Poverty and Conformity

How to cite

Roth, Christopher; Haushofer, Johannes; Grigorieff, Alexis. 2015. "Poverty and Conformity." AEA RCT Registry. AEARCTR-0000809. https://doi.org/10.1257/rct.809-1.0

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

Christopher Roth* , Johannes Haushofer , Alexis Grigorieff

* Primary investigator

Registry ID
AEARCTR-0000809
DOI
https://doi.org/10.1257/rct.809-1.0
First registered
Last updated
Status
In development
Trial dates
2015-08-12 – 2015-08-13
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/809

Abstract

This document describes the analysis plan for a randomized experiment examining whether people who receive a poverty prime are more likely to imitate others in risky choices and inter-temporal choices. We will recruit 1012 respondents from Amazon Mechanical Turk. Half of the sample will receive a prime which triggers financial worries (poverty prime), while the other half will not (placebo). Then, half of the participants who received the poverty prime and half of those who received the placebo will get some information about how previous MTurk participants answered the questions on risk preferences and time discounting, while the other participants will not receive any information about previous participants' behavior. In our experiment we can therefore identify (i) how the poverty prime affects risk preferences and time discounting, (ii) how information about others' behavior affects the respondents' behavior and (iii) whether the poverty prime makes people more likely to imitate others in risky choices and inter-temporal choices. This plan outlines the design of the experiment, the outcomes of interest, the econometric approach and the dimensions of heterogeneity we intend to explore.

Experimental design

Poverty Prime:


We have adapted the poverty prime developed by Mani et al. (2013) to the MTurk environment. Participants are randomly assigned to either get the poverty prime or to get the placebo.

Participants who receive the poverty prime (placebo) need to explain how they would deal with an income decrease of 20% (5%). We then ask them a variety of questions on whether this income shock would substantially affect their situation and what kind of sacrifices they would need to make. Then, respondents are asked how they would come up with $3000 ($150) on short notice. The order with which these hypothetical scenarios is presented is randomized. Respondents write down how they might deal with the different financial scenarios. The aim of the hard financial scenarios is to trigger feelings of poverty.

We have made two main changes to the primes used by Mani et al. (2013): first, we increased the amounts used in the poverty prime. Second, we removed two financial scenarios because they did not seem well-suited for the MTurk population. We have conducted a pilot study with a sample of 210 participants on August 1st in which we document that our poverty prime makes people more worried with their financial situation. In particular, poorer individuals from our sample are quite strongly affected by our treatment: they are significantly more worried about their financial situation and they are less satisfied with their income level. The primes are further described in Appendix A. Moreover, at the very end of the document, we attach the exact experimental instructions.

Peer Information:

We ran a survey on Amazon Mechanical Turk on July 30th in order to get incentivized preference data that we could use as peer group information in the experiment outlined in this pre-analysis plan. We recruited 223 participants out of which 210 participants completed our survey. We asked respondents various questions on their level of risk aversion and time discounting. In particular, we find that 50.25 percent of the sample chose $100 now and 49.75 percent chose $110 in one week. 55 percent of the participants chose $100 for sure, and 45 percent chose the lottery between $70 and $160. We chose those time and risk preference measures for which the preferences over the two different options were most evenly distributed.

Half of our respondents will be randomly assigned to receive information about which option the majority of the respondents in the survey chose. In particular, they will receive the following message on the page before the behavioral measures:

On the 30th July 2015, we conducted a survey with more than 200 American MTurkers to get some information about their preferences. Like you, these MTurkers come from the US, have completed more than 500 HITs, and have an approval rating of more than 95\%. In the survey, we asked them the same two questions as the ones you will answer now. For each question, we will let you know which option most MTurkers chose. We advertised the survey on the Reddit page "HitsWorthTurkingFor" on July 30th. You can check that the survey actually took place by clicking on the following link: https://www.reddit.com/r/HITsWorthTurkingFor/comments/3f5typ/us_5_minute_experiment_on_economic_preferences.

Subsequently on the page of the time-preference task, the treatment group will be told: Most of the MTurkers who took our previous survey preferred getting $100 today.

Then on the page of the risk-preference task, the treatment group will be told: Most of the MTurkers who took our previous survey preferred getting $100 for sure.

Design details

Randomization method
Randomization is done using a computer.
Randomization unit
Individual.
Clusters
1012 individuals.
Observations
1012 individuals.
Arms
256 receive the easy financial scenario and get no information about the behavior of other MTurkers. 256 receive the hard financial scenario and get no information about the behavior of other MTurkers. 256 receive the easy financial scenario and also get information about the behavior of other MTurkers. 256 receive the hard financial scenario and also get information about the behavior of other MTurkers.
Primary outcomes
Risk Preferences: Choose between $100 for sure and a lottery: receiving $70 with 50% and $160 with 50%. Time Preferences: Choosing between $100 on the day of the experiment and $110 one week later. Manipulation Checks - Financial Worries: This 4-item questionnaire provides an addition manipulation check for our poverty primes. We ask respondents to self-report on a Likert scale how worried they are about their financial situation. Satisfaction with income. Mechanisms Confidence. Self-esteem. Individualism/collectivism. Sense of control. Social comparisons and relative concerns. Social desirability. Cognitive function: Number of correct answers and response time to the Raven Progressive Matrices.

Cited by

No citation data available for this registration yet.

Keywords