Evaluating impact of informational framing of childcare benefits on prioritizing childcare as a community investment

The intervention consists of an online informational module embedded in a questionnaire administered to residents of randomly assigned villages, with villages serving as the unit of randomization across three arms. The control group receives a standard questionnaire covering demographic and labor questions alongside a question about their views on kindergarten attendance. The two treatment arms receive the same baseline questionnaire augmented with a short informational text: Treatment 1 emphasizes the child development benefits of childcare attendance — including school readiness, cognitive development, and health monitoring — citing statistics from UNICEF and the Kyrgyz Ministry of Education; Treatment 2 emphasizes the maternal employment and household welfare benefits of affordable childcare, citing World Bank research on the effects of childcare availability on maternal labor force participation and household income. Both treatments are followed by an additional question gauging whether the information shifted respondents' views. The informational texts were matched in length and tone across treatment arms, translated into Kyrgyz by a social scientist with extensive regional fieldwork experience, and piloted with field staff recruited from the intervention areas to ensure clarity and ease of completion.

2025-12-08

2025-12-12

There are three outcome variables which capture prioritization of childcare vis-à-vis alternative investments at different levels: focus-groups across villages, village, and individual. These outcomes map onto the sequential stages of the CDD process as follows: first, several focus groups (such as women, vulnerable, youth, farmers, elderly) meet to discuss priorities. Second, all potential investments identified as relevant for the village by different groups are included in the list from which individuals can choose. Third, individuals vote through an online platform and select priority project.
To capture this process, we define 3 dummy variables:
• ChildcareFG_vg is equal to 1 if a focus group g in village v included childcare in its list of priorities.
• ChildcareSelected_v is equal to 1 if childcare was selected by majority vote in village v to receive investment through the CDD project, from the consolidated list of priorities.
• ChildcareVote_i is equal to 1 if an individual i cast her/his vote for childcare.

339 villages, which participate in the NCIP CDD project, are randomly assigned to one of the three study arms, with 113 villages in each arm. All individuals within a given village, who registered to be a part of village project WhatsApp group, receive the same version of one of the three questionnaires (one control and two information treatments) corresponding to the village’s assignment. 60,000 registered and 37,897 completed the questionnaires.
Randomization at the village level minimizes information spillovers between treatment arms that could arise if households within the same village were exposed to different messages. Balance across treatment arms will be assessed using village level characteristics, available from administrative data on village population, number of children aged 0 to 5, and availability of childcare. No re randomization or re weighting will be conducted.

Not available

Randomization done in office by a computer in Excel.

339 villages

Yes

339

186,838 individuals

Each treatment arm includes 113 villages. The number of voters per village has not been predetermined and is beyond experimenter's direct control.