QUESTION CODE/RESPONSE SKIP In the last year, to what extent have you been able to make the following kinds of decisions? [INSTRUCTION: First read the possible responses. Then list each item (K2.1 – K2.12) one-by-one and record code number of response.] Code list 1 = Can decide alone 2 = Can decide with husband or other adult male family member 3 = Husband makes decision after discussion with wife 4 = Not involved in decision 5 = Not applicable K2.1 Buying small food items, groceries, toiletries K2.2 Buying clothing for yourself and your children K2.3 Spending money that you yourself have earned K2.4 Buying or selling major household assets (land, livestock, crops) K2.5 Buying or selling jewelry K2.6 Use of loans or savings K2.7 Expenses for your children’s education K2.8 Expenses for your children’s marriage K2.9 Medical expenses for yourself or your children K2.10 Expenses for family planning (contraceptives) K2.11 To move to shelter during time of disaster K2.12 Actively participate and involved in salish decision making 92 QUESTION CODE/RESPONSE SKIP K2.13 Imagine ten steps, where on the bottom, the first step, is a woman who can make none of these types of decisions, and on the highest step, the 10th, is a person who can make all of them. [INSTRUCTION: Show picture of steps]. On a usual day in the last month, which step were you on? [____] K2.14 Which step were you on five years ago (before SHOUHARDO II began)? [____] Freedom of movement [INSTRUCTION: Circle code number of response.] QUESTION CODE/RESPONSE SKIP K2.13 Are you allowed to travel to the local market to buy things Yes.................................................. 1 No................................................... 2 2→ K2.15 K2.14 Can you go alone? Yes.................................................. 1 No................................................... 2 K2.15 Are you allowed to travel to a local health center or doctor Yes.................................................. 1 No................................................... 2 2→ K2.17 K2.16 Can you go alone? Yes.................................................. 1 No................................................... 2 K2.17 Are you allowed to travel to homes of friends in the neighborhood Yes.................................................. 1 No................................................... 2 2→ K2.19 K2.18 Can you go alone? Yes.................................................. 1 No................................................... 2 K2.19 Are you allowed to travel to a nearby mosque/shrine Yes.................................................. 1 No................................................... 2 2→ K2.21 93 QUESTION CODE/RESPONSE SKIP K2.20 Can you go alone? Yes.................................................. 1 No................................................... 2 Earning of Cash Income QUESTION CODE/RESPONSE SKIP K2.21 As you know, some women take up jobs for which they are paid in cash. Others sell things, have a small business or work on the farm or in the family business. In the last 12 months, have you done any of these things? Yes.................................................... 1 No .................................................... 2 2→ K2.23 K2.22 If yes, did you earn any money from your work in the last 12 months? Yes.................................................... 1 No .................................................... 2 Attitudes about family life [INTRODUCTION: Now I would like to get your opinion on some aspects of family life. Please tell me if you agree or disagree with each statement.] QUESTION CODE/RESPONSE SKIP K2.23 The important decisions in the family should be made only by the men of the family. Agree ................................................ 1 Disagree............................................ 2 DNK/depends................................... 8 K2.24 If the wife is working outside the home, then the husband should help her with household chores. Agree ................................................ 1 Disagree............................................ 2 DNK/depends................................... 8 K2.25 A married woman should be allowed to work outside the home if she wants to. Agree ................................................ 1 Disagree............................................ 2 DNK/depends................................... 8 94 K2.26 The wife has a right to express her opinion even when she disagrees with what her husband is saying. Agree ................................................ 1 Disagree............................................ 2 DNK/depends................................... 8 K2.27 A wife should tolerate being beaten by her husband in order to keep the family together. Agree ................................................ 1 Disagree............................................ 2 DNK/depends................................... 8 K2.28 It is better to send a son to school than it is to send a daughter. Agree ................................................ 1 Disagree............................................ 2 DNK/depends................................... 8 Domestic Violence QUESTION CODE/RESPONSE SKIP K2.29 Sometimes a husband is annoyed or angered by things his wife does. In your opinion, is a husband justified in hitting or physically abusing his wife in the following situations? a. If she goes out without telling him Yes.....................................................1 No .....................................................2 DNK ...................................................8 b. If she neglects the children Yes.....................................................1 No .....................................................2 DNK ...................................................8 c. If she argues with him Yes.....................................................1 No .....................................................2 DNK ...................................................8 d. If she refuses to have sex with him Yes.....................................................1 No .....................................................2 DNK ...................................................8 e. If she burns the food Yes.....................................................1 No .....................................................2 DNK ...................................................8 f. If she does not obey elders Yes.....................................................1 No .....................................................2 DNK ...................................................8 K2.30 Did any female member of your household experience being yelled at or struck during the previous year? Yes.....................................................1 No .....................................................2 DNK ...................................................8 Refuse to answer ..............................9 2,8,9→K2.35 95 K2.31 What was the nature of this yelling or striking? Physical ............................................ 1 Verbal............................................... 2 Both physical and verbal.................. 3 K2.32 How often did incidences like this occur? One time only .................................. 1 Several times.................................... 2 Often................................................ 3 K2.33 Was any assistance sought after these incidents? Yes.................................................... 1 No .................................................... 2 2→K2.35 K2.34 Did you get assistance from ….? a. A medical facility? Yes.................................................... 1 b. The police NoYes................................ .................................................... .................... 21 No .................................................... 2 c. A lawyer or legal firm Yes.................................................... 1 d. A relative, friend or neighbor NoYes................................ .................................................... .................... 21 e. A women’s support group NoYes................................ .................................................... .................... 21 f. Other NoYes................................ .................................................... .................... 21 No .................................................... 2 Women’s Participation in Community Groups and Local Institutions Which of the following groups are you a member of or in a responsible position? Example: Chairman, Secretary, Cashier, Committee member. QUESTION CODE/RESPONSE SKIP K2.35 Savings or credit group? Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.36 Community agriculture (including fishery group) or homestead garden group or IGA group? Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.37 Community health group Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.38 Parent-Teacher Association or School Management Committee Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.39 Mother’s Group Member........................................... 1 Committee Member/Officer............ 2 Not a member.................................. 3 96 K2.40 Women’s support group (EKATA, VDC, ECCDMC, MCHN mothers, Women Empowerment, EVAWF Forum, BRAC PalliSamaj, NariPokkhoDorbar network etc.) Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.41 UP Special Committee Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.42 UP Standing Committee Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.43 UP disaster committee Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.44 Other_____________________________ Member........................................... 1 Committee Member/responsible position ............................................ 2 Not a member.................................. 3 K2.45 Have you ever attended a Salish meeting in your village? Yes.................................................... 1 No .................................................... 2 2→Module L K2.46 Did you speak at the meeting? Yes.................................................... 1 No .................................................... 2 97 PART III. INFORMATION ON CHILDREN 0-59 MONTHS OLD AND THEIR MOTHERS MODULE L. RESPONDENT AND CHILD IDENTIFICATION FOR PART III The purpose of this module is to select the index child for administration of Modules M, N, O, P, Q, and R to identify the children to be weighed and measured for Module T (all children in the household under five). INSTRUCTION: List all of the children living in the household who are under 6 years of age, that is, those with circled line numbers in column C8 of Module C. ***If no child is under 6 years, then the survey is complete *** Names and line numbers of children identified to be under 6years from Module C. L1 Child 1__________________________________________________________ 1. Line No____ Child 2__________________________________________________________ 2. Line No____ Child 3__________________________________________________________ 3. Line No____ Child 4__________________________________________________________ 4. Line No____ Child 5__________________________________________________________ 5. Line No____ INSTRUCTION: Then read the children’s names to present household members and ask: “Are these all of the children younger than 6 years living here?” If there are more children, add their names and line numbers. Age verification of children and selection of children for Module T I would like to ask you some questions about (NAME). I will need (NAME’S) vaccination or birth card. 98 L2.1 CHILD 1 What is (NAME’s) birth date? Verify birth date on vaccination or birth card and fill in the day, month and year. If day is not known, enter ‘99’. If a vaccination or birth card is not available, ask the mother if she knows the birth date and if she does, enter it. If there is no birth date and the mother does not know it, use the local calendar of events provided in the training to approximate the month and year of birth (leave day blank). Day ................................................ Month ........................................... Year..................................... L2.2 Check birth date. Was child born on TODAY’s DATE 2009 or later? Yes.................................................................1 No..................................................................2 1→ Enter Line number L2.3 CHILD 2 What is (NAME’s) birth date? Day ................................................ Month ........................................... Year..................................... L2.4 Check birth date. Was child born on TODAY’s DATE 2009 or later? Yes.................................................................1 No................................................................. 2 1→ Enter Line number L2.5 CHILD 3 What is (NAME’s) birth date? Day ................................................ Month ........................................... Year..................................... L2.6 Check birth date. Was child born on TODAY’s DATE 2009 or later? Yes.................................................................1 No ................................................................ 2 1→ Enter Line number L2.7 CHILD 4 What is (NAME’s) birth date? Day ................................................ Month ........................................... Year..................................... L2.8 Check birth date. Was child born on TODAY’s DATE 2009 or later? Yes.................................................................1 No................................................................. 2 1→ Enter Line number L2.9 CHILD 5 What is (NAME’s) birth date? Day ........................................................ Month ................................................... Year .............................................. 99 L2.10 Check birth date. Was child born on TODAY’s DATE 2009 or later? Yes.................................................................1 No..................................................................2 1→ Enter Line number 100 Selection of index child INSTRUCTION: If there is only one child listed, this is the index child. If there is more than one child, use the “numbered papers in a container” method to randomly choose the index child and record the information about the child below. Name of index child: _______________________________________ L3 Line number of index child L3a Date of birth of index child Day ........................................................ Month.................................................... Year ............................................. *** The index child’s mother (or caretaker) is the respondent for the remainder of the questionnaire *** Information on respondent for Part III: General L4 Name of respondent: _______________________________________ 101 L5 Line number of respondent Record line number from Module C ................................................... L6 Relationship to household head (see codes below) ................................................... L7 How old are you? (years) ................................................... L8 Are you currently pregnant? Yes..........................................................1 No ..........................................................2 DNK........................................................8 L9 What is your level of education? (see codes below) Code........................................... L10 If the respondent is the caretaker (but not the mother) of the child, write 99 in the box. Code..................................... Codes for L6: 1= Household head, 2= Wife of household head, 3= Daughter, 4=Granddaughter, 5=Niece, 6=Mother, 7= Daughter in law, 8=Sister, 9=Sister-in-law, 10=Brother’s wife. Codes for L9: 0 = No class, 1 = Class 1, 2 = Class 2, 3 = Class 3, 4 = Class 4, 5 = Class 5, 6 = Class,6 7 = Class 7, 8 = Class 8, 9 = Class 9, 10 = SSC pass, 11 = HSC pass, 12 = Graduate, 13 = Masters. 102 MODULE M. ANTENATAL CARE Now I would like to ask you some questions about your current or most recent pregnancy. (Interviewer: See question L8 to determine if respondent is pregnant) QUESTION CODE/RESPONSE SKIP M1 Did you see anyone for antenatal care during your pregnancy? Yes ................................................... 1 No .................................................... 2 2→M8 M2 Whom did you see? Doctor ............................................. 1 Nurse/midwife................................. 2 Traditional birth attendant ............. 3 Skilled Birth Attended (SBA)............ 4 Community/village health worker... 5 Other ............................................... 6 M3 Where did you receive antenatal care during your pregnancy? Your home ....................................... 1 Other home ..................................... 2 Government hospital ...................... 3 Other Government health facility ... 4 Private hospital/clinic ...................... 5 Community Clinic............................. 6 Other Private health facility ............ 7 Other ............................................... 8 M4 How many months pregnant were you when you first received antenatal care? Number of months............... DNK.................................................98 M5 How many check-ups did you have during your pregnancy? Number of visits ................... M6 Do you have an antenatal card or a prescription sheet for your most recent pregnancy? If yes: May I see it please? Yes, Seen...........................................1 Yes, Not Seen....................................2 103 No Card.............................................3 2,3 →M8 M7 Interviewer: Verify number of antenatal visits (Is the number of documented visits on the card different than the stated number of visits in M5?) Same as stated..................................1 Different than stated ........................2 Note number of documented visits...................................... M8 During your pregnancy, do/did you take the same amount of food as you usually take or do/did you take more or less food than you usually take? More food ........................................1 Less food ..........................................2 Same as usual ..................................3 M9 During your pregnancy, do/did you take as much daytime rest as you usually take? More rest .........................................1 Less rest ...........................................2 Same as usual ..................................3 M10 Did you receive Vitamin A Capsule within 42 days (6 weeks) of delivery of the child? [INTERVIEWER: Show her the red vitamin A capsule] Yes ....................................................1 No .....................................................2 Can’t Recall .......................................3 Not applicable (still pregnant)…. 4 M11 In your last pregnancy, did you take any iron and folic acid tablets like this? [INTERVIEWER: Show her the iron tablet] Yes ....................................................1 No .....................................................2 2→ M13 M12 For how many months during your last pregnancy did you take iron and folic acid tablets? 1-2.....................................................1 3-4.....................................................2 5-6…..................................................3 > 6.....................................................4 M13 Have you taken an iron and folic acid tablet in the last 7 days/ latest 7 days in pregnancy during your current or index child’s name pregnancy? Yes ....................................................1 No .....................................................2 Can’t Recall .......................................3 104 105 MODULE N. FOOD CONSUMPTION OF MOTHER Now I would like to ask you (mother) about the types of foods that you (mother) ate yesterday during the day or at night. Please include all foods, including the foods eaten here at your house or somewhere else. QUESTION CODE/RESPONSE SKIP N1 Any cereals, e.g. rice, bread, wheat, wheat bread, rice flakes, puffed rice, barley, wheat grain, popcorn? Yes...................................................... 1 No ...................................................... 2 N2 Any pumpkin, carrots, squash, or sweet potatoes or vegetables that are yellow or orange inside? Yes...................................................... 1 No ...................................................... 2 N3 Any white potatoes, white yams or other foods made from roots and tubers? Yes...................................................... 1 No ...................................................... 2 N4 Any dark green, leafy vegetables, e.g. ipomoea, amaranth, spinach, parwar sag, and drumstick leaves? Yes...................................................... 1 No ...................................................... 2 N5 Any other vegetables, e.g. cucumber, radish, pepper, string beans, cabbage, cauliflower, radish, onion? Yes...................................................... 1 No ...................................................... 2 N6 Any ripe papaya, mangoes or other fruits that are yellow or orange inside? Yes...................................................... 1 No ...................................................... 2 N7 Any other fruits, e.g. banana, papaya, sithphal, grapefruit, apple, orange, jackfruit, jambu fruit, plums, melon, tomato, date, lemon, etc. ? Yes...................................................... 1 No ...................................................... 2 N8 Any meat, such as, liver, beef, poultry, lamb, pork, etc.? Yes......................................................1 No ......................................................2 N9 Any eggs? Yes......................................................1 No ......................................................2 N10 Any fresh or dried fish or shellfish? Yes......................................................1 No ......................................................2 N11 Any legumes/pulses, e.g. Bengal gram, black gram dal, lentil, Khesarl? Yes......................................................1 No ......................................................2 N12 Any Milk or Milk products, e.g. cow milk, buffalo milk, goat milk, yogurt, curd, cheese? Yes......................................................1 No ......................................................2 106 N13 Any foods prepared using fat, e.g., oil, butter, dalda or ghee? Yes......................................................1 No ......................................................2 N14 Any sugar or honey?Example:sugar, gur, honey, chocolate Yes......................................................1 No ......................................................2 MODULE O. MOTHER’S HAND WASHING HABITS AND DISPOSAL OF CHILD’S FECES QUESTION CODE/RESPONSE SKIP Hand washing habits O1 Please mention all of the times when it is important to wash your hands. Any other time besides this? Any other time besides this? [DO NOT READ THE ANSWERS] Before eating ................................... 01 After eating...................................... 02 Before praying ................................. 03 Before breastfeeding or feeding a child .............................................. 04 Before cooking or preparing food ... 05 After defecation/urination .............. 06 After cleaning a child that has defecated/changing a child’s diaper .............................................. 07 When my hands are dirty ................ 08 After cleaning the toilet or potty..... 09 Other (specify)................................. 10 DNK.................................................. 98 107 O2 Can you please show me where members of your household most often wash their hands? [OBSERVE AND CIRCLE RESPONSE CODE] Inside/within 10 paces of the toilet facility ................................................ 1 Inside/within 10 paces of the kitchen/cooking place ....................... 2 Elsewhere in home or yard................ 3 Outside yard ...................................... 4 No specific place................................ 5 No permission to see......................... 6 5,6→O5 O3 Is water present at the place? [OBSERVE:If there is a tap or pump see if water comes out. If there is a container, see if water is in it.] Yes ..................................................... 1 No ...................................................... 2 O4 Is soap, detergent, ash or clay present at the place? [OBSERVE AND CIRCLE ALL RESPONSE CODES] None .................................................. 1 Bar soap............................................. 2 Detergent (powder/liquid/paste)...... 3 Liquid soap (including shampoo)....... 4 Ash or clay ......................................... 5 Disposal of child’s feces O5 The last time (NAME) passed stool, where did he/she defecate? Used potty ......................................... 1 Used washable diaper ....................... 2 Used disposable diaper ..................... 3 Went in his/her clothes..................... 4 Went in house/yard........................... 5 Went outside of house/yard ............. 6 Used latrine ....................................... 7 Other ................................................. 8 7→ P1 108 O6 The last time (NAME) passed stool, where were his/her feces disposed? Dropped into toilet facility/latrine .... 1 Buried ................................................ 2 Put into container for trash............... 3 In yard................................................ 4 In sink or tub...................................... 5 Thrown into waterway ...................... 6 Washed or rinsed away ..................... 7 Left at the same place where the child defecated ................................. 8 Threw it away to bush/outside of the house....................................... 9 7→ O7 All Others→P1 O7 If “washed or rinsed away”, probe where the waste water was disposed. Dropped into toilet facility ................ 1 Put into container for trash............... 2 In yard................................................ 3 Outside of yard .................................. 4 Into sink or tub .................................. 5 Thrown into waterway ...................... 6 109 MODULE P. FEEDING OF CHILDREN 0-23 MONTHS ***For index child if 0-23 months*** OR ***If the index child is not 0-23 months, and another child in the household is 0-23 months, ask Module P and Module Q questions for the child that is 0-23 months*** Check the index child’s birth date (question L3a). Was the index child (or any child in the household) born on today’s date in 2012 or later? If so, he/she is 0-23 months. If no children are 0-23 months, skip to Module R. To mother: Now I would like to ask you about what your child eats and drinks. QUESTION CODE/RESPONSE SKIP P0 Is this the index child? Yes......................................................1 No.......................................................2 1  P1 P0.1 Enter child’s roster number from Module C P1 Has (NAME) ever been breastfed? Yes......................................................1 No.......................................................2 DNK ....................................................8 2,8→ P5 P2 Do you currently breastfeed your child? Yes......................................................1 No.......................................................2 Not applicable....................................3 DNK ....................................................8 2,3,8→ P5 P3 Was (NAME) breastfed yesterday during the day or at night? Yes......................................................1 No.......................................................2 DNK ....................................................8 P4 Did you feed any other food even water your child without breastfeeding in last 24 hours? Yes......................................................1 No.......................................................2 110 Not applicable....................................3 DNK ....................................................8 P5 [Now I would like to ask you about some medicines and vitamins that are sometimes given to infants.] Was (NAME) given any vitamin drops or other medicines as drops yesterday during the day or at night? Yes......................................................1 No.......................................................2 DNK ....................................................8 NR.......................................................9 P6 Was (NAME) given any oral rehydration solution yesterday during the day or night? (salt/sugar saline, Labon-gur, packet saline, rice poser)? Yes......................................................1 No.......................................................2 DNK ....................................................8 Child’s consumption of liquids (0-23 Months) Read the questions below. Read the list of liquids one by one and circle applicable code. Next I would like to ask you about some liquids that (NAME) may have had yesterday during the day or night. P7:Did (NAME) have the following water or liquids? P8:How many times yesterday during the day or at night did (NAME) consume any….? a. Plain water? Yes.................. 1 No .................. 2 DNK................ 8 b. Infant formula/baby formula bought with money? Yes.................. 1→ No .................. 2 DNK................ 8 Times............................... DNK........................................... 98 c. Milk, such as tinned, powdered or fresh animal milk? Yes.................. 1→ No .................. 2 DNK................ 8 Times............................... DNK........................................... 98 111 d. Juice or juice drinks? Yes ………….1 No .................. 2 DNK................ 8 e. Clear broth? Yes ………….1 No .................. 2 DNK................ 8 f. Yogurt? Yes.................. 1→ No .................. 2 DNK................ 8 Times............................... DNK........................................... 98 g. Thin porridge? Yes ………….1 No .................. 2 DNK................ 8 h. Any other liquids? Yes ………….1 No .................. 2 DNK................ 8 Child’s consumption of solids (0-23 Months) As the respondent recalls foods, in the table below circle the response in the column next to the food group. If the food is not listed, write it down in the “OTHER FOODS” box (P9). Ask the mother: Please describe everything that (NAME) ate yesterday during the day or night, whether at home or outside the home. Use these probes. (a) Think about when (NAME) first woke up yesterday. Did (NAME) eat anything then? IF YES: Please tell me everything (NAME) ate then … anything else? 112 Keep asking until respondent indicates nothing else. Then continue to question (b). (b) What did (NAME) do after that? Did (NAME) eat anything then? IF YES: Please tell me everything (NAME) ate then… anything else? Keep asking until respondent indicates nothing else. Keep repeating question (b) until the respondent says the child went to sleep until the next day. If respondent mentions anything like a porridge, sauce or stew, probe: (c) What ingredients were in that (MIXED DISH)? Anything else? P9 OTHER FOODS: Write in other foods mentioned by mother, not listed below, here. ------------------------------- Yes ..................................................... 1 No ...................................................... 2 P10A Any cereals: porridge, bread, rice, noodles, or other foods made from cereals Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10B Pumpkin, carrots, squash or sweet potatoes that are yellow or orange inside Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10C White potatoes, white yams, manioc, cassava, or any other foods made from roots Yes ..................................................... 1 No ...................................................... 2 113 DNK.................................................... 8 P10D Any dark green leafy vegetables, such as ipomoea, amaranth, spinach, parwar sag, and drumstick leaves Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10E Ripe mangoes, ripe papayas or other fruits that are yellow or orange inside Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10F Any other fruits or vegetables Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10G Liver, kidney, heart or other organ meats Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10H Any meat, such as beef, pork, lamb, goat, chicken, or duck Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10I Eggs Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10J Fresh or dried fish, shellfish, or seafood Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10K Any foods made from beans, peas, lentils, nuts or seeds, such as Bengal gram, black gram, dal, Khesari Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10L Cheese, yogurt, curd or other milk products Yes ..................................................... 1 114 No ...................................................... 2 DNK.................................................... 8 P10M Any oil, butter, dalda or ghee or foods made with any of these Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10N Any sweet foods such as honey, chocolates, sweets, candies, pastries, cakes or biscuits Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10O Condiments for flavor, such as chilies, spices, herbs, or fish powder Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10P Grubs, snails, or insects Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 P10Q Foods made with red palm oil, red palm nuts, or red palm nut pulp sauce Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 INTERVIEWER: CHECK CATEGORY A-Q. IF NO IN ALL CATEGORIES, THEN GO TO P11.IF ONLY ONE CATEGORY IS CIRCLED AS YES OR ALL CATEGORY AS DNK, THEN GO TO P12. P11 Did (NAME) eat any solid, semi-sold, or soft foods yesterday during the day or at night? Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 If yes, probe: What kinds of foods did (NAME) eat? Go back to P7 and record foods eaten. Then continue with P10. 2,8→ P13 P12 How many times did (NAME) eat solid, semi-solid, or soft foods other than liquids Number of times..................... 115 yesterday during the day or night? DNK.................................................. 98 P13 Did (NAME) drink anything from a bottle with a nipple yesterday during the day or night? Yes ..................................................... 1 No ...................................................... 2 DNK.................................................... 8 116 MODULE Q. IMMUNIZATION OF CHILDREN 0-23 MONTHS AND MICRONUTRIENT SUPPLEMENTATION (MONIMIX/OTHER SPRINKLES) ***If the index child is not 0-23 months, and another child in the household is, ask Module P and Module Q questions for the child that is 0-23 months*** Circle the code number of the response and follow the arrowed skip codes. QUESTION CODE/RESPONSE SKIP Q1 Does the mother have a vaccination card for (NAME)? Have you seen it? Yes, Seen.......................................................1 Yes, Not Seen................................................2 No Card.........................................................3 2,3→ Q4 117 QUESTION CODE/RESPONSE SKIP Q2 (1) Copy vaccination date for each vaccine from the card. (2) Write “44” in “Day” column if card shows that a vaccination was given but no date is recorded. BCG PO (OPV) P1 (OPV) P2 (OPV) P3 (OPV) DPT/Penta-1 DPT/Penta-2 DPT/Penta-3 MEA/MM Day Month Year Skip to Q10 If all vaccines given and recorded in card 118 QUESTION CODE/RESPONSE SKIP Q3 Has (NAME) received any vaccinations that were not recorded on this card? Record “Yes” only if respondent mentions BCG, POLIO 1-3, Pentavalent 1-3, and/or measles/MMR vaccine(s) Probe for vaccinations and write “66” in the corresponding day column in Question Q2 Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 2,8→Q10 Q4 Please tell me if (NAME) received any of the following vaccinations: A BCG vaccination against tuberculosis, that is, an injection in the left shoulder that caused a scar? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 Q5 Polio vaccine that is, drops in the mouth? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 2,8→ Q7 Q6 How many times did (NAME) receive polio vaccine: Times ........................................................... Q7 DPT /Pentavalent vaccination, that is, an injection given in the thigh or buttocks, sometimes at the same time as polio drops? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 2,8→ Q9 Q8 How many times? Number of times.......................................... Q9 An injection given to prevent measles/MMR after 9 months of age? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 119 QUESTION CODE/RESPONSE SKIP Q10 Has (NAME) received a vitamin A capsule like this in the last 6 months? Interviewer: Check vaccination card if available. Show blue and red Vitamin A capsules as either may have been given depending on child’s age. Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 120 Q11 Are you or someone else adding any Moni-mix or other sprinkles packets into (NAME’s) food? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 Q12 Do you think that children of age 6 to 59 months (less than 5 years) can suffer from micronutrient deficiencies? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 Q13 Do you give your child any drop/tablet/ syrup/packet so that the child does not suffer from iron deficiency? Yes ................................................................... 1 No .................................................................... 2 DNK.................................................................. 8 2,8→ R1 Q14 What do you give to child to prevent iron deficiency? [MULTIPLE RESPONSE] Iron tablet...................................................... 01 Iron syrup....................................................... 02 MNP............................................................... 03 Pustikona ....................................................... 04 Monimix......................................................... 05 Mymix ............................................................ 06 Multivitamin tablet........................................ 07 Multivitamin syrup ........................................ 08 Homeopathik medicine ................................. 09 Quack medicine ............................................ 10 Other (specify)............................................... 11 DNK................................................................ 98 MODULE R. DIARRHEA AMONG CHILDREN 0-59 MONTHS INSTRUCTIONS: Administer module R for the selected index child between 0-59 months QUESTION CODE/RESPONSE SKIP 121 QUESTION CODE/RESPONSE SKIP R1 Has (NAME) had diarrhea (having 3 or more loose stools in 24 hours) in the last 2 weeks? Yes............................................................1 No.............................................................2 2→S1 R2 Now I would like to know how much (NAME) was given to drink during the diarrhea (including breastmilk). Was he/she given less than usual to drink, about the same amount, or more than usual to drink? ++ If “less”, probe: Was he/she given much less than usual to drink or somewhat less? Much less .................................................1 Somewhat less .........................................2 About the same........................................3 More.........................................................4 Nothing to drink .......................................5 DNK ..........................................................8 R3 When (NAME) had diarrhea, was he/she given less than usual to eat (solid/semi-solid food), about the same amount, more than usual, or nothing to eat? If “less”, probe: Was he/she given much less than usual to drink or somewhat less? Much less .................................................1 Somewhat less .........................................2 About the same........................................3 More.........................................................4 Nothing to drink .......................................5 DNK ..........................................................8 R4 Are you still breastfeeding (NAME)? Yes............................................................1 No.............................................................2 2→R6 R5 Did you continue to breastfeed (NAME) during diarrhea? Continued.................................................1 Did not continue ......................................2 R6 Was anything given to (NAME) to treat the diarrhea? [MULTIPLE RESPONSE] A. Homemade (sugar/salt) saline......... 01 B. Homemade (Labon-gur) saline……... 02 C. Packet saline/ORS ............................ 03 D. Rice poser......................................... 04 E. Pill/capsule/syrup ........................... 05 F. Injection ........................................... 06 G. Intravenous...................................... 07 H. Home remedies/herbal medicine/ plants ............................................... 08 I. Plain drinking water......................... 09 J. Others (Specify)................................ 11 K. Did not give anything ....................... 12 MODULE S: Additional information on respondent for Part III 122 Information on respondent for Part III: Participation in SHOUHARDOII QUESTION CODE/RESPONSE SKIP S1 Are you a member of an EKATA group? Yes.................................................... 1 No .................................................... 2 2→ S4 S2 How often does the EKATA group usually meet? Once a week ……………………..1 Once a month …………………….2 Once a year ………………………3 S3 How many EKATA group meetings have you attended in the last year (since last December)? [____] S4 Are you a member of a SHOUHARDO II savings group? Yes.................................................... 1 No .................................................... 2 S5 Have you attended a SHOUHARDO II courtyard session (UthanBoithok) with other mothers on the health and nutrition of mothers and children? Yes.................................................... 1 No .................................................... 2 1→ S8 S6 How many of these courtyard sessions have you attended in the last year (since last December)? [____] S7 What subjects did you learn about at the courtyard sessions? (Read each session type to enumerator) S7a Antenatal care (care for pregnant women) Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 123 S7b Food and nutrition for pregnant women and lactating mothers Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7c Post-natal care for mothers and babies Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7d Breastfeeding and complementary feeding of children Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7e Children’s illnesses: ARI, Diarrhoea Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7f Immunization Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7g Growth monitoring and promotion (GMP) Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7h Good hygiene practices (e.g., hand washing, disposal of feces) Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7i Safe drinking water/Sanitary latrine Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 124 S7j Health in disaster Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S7k Physical and mental torture during pregnancy Yes.................................................... 1 No .................................................... 2 DNK.................................................. 8 S8 Have you attended a SHOUHARDO II cooking and feeding demonstration (khichuri) session? Yes.................................................... 1 No .................................................... 2 S9 Does the SHOUHARDO II project have a place where you can take your child/children to be weighed and measured each month? Yes.................................................... 1 No .................................................... 2 2→S11 S10 How many months in the last year have you taken at least one of your children to be weighed and measured in the last year (since last December?). [____] S11 Have you ever received a food ration (wheat, oil or split peas) from the SHOUHARDO II program? Yes.................................................... 1 No .................................................... 2 2→S15 S12 Did you receive a food ration every month since last December? Yes.................................................... 1 No .................................................... 2 2→S15 S13 How many months since last December did you NOT receive a food ration? [____] S14 Why did you not receive a food ration in every month? Not eligible (not pregnant or lactating) ………………………… 1 Did not pick up the ration …..… 2 Ration not available …………….. 3 125 126 Information on respondent for Part III: Time constraints QUESTION CODE/RESPONSE SKIP S14.1 Is the current respondent the same respondent to respond to Module J2? Yes....................................................1 No .....................................................1→ S29 2 QUESTION CODE/RESPONSE SKIP S15 Do you do any agricultural production, gardening, or animal raising? Yes...............................1 No................................2 2→ S17 S16 On a typical day in the last month, how many hours did you spend on these activities? [____] S17 Do you do any post-harvest activities like threshing? Yes...............................1 No…………………… 2 2→ S19 S18 On a typical day in the last month, how many hours did you spend on post-harvest activities? [____] S19 Do you earn any cash income? Yes...............................1 No…………………… 2 ......... 2→ S21 S20 On a typical day in the last month, how many hours did you spend earning cash income? [____] S21 Do you do any housework, like cooking, washing dishes and clothes, fetching wood and water? Yes...............................1 No…………………… 2 2→ S23 S22 On a typical day in the last month, how many hours did you spend on housework? [____] S23 Do you take care of any children? Yes...............................1 No................................2 2→ S25 S24 On a typical day in the last month, how many hours did you spend taking care of children? [____] S25 During the daytime do you spend time doing other things like socializing, watching TV, taking naps, or reading? Yes...............................1 2→ S27 127 No................................2 S26 On a typical day in the last month, how many hours did you spend in these kinds of “leisure” activities where you were not working? [____] S27 Imagine ten steps, where on the bottom, the first step, is a person who spends no time in the day doing these leisure activities, and on the highest step, the 10th, is a person who spends the whole day doing them [INSTRUCTION: Show picture of steps]. On a typical day in the last month, which step were you on? [____] S28 Which step were you on five years ago (before SHOUHARDO II began)? [____] Information on respondent for Part III: Decision making in household INSTRUCTION: First answer Question S29 to determine whether the respondent should answer the rest of the questions in this module. QUESTION CODE/RESPONSE SKIP S29 Is the respondent for PART III also the respondent for PART II? Yes.................................................... 1 No .................................................... 2 1→ T0 Decision making in household QUESTION CODE/RESPONSE SKIP 128 QUESTION CODE/RESPONSE SKIP In the last year, to what extent have you been able to make the following kinds of decisions? [INSTRUCTION: First read the possible responses. Then list each item (S30.1 – S30.12) one-by-one and record code number of response.] Code list 1 = Can decide alone 2 = Can decide with husband or other adult male family member 3 = Husband makes decision after discussion with wife 4 = Not involved in decision 5 = Not applicable (Decision not made) S30.1 Buying small food items, groceries, toiletries S30.2 Buying clothing for yourself and your children S30.3 Spending money that you yourself have earned S30.4 Buying or selling major household assets (land, livestock, crops) S30.5 Buying or selling jewelry S30.6 Use of loans or savings S30.7 Expenses for your children’s education S30.8 Expenses for your children’s marriage S30.9 Medical expenses for yourself or your children S30.10 Expenses for family planning (contraceptives) S30.11 To move to shelter during time of disaster S30.12 Actively participate and involved in salish decision making 129 QUESTION CODE/RESPONSE SKIP S31 Imagine ten steps, where on the bottom, the first step, is a woman who can make none of these types of decisions, and on the highest step, the 10th, is a person who can make all of them. [INSTRUCTION: Show picture of steps]. On a usual day in the last month, which step were you on? [____] S32 Which step were you on five years ago (before SHOUHARDO II began)? [____] 130 MODULE T. HEIGHT AND WEIGHT OF ALL CHILDREN IN HOUSEHOLD 0-59 MONTHS AND MOTHER *** For ALL children 0-59 months in household *** [INSTRUCTION: Request permission of the respondent to measure her height and weight and that of her children under 5] QUESTION CODE/RESPONSE Child 1 Child 2 Child 3 T0 Interviewer: Write in the line number of the child from the right hand column of Question L2. T1 Interviewer: Write in the birth date of the child from Question L2 Day................. Month............ Year.. Day ................ Month............ Year .. Day................. Month............ Year . Interviewer: If the child was born on TODAY’s DATE in 2012 or later, he/she is 0-23 months. Measure length of child lying down. If the child was born before TODAY’s DATE in 2012 or earlier, he/she is 24 to 59 months. Measure standing height. T2 Sex of child Male.......................... 1 Female ...................... 2 Male ..........................1 Female.......................2 Male ..........................1 Female.......................2 T3 Height or length of child in centimeters 988 = NOT MEASURED . cm . cm . cm T4 Height of mother or caregiving centimeters . cm . cm . cm 131 QUESTION CODE/RESPONSE Child 1 Child 2 Child 3 T5 Weight of child (in kilograms) . kg . kg . kg T6 Weight of the mother or caregiver(in kilograms) . kg . kg . kg T7 Date of height measured/weighed (today) Day................. Month............ Year.. Day ................ Month............ Year .. Day................. Month............ Year . T8 Result for child Child measured......... 1 Child sick ................... 2 Child not present...... 3 Child refused............. 4 Mother refused ........ 5 Other (Specify).......... 6 Child measured .........1 Child sick....................2 Child not present.......3 Child refused .............4 Mother refused .........5 Other (Specify) ..........6 Child measured..........1 Child sick....................2 Child not present.......3 Child refused .............4 Mother refused .........5 Other (Specify) ..........6 132 QUESTION CODE/RESPONSE Child 4 Child 5 T0 Interviewer: Write in the line number of the child from the right hand column of Question L2. T1 Interviewer: Write in the birth date of the child from Question L1 Day................. Month............ Year.. Day ................ Month............ Year .. T2 Sex of child Male.......................... 1 Female ...................... 2 Male ..........................1 Female.......................2 T3 Height or length of child in centimeters 988 = NOT MEASURED . cm . cm T4 Height of mother or caregiverin centimeters . cm . cm T5 Weight of child (in kilograms) . kg . kg T6 Weight of the mother or caregiver(in kilograms) . kg . kg T7 Date of height measured/weighed Day................. Day ................ 133 QUESTION CODE/RESPONSE Child 4 Child 5 (today) Month............ Year.. Month............ Year .. T8 Result for child Child measured......... 1 Child sick ................... 2 Child not present...... 3 Child refused............. 4 Mother refused ........ 5 Other (Specify).......... 6 Child measured .........1 Child sick....................2 Child not present.......3 Child refused .............4 Mother refused .........5 Other (Specify) ..........6 RECORD TIME THE INTERVIEW ENDED. HOUR ......................................... MINUTE...................................... 134 Annex 4: Sample Weights Calculation of sampling weights starts with construction of the design weight for units within each strata, where strata are denoted i=1,..,8. The design weights are the reciprocal of each unit’s probability of selection into the sample: where pi is the probability of selection. A. Household-level sampling weights The household sampling weights are constructed by first calculating the design weight and response rate (rr) for households in each stratum as follows. The final weight is the design weight divided by the response rate: . B. Child sampling weights Construction of the child-level sampling weights starts with the household-level design weight as an approximation of the child-level design weight.4 . The response rate takes into account both the household response rate and the fact that all under-5 children listed in a household won’t be available for anthropometric measurements. The final weight is the design weight divided by the response rate: 4 The proportion of households with children under five in the population and sample is not known in advance. 135 Annex 5: IPTT matrix and SAPQ Indicator Target IPTT Baseline (computed baseline)‡ Midterm Endline Percent Difference (Endline￾Baseline) Significance level Target met? Percent difference (Endline-Target) SO 1: "Availability of" and "access to" nutritious foods enhanced and protected for 370,000 poor & extreme poor (PEP) households Average household dietary diversity score 6.0 5.1 (4.8) 5.9 8.7 81.3 * 45.0 Average number of months of adequate household food provisioning 8.0 6.8 (5.9) 9.9 11.0 86.4 * 37.5 SO 2: Improved health, hygiene and nutrition status of 176,706 children under 2 years of age Percent of underweight children under five years of age (0-59 months) 38.4 40.8 (42.2) 34.4 35.9 -14.9 * -8.1 Females 34.6 36.6 (37.2) 34.9 36.4 -2.2 5.2 Males 42.2 44.6 (46.7) 33.4 34.3 -26.6 * -18.7 Percent of stunted children under five years of age (6-59 months) 55.1 61.7 52.7 50.4 -18.3 * -11.4 Females 51.2 56.5 52.2 47.8 -15.3 * -6.6 Males 58.9 66.1 53.2 49.7 -24.9 * -15.6 Percent of children under age two who had diarrhea in the prior two weeks 10.3 14.7 (13.8) 11.6 8.1 -41.3 -21.4 Percent of children underweight under the age of 2 years (0- 23 months) 31.1 35.3 (33.2) 32.1 29.5 -11.1 -5.1 Percent of children immunized against 8 diseases by 12 months of age 82.0 59.9 (58.6) 74.9 73.0 24.6 * -11.0 Percent of pregnant and lactating women taking iron supplements in last 7 days 45.0 16.1 (15.2) 29.8 52.0 242.1 * 15.6 Prevalence of exclusive breast feeding of children under six months 66.2 62.2 (64.1) 66.4 62.2 -3.0 -6.0 136 Percent of mothers who feel it is important to wash hands at five critical times£ 35.0 na (9.8) 10.9 29.8 204.1 * -14.9 Number of people in target areas with improved access to sanitation facilities 60.0 26.3 (20.3) 32.4 52.8 160.1 * -12.0 Percent of children 6–23 months of age who receive a minimum acceptable diet 20.0 10.6 (8.7) 17.5 47.9 450.6 * 139.5 Number of people in target areas with improved access to drinking water supply 80.0 61.1 (59.5) 76.3 77.9 30.9 * -2.6 SO 3: PEP women and adolescent girls empowered in their families, communities and Union Parishad Percent of women control over economic resources 25.0 10.5 (5.7) 11.3 19.8 247.4 * -20.8 SO 5: Targeted community members and government institutions are better prepared for, mitigate, and respond to disasters and adapt to climate change Percent of PEP households distress selling 7 12.5 (9.6) 9.5 9.8 2.1 40.0 Taka value of distress sale (PEP households) 15000 25,274 (19979) 21304 16231 -18.8 8.2 Note: Stars indicate difference is statistically significant at the 10% (*) level Note: Under column "Target met?", red indicates that the target was not met and green indicates target was met (or that endline and target were not significantly different) Note: There are no impact/outcome indicators in the SHOUHARDO II Indicator Performance Tracking Table (IPTT) under SO4. £ The five critical times are: Before eating, before breastfeeding or feeding a child, before cooking or preparing food, after defecation/urination, and after cleaning a child that has defecated/changing a child's diaper. ‡ Values in parenthesis are the baseline numbers computed by TANGO during midterm and endline analysis. Some of these numbers differ from the baseline numbers in IPTT indicator table. 137 Was your food aid program awarded in FY2011? Yes/No No Yes/No Yes, '11 Yes/No Yes Yes/No Yes FY 11 # households in target areas 584,003 Yes/No Yes 1A: Months of Adequate Food Provisioning (Impact Indicator) wadud@bd.care.org CARE Bangladesh, 20-21 Kawran Bazar (Level 12), Dhaka 1215, Bangladesh 6 5 Are the final data from your program's survey available at this time? Do not answer "Yes" if you have preliminary data only. If final data are not yet available, answer "No" and report on them in next year's SAPQ. Only final data should be entered into the SAPQ. What is the estimated total number of households in your target geographic area? How many households live in your target geographic area? (This is not the sample size, it's the population size) 4 Does your program aim to improve household food access? A program that aims to improve household food access generally promotes agricultural production, agricultural product processing and marketing, microcredit and other income- and employment-generation activities In the survey, did your program measure average number of months of (in)adequate food provisioning, following the standard FANTA methodology for this indicator? See the "Definitions" tab for a description of the standard methodology for this indicator. Program location(s) in the host country Program Expiration Date (mm/dd/yy) 5/31/15 SHOUHARDO II (Strengthening Household Ability to Respond to Development Opportunities) 6/1/10 AID-FFP-A-10-00010 3 Contact Name (person filling out the SAPQ) Answer "Yes, '11" if you conducted a survey in FY11. Answer "Yes, '10" if you conducted a survey in FY10 but you did not report on it in last year's FY10 SAPQ because the final data were not yet available. If you conducted a survey in FY10 and already reported the results in the FY10 SAPQ, choose "No". FFP Standardized Annual Performance Questionnaire (SAPQ) - FY 2011 Award Number Awardee Name(s) Program Start Date (mm/dd/yy) Program Name AWARDEE FOOD AID PROGRAM INFORMATION CARE Bangladesh Bangladesh 11 Distrcits located in the North and Mid Chars, Haor and Coast Host Country (or Countries, for Regional Programs) AWARDEE CONTACT INFORMATION 1 AKM Abdul Wadud Baseline Survey Contact Email Which type of quantitative survey did your program conduct in FY11 (or in FY10 and you did not report the results in last year's SAPQ)? Choose your answer from the drop down menu. Contact Phone SECTION 1: Data from a Representative Population-based Survey This section asks for impact data coming from a quantitative survey such as a baseline or final evaluation +880-2-9112315 Contact Address 2 0 Did your food aid program conduct a quantitative, population-based, statistically representative survey such as a baseline or final evaluation in FY11 (or in FY10 and you did not report the results in last year's SAPQ)? 138 # months From w hich FY? # months What FY is the final evaluation? # of months 7 FY # FY15 8 Yes/No Yes # of food groups From w hich FY? # of food groups What FY is the final evaluation? # of food groups 5 FY # FY15 6 Yes/No Yes Yes/No Yes % underw eight From w hich FY? % underw eight What FY is the final evaluation? % underw eight 41% FY # FY15 38% FY 11 # 0-59 mo 319,683 What is the estimated total number of children 0-59 months of age, living in your target geographic area? How many children 0-59 months live in your geographic area? (This is the population size, not the sample size) FY 11 13 12 Final Evaluation Target Most recent FY prior to FY11 (enter n/a if FY11 w as the baseline) Final Evaluation Target 8 In the survey, did your food aid program measure the prevalence of underweight (WAZ <-2) in children 0 - 59 months of age? If you measured underweight for a different age group, or you used a different measure or cutoff, answer NO. 1C: Underweight (Impact Indicator) 1B: Household Dietary Diversity (Impact Indicator) 9 FY 11 Most recent FY prior to FY11 (enter n/a if FY11 w as the baseline) Average household dietary diversity score Indicator Average number of months of ADEQUATE food provisioning 7 What was the average number of months of adequate food provisioning ? Fill out the table below with the final data from your survey. If you measured INADEQUATE instead of ADEQUATE months, convert your data to ADEQUATE months (12 - number of inadequate months). Only provide data if you used the standard FANTA methodology. If this is a baseline survey, please also provide your final evaluation target for this indicator. If this survey is a final evaluation, please also provide the average number of months of adequate food provisioning data from the most recent population￾based survey prior to FY11 (probably from your baseline survey). Indicate the year in which the data were collected. Most recent FY prior to FY11 (enter n/a if FY11 w as the baseline) 10 What was the average household dietary diversity score ? Fill out the table below with the data from your survey. Only provide data if you used the standard FANTA methodology. If this is a baseline survey, please also provide your final evaluation target for this indicator. If this survey is a final evaluation, please also provide the average dietary diversity score from the most recent population-based survey prior to FY11 (probably from your baseline survey). Indicate the year in which the data were collected. Indicator % of underweight (WAZ<-2) children 0-59 months of age What was the prevalence of underweight (WAZ <-2) in children 0 - 59 months of age? If this survey is a baseline survey, please also provide your final evaluation target for this indicator. If this survey is a final evaluation, please also provide the underweight data from the most recent population-based survey prior to FY11 (probably from your baseline survey). Indicate the year in which the data were collected. Does your program aim to improve the nutritional status of children 0-24 months old? Final Evaluation Target Indicator FY 11 In the survey, did your food aid program measure household dietary diversity, following the standard FANTA methodology for this indicator? See the "Definitions" tab for a description of the standard methodology for this indicator. 11 139 Yes/No Yes % stunted From w hich FY? % stunted What FY is the final evaluation? % stunted 59% FY # FY15 55% FY 11 # 6-59 mo 292,591 Yes/No Yes Yes/No Yes FY 11 FY 10 actual % actual % 19.1 - decrease % % 19.2 % % 19.3 % % 19.4 % % 14 Indicator Final Evaluation Target 2A: Anthropometry (Monitoring Indicators) 17 What is the estimated total number of children 6-59 months of age, living in your target geographic area? How many children 6-59 months of age live in your geographic area? (This is the population size, not the sample size. ) % of stunted (HAZ<-2) children 6-59 months of age Most recent FY prior to FY11 (enter n/a if FY11 w as the baseline) 18 % of children 0 - 23 mo old with WAZ < -2 Indicators Desired direction (+ / -) 1D: Stunting (Impact Indicator) 19 15 SECTION 2: Annual Monitoring Data This section asks for data about direct beneficiaries, coming from your routine monitoring system 16 Did your food aid program implement activities (deliver goods and services (assistance) to beneficiaries) in FY11? Did your food aid program implement activities to maintain or improve the nutritional status of beneficiaries in FY11? In the survey, did your food aid program measure the prevalence of stunting (HAZ <- 2) in children 6 - 59 mo of age? If you measured stunting for a different age group, or you used a different measure or cutoff, answer NO. What was the prevalence of stunting (HAZ <-2) in children 6 - 59 mo of age? If this survey is a baseline survey, please also provide your final evaluation target for this indicator. If this survey is a final evaluation, please also provide the stunting data from the most recent population-based survey prior to FY11 (probably from your baseline survey). Indicate the year in which the data were collected. What anthropometric indicators does your program use for regular monitoring of the nutritional status of beneficiaries? For each indicator, fill in the desired direction of change (increase or decrease) and the data for FY 11 and the previous year, FY 10. It is OK to leave prior year data blank if you do not have beneficiary data from the prior year. Fill out the table below with the ANTHROPOMETRIC indicators used by your program for annual monitoring of the nutritional status of your program's beneficiaries. Please write the precise definition for each indicator, including the measure used and the age group (e.g. % of children 0 - 23 mo old with WAZ < -2). In other words, do not write simply "Malnutrition rate" or "Recovery rate" or "% graduating from feeding program" without explaining what anthropometric measure and cutoff is used. Please provide only ANTHROPOMETRIC indicators which are a measure of the physical body. Acceptable anthropometric measures include prevalence of stunting (height for age Z - HfA), underweight (weight for age - WfA), wasting (weight for height WfH), weight gain, growth faltering (trend of weight gain), body mass index (BMI), middle-upper arm circumference (MUAC); average HfA Z score (HAZ), WfA Z score (WAZ), WfH Z score (WHZ); proportion of children/adults recuperating to defined cutoffs (e.g. WAZ 80% median). Measures such as breastfeeding, vaccination rates, or numbers of ration recipients are NOT anthropometric. Only include data for indicators that you monitor annually among direct beneficiaries. These data will be based on regular monitoring of your program beneficiaries and not on a representative sample survey of a broader population. DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION FY 11 140 Yes/No Yes FY 12 FY 13 FY 14 FY 15 actual % # beneficiaries target % target % target % target % 21.1 % # 63% 64% 65% 66% 21.2 % # 18% 20% 23% 25% 21.3 % # % % % % 21.4 % # % % % % Yes/No Yes FY 12 FY 13 FY 14 FY 15 # farmers # farmers # farmers # farmers # farmers 4,580 60,425 75,839 75,839 75,839 # technologies 4 25 25.1 25.2 25.3 25.4 25.5 25.6 25.7 25.8 25.9 25.10 # technologies 3 Organic fertilizer 26 2C: Agricultural Extension (Monitoring Indicator) 22 23 24 Did your program implement activities to improve the health, nutrition or hygiene 20 behaviors of beneficiaries in FY11? % of women taking iron supplements in last 7 days % children 0-6 months of age exclusively breastfed Future Targets Indicators What behavior change indicators does your program use for regular monitoring of beneficiaries? For each indicator, fill in data on the FY 11 indicator value (i.e. the result achieved) and the number of beneficiaries reached in FY11. Please provide future year targets for the indicator, as applicable. Use the drop down menu to select the indicator on which you are reporting. Give the percentage (%) of beneficiaries adopting the improved health, nutrition or hygiene behaviors. You may take a census or a sample of your beneficiaries. DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION. Only the indicators on the drop down menu can be included. See FFP Information Bulletin 07-02 (http://www.usaid.gov/our_work/humanitarian_assistance/ffp/fy08_ffpib_new_reporting.pdf) for further information on these indicators. For indicators with an *, the specific behaviors that comprise these indicators are to be defined by the awardee. See the "Definitions" tab for a definition of "beneficiaries". FY 11 21 2B: Behavior Change: Health, Nutrition, Hygiene (Monitoring Indicators) How many sustainable agricultural technologies did your program transfer in FY11? See the "Definitions" tab for more information about "agricultural technologies" Future Targets What are the sustainable agricultural technologies your program made available for transfer in FY11? If you transferred more than 10 technologies, you can list the others in the comments column to the right. Did your food aid program provide farmers with extension/outreach services in FY11? How many farmers (individuals, not households) received extension/outreach services in FY11? Please provide future year targets for number of farmer beneficiaries, as applicable. Improved bed system Improved pit/heap systems Compost preparation FY 11 What is the minimum number of sustainable agricultural technologies your program would like an individual farmer to use/adopt as a result of your program's assistance? See the "Definitions" tab for a definition of "minimum number." This number should be less than the response to question 22. 141 FY 12 FY 13 FY 14 FY 15 % beneficiary farmers % beneficiary farmers % beneficiary farmers % beneficiary farmers % beneficiary farmers % 44% 50% 55% 55% Yes/No No # communities FY 12 FY 13 FY 14 FY 15 # communities # communities # communities # communities # communities #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! 31 Yes/No Yes # communities 225 33 33.1 33.2 33.3 33.4 33.5 FY 12 FY 13 FY 14 FY 15 # communities # communities # communities # communities # communities 16 76 136 181 225 7% 34% 60% 80% 100% 2D: Disaster Early Warning Systems (Monitoring Indicator) Future Targets 27 How many of your program's targeted communities had improved physical infrastructure to mitigate the impact of shocks in FY11 as a result of your program's assistance? Please provide the future year targets for number of communities, as applicable. Future targets should be cumulative. For instance, if 25 communities have infrastructure in place in Year 1 and another 25 are added in Year 2, then the Year 2 target would be 50, not 25. DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION. 30 28 2E: Infrastructure To Mitigate Shocks (Monitoring Indicator) How many of your program's targeted communities had disaster early warning and response systems in place in FY11 as a result of your program's assistance? Please provide the future year targets for # of communities, as applicable. Future targets should be cumulative. For instance, if 25 communities have early warning systems in Year 1 and another 25 are added in Year 2, the Year 2 target would be 50, not 25. DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION. Future Targets FY 11 (Automatic Calculation) % of communities with disaster early warning systems in place 32 Community Place Development School Maintenance/Renovation/ Ground Raising Flood Shelter cum School Development 34 FY 11 How many communities does your program plan to assist to improve or develop infrastructure to mitigate the impact of shocks over the life of the award? What percentage (%) of program beneficiaries (farmers) adopted the minimum number of technologies in FY11? Please provide the future year targets, as applicable. DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION. Future Targets How many communities does your program plan to assist to develop disaster early warning and response systems over the life of the award? (Automatic Calculation) % of communities with disaster early warning systems in place Did your food aid program assist communities to improve or develop physical infrastructure to mitigate the impact of shocks in FY11? See the "Definitions" tab for a definition of "infrastructure" What kinds of physical infrastructure did your program improve or develop in FY11? If there are more than 5 kinds of infrastructure, you can list the others in the comments column to the right. FY 11 Did your food aid program assist communities to develop disaster early warning and response systems in FY11? See the "Definitions" tab for a definition of "disaster early warning and response system". 29 142 Yes/No Yes # communities 905 FY 12 FY 13 FY 14 FY 15 # communities # communities # communities # communities # communities 0 452 814 905 905 0% 50% 90% 100% 100% Yes/No Yes # communities 1,509 40 40.1 40.2 40.3 40.4 40.5 40.6 40.7 40.8 40.9 40.10 FY 12 FY 13 FY 14 FY 15 # communities # communities # communities # communities # communities 1,509 1,509 1,509 1,509 1,509 100% 100% 100% 100% 100% Analysis and planning capacity 37 Did your food aid program assist communities to strengthen community capacity in FY11? Community capacity refers to a community's ability to govern itself; to organize, analyze, plan, manage, problem-solve, implement actions, and represent its interests and participate in broader fora. This goes beyond targeted efforts to strengthen communities in nutrition, agriculture, infrastructure, early warning, or other topics covered elsewhere in the SAPQ. How many communities does your program plan to assist to strengthen safety nets over the life of the activity? Future Targets (Automatic Calculation) % of communities with disaster early warning systems in place What are the components of community capacity that your program strengthened in FY11? Select from the drop down menu. If there are more than 10 components, you can list the others in the comments column to the right. How many of your program's targeted communities had strengthened community capacity in FY11 as a result of your program's assistance? Please provide the future year targets for number of communities, as applicable. Future targets should be cumulative. For instance, if 25 communities have strengthened capacity in Year 1 and another 25 are added in Year 2, then the Year 2 target would be 50, not 25 DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION. 39 35 36 38 How many communities does your program plan to assist to strengthen community capacity over the life of the award? Implementation capacity Congratulations! You have finished the SAPQ How many of your programs targeted communities that had safety nets in place in FY11 as a result of your program's assistance? Please provide the future year targets for number of communities, as applicable. Future targets should be cumulative. For instance, if 25 communities have safety nets in place in Year 1 and another 25 are added in Year 2, then the Year 2 target would be 50, not 25. DO NOT PROVIDE DATA FROM A POPULATION BASED SURVEY SUCH AS A BASELINE OR FINAL EVALUATION. Did your food aid program assist communities to strengthen safety nets to address the needs of their most vulnerable members in FY11? A community-based safety net supported under a Title II development program can be a broadly defined system for addressing the food security needs of a community's most vulnerable members during a shock. A community￾based safety net is: managed and maintained by the community; internally resourced, at least in part; and can be year round or seasonal. Examples include community food banks or insurance schemes. Future Targets 2F: Safety Nets (Monitoring Indicator) 41 FY 11 (Automatic Calculation) % of communities with disaster early warning systems in place FY 11 2G: Community Capacity (Monitoring Indicator) SHOUHARDO II Final QPE Annex 6: Quantitative Tables SO1 Tables SO1 PROGRAM INDICATORS Table 1. SO1 key indicators, by region, well-being category and sex of household head Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline Average number of months of adequate household food provisioning All households 5.9 9.9 11.0 86.4 *** 7,085 6,916 7,148 Region Coast 5.4 10.1 10.8 100.0 *** 1,885 1,852 1,907 Haor 5.5 10.3 11.1 101.8 *** 1,672 1,701 1,689 Mid Char 7.0 9.9 10.8 54.3 *** 1,798 1,745 1,652 North Char 5.9 9.4 11.0 86.4 *** 1,730 1,618 1,900 Well-being category Extreme poor 4.6 9.3 10.6 130.4 *** 1,195 1,237 1,333 Poor 6.0 10.0 11.1 85.0 *** 4,566 4,435 4,511 Middle 7.7 10.6 11.5 49.4 *** 638 599 577 Middle rich 9.1 11.2 11.6 27.5 *** 428 409 452 Rich 10.7 11.5 11.9 11.2 *** 252 235 274 Sex of headship Female 4.2 9.0 10.6 152.4 *** 957 1,009 1,221 Male 6.2 10.1 11.1 79.0 *** 6,128 5,907 5,927 Average household dietary diversity score All households 4.8 5.9 8.7 81.3 *** 7,084 6,916 7,146 Region Coast 4.7 5.7 8.1 72.9 *** 1,885 1,852 1,907 Haor 4.8 6.8 9.0 88.9 *** 1,672 1,701 1,689 Mid Char 4.8 5.0 8.2 71.9 *** 1,797 1,745 1,652 North Char 4.9 5.4 8.6 76.8 *** 1,730 1,618 1,898 Well-being category Extreme poor 4.5 5.6 8.3 85.0 *** 1,195 1,237 1,333 Poor 4.8 6.0 8.7 82.6 *** 4,565 4,435 4,509 Middle 5.4 6.4 9.1 69.2 *** 638 599 577 Middle rich 6.0 6.6 9.3 55.1 *** 428 409 452 Rich 6.9 7.4 10.0 44.6 *** 252 235 274 Sex of headship Female 4.4 5.3 7.8 77.3 *** 957 1,009 1,221 Male 4.9 6.0 8.8 81.1 *** 6,127 5,907 5,925 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. IR 1.1 Table 2. Indicators of agricultural production Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Percent of households cultivating field crops in the previous year 33.4 27.9 44.1 32.0 * Percent of households growing vegetables in the previous year 16.5 63.8 63.0 281.8 * Percent of households raising livestock in the previous year 61.4 78.7 86.0 40.1 * Percent of households raising fish in previous year 2.4 7.2 9.3 287.5 * Total number of (the four) agricultural activities engaged in in previous year 1.1 1.8 2.0 81.8 * Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 3. Percent of field crop producers adopting the minimum number of technologies (three) Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 42.1 71.0 92.9 120.7 *** 2,411 2,051 2,976 Region Coast 15.4 67.7 83.7 444.1 *** 296 334 517 Haor 41.6 65.0 92.4 121.9 *** 724 610 856 Mid Char 41.7 68.1 90.2 116.3 *** 714 689 727 North Char 44.8 84.8 95.8 113.8 *** 677 418 876 Well-being category Extreme poor 42.0 64.9 91.5 118.0 *** 182 181 373 Poor 41.1 69.8 92.6 125.3 *** 1,286 1,212 1,804 Middle 40.8 78.3 97.2 138.5 *** 411 273 314 Middle rich 46.9 81.0 96.0 104.7 *** 325 231 278 Rich 54.1 86.1 99.0 83.1 *** 206 153 207 Sex of headship Female 50.4 60.9 81.6 61.8 *** 120 119 233 Male 41.6 71.6 93.7 125.0 *** 2,291 1,932 2,743 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 4. Average number of field crops produced in previous year Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 1.9 2.1 3.1 63.2 *** 2,411 2,051 2,976 Region Coast 1.3 1.7 1.8 42.1 296 334 517 Haor 1.8 1.7 2.8 54.8 *** 724 610 856 Mid Char 1.9 2.4 3.6 88.6 *** 714 689 727 North Char 2.1 2.6 3.6 74.5 *** 677 418 876 Well-being category Extreme poor 1.7 1.8 3.0 72.5 *** 182 181 373 Poor 1.8 2.1 3.1 67.8 *** 1,286 1,212 1,804 Middle 1.9 2.1 3.3 70.4 *** 411 273 314 Middle rich 2.2 2.3 3.4 53.6 *** 325 231 278 Rich 2.6 2.8 3.4 30.9 *** 206 153 207 Sex of headship Female 1.5 1.8 2.6 77.3 *** 120 119 233 Male 1.9 2.1 3.2 66.7 *** 2,291 1,932 2,743 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 5. Percent of vegetable producers adopting the minimum number of improved technologies (three) Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 39.3 72.5 92.8 136.1 *** 1,178 4,286 4,210 Region Coast 41.9 60.8 91.9 119.5 *** 235 948 938 Haor 35.6 74.5 91.1 156.2 *** 304 1,212 1,225 Mid Char 48.3 57.6 91.9 90.1 *** 249 1,054 802 North Char 40.2 78.3 95.5 137.7 *** 390 1,072 1,245 Well-being category Extreme poor 22.1 65.5 90.2 308.9 *** 101 641 688 Poor 39.1 73.6 93.4 139.0 *** 630 2,731 2,705 Middle 53.6 75.4 94.3 75.9 *** 154 411 373 Middle rich 48.6 75.6 94.4 94.4 *** 161 308 275 Rich 54.5 76.7 94.5 73.3 *** 131 195 188 Sex of headship Female 17.3 66.4 91.2 427.2 *** 56 483 544 Male 40.8 73.1 93.1 127.9 *** 1,122 3,803 3,666 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 6. Average number of vegetable crops produced in previous year Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 3.6 5.0 8.4 133.3 *** 1,183 4,293 4,212 Region Coast 3.9 4.9 7.8 99.0 *** 235 950 940 Haor 3.7 5.0 8.4 125.3 *** 305 1,213 1,225 Mid Char 3.1 4.5 6.5 107.4 *** 249 1,056 802 North Char 3.6 5.3 9.2 153.3 *** 394 1,074 1,245 Well-being category Extreme poor 2.7 4.6 8.4 207.1 *** 104 641 668 Poor 3.5 5.0 8.4 143.1 *** 632 2,736 2,707 Middle 4.4 4.9 9.0 105.0 *** 154 412 373 Middle rich 4.7 5.5 8.5 79.0 *** 161 309 275 Rich 5.6 6.7 9.2 65.1 *** 131 195 188 Sex of headship Female 2.7 4.7 7.5 177.8 *** 57 483 544 Male 3.7 5.0 8.5 130.8 *** 1,126 3,810 3,668 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 7. Percent of livestock producers adopting minimum number of improved technologies (three) Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 2.6 12.3 44.7 1619.2 4,468 5,150 6,117 Region Coast 1.4 3.9 21.5 1435.7 1,158 1,139 1,668 Haor 2.4 10.0 42.1 1654.2 946 1,380 1,489 Mid Char 6.5 14.9 41.8 543.1 1,098 1,276 1,265 North Char 1.3 14.7 52.1 3907.7 1,265 1,355 1,693 Well-being category Extreme poor 1.7 7.7 39.4 2217.6 *** 606 826 1,080 Poor 2.2 13.2 46.1 1995.5 *** 2,832 3,302 3,909 Middle 4.8 11.7 46.2 862.5 *** 498 489 503 Middle rich 9.8 15.0 48.8 398.0 *** 318 333 385 Rich 10.3 24.7 49.1 376.7 *** 213 199 239 Sex of headship Female 1.7 7.7 32.7 1823.5 *** 480 625 965 Male 2.7 12.9 46.8 1633.3 *** 3,988 4,525 5,152 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 8. Percent of fish producers adopting minimum number of improved technologies (three) Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 35.0 71.9 87.3 149.4 *** 173 390 575 Region Coast n/a 50.0 80.0 11 2059 938 Haor 28.7 67.5 81.7 184.7 *** 75 200 222 Mid Char n/a 63.2 95.7 20 43 79 North Char 43.5 81.7 94.4 117.0 *** 67 27 215 Well-being category Extreme poor n/a 67.8 85.0 5 29 57 Poor 23.5 72.3 87.3 271.5 *** 49 171 291 Middle 23.5 65.5 90.2 283.8 *** 31 54 81 Middle rich 50.0 71.8 94.4 88.8 *** 42 74 76 Rich 61.5 81.8 81.8 33.0 *** 47 62 69 Sex of headship Female n/a 73.2 86.9 24 102 171 Male 45.6 69.5 88.0 93.0 *** 149 288 404 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 9. Percent of households accessing agri-inputs, finances and services Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline All households 93.1 96.5 98.4 5.7 *** 2,411 2,051 2,976 Region Coast 82.7 91.9 94.6 14.4 ** 296 334 517 Haor 89.7 97.3 98.5 9.8 *** 724 610 856 Mid Char 94.1 95.6 96.9 3.0 ** 714 689 727 North Char 98.4 96.2 99.3 0.9 ** 677 418 876 Well-being category Extreme poor 94.6 98.0 96.9 2.4 182 181 373 Poor 92.4 95.9 98.6 6.7 *** 1,286 1,212 1,804 Middle 95.7 98.3 100.0 4.5 ** 411 273 314 Middle rich 94.5 98.0 99.2 5.0 ** 325 231 278 Rich 94.9 98.6 99.0 4.3 206 153 207 Sex of headship Female 97.5 95.5 94.8 -2.8 120 119 233 Male 92.9 96.5 98.6 6.1 *** 2,291 1,932 2,743 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. IR 1.2 Table 10. Indicators of income and wealth, by region, well-being category, and sex of household head Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline Mean household monthly income per capita (2014 Taka) All households 647.1 951.0 980.5 51.5 *** 7,084 6,888 7046 Region Coast 578.3 1015.8 1082.9 87.3 *** 1,885 1,841 1,872 Haor 667.7 973.0 948.7 42.1 *** 1,672 1,697 1,673 Mid Char 652.6 948.5 991.0 51.9 *** 1,797 1,738 1,625 North Char 630.5 917.9 998.2 58.3 *** 1,730 1,612 1,876 Well-being category Extreme poor 557.5 848.5 901.4 61.7 *** 1,195 1,237 1,324 Poor 645.0 946.6 967.4 50.0 *** 4,566 4,418 4,465 Middle 742.4 1096.3 1,123.0 51.3 *** 638 594 563 Middle rich 874.8 1231.6 1,287.3 47.2 *** 428 405 436 Rich 1229.4 1674.3 1,763.5 43.4 ** 251 233 257 Sex of headship Female 612.3 863.4 993.3 62.2 *** 957 1,009 1,221 Male 652.3 965.2 903.4 38.5 *** 6,121 5,907 5,927 Mean number of months of hh employment (per capita) All households 2.70 3.08 3.35 24.0 *** 7,085 6,916 7148 Region Coast 2.12 2.49 2.59 22.1 ** 1,885 1,852 1,907 Haor 2.70 3.11 2.93 8.5 *** 1,672 1,701 1,689 Mid Char 2.52 2.71 3.48 38.0 *** 1,798 1,745 1,652 North Char 2.87 3.30 3.88 35.2 *** 1,730 1,618 1,900 Well-being category Extreme poor 3.20 3.79 3.99 24.8 *** 1,195 1,237 1,333 Poor 2.66 2.97 3.26 22.4 *** 4,566 4,435 4,511 Middle 2.09 2.53 2.53 20.8 * 638 599 577 Middle rich 1.69 1.93 2.17 28.1 * 428 409 452 Rich 1.58 2.02 2.01 26.9 252 235 274 Sex of headship Female 3.72 4.29 4.50 21.1 *** 957 1,009 1,221 Male 2.55 2.88 3.14 23.2 *** 6,121 5,907 5,927 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 11. Ownership of various assets Type of practice Baseline Midterm Endline Percent difference (Endline - Baseline) Domestic Assets 25.1 32.3 39.1 55.8 * Productive Assets 15.9 21.5 27.9 75.5 * Animal Assets 17.5 21.9 30.0 71.4 * Tree Assets 137.5 224.8 443.1 222.3 * Total Assets 196.0 300.8 540.1 175.6 * Note: Stars indicate endline-baseline difference is statistically significant at the 5% level. Table 12. Domestic Asset Index, by region, well-being category, program approach, and sex of household head Baseline Midterm End line Percent difference (Endline - Baseline) p-value for difference Number of observations Indicator Baseline Midterm Endline All households 25.1 32.6 39.0 55.5 0.000 *** 7,085 6,916 7,148 All PEP households 22.5 30.1 36.3 61.1 0.000 *** 5,765 5,672 5,844 Region Coast 27.0 31.2 40.7 50.8 0.000 *** 1,885 1,852 1,907 Haor 23.1 30.8 37.4 61.8 0.000 *** 1,672 1,701 1,689 Mid Char 29.6 36.6 43.8 47.9 0.000 *** 1,798 1,745 1,652 North Char 24.9 33.0 38.5 54.6 0.000 *** 1,730 1,618 1,900 Well-being category Extreme poor 16.5 22.5 22.4 35.7 0.000 *** 1,195 1,237 1,333 Poor 24.1 32.1 32.1 33.5 0.000 *** 4,566 4,435 4,511 Middle 42.3 46.9 46.9 10.9 0.000 *** 638 599 577 Middle rich 53.3 65.1 65.1 22.2 0.000 *** 428 409 452 Rich 91.1 98.6 98.6 8.3 0.076 * 252 235 274 Program approach MCHN/PEP (PEP only) 22.7 30.5 36.0 58.4 0.000 *** 2,888 2,766 2,945 PM2A 33.0 39.6 49.5 50.0 0.000 *** 4,197 4,150 4,203 Sex of headship Female 15.8 21.5 25.7 62.5 0.000 *** 957 1,009 1,221 Male 26.5 34.4 41.5 56.5 0.000 *** 6,128 5,907 5,927 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5% (**) or 1% (***) levels. Table 13. Mean household remittance per year (2010 Taka) Indicator Baseline Midterm End line Percent difference (Endline - Baseline) p-value for difference Number of observations Baseline Midterm End line All households 1,666 2,454 6,185 271.3 0.000 *** 7,085 6,911 7,148 All PEP households 1,285 2,230 4,868 278.8 0.000 *** 5,765 5,672 5,844 Region Coast 4,686 5,768 27,621 489.5 0.000 *** 1,885 1,849 1,907 Haor 1,670 2,571 5,417 224.4 0.002 *** 1,672 1,701 1,689 Mid Char 2,275 3,184 6,029 165.0 0.004 *** 1,798 1,745 1,652 North Char 963 1,512 4,369 353.7 0.001 *** 1,730 1,616 1,900 Well-being category Extreme poor 939 1,811 1,811 92.8 0.001 *** 1,195 1,237 1,333 Poor 1,371 2,095 2,342 70.8 0.000 *** 4,566 4,432 4,511 Middle 3,311 4,876 6,816 105.9 0.011 ** 638 598 577 Middle rich 8,837 8,202 13,324 50.8 0.083 * 428 408 452 Rich 8,427 12,589 20,323 141.2 0.024 ** 252 235 274 Program approach MCHN/PEP (PEP only) 1,266 2,055 4,334 242.3 0.002 *** 2,888 2,764 2,945 PM2A 2,977 3,767 12,457 318.4 0.000 *** 4,197 4,147 4,203 Sex of headship Female 3,272 5,463 11,417 248.9 0.000 *** 957 1,008 1,221 Male 1,423 1,969 5,224 267.2 0.000 *** 6,128 5,903 5,927 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5% (**) or 1% (***) levels. Table 14. Mean household monthly remittances per capita (2010 Taka) Indicator Baseline Midterm End line Percent difference (Endline - Baseline) p-value for difference Number of observations Baseline Midterm End line All households 34.82 59.91 80.41 130.1 0.000 *** 7,085 6,916 7,148 All PEP households 29.21 47.15 71.89 146.2 0.000 *** 5,765 5,672 5,844 Region Coast 81.47 107.55 282.28 245.9 0.000 *** 1,885 1,852 1,907 Haor 33.67 57.37 72.45 114.1 0.002 *** 1,672 1,701 1,689 Mid Char 51.88 77.43 106.59 106.7 0.010 ** 1,798 1,745 1,652 North Char 21.52 47.83 51.71 140.4 0.008 *** 1,730 1,618 1,900 Well-being category Extreme poor 29.14 58.04 74.87 155.2 0.000 *** 1,195 1,237 1,333 Poor 29.09 51.1 70.87 142.8 0.000 *** 4,566 4,435 4,511 Middle 54.77 101.96 154.80 182.6 0.023 ** 638 599 577 Middle rich 143.13 164.87 211.69 47.9 0.126 428 409 452 Rich 141.81 232.26 199.85 40.9 0.269 252 235 274 Program approach MCHN/PEP (PEP only) 28.74 52.94 65.91 128.3 0.000 *** 2,888 2,766 2,945 PM2A 54.75 82.83 129.79 137.1 0.000 *** 4,197 4,150 4,203 Sex of headship Female 89.53 184.3 189.28 111.0 0.000 *** 957 1,009 1,221 Male 26.54 39.79 60.53 127.2 0.000 *** 6,128 5,907 5,927 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5% (**) or 1% (***) levels. Table 15. Indicators of economic vulnerability Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Migration 32.2 34.4 33.8 5.0 * Selling labor in advance 9.8 7.9 5.2 -46.9 * Taking out interest bearing loan 21.3 19.6 16.7 -21.6 * Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5% (**) or 1% (***) levels. Table 16. Indicators of access to markets Indicator Baseline Midterm Endline Percent difference (Endline - Baseline) Number of observations Baseline Midterm Endline Percent of households that purchase food for whom the nearest food vendor is close by (<30 minute walk away) All households 57.8 64.1 64.3 11.3 *** 6,802 6,643 6,938 Region Coast 57.5 43.8 47.0 -18.2 *** 1,842 1,800 1,853 Haor 49.6 60.7 60.0 21.0 *** 1580 1642 1,633 Mid Char 65.0 59.5 72.0 10.7 *** 1728 1673 1,595 North Char 63.3 73.3 67.9 7.3 *** 1652 1528 1,857 Well-being category Extreme poor 58.8 60.3 58.6 -0.3 1130 1185 1,284 Poor 57.8 65.3 66.1 14.3 *** 4434 4290 4,396 Middle 54.3 62.6 65.4 20.5 *** 623 568 562 Middle rich 58.3 58.8 66.9 14.8 393 387 433 Rich 50.5 68.8 63.9 26.5 * 216 212 262 Sex of headship Female 60.4 61.7 67.3 11.4 *** 910 970 1,168 Male 57.4 64.5 63.8 11.2 *** 5892 5673 5,770 Percent of households that sell food for whom the nearest point of sale is close by (<30 minute walk away) All households 39.5 48.9 51.5 30.2 *** 1206 1572 2,781 Region Coast 43.8 20.4 38.3 -12.5 179 298 526 Haor 45.5 48.4 50.7 11.4 * 328 466 806 Mid Char 34.2 39.5 53.6 56.8 *** 372 418 673 North Char 34.4 57.8 52.8 53.4 ** 327 390 776 Well-being category Extreme poor 45.5 39.7 43.2 -5.0 87 132 377 Poor 37.2 48.7 53.1 42.7 *** 558 880 1,654 Middle 45.1 53.2 55.9 24.1 * 206 211 295 Middle rich 42.6 49.4 54.5 28.1 * 207 193 268 Rich 42.9 64.4 56.5 31.8 147 155 196 Sex of headship Female 35.2 34.0 55.9 58.9 *** 53 98 241 Male 39.8 50.0 51.2 28.7 *** 1153 1474 2,540 Percent of households that purchase agricultural inputs for whom the nearest vendor is close by (<30 minute walk away) All households 45.3 58.0 53.4 17.9 *** 2,297 3,022 3,996 Region Coast 35.4 29.5 35.0 -1.2 271 577 783 Haor 44.0 54.3 50.2 14.2 *** 674 904 1,140 Mid Char 42.1 49.9 61.1 45.3 *** 703 797 928 North Char 49.8 70.3 55.9 12.3 *** 649 744 1,145 Well-being category Extreme poor 51.1 55.2 41.8 -18.2 *** 169 302 573 Poor 45.0 59.1 56.1 24.7 *** 1,226 1,878 2,461 Middle 41.8 54.3 55.3 32.3 *** 392 367 407 Middle rich 47.1 46.5 55.2 17.3 306 287 327 Rich 40.6 63.6 56.1 38.1 * 203 187 227 Sex of headship Female 38.1 51.6 59.6 56.3 *** 106 203 350 Male 45.7 58.5 52.9 15.8 *** 2,191 2,819 3,646 Note: Stars indicate endline-baseline difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. BY PEP/NON-PEP Table 17. SO1 key indicators (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non-PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Food security indicators Average number of months of adequate food provisioning 8.8 11.6 24.1 * 5.7 11.0 48.2 * 24.0 Average household dietary diversity score 5.9 9.4 37.2 * 4.7 8.6 45.3 * 8.1 Percent of households with hunger Agricultural production indicators Adoption of minimum number of technologies Field crops 45.9 97.3 112.0 41.2 92.3 124.0 12.0 Vegetable production 51.9 94.4 81.9 36.6 92.7 153.3 71.4 * Livestock rearing 7.5 47.7 536.0 2.1 44.5 2019.0 1483.0 * Fish rearing 50.7 87.8 73.2 20.4 87.5 328.9 255.7 Percent of households accessing agri-inputs, finances and services 95.1 99.2 4.3 92.7 98.3 6.0 1.7 * Indicators of income and market access Mean household monthly income per capita 884.26 1493.7 40.8 * 626.5 1250.4 49.9 * 9.1 Mean number of months of household employment (per capita) 1.86 2.29 18.8 * 2.8 3.4 19.5 * 0.7 Percent of purchasing and selling households with nearest vender close by Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. BY MALE/FEMALE HHH Table 18. SO1 key indicators, by sex of household head Male headed households Female headed households Difference (Female vs. male) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Food security indicators Average number of months of adequate food provisioning 6.2 11.1 79.0 * 4.2 10.6 152.4 * 73.3 Average household dietary diversity score 4.9 8.8 79.6 * 4.4 7.8 77.3 * -2.3 Percent of households with hunger * Agricultural production indicators Adoption of minimum number of technologies Field crops 41.6 93.7 125.2 50.4 81.6 61.9 -63.3 Vegetable production 40.8 93.1 128.2 17.3 91.2 427.2 299.0 * Livestock rearing 2.7 49.8 1744.4 1.7 32.7 1823.5 79.1 * Fish rearing 45.6 88.0 93.0 20.7 86.9 319.8 226.8 Percent of households accessing agri-inputs, financies and services 92.9 98.6 6.1 97.5 94.8 -2.8 -8.9 * Indicators of income and market access Mean household monthly income per capita 652.3 1290.8 49.5 612.3 1152.9 46.9 * -2.6 Mean number of months of household employment (per capita) 2.55 3.14 18.79 3.72 4.50 17.33 * -1.5 Percent of purchasing and selling households with nearest vender close by Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. SO2 Tables SO2 PROGRAM INDICATORS Table 19. Overview: Moderate and severe malnutrition among children under five and children under two Indicator Baseline Midterm Endline – Index child only Endline – All children Percent difference (Endline index child only - Baseline) Under fives Stunting (6-59m) 61.7 52.7 48.8 50.4 -21.0 *** Wasting (6-59m) 15.6 9.0 12.2 11.2 -21.5 Underweight (0-59m) 42.2 34.2 35.3 35.9 -16.3 *** Severe stunting (6-59m) 30.8 22.6 18.4 19.6 -40.4 *** Severe wasting (6-59m) 6.3 1.6 3.1 2.7 -51.6 *** Severe underweight (0-59m) 13.5 9.8 10.1 10.3 -25.4 ** Under twos Stunting (6-23m) 55.8 52.0 42.9 43.7 -23.1 *** Wasting (6-23m) 18.2 11.7 14.2 13.0 -21.9 Underweight (0-23m) 33.2 32.1 29.5 29.5 -11.1 Severe stunting (6-23m) 31.4 23.2 15.1 16.7 -51.9 *** Severe wasting (6-23m) 8.1 2.5 5.0 4.4 -38.9 Severe underweight (0-23m) 10.9 9.3 10.8 10.4 -1.1 Note: Stars indicate endline-baseline difference is statistically significant at the 5% level. Table 20. Child malnutrition indicators (moderate), stunting by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under five stunted (6-59m) All households 61.7 52.7 48.8 -21.0 *** 2,384 2,167 2,143 Sex of child Female 56.5 52.2 47.8 -15.3 ** 1,145 1,027 1,016 Male 66.1 53.2 49.7 -24.9 *** 1,239 1,140 1,127 Region Coast 58.1 54.0 44.5 -23.4 *** 753 718 707 Haor 65.3 52.8 57.9 -11.3 * 548 583 517 Mid Char 53.1 47.8 44.6 -16.0 ** 566 454 468 North Char 62.8 54.8 39.6 -37.0 *** 517 412 451 Well-being category Extreme poor 64.2 62.2 50.9 -20.7 ** 272 271 296 Poor 62.0 51.5 49.8 -19.8 *** 1,704 1,545 1,493 Middle 52.9 50.6 38.1 -28.0 ** 197 180 168 Middle rich 52.1 52.4 25.3 -51.5 *** 133 106 119 Rich 59.5 40.6 29.2 -51.0 *** 77 64 67 Program approach MCHN/PEP 62.2 52.7 49.6 -20.2 *** 961 871 918 PM2A 59.9 52.6 45.2 -24.6 *** 1423 1296 1225 Sex of headship Female 72.9 54.4 38.7 -46.9 *** 170 165 256 Male 61.0 52.6 49.5 -18.8 2,214 2,002 1,887 Percent of children under two stunted (6-23m) All households 55.8 52.0 42.9 -23.1 *** 727 782 845 Sex of child Female 50.4 46.1 41.2 -18.2 359 376 408 Male 61.0 58.5 45.0 -26.2 *** 368 406 437 Region Coast 44.0 47.8 33.2 -24.7 ** 243 277 314 Haor 59.3 48.8 50.1 -15.5 180 208 208 Mid Char 50.1 45.2 32.9 -34.4 ** 142 162 175 North Char 55.9 61.7 40.6 -27.3 ** 162 135 148 Well-being category Extreme poor 63.0 54.8 42.5 -32.6 ** 72 83 115 Poor 56.3 51.3 44.9 -20.2 ** 518 565 594 Middle 41.1 58.0 24.9 -39.5 ** 65 77 63 Middle rich 33.9 52.8 17.1 -49.5 48 39 46 Rich ‡ ‡ ‡ ‡ 24 18 27 Program approach MCHN/PEP 57.9 52.1 44.1 -23.8 *** 285 296 357 PM2A 49.4 51.6 39.0 -21.0 ** 442 486 488 Sex of headship Female 68.8 35.6 41.2 -40.1 * 32 54 100 Male 55.4 52.6 43.1 -22.3 *** 695 728 745 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 21. Child malnutrition indicators (severe), stunting by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under five severely stunted (6-59m) All households 30.8 22.6 18.4 -40.4 *** 2,384 2,167 2,143 Sex of child Female 27.1 22.4 18.3 -32.6 *** 1,145 1,027 1,016 Male 34.0 22.8 18.5 -45.6 *** 1,239 1,140 1,127 Region Coast 30.0 23.4 17.3 -42.2 *** 753 718 707 Haor 36.7 25.6 22.5 -38.7 *** 548 583 517 Mid Char 22.1 20.3 15.1 -31.6 * 566 454 468 North Char 28.9 19.2 14.8 -48.8 *** 517 412 451 Well-being category Extreme poor 37.1 27.5 20.6 -44.5 *** 272 271 296 Poor 29.8 22.2 18.5 -37.8 *** 1,704 1,545 1,493 Middle 28.4 22.2 12.6 -55.5 *** 197 180 168 Middle rich 31.1 14.3 10.2 -67.2 *** 133 106 119 Rich 33.3 16.4 5.6 -83.1 *** 77 64 67 Program approach MCHN/PEP 32.5 21.8 19.2 -36.7 *** 961 871 918 PM2A 30.3 22.9 15.7 -51.9 *** 1423 1296 1225 Sex of headship Female 35.7 26.7 14.4 -59.6 *** 170 165 256 Male 30.5 22.4 18.7 -38.8 *** 2,214 2,002 1,887 Percent of children under two severely stunted (6-23m) All households 31.4 23.2 15.1 -51.9 *** 727 782 746 Sex of child Female 25.1 20.0 14.6 -42.0 ** 359 376 388 Male 37.5 26.8 15.8 -57.9 *** 368 406 358 Region Coast 24.2 22.1 13.2 -45.4 *** 243 277 261 Haor 35.6 23.8 15.1 -57.7 *** 180 208 181 Mid Char 22.1 17.8 9.1 -59.0 ** 142 162 162 North Char 31.4 25.6 18.8 -40.0 * 162 135 142 Well-being category Extreme poor 43.6 32.1 11.7 -73.1 *** 72 83 103 Poor 30.3 22.3 16.8 -44.6 *** 518 565 518 Middle 22.0 22.7 6.0 -72.6 ** 65 77 57 Middle rich 17.4 17.4 10.2 -41.6 48 39 45 Rich ‡ ‡ ‡ ‡ 24 18 23 Program approach MCHN/PEP 32.6 23.3 15.9 -51.1 *** 285 296 305 PM2A 27.8 23.0 12.5 -55.1 *** 442 486 441 Sex of headship Female 39.5 11.3 7.2 -81.9 * 32 54 88 Male 31.2 23.6 15.8 -49.5 *** 695 728 568 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 22. Child malnutrition indicators (moderate), wasting by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under five wasted (6-59m) All households 15.6 9.0 12.2 -21.5 2,384 2,167 2,143 Sex of child Female 15.4 9.2 11.1 -28.1 * 1,145 1,027 1,016 Male 15.7 8.8 13.4 -14.8 1,239 1,140 1,127 Region Coast 15.5 13.9 8.6 -44.3 *** 753 718 707 Haor 20.3 9.2 10.5 -48.2 ** 548 583 517 Mid Char 11.1 9.8 12.0 7.9 566 454 468 North Char 12.6 7.2 15.3 21.7 517 412 451 Well-being category Extreme poor 20.0 9.9 13.4 -32.9 272 271 296 Poor 15.1 8.8 12.3 -18.7 1,704 1,545 1,493 Middle 13.4 11.1 7.3 -45.3 * 197 180 168 Middle rich 10.3 9.7 8.9 -14.1 133 106 119 Rich 13.3 4.0 11.1 -16.8 77 64 67 Program approach MCHN/PEP 15.5 8.1 12.8 -17.7 961 871 918 PM2A 15.9 11.7 10.4 -34.6 *** 1423 1296 1225 Sex of headship Female 14.2 13.5 15.2 6.9 170 165 256 Male 15.6 8.7 12.0 -23.0 * 2,214 2,002 1,887 Percent of children under two wasted (6-23m) All households 18.2 11.7 14.2 -21.9 727 782 746 Sex of child Female 22.8 10.9 13.1 -42.8 ** 359 376 388 Male 13.8 12.5 15.7 13.6 368 406 358 Region Coast 18.0 15.7 11.5 -36.1 * 243 277 261 Haor 22.0 13.4 13.9 -36.8 180 208 181 Mid Char 12.8 10.7 12.2 -4.8 142 162 162 North Char 15.6 8.4 16.3 4.6 162 135 142 Well-being category Extreme poor 13.0 10.6 14.8 14.1 72 83 103 Poor 19.7 12.1 15.0 -24.1 518 565 518 Middle 13.6 8.3 4.1 -70.1 ** 65 77 57 Middle rich 16.4 9.2 7.7 -52.9 48 39 45 Rich ‡ ‡ ‡ 24 18 23 Program approach MCHN/PEP 18.6 10.6 15.6 -16.3 285 296 305 PM2A 17.1 14.6 10.0 -41.6 *** 442 486 441 Sex of headship Female 5.1 5.9 16.7 227.8 32 54 88 Male 18.6 11.9 14.0 -24.7 695 728 658 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 23. Child malnutrition indicators (severe), wasting by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under five severely wasted (6-59m) All households 6.3 1.6 3.1 -51.6 *** 2,384 2,167 2,143 Sex of child Female 7.7 1.7 3.1 -60.1 *** 1,145 1,027 1,016 Male 5.1 1.5 3.0 -40.8 * 1,239 1,140 1,127 Region Coast 5.2 5.4 1.7 -66.7 *** 753 718 707 Haor 9.5 1.1 2.4 -75.1 *** 548 583 517 Mid Char 4.2 1.4 3.6 -14.3 566 454 468 North Char 3.9 1.6 3.9 0.0 517 412 451 Well-being category Extreme poor 7.7 2.3 2.5 -67.8 ** 272 271 296 Poor 6.4 1.3 3.3 -48.1 *** 1,704 1,545 1,493 Middle 2.9 4.0 n/a 197 180 168 Middle rich 3.2 2.0 3.1 -2.2 133 106 119 Rich 2.7 0.4 3.4 24.1 77 64 67 Program approach MCHN/PEP 6.4 1.1 3.1 -51.4 *** 961 871 918 PM2A 6.0 3.1 2.8 -53.2 *** 1423 1296 1225 Sex of headship Female 6.0 5.1 3.9 -35.2 170 165 256 Male 6.3 1.4 3.0 -52.7 *** 2,214 2,002 1,887 Percent of children under two severely wasted (6-23m) All households 8.1 2.5 5.0 -38.9 727 782 746 Sex of child Female 11.2 2.5 6.4 -43.3 359 376 388 Male 5.2 2.5 3.2 -38.5 368 406 358 Region Coast 5.4 4.8 2.5 -54.3 243 277 261 Haor 12.0 2.4 5.2 -57.1 * 180 208 181 Mid Char 4.6 2.0 1.8 -60.7 142 162 162 North Char 5.0 2.3 6.9 38.0 162 135 142 Well-being category Extreme poor 3.8 n/a 4.8 25.3 72 83 103 Poor 9.4 2.8 5.4 -42.2 518 565 518 Middle 4.5 3.5 n/a 65 77 57 Middle rich 5.4 1.8 1.8 -67.4 48 39 45 Rich ‡ ‡ ‡ ‡ 24 18 23 Program approach MCHN/PEP 8.8 1.9 5.6 285 296 305 PM2A 6.0 4.2 2.9 -52.2 * 442 486 441 Sex of headship Female 3.7 4.6 1.6 -57.8 32 54 88 Male 8.3 2.4 5.2 -37.0 695 728 658 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 24. Child malnutrition indicators (moderate), underweight by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under five underweight (0-59m) All households 42.2 34.2 35.3 -16.3 *** 2,594 2,380 2,326 Sex of child Female 37.2 34.9 36.4 -2.2 1,265 1,132 1,093 Male 46.7 33.4 34.3 -26.6 *** 1,329 1,248 1,233 Region Coast 40.2 36.9 31.2 -22.4 *** 827 789 764 Haor 46.0 36.3 38.3 -16.8 ** 594 633 574 Mid Char 35.3 32.4 35.9 1.6 616 504 497 North Char 42.0 31.2 31.7 -24.5 ** 557 454 491 Well-being category Extreme poor 40.8 39.1 38.5 -5.5 310 301 324 Poor 43.0 33.6 35.7 -17.0 *** 1,851 1,695 1,613 Middle 41.6 34.4 23.5 -43.5 *** 210 197 188 Middle rich 32.9 31.2 21.8 -33.8 * 141 118 127 Rich 33.6 24.8 24.0 -28.6 81 68 74 Program approach MCHN/PEP 42.6 34.4 36.3 -14.7 ** 1053 963 989 PM2A 41.0 33.3 31.9 -22.2 *** 1541 1417 1337 Sex of headship Female 46.5 30.1 31.1 -33.2 * 187 182 276 Male 42.0 34.4 35.7 -15.1 *** 2,407 2,198 2,050 Percent of children under two underweight (0-23m) All households 33.2 32.1 29.5 -11.1 937 995 929 Sex of child Female 29.6 29.4 29.5 -0.2 479 481 465 Male 36.8 34.9 29.5 -19.8 458 514 464 Region Coast 28.9 31.7 24.6 -15.0 317 348 318 Haor 36.0 34.4 31.4 -12.9 226 258 238 Mid Char 22.2 28.5 26.0 17.0 192 212 191 North Char 35.6 30.5 29.6 -16.9 202 177 182 Well-being category Extreme poor 33.4 32.4 29.3 -12.3 110 113 131 Poor 34.0 32.4 31.4 -7.6 665 715 638 Middle 30.7 30.4 8.7 -71.7 *** 78 94 77 Middle rich 19.4 21.4 16.4 -15.6 56 51 53 Rich ‡ ‡ 13.8 ‡ 28 22 30 Program approach MCHN/PEP 34.1 32.9 31.0 -9.0 377 388 376 PM2A 29.9 29.6 24.8 -16.9 560 607 553 Sex of headship Female 14.8 9.2 32.6 120.4 49 71 108 Male 33.8 33.2 29.2 -13.6 888 924 821 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 25. Child malnutrition indicators (severe), underweight by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under five severely underweight (0-59m) All households 13.5 9.8 10.1 -25.4 ** 2,594 2,380 2,326 Sex of child Female 12.9 11.2 11.2 -13.1 1265 1132 1093 Male 14.1 8.5 9.0 -36.3 ** 1,329 1,248 1,233 Region Coast 14.2 11.6 7.8 -45.0 *** 827 789 764 Haor 17.8 12.1 11.5 -35.2 ** 594 633 574 Mid Char 8.8 5.4 7.1 -19.4 616 504 497 North Char 11.0 8.3 10.0 -9.0 557 454 491 Well-being category Extreme poor 14.8 12.5 14.3 -3.6 310 301 324 Poor 13.5 9.5 9.4 -30.1 ** 1,851 1,695 1,613 Middle 11.5 8.9 5.6 -51.2 ** 210 197 188 Middle rich 15.0 8.4 5.6 -62.5 ** 141 118 127 Rich 8.3 6.1 4.5 -45.7 81 68 74 Program approach MCHN/PEP 12.7 10.0 10.7 -15.6 1053 963 989 PM2A 16.2 8.0 7.8 -51.6 *** 1541 1417 1337 Sex of headship Female 12.3 10.1 14.2 15.6 187 182 276 Male 13.6 9.8 9.7 -28.4 2,407 2,198 2,050 Percent of children under two severely underweight (0-23m) All households 10.9 9.3 10.8 -1.1 937 995 929 Sex of child Female 11.8 8.5 12.6 6.9 479 481 465 Male 10.0 10.2 8.7 -12.7 458 514 464 Region Coast 7.0 8.4 8.4 20.6 317 348 318 Haor 12.9 11.7 12.3 -5.0 226 258 238 Mid Char 7.4 4.9 5.8 -21.1 192 212 191 North Char 10.6 8.4 11.6 9.3 202 177 182 Well-being category Extreme poor 9.8 8.4 10.9 11.4 110 113 131 Poor 11.4 9.7 11.7 2.6 665 715 638 Middle 7.0 8.4 1.4 -80.1 78 94 77 Table 26. Child malnutrition indicators (moderate, under five), by age of mother Indicator Baseline Midterm Endline – Index child only Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Percent of children under five Stunted (6-59m) - All 61.7 52.7 48.8 -21.0 *** 2,384 2,167 2,143 Age of mother 15-25 years 62.3 53.9 47.8 -23.3 *** 1,237 1,059 946 26-49 years 60.7 51.2 49.5 -18.5 *** 1,114 1,062 1,139 Wasted (6-59m) - All Age of mother 15-25 years 16.0 10.3 12.6 -21.2 1,237 1,059 946 26-49 years 15.0 7.9 11.9 -20.8 1,114 1,062 1,139 Underweight (0-59m) - All 42.2 34.2 35.3 -16.3 *** 2,594 2,380 2,326 Age of mother 15-25 years 43.2 32.6 36.9 -14.5 * 1,398 1,190 1,049 26-49 years 40.9 35.6 34.2 -16.3 ** 1,171 1,142 1,217 Percent of children under two Baseline Midterm Endline Stunted (6-23m) - All 55.8 52.0 42.9 -23.1 *** 727 782 746 Age of mother 15-25 years 55.6 49.0 42.2 -24.0 ** 466 472 385 26-49 years 56.3 55.4 44.6 -20.9 * 257 298 348 Wasted (6-23m) - All 8.1 2.5 5.0 -38.9 727 782 746 Age of mother 15-25 years 16.3 13.8 13.5 -16.9 466 472 385 26-49 years 21.2 8.9 14.4 -31.9 257 298 348 Underweight (0-23m) - All 33.2 32.1 29.5 -11.1 937 995 929 Age of mother 15-25 years 34.9 31.1 28.1 -19.6 618 603 488 26-49 years 30.0 33.2 30.4 1.2 314 378 426 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Middle rich 13.4 8.3 3.4 -74.9 ** 56 51 53 Rich ‡ ‡ 1.0 ‡ 28 22 30 Program approach MCHN/PEP 10.9 9.6 12.0 10.5 377 388 376 PM2A 11.0 8.7 6.9 -37.7 * 560 607 553 Sex of headship Female 0.5 2.5 10.9 2086.0 * 49 71 108 Male 11.3 9.7 10.8 -4.7 888 924 821 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 27. Percent of children 6-23 months with diarrhea in the last two weeks Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline All 13.8 11.6 8.1 -41.1 ** 848 902 905 PEP households 13.9 11.5 8.1 -42.1 ** 684 750 755 Region Coast 23.3 15.2 5.4 -76.8 *** 282 319 329 Haor 14.2 10.9 8.6 -40.0 225 241 206 Mid Char 15.3 10.0 5.7 -62.7 *** 161 185 187 North Char 10.6 12.7 9.4 -11.8 180 157 183 Well-being category Extreme poor 13.9 13.9 11.3 -18.9 78 98 125 Poor 13.9 11.2 7.2 -48.4 *** 606 652 630 Middle 13.6 8.2 7.3 -46.5 77 85 69 Middle rich 10.7 23.3 12.3 14.3 61 45 51 Rich ‡ ‡ ‡ ‡ ‡ 25 22 30 Program approach MCHN/PEP 13.2 10.7 8.6 -35.0 345 352 391 PM2A 16.1 14.2 6.6 -58.8 *** 503 550 514 Sex of headship Female 34.2 17.2 14.1 -58.9 38 66 105 Male 13.2 11.3 7.6 -42.9 ** 810 836 800 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 28. Moderate and severe malnutrition among children under five and children under two, by sex of household head Male headed households Female headed households Difference (Female vs. male) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Under fives Stunting 61.0 49.5 -18.8 72.9 38.7 -46.9 * 148.9 Wasting 15.6 12.0 -23.0 * 14.2 15.2 6.9 -130.0 Underweight 42.0 35.7 -15.1 * 46.5 31.1 -33.2 * 120.5 Under twos Stunting 55.4 43.1 -22.2 * 68.8 41.2 -40.1 * 80.7 Wasting 18.6 14.0 -24.7 5.1 16.7 227.5 -1019.7 Underweight 33.8 29.2 -13.6 14.8 32.6 120.3 -983.7 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 10% level. Table 29. Percent of women underweight, by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline All 36.8 32.2 30.7 -16.5 * 2,881 2,485 2,467 PEP households 37.2 32.5 31.5 -15.4 * 2,401 2,088 2,073 Region Coast 21.4 14.5 13.2 -38.1 *** 896 799 843 Haor 37.1 36.5 39.1 5.5 706 664 589 Mid Char 39.8 28.5 25.7 -35.3 *** 673 533 506 North Char 37.5 31.6 24.6 -34.4 ** 606 489 529 Well-being category Extreme poor 37.4 33.9 42.9 14.7 345 306 343 Poor 37.2 32.3 28.5 -23.4 ** 2,056 1,782 1,730 Middle 35.9 30.3 24.1 -33.0 * 232 201 197 Middle rich 29.3 29.6 23.5 -20.0 161 1,213 125 Rich 25.2 20.8 11.6 -53.9 85 123 72 Program approach MCHN/PEP 36.9 33.0 31.8 -13.8 1,188 1,014 1,054 PM2A 36.4 29.8 27.1 -25.4 *** 1,693 1,471 1,413 Sex of headship Female 32.8 27.1 40.5 23.6 212 204 283 Male 37.0 32.6 29.9 -19.1 ** 2,669 2,281 2,184 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. IR 2.1 Table 30. Caring practices for children under two – Immunizations (Percent) Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Immunized against 8 diseases by 12 months (12-23 months) All 58.6 74.9 73.0 24.6 ** 392 480 489 PEP households 58.7 75.3 72.3 23.2 ** 307 393 419 Region Coast 64.0 74.0 66.5 3.9 118 150 172 Haor 56.1 72.1 70.3 25.3 88 133 112 Mid Char 58.4 65.2 73.5 25.9 * 75 105 106 North Char 60.0 83.1 78.1 30.1 * 111 92 99 Well-being category Extreme poor ‡ ‡ ‡ ‡ ‡ 34 49 72 Poor 57.1 76.3 74.2 29.9 *** 273 344 347 Middle 45.8 67.9 ‡ ‡ ‡ 40 56 31 Middle rich ‡ ‡ ‡ ‡ ‡ 28 18 26 Rich ‡ ‡ ‡ ‡ ‡ 16 13 13 Program approach MCHN/PEP 58.6 76.6 72.9 24.3 ** 160 181 210 PM2A 58.3 69.8 73.6 26.1 *** 232 299 279 Sex of headship Female ‡ ‡ ‡ ‡ ‡ 16 38 56 Male 58.6 74.6 75.1 28.2 *** 376 442 433 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 31. Overview: Access to and utilization of health and nutrition services improved to caregivers of children under 2 years of age Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of women underweight 36.8 32.2 30.8 -6.0 * 2,881 2,485 2,467 Dietary diversity scores of mothers of children under five (out of 14 food groups) 4.6 5.6 8.1 76.3 *** 3,105 2,780 2,861 Percent of mothers receiving antenatal care 47.1 64.5 85.3 38.2 *** 3,119 2,802 2,885 Percent of mothers receiving antenatal care in a medical facility a/ 23.5 35.5 63.3 42.1 *** 3,120 2,797 2,834 Percent of mothers taking more food during pregnancy 12.5 25.2 57.6 45.1 *** 3,119 2,797 2,835 Percent of mothers taking more rest during pregnancy 23.5 35.5 63.3 42.1 *** 3,120 2,797 2,834 Percent of pregnant and lactating women taking iron/folic acid in the last 7 days 15.2 29.8 52.0 36.8 *** 1,255 1,214 1,346 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. a/ Reported only for mothers receiving any antental care. Table 32. Indicators of caring practices for mothers during pregnancy, by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of pregnant and lactating women taking iron/folic acid in the last 7 days All 15.2 29.8 52.0 242.1 *** 1,255 1,214 1,346 PEP households 23.3 35.5 51.8 122.3 *** 1,043 1,024 1,139 Region Coast 13.9 23.0 50.8 265.7 *** 450 433 502 Haor 11.6 26.2 48.7 320.1 *** 332 330 333 Mid Char 16.0 31.4 43.1 169.2 *** 226 238 260 North Char 20.5 36.3 62.7 206.1 *** 247 213 251 Well-being category Extreme poor 12.4 18.4 41.9 237.7 *** 136 147 202 Poor 15.1 31.4 54.7 262.5 *** 907 877 937 Middle 19.1 34.4 56.9 100 108 97 Middle rich 19.5 22.0 52.6 169.5 ** 77 57 66 Rich 37.5 ‡ 53.2 ‡ 35 25 44 Program approach MCHN/PEP 15.1 30.2 51.7 242.6 *** 529 508 588 PM2A 15.5 28.5 52.9 241.4 *** 726 706 758 Sex of headship Female 15.2 30.1 52.0 241.6 *** 69 85 150 Male 15.4 30.6 51.9 237.1 *** 1,186 1,129 1,196 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 33. Indicators of caring practices for mothers during pregnancy, by region, well-being category, program approach and sex of household head (continued) Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of mothers receiving antenatal care All 47.1 64.5 85.3 81.0 *** 3,119 2,802 2,885 PEP households 45.8 64.0 83.8 83.0 *** 2,602 2,367 2,408 Region Coast 45.4 58.1 79.6 75.2 *** 1,013 926 972 Haor 41.4 55.7 83.5 101.6 *** 777 751 705 Mid Char 46.0 60.5 84.9 84.6 *** 695 598 584 North Char 55.7 80.7 89.3 60.3 *** 634 527 624 Well-being category Extreme poor 48.0 56.6 80.9 68.5 *** 370 356 404 Poor 45.5 65.2 86.0 89.0 *** 2,232 2,011 2,004 Middle 57.4 65.5 92.9 61.8 *** 249 222 227 Middle rich 62.8 70.9 88.6 41.0 *** 175 136 154 Rich 74.2 80.5 95.2 28.3 *** 91 76 96 Program approach MCHN/PEP 45.8 64.2 84.0 83.4 *** 1,299 1,173 1,244 PM2A 51.2 65.4 89.7 75.2 *** 1,820 1,629 1,641 Sex of headship Female 53.3 64.9 80.4 50.8 *** 236 228 333 Male 46.7 64.4 85.7 83.4 *** 2,883 2,574 2,552 Continued. Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of mothers receiving antenatal care in a medical facility£ All 69.8 65.2 73.7 5.5 1,526 1,769 2,232 PEP households 68.6 64.5 73.1 6.6 1,192 1,457 1,827 Region Coast 87.6 90.8 97.2 11.0 *** 476 540 634 Haor 62.1 49.3 97.2 56.5 *** 329 424 581 Mid Char 70.4 76.8 84.6 20.1 ** 334 365 471 North Char 74.5 73.1 81.9 9.9 387 440 546 Well-being category Extreme poor 65.4 66.2 71.8 9.8 167 205 297 Poor 69.2 64.3 73.5 6.2 1,025 1,252 1,530 Middle 76.3 62.4 76.3 0.0 148 150 197 Middle rich 84.4 83.5 84.9 0.6 115 98 125 Rich 92.7 87.9 82.3 -11.2 * 69 63 83 Program approach MCHN/PEP 69.9 65.8 73.7 5.4 572 720 922 PM2A 69.6 63.3 73.7 5.9 954 1,049 1,310 Sex of headship Female 56.9 70.9 73.5 29.2 120 142 229 Male 70.7 64.8 75.7 7.1 1,406 1,627 2,003 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. £ Reported only for mothers receiving any antental care. Table 34. Indicators of caring practices for mothers during pregnancy, by region, well-being category, program approach and sex of household head (continued) Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of mothers taking more food during pregnancy All households 12.5 25.2 57.6 361.0 *** 3,119 2,797 2,835 PEP households 12.2 24.9 57.2 368.9 *** 2,602 2,363 2,375 Region Coast 7.1 10.4 38.4 441.4 *** 1,014 925 963 Haor 11.4 24.9 60.0 426.7 *** 775 750 692 Mid Char 14.2 19.0 51.9 265.2 *** 695 597 569 North Char 14.0 32.0 61.0 335.6 *** 635 525 611 Well-being category Extreme poor 10.7 23.6 57.5 437.6 *** 372 356 398 Poor 12.4 25.1 57.1 360.7 *** 2,230 2,007 1,977 Middle 13.0 29.8 68.4 426.2 *** 249 221 221 Middle rich 15.2 26.5 54.9 261.1 *** 175 136 148 Rich 30.4 36.1 64.4 111.8 *** 91 76 91 Program approach MCHN/PEP 12.6 25.2 57.1 352.9 *** 1,297 1,170 1,229 PM2A 11.9 25.2 61.1 413.1 *** 1,822 1,627 1,606 Sex of headship Female 11.6 20.1 58.2 401.9 *** 237 228 323 Male 12.5 25.6 50.1 300.9 *** 2,882 2,569 2,512 Continued. Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of mothers taking more rest during pregnancy All households 23.5 35.5 63.3 169.5 *** 3,120 2,797 2,834 PEP households 23.3 35.5 65.5 181.0 *** 2,603 2,363 2,374 Region Coast 26.8 22.1 53.8 100.8 *** 1,014 926 964 Haor 18.8 31.2 61.2 225.3 *** 775 750 690 Mid Char 29.8 38.6 64.1 115.2 *** 696 597 569 North Char 25.6 42.9 75.3 194.1 *** 635 524 611 Well-being category Extreme poor 17.3 26.2 58.5 237.9 *** 372 356 398 Poor 24.3 37.0 67.3 177.0 *** 2,231 2,007 1,976 Middle 22.5 34.8 71.0 215.3 *** 249 221 221 Middle rich 27.6 34.6 61.2 121.8 *** 175 136 148 Rich 32.5 36.7 66.0 102.8 *** 91 76 91 Program approach MCHN/PEP 24.1 36.0 65.1 170.0 *** 1,298 1,171 1,228 PM2A 21.6 33.9 67.2 211.1 *** 1,822 1,626 1,606 Sex of headship Female 22.5 30.3 50.9 126.4 *** 237 228 323 Male 23.5 35.9 66.7 183.9 *** 2,883 2,569 2,511 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Table 35. Dietary diversity of mothers of children under five, by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Dietary diversity score (out of 14 food groups) All 4.6 5.6 8.4 81.6 *** 3,105 2,780 2,861 PEP households 4.5 5.6 8.3 84.9 *** 2,592 2,347 2,387 Region Coast 4.5 5.3 7.6 67.8 *** 1,011 923 967 Haor 4.6 6.3 8.5 84.0 *** 773 746 691 Mid Char 4.6 4.8 7.9 72.6 *** 691 589 581 North Char 4.6 5.2 8.6 86.6 *** 630 522 622 Well-being category Extreme poor 4.4 5.7 8.3 89.1 *** 368 352 404 Poor 4.5 5.6 8.3 84.9 *** 2,224 1,995 1,983 Middle 5.1 6.0 8.6 69.3 *** 248 221 228 Middle rich 5.5 6.1 8.4 53.5 *** 172 136 155 Rich 6.9 7.2 9.6 39.2 *** 91 75 91 Program approach MCHN/PEP 4.5 5.6 8.3 83.9 *** 1,296 1,158 1,234 PM2A 4.8 5.8 8.6 79.2 *** 1,809 1,622 1,627 Sex of headship Female 4.5 5.4 7.7 71.7 *** 235 225 327 Male 4.6 5.7 8.4 82.6 *** 2,870 2,555 2,534 Continued. Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of mothers consuming foods from food groups in the previous 24 hours Cereals, roots and tubers 98.9 98.7 99.0 0.1 3,108 2,780 2,328 Pulses and legumes 16.0 33.1 57.4 258.5 *** 3,107 2,781 2,328 Dairy products 8.2 13.8 39.3 379.4 *** 3,107 2,781 2,328 Meat, fish/seafood & eggs 61.8 72.7 90.7 46.8 *** 3,107 2,781 2,328 Oils and fats 69.5 72.2 92.0 32.4 *** 3,107 2,781 2,328 Vegetables 90.6 93.1 97.0 7.0 *** 3,107 2,781 2,328 Vitamin A -rich vegetables 20.2 22.8 34.8 72.2 *** 3,107 2,781 2,328 Dark green leafy vegetables 39.5 49.4 74.3 88.2 *** 3,108 2,781 2,328 Fruits 6.3 13.4 48.0 662.7 *** 3,108 2,780 2,328 Vitamin A -rich fruits 3.6 7.8 25.5 607.1 *** 3,107 2,781 2,328 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. IR 2.2 Table 36. Caring practices for children under two - Exclusive breastfeeding (Percent) Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Exclusive breastfeeding (0-5 months) All 64.1 66.4 62.2 -3.0 282 247 266 PEP households 65.2 65.9 63.1 -3.3 247 213 218 Region Coast 57.4 63.4 71.6 24.7 ** 101 83 86 Haor 64.0 66.7 50.8 -20.7 70 58 77 Mid Char 61.4 62.1 44.0 -28.4 60 58 45 North Char 67.2 68.9 89.2 32.7 ** 51 48 58 Well-being category Extreme poor 51.0 71.5 63.5 24.4 50 37 42 Poor 68.4 64.5 63.0 -8.0 197 176 176 Middle ‡ ‡ ‡ ‡ 15 17 30 Middle rich ‡ ‡ ‡ ‡ 12 12 8 Rich ‡ ‡ ‡ ‡ 8 5 10 Program approach MCHN/PEP 67.9 65.3 66.8 -1.6 120 118 99 PM2A 50.4 71.1 51.1 1.3 162 129 167 Sex of headship Female ‡ ‡ ‡ ‡ 23 20 24 Male 65.0 66.2 61.0 -6.1 242 227 259 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. Table 37. Indicators of hygiene practices, by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of mothers of children under five who feel it is important to wash their hands at all five critical times £ All 9.8 10.9 29.8 204.4 *** 3,126 2,791 2,908 PEP households 9.7 10.9 29.7 206.3 *** 2,608 2,358 2,423 Region Coast 8.2 6.5 15.3 87.0 *** 1,014 923 976 Haor 8.1 10.7 28.7 254.0 *** 777 748 715 Mid Char 18.1 18.7 26.1 44.0 * 699 596 591 North Char 7.8 7.7 36.6 368.6 *** 636 524 626 Well-being category Extreme poor 6.9 14.8 23.6 242.3 *** 372 355 408 Poor 10.1 10.3 31.3 209.6 *** 2,236 2,003 2,015 Middle 10.5 8.8 35.0 232.9 *** 249 221 231 Middle rich 9.9 12.5 26.2 164.7 ** 176 136 158 Rich 13.0 7.7 30.9 138.3 ** 91 75 96 Program approach MCHN/PEP 10.0 11.3 29.3 192.9 *** 1300 1167 1251 PM2A 9.0 9.3 31.7 252.4 *** 1,826 1,624 1,657 Sex of headship Female 5.6 5.7 19.4 247.0 *** 237 227 339 Male 10.0 11.2 30.7 207.0 *** 2,889 2,564 2,569 Continued. Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of children under three years whose feces are disposed of safely All 41.4 68.2 60.3 45.5 *** 1,620 1,547 1,447 PEP households 41.2 68.0 59.9 45.5 *** 1,365 1,303 1,216 Region Coast 32.7 53.5 56.3 72.0 *** 566 571 530 Haor 35.7 68.8 54.9 53.9 *** 408 401 360 Mid Char 40.2 58.2 57.6 43.3 *** 323 305 269 North Char 51.6 75.7 71.1 37.8 *** 323 270 288 Well-being category Extreme poor 35.9 63.5 59.2 64.8 *** 189 179 210 Poor 42.1 68.8 60.2 42.9 *** 1,176 1,124 1,006 Middle 45.6 66.0 65.2 42.9 ** 123 127 107 Middle rich 45.6 66.0 65.3 43.1 ** 85 79 76 Rich 32.2 69.5 58.6 81.7 ** 46 38 48 Program approach MCHN/PEP 42.5 68.9 59.0 38.8 *** 678 625 608 PM2A 38.0 66.0 64.3 69.2 *** 942 922 839 Sex of headship Female 38.3 67.7 51.0 33.1 111 123 161 Male 41.6 68.2 61.0 46.7 *** 1,509 1,424 1,286 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. £ The five critical times are: Before eating, before breastfeeding or feeding a child, before cooking or preparing food, after defecation/urination, and after cleaning a child that has defecated/changing a child's diaper. Table 38. Caring practices for children under two (percent) Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Exclusive breastfeeding (0-5 months) All households 64.1 66.4 62.2 -3.0 282 247 266 PEP households 65.2 65.9 63.1 -3.3 247 213 218 Program Approach MCHN/PEP 67.9 65.3 66.8 -1.6 120 118 99 PM2A 50.4 71.1 51.1 1.3 162 129 167 Sex of headship Female ‡ ‡ ‡ ‡ 23 20 24 Male 65.0 66.2 61.0 -6.1 242 227 259 Minimum dietary diversity (6-23 months) All households 12.7 25.9 56.9 348.2 *** 836 882 903 PEP households 11.2 25.0 56.7 406.1 *** 672 733 753 Program Approach MCHN/PEP 11.3 25.3 56.4 399.6 *** 336 344 391 PM2A 17.2 27.8 58.9 242.0 *** 500 538 512 Sex of headship Female 26.0 30.8 46.5 78.9 40 64 104 Male 12.3 25.7 57.9 371.3 *** 796 818 799 Minimum meal frequency (breastfed, 6-23 months) All households 47.7 45.6 65.7 37.7 *** 669 683 747 PEP households 47.8 45.6 65.3 36.5 *** 538 573 628 Program Apporach MCHN/PEP 49.1 46.7 64.7 31.7 ** 263 264 319 PM2A 43.0 42.3 69.3 61.2 *** 406 419 428 Sex of headship Female 47.6 35.7 61.4 28.9 31 45 85 Male 47.7 46.0 66.1 38.6 *** 638 638 662 Continued. Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Minimum acceptable diet (breastfed, 6-23 months) All households 8.7 17.5 47.9 451.0 *** 669 683 747 PEP households 7.5 17.0 47.8 537.7 *** 538 573 628 Program Apporach MCHN/PEP 7.1 17.8 47.9 577.0 *** 263 264 319 PM2A 13.8 16.5 47.9 247.9 *** 406 419 428 Sex of headship Female 22.5 8.6 48.0 112.9 31 45 85 Male 8.2 17.8 47.9 481.7 *** 638 638 662 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Note: Breakdowns by region and well-being category are not possible due to insufficient number of observations. ‡ Insufficient number of observations. Table 39. Caring practices for children under two (percent) Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Monomix or other vitamin sprinkles (6-23 months) All households 2.5 6.6 31.7 1166.8 *** 835 891 870 PEP households 2.4 5.9 31.9 1227.5 *** 672 742 729 Program Apporach MCHN/PEP 2.7 6.0 31.5 1050.0 *** 334 350 374 PM2A 1.9 8.4 32.2 1642.2 *** 501 541 496 Sex of headship Female 13.1 3.1 15.3 40 65 100 Male 2.2 6.8 33.11 1405.0 *** 795 826 770 Continued. Indicator Baseline Midterm Endline Percent Difference Number of observations Baseline Midterm Endline Vitamin A capsule in last 6 months (6-23 months) All households 58.3 67.3 85.2 46.1 *** 849 888 873 PEP households 58.4 67.7 85.3 46.1 *** 684 738 733 Program Apporach MCHN/PEP 58.2 69.0 86.5 48.5 *** 342 351 375 PM2A 58.5 62.4 80.4 37.5 *** 507 537 498 Sex of headship Female 82.8 70.2 83.8 40 63 100 Male 57.6 67.1 85.3 48.2 *** 809 825 773 Immunized against 8 diseases by 12 months (12-23 months) All households 58.6 74.9 73.0 24.6 ** 392 480 489 PEP households 58.7 75.3 72.3 23.2 ** 307 393 419 Program Apporach MCHN/PEP 58.6 76.6 72.9 24.3 ** 160 181 210 PM2A 58.3 69.8 73.6 26.2 *** 232 299 279 Sex of headship Female ‡ 81.6 51.6 ‡ 16 38 56 Male 58.6 74.6 75.1 28.2 *** 376 442 433 Percent receiving ORS for diarrhea (6-24 m olds with diarrhea in last two weeks) All households 61.1 86.0 97.7 59.9 *** 147 120 55 PEP households 59.3 85.4 97.5 64.5 *** 125 101 46 Program Apporach MCHN/PEP 55.3 84.6 97.2 75.7 *** 58 46 32 PM2A 76.5 89.0 100.0 30.8 *** 89 74 23 Sex of headship Female ‡ ‡ ‡ ‡ 7 15 8 Male 61.4 87.5 97.3 58.5 *** 140 105 47 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Note: Breakdowns by region and well-being category are not possible due to insufficient number of observations. ‡ Insufficient number of observations. WATER AND SANITATION Table 40. Improved drinking water and sanitation facility, by region, well-being category, program approach, and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of households with access to an improved drinking water source All 59.5 76.3 77.9 30.8 *** 7,085 6,916 7146 PEP households 59.2 75.9 77.4 30.9 *** 5,761 5,672 5842 Region Coast 66.4 76.2 71.4 7.4 * 1,885 1,852 1907 Haor 59.3 75.2 77.2 30.3 *** 1672 1701 1689 Mid Char 61.9 74.0 73.5 18.9 *** 1798 1745 1652 North Char 57.8 78.7 81.0 40.0 *** 1730 1618 1898 Well-being category Extreme poor 54.7 73.1 79.4 45.0 *** 1195 1237 1333 Poor 60.4 76.7 76.8 27.2 *** 4566 4435 4509 Middle 58.5 81.6 77.3 32.1 *** 638 599 577 Middle rich 69.0 79.6 84.2 22.0 *** 428 409 452 Rich 68.6 80.8 78.1 13.8 * 252 235 274 Program approach MCHN/PEP 59.3 76.1 77.2 30.2 *** 2888 2766 2944 PM2A 60.4 77.0 79.0 30.7 *** 4197 4150 4202 Sex of headship Female 58.0 73.2 77.0 32.7 *** 957 1009 1221 Male 59.8 76.8 77.7 30.0 *** 6128 5907 5925 Continued. Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of households with access to an improved sanitation facility All 20.3 32.4 52.8 160.1 *** 7085 6916 7148 PEP households 19.2 30.9 51.6 169.2 *** 5761 5672 5844 Region Coast 18.9 31.3 46.0 142.7 *** 1885 1852 1907 Haor 16.3 30.6 51.3 214.8 *** 1672 1701 1689 Mid Char 25.4 33.9 44.1 73.8 *** 1798 1745 1652 North Char 22.5 34.1 59.9 166.4 *** 1730 1618 1900 Well-being category Extreme poor 13.6 27.7 49.5 265.1 *** 1195 1237 1333 Poor 20.7 31.8 52.3 153.0 *** 4566 4435 4511 Middle 27.5 39.3 62.6 127.8 *** 638 599 577 Middle rich 30.4 54.9 68.1 124.0 *** 428 409 452 Rich 53.7 70.0 76.9 43.1 *** 252 235 274 Program approach MCHN/PEP 19.9 31.4 51.2 157.1 *** 2888 2766 2945 PM2A 21.7 35.8 58.3 169.0 *** 4197 4150 4203 Sex of headship Female 12.9 23.8 47.3 266.1 *** 957 1009 1221 Male 21.4 33.8 53.9 151.2 *** 6128 5907 5927 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. BY PROGRAM APPROACH Table 41. Moderate and severe malnutrition among children under five and children under two, by program approach Baseline Endline Indicator MCHN/PEP PM2A Difference Percent difference (PM2A￾MCHN/PEP) MCHN/PEP PM2A Difference Percent difference (PM2A￾MCHN/PEP) Under fives Stunting (6-59m) 62.2 59.9 2.3 -3.7 49.6 45.2 4.5 -9.0 Wasting (6-59m) 15.5 15.9 -0.4 2.6 12.8 10.4 2.4 -18.5 Underweight (0-59m) 42.6 41.0 1.6 -3.7 36.3 31.9 4.4 -12.2 * Under twos Stunting (6-23m) 57.9 49.4 8.5 -14.6 * 44.1 39.0 5.1 -11.6 Wasting (6-23m) 18.6 17.1 1.4 -7.8 15.6 10.0 5.6 -35.9 * Underweight (0-23m) 34.1 29.9 4.2 -12.2 31.0 24.8 6.2 -19.9 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. BY PEP/NON-PEP Table 42. PEP participation by region Baseline Midterm Endline Number of PEP participants PEP Non-PEP PEP Non-PEP PEP Non-PEP Region Coast 200 1,685 183 1,669 179 1,728 Haor 330 1,340 329 1,372 335 1,354 Mid Char 366 1,431 336 1,408 354 1,298 North Char 422 1,305 395 1,223 435 1,464 Table 43. Moderate and severe malnutrition among children under five and children under two (PEP versus non￾PEP) Non-Poor PEP Difference (PEP vs. non￾PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Under fives Stunting 53.8 32.3 -39.9 * 62.3 50.0 -19.8 * 20.2 Wasting 12.3 8.5 -30.8 15.8 12.5 -21.0 9.8 Underweight 37.0 23.1 -37.7 * 42.7 36.3 -15.0 * 22.8 Under twos Stunting 38.9 23.7 -39.1 * 57.3 44.4 -22.5 * 16.6 Wasting 13.6 5.3 -61.1 * 18.6 14.9 -19.9 41.2 Underweight 24.3 12.1 -50.4 * 33.9 31.0 -8.4 42.0 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 10% level. Table 44. Indicator of diarrhea incidence (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non-PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Percent of children 6-23 months with diarrhea in the last two weeks 12.8 9.2 -28.0 13.9 8.1 -42.3 * 14.22 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. Table 45. Access to and utilization of health and nutrition services improved to caregivers of children under 2 years of age (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non￾PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Percent of women underweight 31.3 21.4 -31.7 * 37.2 31.5 -15.2 * 16.4 31.3 Dietary diversity scores of mothers of children under five (out of 14 food groups) 5.6 8.4 51.8 4.5 8.3 83.6 31.8 5.6 Percent of mothers receiving antenatal care 64.1 89.6 39.7 45.8 83.8 83.0 43.3 64.1 Percent of mothers receiving antenatal care in a medical facility£ 84.6 80.0 -5.5 * 68.6 73.1 6.6 * 12.1 84.6 Percent of mothers taking more food during pregnancy 17.2 63.2 266.5 * 12.2 57.2 368.9 * 102.4 17.2 Percent of mothers taking more rest during pregnancy 27.5 66.7 142.6 * 23.3 65.5 181.0 * 38.3 27.5 Percent of pregnant and lacatating women taking iron/folic acid in the last 7 days 22.5 55.0 143.8 * 14.5 51.8 257.8 * 114.0 22.5 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. £ Reported only for mothers receiving any antental care. Table 46. Indicators of hygiene practices (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non-PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Percent of mothers of children under five who feel it is important to wash their hands at all five critical times £ 10.8 31.5 192.5 * 9.7 29.7 206.9 14.45 Percent of children under three years who feces are disposed of safely 45.3 64.1 41.5 * 41.2 59.9 45.6 * 4.11 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. £ The five critical times are: Before eating, before breastfeeding or feeding a child, before cooking or preparing food, after defecation/urination, and after cleaning a child that has defecated/changing a child's diaper. Table 47. Caring practices for children under 2 (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non-PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Exclusive breastfeeding (0-5 months) 45.7 53.9 17.9 65.2 63.1 -3.3 21.25 Minimum dietary diversity (6- 23 months) 28.8 60.1 108.9 * 11.2 56.7 406.1 * 297.17 Minimum meal frequency (Breastfed, 6-23 months) 46.9 71.6 52.7 * 47.8 65.3 36.7 * 16.00 Minimum acceptable diet (breastfed, 6-23 months) 21.5 49.4 129.4 * 7.5 47.8 542.0 * 412.65 Monomix or other vitamin sprinkes (6-23 months) 4.5 29.2 546.3 * 2.4 31.9 1255.7 * 709.40 Vitamin A capsule in last 6 months (6-23 months) 56.7 82.7 46.0 * 58.4 85.3 46.1 0.13 Immunized against 8 diseases by 12 months (12-23 months) 56.7 85.7 51.2 * 58.7 72.3 23.2 * 27.94 Percent receiving ORS for diarrhea (6-24 m olds with diarrhea in last two weeks) ‡ ‡ ‡ 59.3 97.5 64.5 * ‡ Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. ‡ Insufficient number of observations. Table 48. Indicators of improved water and sanitation (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non-PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Percent of households with access to an improved drinking water source 63.9 79.7 24.7 * 59.2 77.4 30.9 * 6.17 Percent of households with access to an improved sanitation facility 33.8 67.6 100.3 * 19.2 51.6 169.2 * 68.94 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. BY MALE/FEMALE HHH Table 49. Moderate and severe malnutrition among children under five and children under two, by sex of household head Male headed households Female headed households Difference (Female vs. male) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Under fives Stunting 61.0 50.9 -16.6 72.9 44.1 -39.5 138.6 Wasting 15.6 11.0 -29.5 14.2 14.9 4.9 -116.7 Underweight 42.0 36.1 -14.0 46.5 33.8 -27.3 94.4 Under twos Stunting 55.4 43.7 -21.1 68.8 43.9 -36.2 71.4 Wasting 18.6 12.7 -31.7 5.1 17.0 233.3 -835.6 Underweight 33.8 29.4 -13.0 14.8 31.5 112.8 -966.8 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. Table 50. Access to and utilization of health and nutrition services improved to caregivers of children under 2 years of age, by head of household Male headed households Female headed households Difference (Female vs. male) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Percent of women underweight 37.0 30.1 -18.7 32.8 40.0 21.9 40.6 Dietary diversity scores of mothers of children under five (out of 14 food groups) 4.6 7.5 63.9 * 4.5 7.4 64.0 * 0.1 Percent of mothers receiving antenatal care 46.7 85.7 83.4 * 53.3 80.4 50.8 * -32.6 Percent of mothers receiving antenatal care in a medical facility£ 70.7 75.7 7.1 56.9 73.5 29.2 22.1 Percent of mothers taking more food during pregnancy 12.5 50.1 300.9 * 11.6 58.2 401.9 * 101.0 Percent of mothers taking more rest during pregnancy 23.5 66.7 183.9 * 22.5 50.9 126.4 * -57.5 Percent of pregnant and lactating women taking iron/folic acid in the last 7 days 15.4 51.9 237.1 * ‡ 53.0 ‡ * ‡ Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. £ Reported only for mothers receiving any antenatal care. ‡ Insufficient number of observations. Table 51. Water and sanitation key indicators, by sex of household head Male headed households Female headed households Difference (Female vs. male) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Percent of households with access to an improved drinking water source 59.8 77.7 30.0 * 58.0 77.0 32.7 * 2.7 Percent of households with access to an improved sanitation facility 21.4 53.9 151.3 12.9 47.3 266.3 115.0 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. SO3 Tables SO2 PROGRAM INDICATORS Table X. Indicators of women's empowerment, by region, well-being category, and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of women earning cash income/control over economic resources All 5.7 11.3 19.8 249.65 *** 6492 6795 7146 Region Coast 4.5 8.5 19.5 338.20 *** 1852 1835 1907 Haor 4.3 9.8 14.0 227.04 *** 1639 1654 1689 Mid Char 5.9 8.2 18.4 210.44 *** 1766 1710 1652 North Char 7.2 14.9 27.2 277.92 *** 1685 1596 1898 Well-being category Extreme poor 10.3 14.1 23.4 128.07 *** 1170 1208 1333 Poor 4.8 11.1 19.8 315.55 *** 4481 4363 4509 Middle 2.9 7.9 10.4 259.31 *** 621 590 577 Middle rich 2.9 3.8 7.7 162.24 ** 419 401 452 Rich 1.3 1.4 8.2 530.77 *** 245 232 274 Sex of headship Female 16.4 20.9 31.9 94.57 *** 954 1006 1221 Male 4.0 9.7 17.6 340.35 *** 5988 5789 5925 Continued. Percent of women participating in any group All 20.72 26.97 20.70 -0.10 6943 6795 7146 Region Coast 13.18 11.97 24.03 82.32 *** 1852 1835 1907 Haor 12.67 30.32 16.51 30.31 1640 1654 1689 Mid Char 29.52 28.98 21.34 -27.71 ** 1766 1710 1652 North Char 26.32 24.14 24.82 -5.70 1685 1596 1898 Well-being category Extreme poor 13.58 23.39 18.56 36.67 ** 1170 1208 1333 Poor 22.80 28.13 22.59 -0.92 4482 4363 4509 Middle 20.15 25.66 10.33 -48.73 *** 621 590 577 Middle rich 18.97 19.18 11.50 -39.38 ** 419 401 452 Rich 16.26 25.69 9.66 -40.59 * 245 232 274 Sex of headship Female 12.23 14.14 17.36 41.95 * 954 1006 1221 Male 22.03 29.09 21.31 -3.27 5989 5789 5925 Index of women's freedom of movement All 8.06 8.93 10.48 30.02 *** 6942 6795 7146 Region Coast 7.52 8.29 9.39 24.87 *** 1851 1835 1907 Haor 7.52 8.72 10.4 38.30 *** 1640 1654 1689 Mid Char 8.37 8.98 10.09 20.55 *** 1766 1710 1652 North Char 8.57 9.23 10.92 27.42 *** 1685 1596 1898 Well-being category Extreme poor 8.43 9.24 10.58 25.50 *** 1170 1208 1333 Poor 7.96 8.87 10.46 31.41 *** 4481 4363 4509 Middle 8.14 8.8 10.46 28.50 *** 621 590 577 Middle rich 7.99 8.58 10.46 30.91 *** 419 401 452 Rich 7.6 8.42 10.1 32.89 *** 245 232 274 Sex of headship Female 9.11 9.7 10.72 17.67 *** 953 1006 1221 Male 7.89 8.8 10.44 32.32 *** 5989 5789 5925 Continued. Index of freedom from patriarchal beliefs among women All 1.55 1.70 1.91 23.23 *** 6722 6795 7007 Region Coast 1.21 1.53 1.57 29.75 *** 1803 1835 1871 Haor 1.51 1.75 1.97 30.46 *** 1514 1654 1658 Mid Char 1.45 1.59 1.74 20.00 *** 1748 1710 1622 North Char 1.68 1.73 1.98 17.86 *** 1657 1596 1856 Well-being category Extreme poor 1.60 1.68 1.83 14.38 *** 1119 1208 1299 Poor 1.54 1.70 1.93 25.32 *** 4342 4363 4418 Middle 1.48 1.74 2.06 39.19 *** 606 590 570 Middle rich 1.58 1.78 2.01 27.22 *** 412 401 447 Rich 1.71 1.91 1.99 16.37 *** 237 232 272 Sex of headship Female 1.60 1.66 1.83 14.38 *** 908 1006 1193 Male 1.54 1.71 1.93 25.32 *** 5814 5789 5814 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. IR 3.1 Table 52. Indicators of women's empowerment for IR 3.1, by region, well-being category, and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Decision making score for women in household All 2.28 2.38 2.84 24.56 *** 6528 6440 6612 Region Coast 2.21 2.40 2.66 20.36 *** 1797 1775 1721 Haor 2.24 2.24 2.84 26.79 *** 1588 1550 1596 Mid Char 2.38 2.18 2.74 15.13 *** 1560 1640 1548 North Char 2.29 2.42 2.93 27.95 *** 1583 1475 1747 Well-being category Extreme poor 2.64 2.63 2.98 12.88 *** 1013 1059 1173 Poor 2.19 2.33 2.81 28.31 *** 4264 4200 4217 Middle 2.29 2.31 2.79 21.83 *** 595 567 545 Middle rich 2.30 2.33 2.84 23.48 *** 410 394 422 Rich 2.34 2.32 2.76 17.95 *** 240 219 254 Sex of headship Female 3.17 3.22 3.51 10.73 *** 827 842 1060 Male 2.16 2.26 2.73 26.39 *** 5701 5598 5552 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. IR 3.2 Table 53. Indicators related to domestic violence against women for IR 3.2, by region, well-being category and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Number of observations Baseline Midterm Endline Percent of households in which female member was yelled at or struck during last year All 27.73 12.19 7.18 -74.11 *** 6507 6515 6947 Region Coast 20.54 17.47 10.20 -50.34 *** 1763 1758 1826 Haor 23.10 12.32 5.88 -74.55 *** 1507 1554 1646 Mid Char 23.90 14.53 15.35 -35.77 *** 1656 1626 1621 North Char 35.66 10.25 4.28 -88.00 *** 1581 1577 1854 Well-being category Extreme poor 16.30 11.24 5.75 -64.72 *** 1083 1138 1305 Poor 31.38 12.85 7.84 -75.02 *** 4207 4189 4367 Middle 25.04 9.46 5.15 -79.43 *** 587 577 567 Middle rich 20.18 6.96 7.60 -62.34 *** 393 390 437 Rich 16.90 7.70 3.72 -77.99 *** 232 220 270 Sex of headship Female 7.84 5.85 3.26 -58.42 ** 884 945 1194 Male 30.74 12.21 7.91 -74.27 *** 5623 5570 5753 Percent of women who believe a man is justified in hitting/abusing wife (across 6 situations) She goes out without telling him 50.08 25.68 11.74 -76.56 *** 6837 6737 7047 She neglects the children 51.00 30.99 12.72 -75.06 *** 6856 6730 7047 She argues with him 54.40 34.36 14.74 -72.90 *** 6835 6722 7021 She refuses to have sex with him 20.81 9.22 3.16 -84.81 *** 6404 6360 7004 She burns the food 18.18 11.04 4.28 -76.46 *** 6760 6702 7044 She does not obey elders 61.44 45.75 22.98 -62.60 *** 6823 6697 7045 Continued. Percent of households for which assistance was sought following an incident All 14.17 12.34 7.93 -44.04 ** 1637 837 597 Region Coast 14.56 19.99 11.63 -20.12 353 296 192 Haor 11.27 17.84 9.89 -12.24 349 183 85 Mid Char 23.10 0.50 5.77 -75.02 *** 381 206 244 North Char 13.19 11.33 7.43 -43.67 554 152 76 Well-being category Extreme poor 14.43 14.95 11.14 -22.80 178 120 77 Poor 14.51 11.94 6.78 -53.27 *** 1209 622 429 Middle ‡ ‡ ‡ ‡ 138 51 39 Middle rich ‡ ‡ ‡ ‡ 73 28 40 Rich ‡ ‡ ‡ ‡ 39 16 12 Sex of headship Female ‡ ‡ ‡ ‡ 71 58 66 Male 13.93 11.08 6.90 -50.47 *** 1566 779 531 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. ‡ Insufficient number of observations. BY PEP/NON-PEP Table 54. SO3 key indicators (PEP versus non-PEP) Non-PEP PEP Difference (PEP vs. non￾PEP) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Women’s empowerment indicators Percent of women earning cash income/control over economic resources 2.59 9.03 248.65 * 5.91 20.71 250.42 * 1.77 * Index of women's freedom of movement 7.98 10.38 30.08 * 8.06 10.49 30.15 * 0.07 Index for freedom from patriarchal beliefs among women 1.56 2.03 30.13 * 1.55 1.90 22.58 * -7.55 * Decision making score for women 2.30 2.80 21.74 * 2.28 2.85 25.00 * 3.26 Domestic violence indicators 0.0 Percent of households in which a female member was yelled at or struck during last year 21.80 5.80 -73.39 * 28.26 7.30 -74.17 * -0.77 Percent of households for which assistance was sought following an incident 9.09 11.74 29.15 14.50 7.67 -47.10 * -76.26 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. BY MALE/FEMALE HHH Table 55. SO3 key indicators, by sex of household head Male headed households Female headed households Difference (Female vs. male) Indicator Base￾line End￾line Percent difference (Endline - Baseline) Base￾line End￾line Percent difference (Endline - Baseline) Women’s empowerment indicators Percent of women earning cash income/control over economic resources 3.99 17.57 340.35 * 16.38 31.87 94.57 * -245.78 * Index of women's freedom of movement 7.89 10.44 32.32 * 9.11 10.72 17.67 * -14.65 * Index for freedom from patriarchal beliefs among women 1.54 1.93 25.32 * 1.60 1.83 14.38 * -10.95 * Decision making score for women 2.16 2.73 26.39 * 3.2 3.5 10.73 * -15.66 * Continued. Domestic violence indicators Percent of households in which a female member was yelled at or struck during last year 30.73 7.91 -74.26 * 7.84 3.26 -58.42 * 15.84 Percent of households for which assistance was sought following an incident 13.93 6.90 -50.47 * 20.22 21.37 5.69 56.15 Note: Stars indicate different in the endline-baseline (percent) difference is statistically significant at the 5% level. SO4 Tables Note: Baseline and midline values were changed due to a coding error. IR 4.1 (no outcome indicators) IR 4.2 Continued. Table 56. Utilization of services, by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Average number of services utilized All households 5.3 6.8 8.4 58.1 *** 7,085 6,916 7,148 PEP households 5.2 6.8 8.3 59.4 *** 5,765 5,672 5,844 Region Coast 5.1 6.5 6.6 28.5 *** 1,885 1,852 1,907 Haor 5.2 7.0 7.9 54.2 *** 1,672 1,701 1,689 Mid Char 5.8 7.2 8.8 52.6 *** 1,798 1,745 1,652 North Char 5.2 6.5 8.8 69.9 *** 1,730 1,618 1,900 Well-being category Extreme poor 4.5 6.0 7.4 65.2 *** 1,195 1,237 1,333 Poor 5.4 7.0 8.6 59.4 *** 4,566 4,435 4,511 Middle 6.2 6.8 9.2 49.9 *** 638 599 577 Middle rich 6.2 7.1 9.0 45.2 *** 428 409 452 Rich 6.5 7.2 9.1 40.8 *** 252 235 274 Program approach MCHN/PEP 5.2 6.9 8.3 59.4 *** 2,888 2,766 2,945 PM2A 5.6 6.7 8.7 54.6 *** 4,197 4,150 4,203 Sex of headship Female 3.6 5.1 5.8 61.8 *** 957 1,009 5,927 Male 5.5 7.1 8.8 59.2 *** 6,128 5,907 1,221 Percent of households utilizing various services a/ Primary Health Care Services 85.7 86.4 92.6 8.0 *** 6,344 6,718 6,815 Union Parishad 75.2 87.8 86.6 15.1 *** 6,926 6,880 7,031 Gov. Family Planning 73.6 83.4 75.9 3.2 6,167 6,555 6,686 Family Planning Services 73.3 80.5 75.8 3.5 6,378 6,887 6,856 Grammo Shalish 64.0 65.5 71.9 12.4 *** 6,439 6,815 6,792 Gov. Land Office 53.3 52.0 66.7 25.3 *** 3,689 5,367 5,655 Gov. Immunization Services 69.4 74.3 65.9 -5.1 * 6,591 6,838 6,966 Dept. of Livestock 23.6 30.7 62.8 166.6 *** 1,158 2,582 4,652 Dept. of Ag Extension 22.5 40.0 61.5 172.7 *** 911 3,614 5,292 BADC Seed Department 33.7 34.2 59.1 75.7 *** 679 1,920 3,910 Primary School 56.4 59.0 58.7 4.1 6,994 6,883 7,066 Dept. of Social Services 24.4 26.3 55.5 127.3 *** 683 1,703 4,331 Dept. of Women's Affairs 18.9 24.8 54.4 188.1 *** 596 1,270 4,032 Dept. of Cooperatives 19.3 30.7 39.0 102.0 *** 445 688 2,945 Dept. of Fisheries 11.4 22.9 35.5 210.6 *** 610 2,498 4,155 Dept. of Youth Development 15.9 19.4 34.0 114.3 *** 434 904 3,019 Pre-School 32.0 35.0 33.4 4.6 5,795 6,667 6,591 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. a/ Calculated only for households for which each service is available. Table 57. Utilization of safety nets, by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Average number of safety nets engaged in (out of 11) a/ All households 0.2 0.4 0.6 255.7 *** 7,085 6,916 7,148 PEP households 0.2 0.4 0.7 216.2 *** 5,765 5,672 5,844 Region Coast 0.1 0.3 0.5 455.1 *** 1,885 1,852 1,907 Haor 0.1 0.4 0.6 386.6 *** 1,672 1,701 1,689 Mid Char 0.2 0.3 0.4 94.3 *** 1,798 1,745 1,652 North Char 0.2 0.5 0.8 233.1 *** 1,730 1,618 1,900 Well-being category Extreme poor 0.3 0.5 0.8 188.7 *** 1,195 1,237 1,333 Poor 0.2 0.4 0.6 282.4 *** 4,566 4,435 4,511 Middle 0.1 0.2 0.3 189.4 ** 638 599 577 Middle rich 0.0 0.1 0.2 436.2 ** 428 409 452 Rich 0.1 0.1 0.0 -50.0 252 235 274 Program approach MCHN/PEP 0.2 0.4 0.7 260.9 *** 2,888 2,766 2,945 PM2A 0.1 0.3 0.5 229.5 *** 4,197 4,150 4,203 Sex of headship Female 0.3 0.5 0.9 231.7 *** 957 1,009 5,927 Male 0.2 0.4 0.6 257.0 *** 6,128 5,907 1,221 Continued. Continued. Percent of households engaged in various safety nets SHOUHARDO IIb/ b/ b/ 75.5 -- -- -- -- 7,146 Government VGF 4.0 13.3 39.6 879.0 *** 7,085 6,916 7,146 Aged allowance 5.2 7.8 8.0 52.7 *** 7,085 6,916 7,146 Government VGD 4.0 6.0 6.3 57.8 ** 7,085 6,916 7,146 40 days work a/ 4.6 4.0 -- -- -- 6,916 7,146 Widow allowance 2.3 2.5 2.8 -- 7,085 6,916 7,146 100 days work 1.9 0.8 1.7 -7.4 7,085 6,916 7,146 Govt. cash-for-work 0.5 0.7 1.4 167.1 *** 7,085 6,916 7,146 Non-Govt. cash-for￾work 0.3 0.5 1.0 220.4 *** 7,085 6,916 7,146 Self-help groupb/ b/ b/ 0.9 -- -- -- -- 7,146 Comm. based savings group 0.1 0.8 0.9 1690.4 *** 7,085 6,916 7,146 Other 0.9 3.1 0.8 -5.9 7,085 6,916 7,146 Maternal allowance a/ 0.5 0.7 -- -- -- 6,916 7,146 Non-Govt. food-for￾work 0.4 0.7 0.6 32.0 7,085 6,916 7,146 Disability allowance 2.2 2.5 0.5 -77.5 7,085 6,916 7,146 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. a/ The midterm survey included two additional safety nets not in the baseline: widow allowance and 40 days of work. b/ The end line survey included two safety nets, SHOUHARDO II and self-help groups, neither of which was in baseline or midterm. Table 58. Utilization of common property resources by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Average number of common property resources utilized (out of 12) All households 0.75 1.05 1.84 146.3 *** 7,085 6,916 7,148 PEP households 0.75 1.06 1.86 148.0 5,765 5,672 5,844 Region Coast 1.03 1.50 1.78 73.8 *** 1,885 1,852 1,907 Haor 0.37 1.14 1.83 396.1 *** 1,672 1,701 1,689 Mid Char 1.39 1.33 1.88 35.6 *** 1,798 1,745 1,652 North Char 0.82 0.76 1.84 125.4 *** 1,730 1,618 1,900 Well-being category Extreme poor 0.64 1.06 1.79 177.7 *** 1,195 1,237 1,333 Poor 0.77 1.05 1.88 145.9 *** 4,566 4,435 4,511 Middle 0.91 1.00 1.76 92.6 *** 638 599 577 Middle rich 0.80 1.02 1.55 93.6 *** 428 409 452 Rich 0.71 1.25 1.47 107.6 *** 252 235 274 Program approach MCHN/PEP 0.76 1.05 1.86 146.3 *** 2,888 2,766 2,945 PM2A 0.72 1.08 1.76 145.7 *** 4,197 4,150 4,203 Sex of headship Female 0.59 0.73 1.39 134.7 *** 957 1,009 1,221 Male 0.77 1.11 1.92 149.5 *** 6,128 5,907 5,927 SO5 Tables Continued. Percent of households using various resources a/ (type of common property) River/Canal 44.6 48.8 67.9 52.2 *** 4,152 5,365 5,414 Beel/Haor 41.5 42.5 63.4 52.9 2,813 4,062 3640 Forest land 34.3 45.3 61.9 80.4 *** 274 273 1,034 Grazing Land 32.5 34.0 59.4 83.0 700 1,302 1,900 Hills 42.7 51.7 50.8 19.1 391 358 1,549 Roadside Sloping 23.2 19.7 46.0 98.6 *** 2,782 4,160 5,780 CBO Water Body 12.2 25.7 45.1 271.4 *** 275 282 1,012 Embankments 32.8 27.5 42.1 28.3 * 2,002 2,152 1,985 Khas Pond 17.7 13.4 32.6 84.8 *** 596 1,016 1,098 Khas Land 21.8 15.9 31.8 46.2 *** 1,340 2,532 2,448 Railway Grounds 11.9 19.0 29.7 148.8 * 370 670 280 Other ‡ ‡ ‡ ‡ ‡ 39 141 296 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. a/ Calculated only for households for which each common property is available. ‡ Insufficient number of observations. Table 59. Households sold assets to meet urgent household needs and taka value of distress sales, by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Percent of households distress selling All households 10.0 10.1 9.8 -1.5 649 689 638 PEP households 9.6 9.5 9.8 1.7 473 505 507 Region Coast 6.9 9.2 7.9 14.0 131 161 145 Haor 8.6 9.7 12.6 46.2 140 190 200 Mid Char 10.9 6.3 9.5 -12.9 184 115 146 North Char 11.4 12.5 7.2 -36.8 ** 194 223 147 Well-being category Extreme poor 8.5 7.4 8.6 1.1 76 96 114 Poor 9.9 10.1 10.2 2.7 397 409 393 Middle 11.8 14.1 9.7 -18.0 67 79 54 Middle rich 14.5 19.2 9.8 -32.2 61 66 40 Rich 20.3 18.5 14.2 -30.0 47 39 37 Program approach MCHN/PEP (PEP only) 10.2 9.6 10.0 -1.7 274 262 277 PM2A 9.4 11.8 9.3 -1.2 375 427 361 Sex of headship Female 8.5 9.0 7.8 -8.3 58 87 85 Male 10.2 10.3 10.2 0.0 591 602 553 2010 taka value of household assets sold to meet urgent household need a/ All households 19,979 21,304 16,231 46.0 649 689 637 PEP households 17,132 18,512 12,559 -26.7 473 505 506 Region Coast 34,427 26,207 19,453 -43.5 131 161 145 Haor 30,559 30,522 13,446 -56.0 ** 140 190 200 Mid Char 14,200 21,423 17,861 25.8 184 115 146 North Char 13,392 12,210 20,202 50.9 194 223 146 Well-being category Extreme poor 15,360 13,038 12,930 -15.8 76 96 114 Poor 17,546 19,581 12,451 -29.0 397 409 392 Middle 25,488 32,583 38,664 51.7 67 79 54 Middle rich 40,555 47,240 42,270 4.2 61 66 40 Rich 64,427 42,609 93,430 45.0 47 39 37 Program approach MCHN/PEP (PEP only) 14,706 19,511 12,277 -16.5 274 262 276 PM2A 38,201 25,994 30,544 -20.0 375 427 361 Sex of headship Female 6,462 16,535 8,351 29.2 591 602 552 Male 21,735 21,981 17,346 -20.2 58 87 85 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Note: Two extreme cases (more than 4 times greater than the next largest value) of value of assets sold were excluded from the analysis as outliers. a/ Includes only households reporting sale of assets. Table 60. Information on natural disasters collected from endline survey respondents, by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Percent of households that experienced any disaster in the last 12 months All households 64.2 62.3 74.9 16.6 *** 7,085 6,916 7,106 PEP households 65.0 62.4 74.9 15.2 *** 5,694 5,644 5,806 Region Coast 95.0 75.2 15.6 -83.6 *** 1,885 1,852 1,881 Haor 46.9 44.5 74.1 58.0 *** 1,672 1,701 1,686 Mid Char 73.3 77.6 79.9 9.0 1,798 1,745 1,647 North Char 75.0 73.9 80.3 7.0 1,730 1,618 1,892 Well-being category Extreme poor 71.9 64.4 75.9 5.6 1,195 1,237 1,324 Poor 62.1 62 74.5 20.0 4,566 4,435 4,482 Middle 68.3 61.2 73.1 7.1 638 599 575 Middle rich 62.1 58.9 69.0 11.1 428 409 450 Rich 61.5 59.9 72.3 17.6 252 235 274 Program approach MCHN/PEP (PEP only) 64.0 62.0 76.7 19.8 *** 2,888 2,766 2,917 PM2A 65.1 63.4 67.7 4.0 4,197 4,150 4,189 Sex of headship Female 65.8 61.3 76.7 16.5 *** 957 1,009 1,212 Male 64.0 62.5 67.7 5.8 6,128 5,907 5,894 Percent of households experiencing various types of disasters (types of disasters) None 34.8 37.3 25.4 (26.9) *** 7,085 6,916 7,106 Floods 37.8 38.9 56.6 49.9 *** 7,085 6,916 7,106 Heavy Rains 23.1 23.2 23.8 2.9 7,085 6,916 7,106 Wind Storms 27.8 18.8 23.1 (16.7) *** 7,085 6,916 7,106 Drought 6.7 7,106 Hail 3.3 7,106 Erosion 2.4 3.4 2.5 4.1 *** 7,085 6,916 7,106 Cold Wave 4.4 2.3 1.7 (61.3) *** 7,085 6,916 7,106 Earthquake 7.9 9.2 1.3 (83.4) * 7,085 6,916 7,106 Wildfire 3.3 4.1 1.2 (64.2) *** 7,085 6,916 7,106 Hurricane 0.1 0.1 0.2 136.9 *** 7,085 6,916 7,106 Cyclone 0.3 0.2 0.2 (27.6) * 7,085 6,916 7,106 Other 0.9 3 0.0 (97.0) -- 7,085 6,916 7,106 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels. Continued. Table 61. Information on natural disasters collected from endline survey respondents (continued), by region, well-being category, program approach and sex of household head Indicator Baseline Midterm Endline Percent Difference (Endline￾Baseline) Significance level Number of observations Baseline Midterm Endline Mean number of disasters experienced All households 1.1 1.0 1.2 11.7 * 7,085 6,916 7,148 PEP households 1.1 1.0 1.2 11.9 * 5,765 5,672 5,844 Region Coast 2.0 1.0 0.2 -90.5 *** 1,885 1,852 1,907 Haor 0.7 0.6 1.2 62.5 *** 1,672 1,701 1,689 Mid Char 1.2 1.3 1.4 18.4 1,798 1,745 1,652 North Char 1.3 1.4 1.3 -1.7 1,730 1,618 1,900 Well-being category Extreme poor 1.2 1.1 1.1 -6.0 1,195 1,237 1,333 Poor 1.1 1.0 1.0 -3.5 ** 4,566 4,435 4,511 Middle 1.1 1.0 1.0 -9.2 638 599 577 Middle rich 1.0 0.9 0.9 -9.6 428 409 452 Rich 1.1 1.0 1.0 -8.8 252 235 274 Program approach MCHN/PEP (PEP only) 1.1 1.0 1.3 16.4 * 2,888 2,766 2,945 PM2A 1.1 1.0 1.0 -5.0 4,197 4,150 4,203 Sex of headship Female 1.1 1.0 1.1 5.6 957 1,009 5,927 Male 1.1 1.0 1.2 13.0 * 6,128 5,907 1,221 Percent of households citing perceived mitigation measures that could reduce the impact of future disasters (types of mitigation measures) None 47.1 41.2 6.7 (85.9) *** 7,085 6,916 7,148 Structural Improvement to Home 3.5 3.3 48.2 1,277.1 *** 7,085 6,916 7,148 Improvement to infrastructure 1.6 1.5 15.9 894.8 *** 7,085 6,916 7,148 Community disaster response plan 3.1 2.5 5.8 87.8 *** 7,085 6,916 7,148 Food stocks 1.5 1.6 38.5 2,464.6 *** 7,085 6,916 7,148 Water stocks 0.1 0.2 17.6 17,538.0 *** 7,085 6,916 7,148 Medical supplies stocks 0 0 7.1 - *** 7,085 6,916 7,148 First aid training 0.1 0.2 2.4 2,327.3 *** 7,085 6,916 7,148 Increased collaboration w/ neighbors 2.2 0.8 6.3 188.1 *** 7,085 6,916 7,148 Increased collaboration w/ communities 0.5 0.4 4.4 770.4 *** 7,085 6,916 7,148 Better forecasting 0.6 0.7 6.2 932.9 *** 7,085 6,916 7,148 Earlier/better warning 0.6 1.6 3.9 542.8 *** 7,085 6,916 7,148 Increased collaboration w/ local gov't 0.4 0.2 1.7 327.4 *** 7,085 6,916 7,148 Diversified income 0.1 0.1 1.6 1,523.2 *** 7,085 6,916 7,148 Evacuation routes/plans 0 0.2 3.4 - *** 7,085 6,916 7,148 Improved modes of communication 0 0.1 0.9 - *** 7,085 6,916 7,148 Other 0.3 0.3 1.8 486.3 *** 7,085 6,916 7,148 Note: Stars indicate difference is statistically significant at the 10% (*), 5%(**) or 1%(***) levels.