Social Preferences of Different Income Groups

Last registered on January 26, 2024


Trial Information

General Information

Social Preferences of Different Income Groups
Initial registration date
March 06, 2023

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 13, 2023, 8:49 AM EDT

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

Last updated
January 26, 2024, 8:53 AM EST

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



Primary Investigator

University of Haifa

Other Primary Investigator(s)

PI Affiliation
FAU Erlangen-Nürnberg, University of Haifa
PI Affiliation
University of Haifa

Additional Trial Information

Start date
End date
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
This study deals with fairness views of different income groups towards different types of inequality. Recent work has concentrated on views regarding inequality that is generated by luck and merit. In this work, we add to this literature and examine fairness views towards inequality that is based on someone else's taste (which we dub taste-based inequality). In an earlier study, we already examined this type of inequality alongside luck- and merit-based inequality on a representative sample of the adult Israeli population (the pre-registration of that study is available on the AEA RCT registry with RCT ID: AEARCTR-0008171).

In that study we reported suggestive evidence according to which people with below-average income treat taste-based inequality differently than people with above-average income. Moreover, it seems that the former group treat this source of inequality in a similar fashion to their treatment of luck while the latter treat it similarly to merit.

Since the previous study was not meant to examine this type of heterogeneity, the results regarding the behavior of the two income groups were underpowered in that study. For this study, we make a power calculation based on the results from that study in order to directly investigate the earlier suggestive findings.

We will run an incentivized online experiment based on the impartial spectator design (as in Almas et al., 2020). Roughly 3000 individuals take part in the experiment, out of which roughly 1140 will play the role of impartial spectators who will make redistribution decisions between two individuals with unequal earnings (570 spectators will belong to the below-average income group and 570 spectators will belong to the above-average income group). There will be three treatments corresponding to three different sources of inequality: Luck, Merit and Taste. Impartial spectators will be randomly selected into one of the three treatments.
External Link(s)

Registration Citation

Cohen, Rachelle, Amnon Maltz and Yuval Ofek-Shanny. 2024. "Social Preferences of Different Income Groups." AEA RCT Registry. January 26.
Experimental Details


Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Implemented inequality by impartial spectators.
Primary Outcomes (explanation)
In each treatment there are two participants who receive initial income that is unequal: one receives 12 ILS and the other receives nothing. The impartial spectators make a redistribution decision. They can redistribute 0,2,4,6,8,10 or 12 ILS from the participant with the higher income to the one with no income. The implemented inequality will be defined as the: (absolute value of the difference between the income of the two participants after the redistribution decision)/12

Secondary Outcomes

Secondary Outcomes (end points)
Distribution of types of fairness views.
Secondary Outcomes (explanation)
See analysis plan.

Experimental Design

Experimental Design
The basic experimental design is similar to the one that appeared in our earlier experiment that studied social preferences over taste-based inequality (the pre-registration is on the AEA RCT registry with RCT ID: AEARCTR-0008171). There are only a few notable differences compared to the previous design. As a result, some chunks of the experimental design described below are taken from the pre-registration of the earlier study.

The general design is as follows:

Following the impartial spectator design in Almas et al. (2020), we will conduct an online experiment in order to capture the social preferences in the Israeli population. The experiment will be run by a panel company and will include roughly 3000 participants. The participants in the final stage of the experiment (the spectators) will be roughly 570 individuals with income that is above average and 570 individuals with an income that is below average (the income classification is based on self-reports of the panelists that are collected and updated regularly by the panel company). All participants will receive a participation fee. Some participants will receive an additional payment according to their performance, luck and decisions made by other participants in the study.

The study consists of three treatments: Luck, Merit and Taste. Using the impartial spectator design (Almas et al., 2020), spectators will be asked to redistribute payments between two other participants in the experiment who are randomly paired together. In each pair, one participant will have 12 ILS and the other will have nothing. The spectator will decide how to redistribute the money in the pair. The difference between treatments lies in the source of the initial inequality. In the first treatment the source will be luck. In the second treatment, it will be merit, i.e., the relative performance level in a task. Finally, in the Taste treatment, the source will be a decision of another participant who earned the money and chose to whom to transfer it based on some personal information provided by the pair.
Experimental Design Details
Detailed Experimental Design

The Luck and Merit treatments include two stages and are quite similar to each other (both are based on Almas et al., 2020). The only difference between them is the source of inequality: In the Merit treatment it is based on performance in a task and in the Luck treatment it is random (each participant in a pair has the same probability to get the money for performing a task). The Taste treatment involves another stage. The details are given below.

Luck Treatment

Stage 1: Participants ("workers") perform a task. Upon completion, they are randomly matched into pairs. In each pair, one worker is randomly chosen (with probability 0.5) to receive 12 ILS for completing the task and the other receives no reward.
Stage 2: Additional participants ("spectators") are randomly assigned to the pairs. Each spectator decides how to redistribute the earnings in the pair (the options are 12-0, 10-2, 8-4, 6-6, 4-8, 2-10 and 0-12).

Merit Treatment

Stage 1: As the Luck treatment but the worker who receives the monetary amount for the task is the one who performed better in the task.
Stage 2: As in the Luck treatment.

Workers will perform two tasks. The first will be the basis for the Luck treatment and will not be graded. The second will be graded and will be the basis of the Merit treatment. Pairs of workers will be matched randomly and separately for each treatment.

Taste treatment

Stage 1: Participants will be asked to describe themselves in a paragraph using up to 200 words. After they write down this description they will be told that they will be randomly matched with another participant who completed the same task. In the second stage of the experiment a different participant (which we call here the "benefactor") will perform a task, earn 12 ILS for her/his performance, and will then be asked to transfer the amount to one of the two participants. The benefactor will be shown the paragraphs written by both participants and, based on them, will have to make his transfer decision.

Note: The participants write down the paragraphs before they read about the next stages of the payment procedure, i.e., they do not know the purpose of the paragraphs at the time of writing.

Stage 2: Benefactors will be recruited. They will perform a task and earn 12 ILS for it. They will then be shown the two paragraphs written by a randomly chosen pair from stage 1. They will be told that these paragraphs are the self-description that was written by the participants before the payment mechanism was explained to them. Based on these paragraphs, the benefactors will be asked to decide to whom to transfer their earnings (they will be told upfront that they cannot keep their earnings but may choose to whom it will be transferred).

Stage 3: Spectators will be given the opportunity to redistribute the money in the pair as in the Luck and Merit treatments.


1. We refer in the pre-registration (but not in the experiment) to the participant who receives the money from the benefactor as the "beneficiary".

2. All the information regarding the different stages and the payment procedure is made clear in the instructions to all participants in the most transparent manner (see instructions that will be made public after the experiment is completed).

3. Spectators will redistribute money only for one pair from one treatment (i.e., the analysis will be held between participants).


Almås, I., Cappelen, A.W., & Tungodden, B. (2020). Cutthroat capitalism versus cuddly
socialism: Are Americans more meritocratic and efficiency-seeking than
Scandinavians? Journal of Political Economy, 128 (5), 1753-1788.
Randomization Method
All randomizations will be made using a computer program.
Randomization Unit
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
No clusters planned.
Sample size: planned number of observations
The number of observations that will be relevant to the data analysis is roughly 1140 (roughly 380 spectators in each treatment). Half of them will be of below-average income and half will be of above-average income.
Sample size (or number of clusters) by treatment arms
We have 4 types of participants:

Participants in the first stage of the Taste treatment - 760
Benefactors in the second stage of the Taste treatment - 380
Workers (first stage of Luck and Merit treatments) - 760
Spectators (380 for each treatment) - 1140

Total - 3040
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The power calculation is attached in the Docs & Materials section.
Supporting Documents and Materials


Document Name
Power Calculation
Document Type
Document Description
Power Calculation

MD5: add493d03ff12794d1d5ef18bc660b8b

SHA1: 652dbaa15542a364577897da52242605b47e1e92

Uploaded At: March 06, 2023

Document Name
Document Type
Document Description

MD5: ee2f47f459f7c34075d04f48c2df5e4f

SHA1: 74fb865a5b535c233f2b6db8c52b887b75d3e44a

Uploaded At: March 06, 2023


Institutional Review Boards (IRBs)

IRB Name
University of Haifa, Faculty of Social Sciences
IRB Approval Date
IRB Approval Number
Analysis Plan

Analysis Plan Documents


MD5: e3b3630d795e5393ac0a0693406baf41

SHA1: 14750f58c02ba554b1054aeb3fec64c3f78ba87e

Uploaded At: March 06, 2023


Post Trial Information

Study Withdrawal

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

Request Information


Is the intervention completed?
Intervention Completion Date
May 05, 2023, 12:00 +00:00
Data Collection Complete
Data Collection Completion Date
May 05, 2023, 12:00 +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
Was attrition correlated with treatment status?
Final Sample Size: Total Number of Observations
Final Sample Size (or Number of Clusters) by Treatment Arms
367 in Taste treatment 390 in Luck treatment 399 in Merit treatment
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials