Adaptive Safety Nets Program 2017-2020, Baseline, Midline and Endline Impact Evaluation Surveys
Baseline, Midline and Endline Impact Evaluation Surveys
Other Household Survey [hh/oth]
This dataset includes processed variables from the baseline (2017), midline (2019), and endline (2020) surveys for the impact evaluation of productive inclusion measures of the Niger Adaptive Social Protection Program.
As part of the Adaptive Safety Net Project, the Government of Niger (with support from the World Bank and the Sahel Adaptive Social Protection Program) launched the implementation of productive inclusion measures to foster more productive livelihoods and improve resilience of cash transfer beneficiary households. This dataset covers three rounds of household surveys from the impact evaluation of these productive inclusion measures among cash transfer beneficiary households. It is published along with the related paper: Bossuroy, Thomas; Goldstein, Markus; Karimou, Bassirou; Karlan, Dean; Kazianga, Harounan; Pariente, William; Premand, Patrick; Thomas, Catherine; Udry, Christopher; Vaillant, Julia; Wright, Kelsey. 2022. "Tackling Psychosocial and Capital Constraints Opens Pathways out of Poverty".
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Households as well as individuals within households.
- v2.1: Edited, anonymous dataset for public distribution.
Niger is one of the poorest countries in the world and faces severe development challenges. It is estimated that 44% percent of the population in Niger lives on less than US$1.25 per day, and 75.23% on less than US$2 per day (WDI, 2010). More than 50 percent of Niger’s population is food insecure, with 22 percent of the population suffering from chronic food insecurity in any given year (World Bank, 2011). The Niger safety net project was rolled out in five regions (Dosso, Maradi, Tahoua, Tillabery, and Zinder), within which beneficiary communes and villages were selected.
The study focuses on a sub-sample of communes in all five regions chosen for the second phase of the Niger Adaptive Safety Net project (Dosso, Maradi, Tahoua, Tillabery, and Zinder). 17 communes were selected for the study, covering 322 villages across the 5 regions where cash transfer beneficiaries were eligible to receive complementary productive inclusion measures. In each sample village, approximately 14 households (maximum 15) were interviewed at baseline.
Only households that are beneficiaries of the national cash transfer, located in communes and villages mentioned above
Producers and sponsors
Sahel Adaptive Social Protection Program
Funded survey data collection
Funded survey data collection
Niger Adaptive Safety Nets Project
Provided sampling frame and implemented program
Collected baseline data
Innovations for Poverty Action
Supervised baseline and midline data collection
Cash transfer beneficiary households were chosen by either proxy means testing, community-based targeting, and a formula to proxy temporary food insecurity (as described in Premand and Schnitzer, 2021). 22,507 cash transfer beneficiary households were later assigned to either a control group or 3 productive inclusion treatment arms (Bossuroy et al., 2022). All three treatment arms include a core package of group savings promotion, coaching, and entrepreneurship training, in addition to the regular cash transfers from the national program. The first variant also includes a lump-sum cash grant (“capital” package). The second variant substitutes the cash grant with psychosocial interventions (“psychosocial” package). The third variant includes the cash grant and the psychosocial interventions (“full” package). The control group only receives the regular cash transfers from the national program. 4,712 households were drawn into a sample for data collection (1206 households in control, 1191 households in capital, 1112 households in psychosocial and 1203 households in full). Before the study, we conducted power calculations assuming an ICC of 0.10 (based on data from Ghana and a Niger national household survey) and equal sized arms. To maximize power, we sampled all villages in this phase. Sampling 15 households per village allowed for minimum detectable sizes of 0.057 SD between arms, before adjusting for baseline outcomes or strata.
Deviations from the Sample Design
The original sample included 4712 households. The baseline, endline, and end-line samples include 4608, 4476, and 4303 households, respectively, and thus completion rates of 97.8%, 95.0%, and 91.3%.
Dates of Data Collection
Data Collection Mode
Data collection was supervised by Innovations for Poverty Action, field coordinators from Sahel Consulting, and the co-authors. The supervision team also worked in collaboration with the safety nets unit. Thorough quality control procedures were put in place, with systematic verifications of questionnaires by enumerators and supervisors. Additional verifications, including household visits, were undertaken by the coordination and quality control teams continuously over the full survey period.
Data Collection Notes
Data used in this study was collected using Android tablets and the SurveyCTO Platform developed by Dobility, Inc, versions 2.0 – 2.6.
Household surveys were collected in 3 survey rounds as described above.
The questionnaires included the following sections:
I. Beneficiary section
II. Household head section
Head of household activities
Livestock and Fish
Education and Health spending
Non food consumption
Questions are generally consistent across rounds.
The data includes process variables, see attachment for variable definitions and Bossuroy et al. (2022) for details.
Survey data are labelled, deduplicated and cleaned. It includes constructed variables. The data is documented in three files. A household panel dataset shows data from the baseline, midline, and end-line surveys where observations missing at in the baseline survey are replaced with strata means. Households are observed in two periods. A household-level file shows select variables from the baseline survey. Finally, a food-level file shows median food prices per food unit.
Variables were constructed according to a pre-analysis plan, registered at https://www.socialscienceregistry.org/versions/52534/docs/version/document, and are further described in Bossuroy, et al (2022).
When citing the data, please make sure to also cite the related paper: Bossuroy, Thomas; Goldstein, Markus; Karimou, Bassirou; Karlan, Dean; Kazianga, Harounan; Pariente, William; Premand, Patrick; Thomas, Catherine; Udry, Christopher; Vaillant, Julia; Wright, Kelsey. 2022. " Tackling psychosocial and capital constraints to alleviate poverty " Nature. https://doi.org/10.1038/s41586-022-04647-8
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.