Can Smart Technology Systems Improve Direct Benefit Transfer Performance and Increase Participation? Evidence from MGNREGA in India

Last registered on December 03, 2024

Pre-Trial

General Information

Title

RCT ID

AEARCTR-0001292

Initial registration date

November 07, 2016

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

November 07, 2016, 3:43 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

December 03, 2024, 1:29 PM EST

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

United States of America

Region

Connecticut

Primary Investigator

Name

Rohini Pande

Affiliation

Yale University

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Eric Dodge

PI Affiliation

Harvard University

Contact Investigator

PI Name

Charity Moore

PI Affiliation

Yale University

Contact Investigator

PI Name

Yusuf Neggers

PI Affiliation

University of Michigan

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2016-11-15

End date

2020-01-31

Keywords

Finance & Microfinance, Governance

Additional Keywords

Technology for development, administrative capacity building, social programs

JEL code(s)

O39

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

The implementation of social protection programs remains a challenge in developing countries, often to the particular detriment of the most vulnerable intended beneficiaries. We will investigate the potential of a new internet- and mobile-based management and monitoring platform, developed in direct collaboration with the Indian Ministry of Rural Development, to improve the administration of a large government welfare scheme. A randomized control trial across multiple states will be conducted in which we provide the platform to different levels of the bureaucratic hierarchy responsible for program administration. The study will determine the extent to which heightening officials' performance incentives versus lowering their costs of information acquisition is effective, and whether complementarities between the two exist, in reducing payment delays and subsequently improving program uptake. We will additionally examine how impacts are mediated by bureaucrats’ own personality traits.

External Link(s)

Registration Citation

Citation

Dodge, Eric et al. 2024. "Can Smart Technology Systems Improve Direct Benefit Transfer Performance and Increase Participation? Evidence from MGNREGA in India." AEA RCT Registry. December 03. https://doi.org/10.1257/rct.1292-3.3

Former Citation

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

This randomized control trial (RCT), which takes place at scale in two large Indian states, investigates how wage payment delays to workers in the government's large workfare program are affected by decreasing costs of acquiring management-relevant information by lower-level bureaucrats and reducing the costs of monitoring by higher-level bureaucrats.

Intervention (Hidden)

Intervention Start Date

2016-11-15

Intervention End Date

2019-12-31

Primary Outcomes

Primary Outcomes (end points)

Our main outcomes of interest are the mean and absolute average deviation in the time to payment for payment sub-steps under the purview of block and district officials, as captured through administrative data used in our system.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

We also will examine MGNREGA-related outcomes available from the MGNREGA MIS, including work requested, number of person-days worked, and total wage and materials expenditure. Contingent on funding availability, we will conduct endline surveys and qualitative interviews with officials to understand changes in their management practices, time allocation, and professional networks, and to obtain additional platform design feedback useful to scale-up.

Secondary Outcomes (explanation)

Experimental Design

We randomized access to PayDash at the district level to the following treatment arms:
1.) Control: Both block and district-level officials overseeing the wage payment process must rely on status quo information sources about payment processing. These sources are readily available but time consuming to compile and examine on a regular basis.
2.) District only PayDash access: Only the highest level of supervisors across the districts receive access to the PayDash platform, ensuring they are better able to monitor payment activity and delays managed by lower-level block officials.
3.) Block only PayDash access: Only the lower-level managers (block officials) receive access to PayDash. Information in the platform helps these officials identify and process payments in their jurisdiction.
4.) District + Block PayDash: Both district and block-level MGNREGA officials have individualized access to the platform.

Experimental Design Details

Randomization Method

The randomization of districts was completed in an office by a computer using STATA.

Randomization Unit

A randomized control trial across multiple states will be conducted in which we provide the platform to different levels of the bureaucratic hierarchy responsible for program administration.

Across the three states, we will randomly assign two cross-cutting treatments at the district level: provision of PayDash to (1) district officials and (2) block officials.

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

73 districts. The randomization strategy includes providing access to the platform to all lower block-level officials in randomly selected districts. In addition to the control arm, our study includes districts where only district-level officials have access to the platform (TD), districts where only lower block-level officials have access (TB), and districts where both district and block-level officials have access (TDB).

Sample size: planned number of observations

73 districts, over 12+ months

Sample size (or number of clusters) by treatment arms

In Madhya Pradesh, 50 districts were split equally across four treatment arms. In the second state, Jharkhand, the four treatment categories were randomly assigned across 23 districts in approximate proportions of Control: 1/3; TD: 1/6; TB: 1/6, TDB: 1/3.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

In Madhya Pradesh, the four treatment categories were randomly assigned across 50 districts, excluding a pilot district, in approximately equal proportions, while across 23 districts in Jharkhand we assigned approximately twice as many districts to the District+Block PayDash and control arms than to the District and Block PayDash arms. This sample size powers us to detect a minimum effect size of 0.187 standard deviations (SD) for comparisons of control to Block+District PayDash (translating into delay decreases of 1.97 days or more in the steps of the payment process under officer purview), of 0.205 SD for comparisons of either control or Block+District PayDash to either Block or District PayDash, and of 0.217 SD for comparisons of Block and District PayDash.

Supporting Documents and Materials

IRB