A Randomized Impact Evaluation of the Graduation Program Strengthen Productive Safety Net Program Phase 4 (PSNP4) Institutions and Resilience (SPIR) in Ethiopia

The impact evaluation evaluates combinations of four SPIR interventions described below; L and N correspond to the primary SPIR interventions around livelihoods and nutrition, respectively, while L* and N* represent enhanced versions of these interventions.

Intervention L: SPIR livelihood activities: starting Village Economic and Social Associations (VESAs), financial literacy training, agriculture and livestock value chain development, home gardening and forage production
Intervention L*: SPIR livelihoods activities plus: (i) Social Analysis and Action (SAA) to improve women’s access to markets, (ii) aspirations promotion activities in randomly selected kebeles (subdistricts), and (iii) targeted poultry or cash livelihood transfers
Intervention N: SPIR nutrition activities: Nutrition Behavior Change Communication (BCC); WASH activities
Intervention N*: SPIR nutrition activities plus: (i) Timed and Targeted Counseling (TTC) (more intensive nutrition BCC), (ii) Community-based Participatory Nutrition Promotion (CPNP), (iii) male engagement in BCC, and (iv) Interpersonal Therapy in Groups (IPT-G) interventions for women and men screened for depression (provided after the midline survey)

All four interventions have the potential to improve women’s empowerment and well-being. The VESA and BCC activities under L and N bring women and men together to provide them with knowledge and training over a wide range of topics, thereby improving social capital and catalyzing women’s empowerment. The L* and N* interventions go a step further by directly addressing depression through the IPT-G and directly addressing gender norms and women’s role in the household through the SAA activities which address constraints on women’s role in intrahousehold decision-making, mobility, choice of livelihood activities, and access to markets that derive from cultural and social norms; targeting women for the poultry and cash transfers; and including men in the nutrition BCC and creating male engagement groups. Positive social interaction and social support as a result of the group interventions and reduced poverty-related stress as a result of higher incomes (through the poultry and cash grants or other livelihood activities) are also pathways through which SPIR may lead to reductions in maternal and paternal depression, and IPV.

2018-03-01

2021-09-30

Child stunting (height-for-age Z-scores (HAZ)<-2) for children under 24 months; Minimally acceptable diet for children under 24 months; food security as measured by the household food gap in months; depression severity score (PHQ-9); gender norms and female empowerment (attitudes toward female mobility; attitudes toward IPV; input into productive decisions, mobility, group membership, respect, self-efficacy; prevalence of intimate partner violence; gender norms); Assets (Total Index, Productive, Livestock); Access to savings (primary male/female membership in Rusacco/VSLA/MFI/Bank Account); Access to credit total value of productive/consumption loan primary male/female took out); Income from livestock and poultry; Aspirations (Current level of education of the oldest child, level of education primary male/female would like the oldest child to receive, level of education primary male/female would like the oldest child to receive in 10 years)

Other anthropometry measures (HAZ, WHZ, WAZ, wasting, SAM, MAM) for children under 24 months; Infant and young child feeding practices; Infant and young child feeding knowledge; Household dietary diversity score; Minimum dietary diversity-Women; Food Security Index; Food Insecurity Experience Scale; Value of household food consumption per adult equivalent; Child health history - ANC during last pregnancy, IFA during last pregnancy, receiving nutrition information or counseling during last pregnancy, receiving breastfeeding information during last pregnancy, birth in a medical facility (last born); Childcare activities (adult male and female); change in depression scale; Net income from farming; Assets (Consumer, Land Index separate); All other access to savings - primary male/female membership to Rusacco/VSLA/MFI/Bank Account, primary male/female deposited money in a Rusacco/VSLA/MFI/Bank Account with a specific frequency in the past 12 months, primary male/female membership with Eqqub, primary male/female membership with Iddir; All other access to credit - primary male/female took out productive/consumption loan in past 12 months, took out loan from primary source (i.e. Rusacco), took out loan from other secondary sources, primary reason for taking out loan (i.e. to buy livestock), other reasons for taking out loans, total value of productive/consumption loan primary male/female took out, total value of loan the primary male/female still owes, primary reason for difficulty in loan repayment, other reasons for difficulty in loan repayment

All households in the study are PSNP4 beneficiaries. The quantitative impact evaluation design compares the impact of combinations of these L, L*, N and N* activity packages by randomly assigning kebeles to one of the following intervention arms:
• Treatment 1: L* + N*
• Treatment 2: L* + N
• Treatment 3: L + N*
• Control: PSNP only
T1, T2 and T3 are each multifaceted graduation programs delivering different combinations of livelihoods and nutrition interventions. For L* activities, SAA was implemented in all L* kebeles, while the aspiration intervention was implemented in a randomly-selected half of the L* kebeles. Poultry or cash was randomly assigned to L* kebeles and provided to the poorest 10 out of 18 of sample households in those kebeles.

Of the 196 study kebeles that were randomized, 49 were assigned to each of the four treatment arms: T1: L*+N*; T2: L*+N; T3: L+N*; and C: PSNP only. Randomization was stratified at the woreda level to provide balance of treatment assignment geographically and because the woreda is the main administrative structure for local government which shapes local public expenditure and public service delivery. Randomization was conducted by computer at IFPRI offices in Washington DC.

Randomization was done at the kebele (subdistrict) level.

Yes

After the initial randomization of 196 kebeles across the four treatment arms, two kebeles were dropped because they had no PSNP beneficiaries and one was dropped for security reasons. The randomization across the 193 remaining kebeles is shown in Table 3.3.2.
Table 3.3.2: Number of kebeles in each treatment arm, by region
Amhara Oromia Total
T1: L*+N* 27 19 46
T2: L*+N 28 21 49
T3: L+N* 29 20 49
C: L+N 29 20 49
Total 113 80 193

In each of 193 kebele, we randomly sampled 18 households, leading to a planned baseline sample of 3,474 households.

The randomization across the 193 kebeles is shown in Table 3.3.2.
Table 3.3.2: Number of kebeles in each treatment arm, by region
Amhara Oromia Total
T1: L*+N* 27 19 46
T2: L*+N 28 21 49
T3: L+N* 29 20 49
C: L+N 29 20 49
Total 113 80 193

Power calculations were conducted using data from the PSNP4 impact evaluation midline survey to estimate the necessary sample size required to measure a detectable effect of the program on the following three primary outcomes: child height-for-age Z-scores (HAZ) for children under 24 months, mother’s nutrition knowledge, and food security as measured by the household food gap in months. The sample size estimation was based on the kebele-level randomization leading to comparisons of each treatment group with the control group. We conducted the power calculations for ANCOVA models, adjusting for autocorrelation from baseline to endline. For HAZ, we assumed an autocorrelation of 0.7. For nutrition knowledge and a household’s food gap we assumed no autocorrelation. We used conventional levels and set the power at 80 percent and the significance level at 0.05. The calculations suggested that a sample of 196 kebele clusters was needed, with 18 households per cluster, for an intended sample of 3,528 households. In this sample, for HAZ, the minimum detectable effect (MDE) size is 0.34 SD; for mother’s nutrition knowledge it is roughly 1.0 point on a mean score of 7.51; and for the food gap, it is 0.89 months from a mean of 2.35 months. Details about the power calculations are also available in the baseline report.

There is evidence that cash transfers reduce intimate partner violence (IPV), but less is known about the impacts of public works and complementary programmes on IPV. Using mixed-methods we examined whether and how the Ethiopian government's public works programme (that includes cash and/or food for work) alongside complementary activities that engage women and men affected IPV. We analysed midline data collected in July–October 2019 from a randomised controlled trial (RCT) designed to measure the added impacts of the complementary programming in the Amhara and Oromia regions. Eligible households for this analysis had at least one child aged 0–35 months and a primary female caregiver who was married, and under 50 years-old (n = 2604). A nested qualitative study was conducted with a sub-sample of RCT participants from February–March 2020; data included seven focus group discussions and 58 in-depth interviews. Male partners of women who reported experiencing IPV were purposively sampled. Ordinary Least Squares regression analyses were used to estimate the average treatment effect of the complementary programming, and sub-analyses were conducted on the poorest ten households from each village who received additional livelihood transfers. Qualitative data were analysed using thematic content analysis. We found no impacts of the complementary programming on IPV in the full sample, but some impacts among the poorest sample. Evidence on pathways found that both the public works and complementary programming decreased poverty-related stress and arguments within relationships and increased emotional wellbeing. There were some impacts on women's empowerment from the complementary programming. However, men's reactions to women's empowerment were mixed. There was strong evidence that engaging men in nutrition behaviour change communication contributed to improving gender relations. Our findings indicate that social protection and complementary programmes have the potential to be gender transformative and prevent the drivers of IPV.

Ranganathan, Meghna; Pichon, Marjorie; Hidrobo, Melissa; Tambet, Heleene; Sintayehu, Wastina; Tadesse, Seifu; and Buller, Ana Maria. 2022. Government of Ethiopia's public works and complementary programmes: A mixed-methods study on pathways to reduce intimate partner violence. Social Science and Medicine 294(February 2022): 114708. https://doi.org/10.1016/j.socscimed.2022.114708

https://doi.org/10.1016/j.socscimed.2022.114708

Poultry has gained renewed attention as a promising value chain for women because it is an asset that is widely accessible to women, has low start-up costs, and provides a good source of nutritious animal-sourced foods for children in chicken meat and, especially, eggs. The current study presents evidence from an experimental intervention that randomly provided women either a poultry package transfer of vaccinated, improved-breed chickens and related inputs, or a cash grant of equivalent value within a sample of households participating in a social safety net program. These transfers were embedded in a set of intensive livelihood and enhanced nutrition interventions as part of a broader experiment in rural Ethiopia. We assess the impact of the poultry package transfer as well as the enhanced nutrition intervention on the consumption of eggs by both children and adult women. We find that the poultry transfer increased the frequency of egg consumption as well as the sale of eggs, falling between the extreme of an autarkic household and one in which production decisions are fully separable from consumption choices.

Alderman, Harold; Gilligan, Daniel O.; Leight, Jessica; Mulford, Michael; and Tambet, Heleene. 2022. The role of poultry transfers in diet diversity: A cluster randomized intent to treat analysis. Food Policy 107(February 2022): 102212. https://doi.org/10.1016/j.foodpol.2021.102212

https://doi.org/10.1016/j.foodpol.2021.102212