Myopic Consumption

How to cite

Vandewalle, Lore; Somville, Vincent. 2014. "Myopic Consumption." AEA RCT Registry. AEARCTR-0000392. https://doi.org/10.1257/rct.392-6.0

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

Lore Vandewalle* , Vincent Somville

* Primary investigator

Registry ID
AEARCTR-0000392
DOI
https://doi.org/10.1257/rct.392-6.0
First registered
Last updated
Status
Withdrawn
Trial dates
2013-09-01 – 2040-12-31
Countries
India ()
JEL codes
D03 D12 G02 O12
Registry record
http://www.socialscienceregistry.org/trials/392

Abstract

Recent theoretical and empirical work on consumers’ behavior emphasizes the importance of self-control problems. While different theories can explain self-control problems, an important part of the literature focuses on “βδ” models of time-inconsistent preferences. In those models, individuals are assumed to be either (i) sophisticated, aware of their preferences, and taking commitment measures to protect their present-self from the behavior of their future-selves; or (ii) partially or fully naïve and ignoring the inconsistency in their preferences when deciding on their present consumption.

Both theoretical and empirical studies alike have so far focused on the “forward looking” part of the consumers: how do they choose current and future consumption, given how much knowledge they have about their own preferences and the available commitment technologies. We on the other hand want to focus on another assumption that we believe is fundamental, but yet not researched. Naïve consumers are typically assumed to ignore not only their true future behavior, but also all their past decisions. The naïve consumer that makes time-inconsistent choices is assumed unable to look back at his past choices and infer from them the true shape of his preferences.

Experimental design

We provide information about past expenditures to households. There are control households and two treatments: some households receive a short overview of their past expenditures, others a more detailed overview.

Design details

Randomization method
The randomizations are done with the software Stata. We followed David McKenzie and Miriam Bruhn’s recommendations in dealing with the uneven numbers in some strata and we followed the Stata code that they shared on the World Bank’s “Development Impact” blog on the 11th of June 2011.
Randomization unit
The unit of observation is the head, or spouse of the head of the household.
Clusters
18 villages
Observations
576 households (32 per village)
Arms
190 control households, 194 households received the short budget sheet, and 192 households received the long budget sheet.
Primary outcomes
All our outcome variables are based on the expenditures information from our surveys. We have this information on a weekly basis for the duration of the experiment. We use both weekly values and averages over the whole course of the experiment. The hypothesis that we want to test is whether the treatment changes the weekly expenditures, and its composition. The outcome variables are related to the first four categories described in the intervention: (i) tobacco and alcohol, (ii) unnecessary cosmetics, (iii) beverages and snacks outside the household, and (iv) necessary goods. We look at: - (Y1) The expenditures made on each of the four categories separately (4 outcome variables) - (Y2) The total expenditures over all the goods in the four categories (1 outcome variable) - (Y3) The share of each of the four categories in the total over the four categories (4 outcome variables) - (Y4) Whether or not the household consumed goods from the first three categories (3 outcome variables) - (Y5) The quantities bought within each of the first three categories.

Cited by

No citation data available for this registration yet.

Keywords