Behavioral Economics | Laboratory Experiment | oTree + R
This repository contains the design, data, and analysis for a controlled laboratory experiment examining how injunctive norms and information shocks influence consumer purchasing decisions in a duopoly market. The experiment was co-designed and implemented as part of independent research at the LEEPS Lab, UC Santa Cruz, in conjunction with Professor Kristian Lopez Vargas.
The core question: how much does revealing unfair labor practices and activating social expectations around boycotts shift consumer willingness to pay a price premium for an ethical seller?
Participants (N = 24) completed a two-block oTree experiment:
Block 1 - Effort Wage Game: Establishes baseline effort and wage preferences to calibrate participant behavior prior to the market game.
Block 2 - Market Game (Duopoly): Participants act as buyers choosing between two sellers, an unethical seller (Seller A, lower cost) and an ethical seller (Seller B, higher cost). Three between-subjects treatments:
| Treatment | Condition |
|---|---|
| T0 (Baseline) | Neutral information - no labor practice details provided |
| T1 (Information Shock) | Participants informed of Seller A's unfair labor practices |
| T2 (Injunctive Norm) | Information shock + participants polled on whether they would support a boycott, activating social expectations |
Logistic regression with participant-level clustered standard errors estimated the treatment effects on the probability of choosing the ethical seller (Seller B):
| Treatment | P(Choose Ethical Seller) | Treatment Effect |
|---|---|---|
| T0 (Baseline) | 12.5% | - |
| T1 (Info Shock) | 75.0% | +57.5 pp |
| T2 (Info + Norm) | 79.2% | +61.5 pp |
The information shock alone accounts for the vast majority of the effect. Adding the injunctive norm condition (T2) yields a modest additional 4 percentage point shift, suggesting that revealed information about labor practices is the dominant mechanism, social norm activation amplifies but does not drive the effect.
Game Architecture: Implemented in oTree (Python-based framework for real-time web experiments), managing participant interactions, randomized treatment assignment, and data collection.
Identification Strategy: Controlled laboratory experiment with random assignment across treatment conditions. The controlled duopoly setting isolates the causal effect of information and norm activation from confounding market factors.
Econometric Approach: Logistic regression modeling the binary choice (ethical vs. unethical seller) as a function of treatment assignment. Clustered standard errors at the participant level account for within-subject correlation across rounds.
Theoretical Framework: Grounded in game theory, the boycott is modeled as a collective action problem where individual utility is a function of both monetary payoff and social image/conformity. Injunctive norms shift the social component of utility, lowering the effective price premium required to induce ethical purchasing.
├── code/
│ ├── probability_modeling.R # Logistic regression, treatment effect estimation
│ └── eda.qmd # Exploratory analysis (Quarto)
├── data/
│ ├── block1_effort_wage_2025-06-04.csv # Effort Wage Game (n=24)
│ └── block2_market_2025-06-04.csv # Market Game with treatment assignments
├── Consumer_Boycott_Results.pdf # Full results writeup
└── README.md
- Sample size: N = 24 pilot participants limits statistical power; results are directionally strong but should be interpreted with caution pending a full study.
- External validity: Participants drawn from the UCSC LEEPS subject pool (student population); effects may differ in real-world purchasing environments where social observation is absent.
- Artificial market structure: Sellers were constructed within the experiment and may not fully replicate competitive dynamics in real markets.
- Lab demand effects: Participants aware of being observed may overstate ethical behavior relative to anonymous real-world settings.
- Lopez Vargas, K. LEEPS Lab, UC Santa Cruz
- oTree: Chen, D.L., Schonger, M., & Wickens, C. (2016). oTree - An open-source platform for laboratory, online, and field experiments. Journal of Behavioral and Experimental Finance.