When Less is More: Experimental Evidence on Information Delivery during India's Demonetization

How to cite

Chandrasekhar, Arun; Golub, Ben; Banerjee, Abhijit; Breza, Emily. 2018. "When Less is More: Experimental Evidence on Information Delivery during India's Demonetization." AEA RCT Registry. AEARCTR-0002881. https://doi.org/10.1257/rct.2881-1.0

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

Arun Chandrasekhar* , Ben Golub , Abhijit Banerjee , Emily Breza

* Primary investigator

Registry ID
AEARCTR-0002881
DOI
https://doi.org/10.1257/rct.2881-1.0
First registered
Last updated
Status
On going
Trial dates
2016-12-01 – 2019-01-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/2881

Abstract

How should policymakers disseminate information: by broadcasting or letting word spread from a small number of “seeds”? While conventional wisdom suggests broadcasting is better, we show theoretically and experimentally that this may not hold when people need to ask questions to fully comprehend the information. During the 2016 Indian demonetization, we varied how official policy information was delivered on two dimensions: how many were initially informed and whether it was disclosed publicly who was initially informed. Our results are consistent with the model: when who is informed is commonly known, telling more individuals leads to fewer conversations and inferior knowledge.

Intervention

We conducted a core 2x2 design crossed with a third arm.

The 2x2 design varied:
1) whether information about demonetization was broadcast to all households or simply 5 households (seeds)
2) whether there was or not common knowledge of this

This was crossed with whether the information pamphlet had 2 facts or 24 facts on it (short or long).

The randomization was done at the village level.

Experimental design

We conducted a core 2x2 design crossed with a third arm.

The 2x2 design varied:
1) whether information about demonetization was broadcast to all households or simply 5 households (seeds)
2) whether there was or not common knowledge of this

This was crossed with whether the information pamphlet had 2 facts or 24 facts on it (short or long).

The randomization was done at the village level.

Design details

Randomization method
Computer
Randomization unit
Village
Clusters
276 villages; 225 for the 8 cells above
Observations
Same as above
Arms
1/8th per cell
Primary outcomes
Number of conversations; Knowledge of demonetization rules; Choice of keeping an old Rs. 500 note vs taking Rs. 200 (delayed)

Cited by

No citation data available for this registration yet.

Keywords