Download StudyGeneral

Study Overview

Improving Smallholder Livelihoods and Nutrition: An Evaluation of WFP’s Nutrition-Sensitive Food for Assets program in Sri Lanka
Study is 3ie funded:
Study ID:
Initial Registration Date:
Last Update Date:
Study Status:
Sri Lanka

To address the immediate and underlying causes of nutritional problems, inputs from multiple sectors are needed. Nutrition-sensitive development programmes that integrate or co-locate nutrition and health components with elements from other sectors such as agriculture, social protection, gender, or water will likely lead to larger impacts than programs that address only one determinant of one nutrition problem. In this impact evaluation, we aim to assess the contribution of the World Food Programme's (WFP) Food for Assets program in Sri Lanka, known as R5N, and of a nutrition-sensitive version of R5N that includes a group-based health promotion process (HPP), on household consumption patterns and individual diets. We conduct multiple rounds of surveys involving a panel of around 1300 households in three study arms: R5N (standard WFP intervention), R5N+HPP (nutrition-sensitive WFP intervention), or control (no R5N or HPP activities). The R5N households are from 30 administrative units (Gram Niladhari's or GNs) across 5 Divisional Secretariat Divisions (DSDs) selected by WFP and the Government of Sri Lanka. Half of these 30 GNs were randomly selected to receive additional HPP activities. A set of 15 matched control GNs within the same DSDs were selected by the evaluation team. Household phone surveys were conducted in Dec 2020-Feb 2021 (baseline), May-July 2021 (midline), and Dec 2021-Feb 2022 (endline) and may also be conducted Dec 2022-Feb 2023 (follow-up sustainability survey). Phone surveys included an individual-level 24-hour dietary recall and modules on household consumption, demographics, program exposure, nutrition knowledge, food security, agriculture activities, and experiences related to COVID-19. Market surveys were conducted monthly and community surveys with key informant interviews were conducted at baseline and endline.


Registration Citation:
Agriculture and Rural Development
Health, Nutrition, and Population
Social Protection
Water and Sanitation
Additional Keywords:
Secondary ID Number(s):

Principal Investigator(s)

Name of First PI:
Deanna Olney
Name of Second PI:
Samuel Scott

Study Sponsor

Agriculture for Nutrition and Health
Study Sponsor Location:
United States

Research Partner

Name of Partner Institution:
Medical Research Institute, Wayamba University, University of Peradeniya, World Food Programme, Johns Hopkins University, University of California Davis
Type of Organization:
Government agency (eg., statistics office, Ministry of Health)
Sri Lanka

Intervention Overview


R5N intervention

The R5N program aims to improve access to water for farming in the dry months and increase household income through a diverse set of activities involving construction and rehabilitation of water infrastructure as well as trainings and hard inputs for agricultural livelihoods. The program also seeks to strengthen community-based organizations, especially those related to agriculture, link producers to markets, build government capacity, and link program households with entitlement schemes during COVID-19. 

HPP intervention 

To make the R5N program more nutrition sensitive, WFP has partnered with the Foundation for Health Promotion (FHP) in Sri Lanka to design and deliver a behavior change intervention through a health promotion process (HPP) to men and women. The HPP includes messaging on: high quality diets, healthy lifestyle, family wellbeing, women's decision-making, cash management, and climate-smart nutrition-sensitive agriculture. Messages are delivered in a group setting to members of women's groups, men's groups, and elderly person groups. Tools were developed such as a money management chart, happiness calendar, containers to measure oil/sugar/salt consumption, kitchen calendar, TV card to measure time spent watching television, and a box containing tape for waist and hip measurements.  Facilitators, trained health promotion graduate students, conducted at least one formal session per month with each community group type. During these sessions, the facilitator used a participatory, community-driven approach whereby the group members were encouraged to identify their own solutions to local challenges related to the key topics. In addition, one to two informal sessions per month were organized at a different forum, depending on the community structure. Due to COVID restrictions, group gatherings were replaced by phone discussions and WhatsApp groups during some months.

Theory of Change:

Program inputs fall into four main categories: water infrastructure, cash transfers, other agriculture inputs and training and health promotion activities. These four categories of inputs contribute to four potentially interrelated program impact pathways. The first is the construction of different types of water infrastructure such as community water reservoir repair, ponds or wells. The inputs for this include the materials for construction and/or cash to buy supplies and training. For the community water reservoirs, participants in the construction activities also received cash transfers in exchange for labor. The outputs of these activities are expected to be improved water sources to support agriculture activities in the participating communities. Some beneficiaries who did not participate in the water-related activities received agriculture inputs and participated in agriculture trainings. These inputs were expected to lead to increased agriculture production and in turn, income. In addition to the agriculture focused activities, some communities were randomly selected to also receive program inputs from the Foundation for Health Promotion (FHP) which included facilitated discussions led by trained health promoters and communication materials. Processes include engagement with community groups and knowledge transfer from FHP staff to community members. Outputs include sharing of messages between community members and increased knowledge of diet, health and nutrition practices. Expected outcomes across the four program impact pathways include increased water availability during the dry months, increased agricultural productivity, increased sale of agricultural outputs, increased income, food security, improved dietary behavior, and increased diet quality. These outcomes are expected to lead to improved physical and mental health in beneficiaries.  

Multiple Treatment Arms Evaluated?

Implementing Agency

Name of Organization:
World Food Programme, Government of Sri Lanka, Foundation for Health Promotion
Type of Organization:
Foreign or Multilateral Aid Agency

Program Funder

Name of Organization:
Korea International Cooperation Agency
Type of Organization:
Public Sector, e.g. Government Agency or Ministry

Intervention Timing

Intervention or Program Started at time of Registration?
Start Date:
End Date:
Evaluation Method

Evaluation Method Overview

Primary (or First) Evaluation Method:
Other (not Listed) Method:
Additional Evaluation Method (If Any):
Difference in difference/fixed effects
Other (not Listed) Method:

Method Details

Details of Evaluation Approach:

For the main impact analysis, we will use matching with difference-in-difference (DID) estimation. Households in the three treatment arms will be matched on baseline demographic characteristics to estimate impact at midline and endline using DID methods. The matching method we will use is nearest neighbor matching (NNM) which matches each beneficiary household to one or more non-beneficiary households based on pre-intervention characteristics. We will conduct robustness checks by varying the number of neighbors over which matching takes place. Matching covariates will include variables likely to affect both selection into treatment and the outcomes of interest, and selection will be guided by theory, knowledge of the intervention, and our previous experience. The final set of variables used for matching will also be chosen based on propensity score densities and balance, ensuring that the common support assumption holds. For dietary outcomes, group differences in dietary diversity will be tested at midline and endline using negative binomial regression. Prevalence of inadequate dietary diversity or dietary quality will be tested using logistic regression. The National Cancer Institute method for modeling usual food and nutrient intakes will be used when evaluating food or nutrient intakes by group. For the food environment sub-study, longitudinal mixed models will be used to assess temporal trends in food environment variables and multilevel models will be used to examine the associations between food environment variables and diet outcomes, with individuals nested within food environments.

Outcomes (Endpoints):

Individual-level outcomes: dietary diversity and overall dietary quality score (primary), and energy and nutrient intakes.  Household-level outcomes: consumption and expenditure (primary), energy and nutrient supply, and food security. Intermediate outcomes: exposure, nutrition knowledge, agriculture activities, asset ownership and perceived COVID-19 impacts. Market survey outcomes: food availability, Nutrition Environment Measures for Stores-Survey score, cost of recommended diet, affordability of healthy diet, relative cost of food groups.

Unit of Analysis:
Individual (diet), household (consumption), cluster/GN (food environment)

The treatment comparisons will be 1) R5N+HPP vs. control (hypothesis: the nutrition-sensitive version of WFP's R5N program benefits outcomes), 2) (R5N + (R5N+HPP)) vs. control (hypothesis: WFP's R5N program benefits outcomes), and 3) R5N+HPP vs. R5N (hypothesis: the nutrition-sensitive version of the program has greater benefits than the standard version). For the food environment sub-study, we have three primary hypotheses: 1) greater availability of nutritious food and higher NEMS-S scores in local food environments will be associated with improved diet intake; 2) lower cost healthy diets and lower relative caloric prices will be associated with improved diet intake; 3) longitudinal variation in food availability and food costs during 2021 will differ from previous years' trends.

Unit of Intervention or Assignment:
Grama Niladhari division (GN)
Number of Clusters in Sample:
Number of Individuals in Sample:
Size of Treatment, Control, or Comparison Subsamples:
30 any R5N GNs (15 R5N GNs, 15 R5N+HPP GNs), 15 control GNs

Supplementary Files

Analysis Plan:
Other Documents:

Outcomes Data

The data will be collected using 1) a household survey at baseline, midline and endline, 2) monthly market surveys to assess food availability in local markets, 3) a community (GN-level) key informant interview at baseline and endline. The household survey and key informant interviews are phone surveys and the market surveys are conducted in person.
Data Already Collected?
Data Previously Used?
Data Access:
Restricted -- Access requires a formal approval process
Data Obtained by the Study Researchers?
Data Approval Process:
The study was approved by the Institutional Review Boards of Wayamba University of Sri Lanka and the International Food Policy Research Institute. Those interested in using the data should contact the PI for permission.
Approval Status:
Yes-obtained approval and have received the data

Treatment Assignment Data

Participation or Assignment Information:
Data Obtained by the Study Researchers?
Data Previously Used?
Data Access:
Data Obtained by the Study Researchers?
Data Approval Process:
Approval Status:

Data Analysis

Data Analysis Status:

Study Materials

Upload Study Materials:

Registration Category

Registration Category:
Prospective, Category 3: Data for measuring impacts have been obtained/collected by the research team but analysis for this evaluation has not started

Completion Overview

Intervention Completion Date:
Data Collection Completion Date:
Unit of Analysis:
Clusters in Final Sample:
Total Observations in Final Sample:
Size of Treatment, Control, or Comparison Subsamples:


Preliminary Report:
Preliminary Report URL:
Summary of Findings:
Paper Summary:
Paper Citation:

Data Availability

Data Availability (Primary Data):
Date of Data Availability:
Data URL or Contact:
Access procedure:

Other Materials

Survey Instrument Links or Contact:
Program Files:
Program Files Links or Contact:
External Link:
External Link Description:
Description of Changes:

Study Stopped