Gender differences and framing in risk preferences

How to cite

Ezquerra, Lara; Jimenez, Natalia; Ortiz, Jose; Espinosa, Maria Paz. 2026. "Gender differences and framing in risk preferences." AEA RCT Registry. AEARCTR-0017190. https://doi.org/10.1257/rct.17190-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Lara Ezquerra* , Natalia Jimenez , Jose Ortiz , Maria Paz Espinosa

* Primary investigator

Registry ID
AEARCTR-0017190
DOI
https://doi.org/10.1257/rct.17190-1.0
First registered
Last updated
Status
In development
Trial dates
2026-02-15 – 2027-03-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/17190

Abstract

This study investigates whether gender differences in risk attitudes emerge only under conditions that trigger intuitive, heuristic-based thinking. Building on dual-process theory (Kahneman & Tversky), we propose that in the absence of framing cues, both men and women rely on analytic, system-2 reasoning and therefore demonstrate comparable levels of risk aversion. However, when even minimal but relevant framing is introduced, decision-makers shift toward system-1 processing, making intuitive judgments shaped by heuristics. Prior research suggests that heuristic tendencies may differ by gender, potentially leading to divergent choices under framed conditions. Drawing on literature examining gender-specific cognitive biases and heuristic use, we test whether framing moderates the relationship between gender and risk behavior. The findings contribute to understanding the cognitive mechanisms behind gendered decision-making, offering theoretical value for behavioral economics and practical implications for policy design, entrepreneurial decision contexts, and financial communication strategies.

Intervention

The experiment consists of a series of incentivized lottery–choice tasks designed to measure risk preferences under different framing conditions. Participants make 10 decisions in a table, choosing between two payoff options in each row. Each option specifies a lottery with outcomes realized over four consecutive periods, and one decision is randomly selected for payment.

Participants are randomly assigned to one of three experimental conditions:

1. Baseline Condition (No Frame / Neutral Risk Task)

Participants choose between Option A and Option B in each of the 10 decisions.

Each option consists of a lottery with specified probabilities and payoff vectors across four periods.

No contextual or emotional framing is provided; options are presented in purely numerical and neutral terms.

This condition is designed to induce analytical, System-2 reasoning.

2. Treatment 1 (Entrepreneurship Frame – Entrepreneurship Highlighted as Risky Option)

Participants choose between being an “employee” or an “entrepreneur.”

Entrepreneur option: variable payoffs across the four periods depending on investment success.

Employee option: safe and constant payoffs across all periods, contingent on firm profitability.

This treatment frames risky choices as entrepreneurial activity and safe choices as employment, aiming to trigger intuitive, heuristic-driven System-1 processing.

3. Treatment 2 (Employeet Frame – Employee Highlighted as Risky Option)

Participants again choose between being an “employee” or an “entrepreneur.”
However, the payoff structure is reversed relative to Treatment 1:

Employee option: stochastic payoff vector depending on firm performance

Entrepreneur option: constant payoff across periods

This treatment presents employment as the riskier choice and entrepreneurship as the safer one, again aiming to activate heuristic-driven System-1 reasoning via framing.

Experimental design

This study investigates decision-making under risk using a series of incentivized choice tasks. Participants complete 10 decisions, each requiring a choice between two options that yield monetary payoffs determined by lotteries. Payoffs are expressed in ECU and converted to euros at the end of the experiment. One decision is selected at random for payment.

The core experimental manipulation consists of three between-subject conditions, to which participants are randomly assigned:

Baseline (Neutral Presentation)
Participants choose between two unlabeled lottery options presented in a neutral, numerical format.

Treatment 1 (Entrepreneurship vs. Employment Frame A)
Participants choose between two options labeled “entrepreneur” and “employee.” One option provides variable outcomes based on investment success, while the other provides constant outcomes based on firm conditions.

Treatment 2 (Entrepreneurship vs. Employment Frame B)
Participants again choose between options labeled “entrepreneur” and “employee,” but the payoff structures are reversed relative to Treatment 1. One option provides variable payoffs tied to firm conditions, and the other provides constant outcomes linked to investment results.

The probability structure and expected values are held constant across conditions; only the contextual framing and interpretation of the options differ. All participants face identical numerical payoff values.

The task is implemented individually and anonymously. Participants receive real monetary compensation based on their choices and chance outcomes.

Design details

Randomization method
computer
Randomization unit
Individual level
Clusters
3 treatments with 100 randomly assigned individuals each
Observations
300
Arms
100
Primary outcomes
Gender, risk aversion

Cited by

No citation data available for this registration yet.

Keywords