Experimental Design Details
5 Design
Throughout the experiment, subjects participate in a real-effort production assignment in four different phases. Aspects of the production assignment will be altered during each phase. The production assignment involves subjects having to correctly solve two two-digit addition problems within a limited amount of time. Subjects’ earnings are based on the correct number of problems they solve. Each phase consists of a blocks. A block consists of five rounds each. Subjects will be paired with another subject in each session. The production assignment consists of four phases: ability, tournament, treatment, and redistribution. After the production assignment, subjects will complete a survey. In the next subsections, we explain details of each phase of the production assignment, instructions given to each subject, and survey details.
5.1 Ability Phase
The ability phase consists of one block (five rounds) and is used to measure subjects’ ability. Subjects have 40 seconds per round to independently solve as many addition problems as possible. At the end of the phase, each subject is randomly paired with another subject from the same session and for the remainder of the experiment.
Ability is measured as the total number of math problems correctly solved by each subject. A subject is considered high (low) ability if they solved more (fewer) problems correctly than the counterpart they are paired with. Subjects are informed of why they were assigned a specific type. For example, a high-ability subject will receive the following prompt, “You are assigned Player A because you solved more problems correctly in Block 1 compared to the other player you are paired with.” In the case subjects solve the same number of problems correctly, we compare who has a higher percentage of correct problems to break the tie, which is the total number of correct problems divided by the total number of problems a subject tried to answer. This is used as an additional measure to reduce the probability of randomly assigning subjects to an ability type due to a tie. If a tie occurs after applying the secondary rubric, then subjects are randomly assigned an ability type and are told their assignment was random.
We do not consider treatment conditions exploring if knowledge about ability creates a priming effect in our experiment due to a lack of replication of social priming experiments across multiple disciplines, including economics (Chivers 2019; Yong 2012). We include priming conditions in our exploratory analysis section as possible extensions in case initial results indicate that priming might need to be investigated.
5.2 Competition Phase
During the second phase of the experiment, each pair engages in a competition to determine who can solve more problems correctly during each round. This phase consists of one block (5 rounds). The winner of a round earns a bonus of 10 ECUs, which equals $1. This earning structure is meant to mimic real-life conditions where individuals who win competitions tend to accumulate a larger piece of the pie. Both subjects have 40 seconds per round during this phase.
5.3 Competition Phase with Treatment
Phase 3 is the same as the competition phase except the treatment is implemented. This phase lasts for one block (5 rounds). The treatments are explained in section 5.6.
5.4 Redistribution Phase
Phase 4 consists of the same competition as phase 3, but it is followed by a redistribution dictator game at the end of this phase. The redistribution phase consists of one blocks (5 rounds). At the end of the block, the subject with the highest-earnings will have an opportunity to transfer any amount from their earnings to the losing subject. We call the subject making the transfer decision the redistributor and the subject receiving the transfer the receiver. The purpose of the dictator game is to capture the winner’s fairness views based on the inequality institution. The rationale for the fairness views is explained in a later section. While the winning subject is making this choice, the losing subject will complete two belief elicitation tasks. One task will gauge how much earnings they believe the winning subject will transfer. The other task will gauge the normative beliefs of how much the losing subject believes should be transferred. The first task is meant to capture the losing subject’s expectations of what they will receive, which is why it is incentivized since we want to see how well subjects can predict the amount they will receive. The second task is meant to capture their fairness view.
Belief elicitation for measuring expected transfers
Please state how much you EXPECT TO RECEIVE from the other player. You will earn 1 ECU if you correctly guess within 1 ECU above or below the amount the other player transferred. [Subject enters an amount ranging from zero to max amount of the other player’s available endowment]
Belief elicitation for measuring fairness view
Please state how much you BELIEVE YOU SHOULD RECEIVE from the other player. You will not earn anything for this question. [Subject enters an amount ranging from zero to max amount of the other player’s available endowment]
5.5 Survey
After the main experiment, subjects will complete non-incentivized survey questions. The survey is used to collect self-reported demographic information, and views on political leaning, equality, competitiveness, confidence, effort, ability, and redistribution. We also ask subjects what they think the objective of the experiment is and how they felt about how time was allocated in the experiment. The survey is in the supplementary material. Some survey questions come from Almås, Cappelen, and Tungodden (2020).