Uncertainty and Attitude to Immigration

Last registered on March 28, 2022

View Trial History

Pre-Trial

Trial Information

General Information

Title

Uncertainty and Attitude to Immigration

RCT ID

AEARCTR-0009154

Initial registration date

March 28, 2022

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

March 28, 2022, 7:18 AM EDT

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

Locations

Country

United Kingdom of Great Britain and Northern Ireland

Region

Primary Investigator

Name

Friederike Mengel

Affiliation

Uni Essex

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Jayant Ganguli

PI Affiliation

PI Name

Adeline Delavande

PI Affiliation

Additional Trial Information

Status

In development

Start date

2022-03-28

End date

2022-09-26

Keywords

Behavior, Lab, Labor

Additional Keywords

Sampling, Contact Theory, Uncertainty, Risk Aversion, Attitude to Immigration

JEL code(s)

C92

Secondary IDs

Prior work

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

Abstract

There is huge interest in the Social Sciences to understand what determines people’s attitudes towards immigrants. One question that has received a lot of popular and scholarly attention is what can be done to address extremely negative sentiment as well as overly negative beliefs. A popular theory to understand these questions has been contact theory, which claims that the more people interact with others of different social identities the less negative or biased their attitudes will become. A huge literature in the Social Sciences and Psychology testing contact theory empirically focuses on shifts in mean attitudes. In this paper we propose an alternative mechanism. Contact reduces uncertainty by increasing the sample based on which people make inference about relevant characteristics of social groups. The experiment pre-registered here is designed to answer two key questions in this context: Do people pay attention to uncertainty when choosing who to interact with? Do people perceive more uncertainty when interacting with foreign nationals?

External Link(s)

Registration Citation

Citation

Delavande, Adeline, Jayant Ganguli and Friederike Mengel. 2022. "Uncertainty and Attitude to Immigration." AEA RCT Registry. March 28. https://doi.org/10.1257/rct.9154-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention Start Date

2022-03-28

Intervention End Date

2022-09-26

Primary Outcomes

Primary Outcomes (end points)

(i) Willingness to pay (wtp) to be matched with a certain group of participants with full information.
(ii) Willingness to pay (wtp) to be matched with a certain group of participants with partial information.
(iii) Beliefs about performance distribution.

Primary Outcomes (explanation)

We are interested in two main questions: (i) How much does uncertainty about ability matter for peoples' partner choice? And (ii) Do people perceive more uncertainty for groups who consist of more foreigners? Question (i) is addressed by eliciting participants’ wtp in several situations. Question (ii) can be addressed indirectly by eliciting the wtp or directly by eliciting beliefs. Hence our two primary outcomes.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Our online experiment elicits three main outcomes described below.

Experimental Design Details

Outcome 1 is participants’ wtp to be matched with someone from a particular group. To elicit this wtp participants are shown two groups of 10 people each. They are shown a nationality indicator for each participant (UK flag for UK nationals, continent symbol for others) as well as the performance distribution in this group. (Performances are from a math and logic test conducted in a different experiment and range between 0 and 20 (integers)). Groups can differ both in the distribution of performances and in the distribution of nationalities.
They are assigned a default group and asked how much they would be willing to pay to be matched with someone from the other group. They indicate the wtp by moving a slider. Participants make 30 choices in the experiment in random order. They are paid based on their performance of person they end up matched with and we use the BDM mechanism to incentivize the wtp.
Outcome 2 is also a wtp elicited in the same way as Outcome 1 with the difference that now only partial information about the distribution of performances is given.
Outcome 3 is a belief about the distribution of performances. To elicit this outcome we show participants again two groups with partial information and then elicit their beliefs about the full distribution. Guesses are incentivized using a binarized scoring rule.
The experiment also collects information on demographics, questions on experience with immigrants and attitudes towards immigration and risk attitude.

Randomization Method

random assignment done by a computer

Randomization Unit

individual

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

1290

Sample size: planned number of observations

27950

Sample size (or number of clusters) by treatment arms

430 individuals per condition

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

-0.1706 Power Analysis: This is the coefficient on “Delta SD” in an OLS regression where we regress WTP on - Delta Mean (the difference in means between the other group and the default group) - Delta SD (the difference in standard deviation between the other group and the default group) - Delta UK (the difference in the share of UK nationals between the other group and the default group) We will incorporate individual fixed effects. As there are no prior studies with effect sizes on which to base our choice of sample size we proceeded as follows. We sampled 61 participants for Condition 1 (WTP with full information) and ran the regression described above. For this small sample we obtained a coefficient size of -0.1706 for “Delta SD”. We then conducted a power analysis using this effect size. This analysis shows that we can detect an effect of -0.1706 with 90% power (at the 5% level) if we use a sample size of 86 participants. As we would like to conduct several analyses of heterogeneity of this main effect using categorical variables with up to 5 values and as there is substantial uncertainty about the effect size in Conditions 2 and 3 we decided on a sample size of N=430 participants (=86*5) per condition.

Supporting Documents and Materials

IRB