81195
                   enGender Impact: The World Bank’s Gender Impact Evaluation Database



Almost Random: Evaluating a Large-Scale Randomized Nutrition Program in the Presence of
                                       Crossover
       Author(s)            Sebastian Linnemayr and Harold Alderman
       Contact              halderman@worldbank.org

       Country              Senegal

  Organizing Theme          Health

        Status              The Project is ongoing but the evaluation is complete

 Intervention Category      Nutrition

        Sector              Health, Nutrition & Population
                            Large-scale randomized interventions have the potential to uncover the causal effect of
                            programs applying to a large population, thereby improving on the insights gained from
                            currently dominant smaller randomized studies. However, the external validity gained
                            through larger interventions typically implies less supervision and often comes at the cost
                            of some deviation from the randomization plan. This paper investigates the impact of the
                            Nutrition Enhancement Program, which aims to improve child nutrition in Senegal based
                            on a large-scale randomized community intervention. The analysis explicitly deals with
                            deviation from the planned treatment and suggests approaches for combining ex-post
       Abstract             adjustments such as propensity score matching with the randomized treatment plan. The
                            authors do not detect a strong overall program impact on the outcome measure of weight-
                            for-age based on planned treatment status, but do find an impact on the youngest children.
                            Moreover, the project impact is clearer when the analysis considers treatment crossover
                            using alternative estimators of two-stage least-squares and propensity score matching. The
                            findings underscore the importance of addressing the shortcomings of large-scale
                            randomization interventions in a systematic manner in order to understand the selection
                            process that can guide further implementation of such projects, as well as to expose the
                            true, causal effect of such programs.
  Gender Connection         Gender Informed Analysis

  Gender Outcomes           Nutrition
                            Clustered Randomized Control Trial, robustness checks are preformed with DID, IV and
      IE Design
                            PSM
                            Monthly growth promotion was provided to 200000 mothers and their children. This
                            consisted of a monthly discussion with mothers. Key community leaders were targeted for
                            involvement. Meetings of pregnant women were organized to generate a forum where
                            women can exchange their ideas and experiences. The intervention was motivated by the
     Intervention           principle of positive deviation where people who show good behavior share their
                            experience with others. The following components are included in the program: exclusive
                            breastfeeding promotion discussion groups, weighing of children and discussion on growth
                            promotion, vitamin A distribution, Iron supplementation, bednet distribution for a fee,
                            deworming and cooking workshops demonstrated the preparation of healthy food.



    Last updated: 14 August 2013                                                                      1
                   enGender Impact: The World Bank’s Gender Impact Evaluation Database


                            The program started in 2002. The main component was the monthly community
  Intervention Period
                            discussion.
                            From a list of 1000 villages, 212 villages were randomly chosen in 3 regions. The nutrition
   Sample population        intervention was randomly assigned to half the villages. A survey was conducted for all
                            children in the villages.
Comparison conditions       The control group received the treatment 18 months after the treatment group.

    Unit of analysis        Individual level
                            A baseline survey was conducted in April 2004, in June 2006 a follow up study was
   Evaluation Period
                            conducted.
                            The intervention increases weight for age by .1 standard deviations. Parent's education
                            improves the nutrition status, and the results do not differ between genders. There was a
        Results             reduction of giving non-breast milk liquids to infants of 11% in the treatment group. After
                            controlling for observable characteristics, there is about a .27 standard deviation increase in
                            weight-for-age score.
                            Because this trial is conducted at such a broad level, there are potential crossover effects
Primary study limitations
                            between the treatment and control.
    Funding Source
                            Linnemayr, S., & Alderman, H. (2011). Almost random: Evaluating a large-scale
      Reference(s)          randomized nutrition program in the presence of crossover. Journal of Development
                            Economics, 96(1), 106-114.
    Link to Studies
                            http://www.sciencedirect.com/science/article/pii/S0304387810000635
       Microdata




    Last updated: 14 August 2013                                                                         2