Intergenerational Perceptions of Fairness

The key manipulation is the source of endowment. In generation 1, participants are randomly assigned to 4 between-subject treatment conditions at group level:
1) High Luck (High endowment earned via Coin Flipping Task / received most number of heads).
2) Low Luck (High endowment earned via Coin Flipping Task / received least number of heads).
3) High Morality (High endowment earned via Token Grabbing Task/ grabbed least tokens from public pool).
4) Low Morality (High endowment earned via Token Grabbing Task/ grabbed most tokens from public pool).

In Generation 2, participants inherit the endowment status (High vs Low) with full information of its source.

2026-03-23

2027-05-31

Our primary outcome is the level of cooperation which will be measured using individual contribution in the public goods game. For analysis, we will use both
1. Absolute contribution: the number of ECU(s) contributed by the participant in each round.
2. Relative contribution: the percentage of the participant’s endowment contributed in that round.

We have several secondary outcomes which we hypothesize will be relevant in influencing levels of cooperation.

1) Participants' fairness judgements.
Using participants' individual decisions on levels of disapproval to assign to others for each potential level of contribution, we will analyse
(a) Patterns of how assigned disapproval varies with others' (potential) contribution.
(b) Realised levels of disapproval based on others' (actual) contribution.

2) Participants' fairness beliefs.
Using pre-task elicitations, we will analyse:
(a) What they think is a fair level of disapproval for each potential level of their contribution.
(b) What they predict will be the amount of disapproval points received for each potential level of their contribution.

Generation 1 Part 1: Endowment-Determining Tasks
Participants first complete two monetary-incentivized tasks, the outcomes of which will determine their endowment.
There are two tasks, actions in which involve luck in one task, and morality in the other.
(1) Token Grabbing Task (Morality). Participants are randomly matched into groups of six and face a shared pool of 60 virtual tokens. Every 6 seconds, each participant has one opportunity to grab one token which is worth 1 ECU. This process is repeated for 10 rounds. Participants’ earnings from this task are calculated as: Earnings=[No. of tokens the participant grabbed + (1.8 x tokens left in pool)/6 ]
(2) Coin Flipping Task (Luck). Every 6 seconds, the computer flips a virtual coin for the participant. This process is repeated 10 times. Participants earn 3 ECUs for each head outcome. Earnings from this task are calculated as: Earnings = No. of heads flipped × 3

Generation 1 Part 2: Public Goods Game
Participants are assigned a family name and placed into fixed groups of three families to play a 16-round public goods game. Within each group, one participant receives a high endowment (30 ECU) and the other two receive low endowments (15 ECU). Endowment assignments remain fixed throughout the game.

Determination of endowments:
Each group's endowments will be randomly assigned based on one of the following four treatments:
1) High Luck (High endowment for member who flipped most heads).
2) Low Luck (High endowment for member who flipped least heads).
3) High Morality (High endowment who grabbed least tokens from public pool).
4) Low Morality (High endowment who grabbed most tokens from public pool).
Treatments 1 (2) mirrors Treatment 4 (3) in that the participant who earns the most (least) in the task is assigned the high endowment. This helps control for any potential differences in perceived income in comparisons of luck and morality. All participants are informed of the reasons behind their own and others' endowments.

Repeated public goods game:
(1) Pre-Rounds. Participants firstly complete two preliminary rounds where participants make contribution decisions under both the high- and low-endowment scenarios.
(2) Ex-Ante Belief Elicitation. Before Round 1 of the public goods game, participants are asked to report the following for each pre-defined contribution range:
i) “What is the fair number of disapproval points do you think you should receive?”
ii) “How many disapproval points do you think you will receive from your group members?”
(3) Main Rounds. Each round proceeds as follows:
i) Disapproval. In Round 1, participants indicate how dissatisfied they would feel (on a scale from 0 to 10) with each teammate’s contribution level across predefined contribution ranges. In subsequent rounds, participants can either retain their previously submitted disapproval schedule or update it.
ii) Contribution Decision. Participants decide how much of their endowment to contribute to the group project.
iii) Feedback Screen. Participants observe each group member’s contribution and the total disapproval points received by each member.
iv) Next Round. Participants proceed to the next round with refreshed endowments and the same group members.
(4) Ex-Post Belief Elicitation. After the last round of the public goods game, participants are again asked to report the following for each pre-defined contribution range: “What is the fair number of disapproval points do you think you should receive?”

Generation 1 Part 3 & Generation 2 Part 1: Family Matching Stage
After Generation 1 participants complete the public goods game, participants from Generation 1 (Session 1) and Generation 2 (Session 2) enter a family matching interface. Each Generation 1 participant selects one Generation 2 participant from an anonymous pool, while Generation 2 participants observe the selection process. To strengthen the perceived intergenerational link, each matched pair completes a three-round cooperative spot-the-difference task together.

Generation 1 Part 4: Post-Questionnaire
After completing the family matching stage, Generation 1 participants proceed to a post-experimental questionnaire.

Generation 2 Part 2: Public Goods Game
After the family matching stage, Generation 2 participants proceed to the public goods game.

Determination of endowments:
All participants inherit the endowments of their family in Generation 1, and are fully informed of how these endowments were determined.

Repeated public goods game:
The flow of the public goods game is exactly the same as in Generation 1.

Generation 2 Part 3: Post-Questionnaire
After completing the public goods game, Generation 2 participants complete a post-experimental questionnaire.

Not available

Randomization is implemented programmatically within the experiment code.

Each group of 3 families across 2 generations.

Yes

Each cluster consists of a group of 3 families across 2 generations or 6 participants.
72 clusters

432 participants.

18 clusters High Luck Condition; 18 clusters Low Luck Condition; 18 clusters High Morality Condition; 18 clusters Low Morality Condition.

Our main outcome is individual contribution in the Public Goods Game (PGG). With 432 participants across four treatment arms (108 per arm), and accounting for clustering with groups of size six (2 generations of 3 families), we have a minimum detectable effect size of 0.5392 standard deviations, assuming an intraclass correlation coefficient (ICC) of 0.20, 80% power, and a two-sided significance level of 0.05. Assuming a typical standard deviation of 25 percentage points which is consistent with typical contribution variance reported in the experimental PGG literature, this corresponds to approximately 13.48 percentage points in absolute terms (0.5392 × 25 pp).