Alternative Transfer Schemes for Transferencias a Primera Infancia (TPI) in Urban Areas

How to cite

García, Arturo; Gray-Lobe, Guthrie; Kremer, Michael; Cooper, Ryan; Vivas, Vanina; Enriquez, Jonathan; Cueva, Selene; Martínez, Maria Adelaida. 2024. "Alternative Transfer Schemes for Transferencias a Primera Infancia (TPI) in Urban Areas." AEA RCT Registry. AEARCTR-0012962. https://doi.org/10.1257/rct.12962-4.0

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

Arturo García* , Guthrie Gray-Lobe , Michael Kremer , Ryan Cooper , Vanina Vivas , Jonathan Enriquez , Selene Cueva , Maria Adelaida Martínez

* Primary investigator

Registry ID
AEARCTR-0012962
DOI
https://doi.org/10.1257/rct.12962-4.0
First registered
Last updated
Status
Completed
Trial dates
2024-05-02 – 2026-01-31
Countries
Peru ()
JEL codes
I10 I12 I38 J13
Registry record
http://www.socialscienceregistry.org/trials/12962

Abstract

This study examines the impact of modifications to the Transferencia a Primera Infancia (TPI) in urban areas. The TPI is a conditional cash transfer (CCT) program available to households that are part of the JUNTOS program (Peru’s flagship CCT program) and have a pregnant woman or a child less than 12 months old. The study will examine the impact of increasing the benefit amount on household welfare, child health, and child cognitive development.

Intervention

JUNTOS is a CCT program in Peru that targets households under the poverty line. The base transfer (TB) amount, unchanged since the program's start in 2005, is 100 soles (33.4 USD in 2005, 26.22 USD in 2024) per month. Starting in 2021, the program expanded to cover all districts, including urban areas. At the same time, JUNTOS introduced the TPI, an additional monthly monetary transfer of 50 soles, specifically for households with pregnant women and newborns.

TPI introduces new conditionalities for households, which, along with those of the base transfer, form a comprehensive health package for pregnant women (including auxiliary exams, iron, and folic acid supplementation) and children up to 12 months old (vaccination schedule, hemoglobin testing, and age-appropriate iron supplementation). The TPI benefit ceases if households fail to meet one of the conditionalities or when the child exceeds 12 months.

The study aims to provide evidence on the marginal impact of an increase in the TPI amount compared to the current scheme. It will evaluate how these schemes affect child-level outcomes like anemia, malnutrition, and cognitive development and household-level outcomes such as food expenditure, consumption, and perceptions of well-being.

Experimental design

There will be two levels of randomization, one at the geographic level (clusters) and the second at the individual level. Cluster boundaries coincide directly with district boundaries. However, some large districts were subdivided into smaller sub-units, resulting in 230 clusters.

1. Cluster-level randomization): Clusters will be divided into two groups
1.1. Treatment (40%, 92 clusters): All households in these clusters will receive 80 soles.
1.2. Control (60%, 138 clusters): Clusters will receive 50 soles. Households can enter a lottery to receive 70 extra soles (see household randomization for details).

2. Household-level randomization: Within each cluster that receives 50 soles (defined as control in the cluster-level randomization), households will be randomized into two groups where treated households get an additional 70 soles and control households the status quo. Households will be divided into two groups.
2.1. Treatment (50% of 60%=30% of sample, 8,400 households): They receive 120 soles monthly.
2.2. Control (50% of 60%=30% of sample, 8,400 households): They receive 50 soles monthly.

Design details

Randomization method
Randomization will be done by a computer.
Randomization unit
Two randomizations will be conducted to vary the transfer amounts. 1. Randomization at the cluster level 2. Randomization at the household level
Clusters
230 clusters. Cluster boundaries coincide mostly with the boundaries of the 212 districts in the study. Large districts were further subdivided, resulting in 230 geographic regions referred to as clusters.
Observations
28,000 households
Arms
At the cluster level, 92 clusters will be assigned to the "high TPI base rate" condition (80 soles) and 138 clusters will be assigned to the status quo TPI base rate (50 soles) At the household level, 8,400 households from the 138 status-quo clusters will receive an additional transfer of 70 additional soles. 8,400 households will continue to receive 50 soles.
Primary outcomes
Household expenditure, household savings, health of children under 1-year-old, children's cognitive development in the long term
Secondary outcomes
Compliance with conditionalities of the JUNTOS program: Including vaccination schedule, hemoglobin testing, and age-appropriate iron supplementation).
Sponsors
Name: UNICEF Peru Location: Lima, Peru Url: https://www.unicef.org/peru/
Partners
Name: Development Innovation Lab, University of Chicago Partner_type: none Url: https://bfi.uchicago.edu/development-innovation-lab/

Cited by

No citation data available for this registration yet.

Keywords