The Impacts of Performance Feedback for Workers of the Integrated Child Development Services Program in India

How to cite

Ravindran, Saravana. 2018. "The Impacts of Performance Feedback for Workers of the Integrated Child Development Services Program in India." AEA RCT Registry. AEARCTR-0002660. https://doi.org/10.1257/rct.2660-1.0

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

Saravana Ravindran*

* Primary investigator

Registry ID
AEARCTR-0002660
DOI
https://doi.org/10.1257/rct.2660-1.0
First registered
Last updated
Status
In development
Trial dates
2018-03-01 – 2021-03-01
Countries
Private
JEL codes
D90, I15, I38
Registry record
http://www.socialscienceregistry.org/trials/2660

Abstract

To help address India's high rates of child malnutrition, the Indian government launched the Integrated Child Development Services (ICDS) program in 1975, and has substantially expanded the program in the last decade. The program provides pre-school education and primary healthcare services to children under 6 years of age. Today, the program is the largest of its kind in the world, reaching about 25% of all children in India under the age of 6 (39.7 million children) by 2010. However, take-up of ICDS services today remains low. In this project I ask, what is the impact of providing feedback on workers' own performance, as measured by utilization of services and malnutrition of children, in addition to feedback on performance of workers they regularly interact with? The proposed experiment is low-cost, can easily be scaled up, and will also help to shed some light on the cases in which performance feedback and social comparisons might be most effective.

Intervention

The intervention consists of mailing one of three types of performance feedback letters to ICDS workers, sent officially by the Department of Women and Child Development, Odisha. In the group with (only) individual feedback, workers will receive a letter every month for the duration of the study informing them of their performance on malnutrition of the children under their center. This malnutrition feedback will be broken down into age groups, and by gender, as well as categories of moderate and severe malnutrition. In the group with (only) feedback on other workers, workers will receive a letter every month informing them of the performance of other ICDS workers in their ICDS project. Lastly, in the group that receives both individual feedback and feedback on other workers, workers will receive a letter that presents both sets of information - i.e. indicators of their performance, as well as indicators of the performance of their peers on malnutrition.

Experimental design

Selection of Districts

The state of Odisha is divided into 3 revenue divisions. The 3 study districts were drawn from each of these revenue divisions so as to form a representative sample - Dhenkanal from the Northern Division, Balasore from the Central Division, and Ganjam from the Southern Division.

Experimental Design

The experimental design cross-randomizes two types of feedback treatments. In the pure control group, business is as usual. In the group with (only) individual feedback, workers will receive a letter every month for the duration of the study informing them of their performance on malnutrition of the children under their center. In the group with (only) feedback on other workers, workers will receive a letter every month informing them of the performance of other ICDS workers in their ICDS project. Lastly, in the group that receives both individual feedback and feedback on other workers, workers will receive a letter that presents both sets of information - i.e. indicators of their performance, as well as indicators of the performance of their peers on malnutrition. These letters will be sent officially by the Department of Women and Child Development, Odisha.

Design details

Randomization method
Randomization to be done in office by a computer.
Randomization unit
Individual level randomization at the level of the ICDS worker (also the level of the ICDS center, since each center is typically staffed by one worker).
Clusters
Approximately 6,000 workers.
Observations
Approximately 6,000 workers.
Arms
Approximately 1,500 workers in the control group, 1,500 workers in the group with only individual feedback, 1,500 workers in the group with only feedback on other health workers, 1,500 workers in the group with both individual feedback and feedback on other workers.
Primary outcomes
First, the rates of utilization and malnutrition in the ICDS centers will be key outcome variables in this experiment. These will be collected separately for children of different ages groups and gender. Workers' knowledge about malnutrition and the effort they exert at work will also be collected. The second set of key outcome variables are derived from a time use survey. This details the amount of time spent on various tasks during the work week. Third, interaction with parents, other workers, and supervisors, as measured by frequency and duration of conversations will be captured in surveys. Fourth, discussions relating to the ICDS at local panchayat meetings will be key outcomes of interest. Fifth, beliefs that workers have about the ICDS program in other centers will be collected. Sixth, workers' tasks during the work week, such as height and weight monitoring will be captured, along with operational information such as opening and closing times of the centers. Lastly, locus of control, competitive preferences, time preferences, risk preferences, and intrinsic motivation will be key outcome variables in this experiment.

Cited by

No citation data available for this registration yet.

Keywords