Investigating the effect of household knowledge of welfare programs eligibility criteria to their investment behaviors

Last registered on June 24, 2019

Pre-Trial

Trial Information

General Information

Title
Investigating the effect of household knowledge of welfare programs eligibility criteria to their investment behaviors
RCT ID
AEARCTR-0001352
Initial registration date
September 10, 2017

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
September 11, 2017, 11:54 AM EDT

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

Last updated
June 24, 2019, 4:30 AM EDT

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

Locations

Region

Primary Investigator

Affiliation
MIT

Other Primary Investigator(s)

PI Affiliation
TNP2K (Indonesia's National Team for the Acceleration of Poverty Reduction)
PI Affiliation
Harvard University
PI Affiliation
MIT

Additional Trial Information

Status
On going
Start date
2015-05-01
End date
2020-12-31
Secondary IDs
Abstract
In November 2014, the Government of Indonesia launched an expansion of existing social protection programs using a new form of beneficiary identification cards: Smart Indonesia Card (KIP), Healthy Indonesia Card (KIS), and Family Welfare Card (KKS). Upon launching these programs, the government recognizes the need to verify and update the existing database from which the social programs' beneficiary list is based. Like many countries, the Statistic Indonesia will update the beneficiary database using a household asset survey--a proxy means test.

One worry is that these forms of asset surveys have distortionary effects--that households do not invest in assets that they are asked about on the proxy means test in order to maintain their eligibility status. Alternatively, households may become worried about how the government will use their data, and these surveys may lead to worse quality on other government datasets. The government of Indonesia seeks to understand if this is indeed the case. Thus, we are working with them to build in an evaluation into their national proxy means test survey (the PBDT) to test this idea.

Using publically-available, de-identified survey data collected by the government (SUSENAS--available in the Harvard library), this research aims to evaluate the effect of receiving the different PBDT surveys on subsequent household asset acquisition.
External Link(s)

Registration Citation

Citation
Banerjee, Abhijit et al. 2019. "Investigating the effect of household knowledge of welfare programs eligibility criteria to their investment behaviors." AEA RCT Registry. June 24. https://doi.org/10.1257/rct.1352-3.0
Former Citation
Banerjee, Abhijit et al. 2019. "Investigating the effect of household knowledge of welfare programs eligibility criteria to their investment behaviors." AEA RCT Registry. June 24. https://www.socialscienceregistry.org/trials/1352/history/48595
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)
The government of Indonesia will randomize different survey forms for the updating of government Unified Database (PBDT) of social assistance beneficiary list in the 34 provinces. In addition, random assignment of questions will produce two equal groups. Control group who were not assigned to receive the question can serve to be the comparison group for those who were randomized to receive the question.
Intervention (Hidden)
We will administer two additional questions in the PMT survey to the household. Each surveyed household will receive one question from each question group below:
1. Questions that may possibly change household’s behavior.
- Total active cell phone number owned by whole household members.
- Total number of flat screen television with 30” screen or larger owned by household members.
2. Questions where unlikely to change behavior, but household may misrepresent.
- Total number of swan-neck closet installed in the household’s dwelling
- Total number of room in the household’s dwelling

Using the above question types, there will be four combinations of questions:
1. Cell phone and TV
2. Cell phone and room
3. Closet and TV
4. Closet and room
The four versions of the questionnaire will allow us to investigate the effect of adding new questions to their subsequent investment behavior.
Intervention Start Date
2015-05-01
Intervention End Date
2015-09-30

Primary Outcomes

Primary Outcomes (end points)
Households' investment behavior as measured by ownership rate for assets being subsequently acquired.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
The research will use a national government-run welfare targeting survey that asks about household assets possession. The government of Indonesia will do province-level randomization of different survey forms for the PBDT in the 33 provinces to evaluate its effect on subsequent household asset acquisition.

Experimental Design Details
From the current utilization of BDT in social programs, there have been reports of exclusion errors and inclusion errors. To address these errors, the selection of asset types that is being used as proxy for welfare eligibility becomes crucial. Furthermore, there is also a need to investigate the effect of surveying the household of their asset ownership on the accuracy of the answer that they provide. Randomized assignment of additional asset questions to households will allow us to investigate the effect of surveying for the following reasons:

- Random assignment of questions will produce two equal groups. Control group who were not assigned to receive the question can serve to be the comparison group for those who were randomized to receive the question. As the only difference between the two groups are whether they were being asked the additional question or not, any difference in the observation of asset ownership rates in PBDT are purely due to the intervention of asking the asset question. For this reason, there is a need to have a control group that do not receive the additional asset question to be the basis of comparison.

- Randomizing by province will ensure that the effect of the questionnaire is not driven by specific characteristics of any given province, Without randomization, teasing out the effect of asking additional question may be harder, as it is possible that there are other reasons that drive any observed differences. Randomization by province also allow us to maximize our statistical power.

There are 4 treatment arms in total:
- Respondents in 8 provinces will receive questions regarding cell phone and television.
- Respondents in 9 provinces will receive questions regarding cell phone and room.
- Respondents in 8 provinces will receive questions regarding swan-neck closet and television.
- Respondents in 8 provinces will receive questions regarding swan-neck closet and room.
Randomization Method
Randomization done by a computer using STATA
Randomization Unit
Province
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
33 provinces in Indonesia
Sample size: planned number of observations
Adults around 26 millions.
Sample size (or number of clusters) by treatment arms
- 8 provinces receiving cell phone and television questions (P0Q0)
- 9 provinces receiving cell phone and room questions (P0Q1)
- 8 provinces receiving cell swan-neck closet and television questions (P1Q0)
- 8 provinces receiving cell swan-neck closet and room questions (P1Q0)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
MIT Committee On the Use of Humans as Experimental Subjects
IRB Approval Date
2015-04-06
IRB Approval Number
1504007043

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials