Impacts of Childcare Vouchers on Women’s Labor Supply, Child Development, and Sibling Schooling in Addis Ababa

How to cite

Bedasso, Biniam; Weldesenbet, Ermias. 2026. "Impacts of Childcare Vouchers on Women’s Labor Supply, Child Development, and Sibling Schooling in Addis Ababa." AEA RCT Registry. AEARCTR-0017932. https://doi.org/10.1257/rct.17932-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Biniam Bedasso* , Ermias Weldesenbet

* Primary investigator

Registry ID
AEARCTR-0017932
DOI
https://doi.org/10.1257/rct.17932-1.0
First registered
Last updated
Status
On going
Trial dates
2025-10-01 – 2026-06-30
Countries
Ethiopia (Addis Ababa)
Registry record
http://www.socialscienceregistry.org/trials/17932

Abstract

This study evaluates whether subsidized access to formal childcare increases women’s economic participation and improves child and household outcomes among low-income families in Addis Ababa, Ethiopia. The intervention offers eligible women a digital childcare voucher that covers 100% of fees to enroll a child aged 6–48 months in a pre-screened network of affordable childcare providers with spare capacity; voucher payments are administered through a digital platform operated by a major bank. The evaluation is a randomized controlled trial in nine sub-cities, targeting low-income women aged 18–49 who (i) have at least one child aged 6–48 months not currently attending external childcare, (ii) live within walking distance of a participating provider, and (iii) are willing to place a child in formal childcare. Approximately 700 women are expected to be enrolled; provider capacity (42 participating providers; up to 788 childcare slots) creates a supply-constrained setting in which not all eligible women can be served. For each provider catchment area, eligible rosters include roughly twice as many candidates as available slots, and women are randomly assigned after baseline surveys to treatment (voucher offer) or control (no offer). Randomization is stratified by childcare-provider catchment area and refugee vs host-community status. The primary estimand is the intention-to-treat effect of voucher offers; treatment-on-the-treated effects will be estimated secondarily by instrumenting actual childcare enrollment with assignment. Primary outcomes are women’s employment status, hours worked, and earnings; maternal psychological wellbeing (WEMWBS); early child development (CREDI scores for voucher-eligible children); and older siblings’ schooling outcomes (attendance and EGRA scores for selected grades). Attrition will be reported by treatment status and, if differential, Lee bounds will be computed for primary outcomes; standard errors will be clustered at the childcare-provider catchment level.

Intervention

Eligible low-income women in nine sub-cities of Addis Ababa are offered a digital childcare voucher that subsidizes the cost of enrolling one eligible child aged 6–48 months into a pre-screened network of participating formal childcare centers that are affordable, have spare capacity, and agree to accept vouchers. Voucher payments to providers are administered through a digital platform operated by Awash International Bank. Childcare supply is limited; the participating network includes 42 providers with capacity for up to 788 children in total.

The voucher is intended to reduce the effective price and uncertainty of formal childcare, increase childcare enrollment/attendance, and thereby relax time constraints on mothers, potentially improving women’s labor market outcomes, maternal wellbeing, child development, and older siblings’ schooling outcomes.

Experimental design

Randomized controlled trial among eligible low-income women in nine sub-cities of Addis Ababa. After baseline data collection and assignment of women to childcare-provider catchment areas based on provider mapping, surveyed women are randomly assigned to: Treatment (offered a childcare voucher) or Control (not offered a voucher). Randomization is stratified by childcare-provider catchment area and by refugee vs host-community status. Primary analysis estimates intention-to-treat impacts with strata fixed effects and standard errors clustered at the catchment level.

Design details

Randomization method
Randomization is conducted in office using a computer-based random assignment, after completion of the baseline survey and definition of childcare-provider catchment areas; assignment is stratified by catchment area and refugee/host status.
Randomization unit
Individual mother/household, with stratification by childcare-provider catchment area (geographic clusters aligned with provider service areas) and refugee vs host-community status.
Clusters
42 childcare providers / childcare-provider catchment areas (clusters for stratification and inference).
Observations
~1,300 eligible women (households) targeted for baseline and randomization.
Arms
Approximately half of eligible women within each catchment area are assigned to voucher offer (treatment) and half to control (because eligible lists are approximately twice available childcare slots). Target overall allocation with N≈1,300 is approximately ~650 treatment, ~650 control (exact numbers vary by catchment-area capacity and achieved sample).
Primary outcomes
1. Women’s labor market outcomes: employment status; hours worked; earnings. 2. Maternal psychological wellbeing: Warwick–Edinburgh Mental Wellbeing Scale (WEMWBS) total score. 3. Young child development (voucher-eligible child): Caregiver Reported Early Development Instruments (CREDI) total score and CREDI z-score. 4. Older siblings’ education: school attendance (self-reported) and Early Grade Reading Assessment (EGRA) scores (2nd and 3rd graders).
Secondary outcomes
- Intra-household relations (secondary family/wellbeing outcome). - Childcare take-up and intensity: enrollment and attendance in formal childcare (mechanism/first-stage). - Heterogeneity-relevant outcomes / subgroups (planned): differential effects by refugee vs host-community status, baseline employment, baseline childcare situation, marital status, education, and eligible child age group (6–24 vs 25–48 months).
Sponsors
Name: IZA - Institute of Labor Economics Location: Bonn, Germany Url: https://www.iza.org/

Cited by

No citation data available for this registration yet.

Keywords