Gen Z and sustainable behavior: a laboratory experiment

Last registered on November 03, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

Gen Z and sustainable behavior: a laboratory experiment

RCT ID

AEARCTR-0017158

Initial registration date

November 01, 2025

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

November 03, 2025, 10:34 AM EST

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

Locations

Country

Brazil

Region

Distrito Federal

Primary Investigator

Name

Ludmila Ferreira Costa

Affiliation

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Soraya Eyzaguirre

PI Affiliation

PI Name

Matheus Melo

PI Affiliation

Additional Trial Information

Status

On going

Start date

2025-10-17

End date

2026-01-31

Keywords

Behavior, Lab

Additional Keywords

Generation Z, Sustainability, Behavioral Economics, Lab Experiment, Intergenerational Dilemmas

JEL code(s)

C72, C92, D62, Q20

Secondary IDs

Prior work

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

Abstract

This study will take place at the Universidade Católica de Brasília (UCB) and will invite undergraduate students to take part in a short session of experimental economics.

In the session, small groups of 3 participants will make decisions on a computer, at the same time, without knowing who the other participants are. In each round, everyone chooses how much “work/effort” to contribute and also makes a guess about what the others will choose. The computer then calculates each person’s payoff based on their own choice and on the choices of the other two group members.

We will compare three experimental conditions (treatments): one with a faster growth of the available resources (FAST), one with slower growth (SLOW), and one in which the resources are reset (RESTART). Each participant only sees their own screen and makes decisions privately.

The goal is to understand how young people (Generation Z students) make economic decisions when:

1. their payoff depends on what others do,
2. the amount of resources available in the future depends on choices made now.

At the end of the session, participants answer a short questionnaire about profile and attitudes, and then receive a certificate of complementary academic activity. No sensitive personal data will be published.

External Link(s)

Registration Citation

Citation

Eyzaguirre, Soraya, Ludmila Ferreira Costa and Matheus Melo. 2025. "Gen Z and sustainable behavior: a laboratory experiment." AEA RCT Registry. November 03. https://doi.org/10.1257/rct.17158-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

We will run a laboratory experimental economics session with undergraduate students at the Catholic University of Brasília (UCB). Participants will be placed in groups of three and will make two decisions in each round: (i) how much effort/work to contribute, and (ii) a prediction about the effort of the other two group members. Payoffs depend on their own choice and on the others’ choices. We will compare three experimental conditions (FAST, SLOW, RESTART) that differ in how the available resources evolve across rounds.

Intervention (Hidden)

The experiment is implemented in z-Tree (v. 5.1.16). Participants are anonymously matched in groups of three, drawn from three computer rooms running the same session time. In every period, each participant chooses an integer “effort” (1–8) and reports a prediction of the sum of the efforts chosen by the other two (2–16). Payoffs follow the production function in Fischer, Irlenbusch & Sadrieh (2004):

if total effort 𝑥<9, 𝐹(𝑥) = 0.6x;
if x≥9, 𝐹(𝑥) = 8.1−0.3x.

Individual return is the participant’s share of 𝐹(𝑥) times the current resource endowment. We add a prediction bonus: 20 – |predicted – actual others’ effort|, bounded below by 0. Treatments differ only in the resource update rule:

* FAST: high benchmark (21) in the resource-growth equation;
* SLOW: low benchmark (6);
* RESTART: resource is reset to the initial endowment every period.

Each participant plays only one session (one generation), but we track the resource that would be passed to a subsequent generation.

Intervention Start Date

2025-10-22

Intervention End Date

2025-11-07

Primary Outcomes

Primary Outcomes (end points)

1. Own effort (individual extraction / work choice) – integer from 1 to 8 entered by the participant in each round.
2. Prediction of others’ effort – participant’s stated belief about the sum of the other two group members’ efforts (integer from 2 to 16).
3. Resource for the next generation – resource level computed by the program for the following generation/round according to the assigned treatment (FAST, SLOW, or RESTART).

Primary Outcomes (explanation)

1. Own effort (individual extraction / work choice): this is the main decision variable. In every period, each participant chooses an integer effort level between 1 and 8. Higher effort increases the participant’s share of the period’s return but also increases the group’s total extraction, which can reduce the resource available later (depending on the treatment).

2. Prediction of others’ effort: after choosing own effort, each participant states a prediction of the sum of the other two participants’ efforts (an integer between 2 and 16). This variable measures beliefs about others’ behavior and is also used to compute an incentivized prediction bonus in the experiment: the closer the prediction is to the actual sum, the higher the bonus.

3.Resource for the next generation: once the group’s total effort is known, the program updates the resource according to the treatment-specific rule. For FAST and SLOW, the resource is adjusted upward or downward as a function of total effort (e.g. using a benchmark of 21 in FAST or 6 in SLOW); for RESTART, the resource is reset to the initial level. We record this computed resource level as an outcome because it captures the intertemporal effect of current choices on the future availability of resources.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Secondary outcomes will capture behavioral categories adapted from Fischer, Irlenbusch and Sadrieh (2004). For each round, we will classify the participant’s decision by comparing (i) own effort, (ii) the best reply to the stated prediction, and (iii) the prediction itself. We will code “intended sacrifice” when effort is lower than the best reply to the prediction, “intended best reply” when effort equals the best reply, and “intended waste” when effort is higher than the best reply. Using the prediction as a coordination anchor, we will also classify “intended gift-giving” when effort is lower than half the prediction, “intended consensus” when effort equals half the prediction, and “intended free-riding” when effort is higher than half the prediction. Beliefs will be classified as optimistic when the prediction is higher than the actual sum of others’ efforts, realistic when it matches that sum, and pessimistic when it is lower. Finally, we will create combined strategy–belief profiles (OG, RG, PG, OC, RC, PC, OF, RF, PF) by crossing the three belief types with the three effort-based categories. These variables will be constructed at the round level and then summarized per participant as the share of rounds in each category.

Experimental Design

This is a lab experiment with between-subject variation in three treatment conditions (FAST, SLOW, RESTART). Students are invited to the lab, seated at separate computers, and randomly assigned to a treatment. They make simultaneous decisions in groups of three, without learning the identities of the others. Each participant takes part in only one session.

Experimental Design Details

The design follows Fischer, Irlenbusch & Sadrieh (2004), adapted to a student sample (Generation Z) and to three growth regimes. We will run multiple lab sessions, each session forming triads across 2–3 computer rooms to avoid within-room contamination. Within a session, participants are pre-assigned to a treatment list so that every triad has 1 FAST, 1 SLOW and 1 RESTART participant interacting through the production function. The sequence of screens in z-Tree is: Identification → Decision → Program (payoff and resource-update computation) → Feedback → Post-experimental questionnaire → Payment screen. Belief elicitation is incentivized via the bonus 20 – |error|.

Randomization Method

Before the session, we prepare printed slips (coupons) that already contain: (i) the room, (ii) the treatment (FAST, SLOW, or RESTART), and (iii) an anonymous participant code. At check-in, each participant blindly draws one slip. This draw determines the room and treatment for that participant and preserves anonymity, because the code on the slip is used during the experiment instead of the participant’s name.

Randomization Unit

Individual participant (each person draws one coupon that assigns room and treatment). After that, participants are grouped in triads for the decision task, so decisions are made in 3-person groups, but the treatment was assigned at the individual level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

60 groups (triads of 3 students each)

Sample size: planned number of observations

180 undergraduate students

Sample size (or number of clusters) by treatment arms

FAST: 60 students (≈ 20 groups)
SLOW: 60 students (≈ 20 groups)
RESTART: 60 students (≈ 20 groups)

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

With 180 students (60 per treatment arm), individual-level randomization, α = 0.05 (two-sided) and power = 0.70, the minimum detectable effect size for pairwise comparisons between treatments is approximately 0.45 SD.

Supporting Documents and Materials

IRB