Gender social norms, expectations, and sanctions: An experimental economics perspective

This intervention is designed as an online experiment to explore the penalization related to non-conformity with gender social norms. It is structured around two main parts: (i) an incentivized question tailored to elicit the social norm about the allocation of time between paid and unpaid household work, following the Krupka & Weber (2013) elicitation method; (ii) a Dictator Game where participants decide on the allocation of resources between themselves and a virtual counterpart. The counterpart’s compliance with gender norms regarding time distribution is the manipulated variable. The deviations are varied along two dimensions: the sex of the virtual counterpart (same sex, opposite sex) and the degree of non-compliance (complies with the gender social norm; does not comply with the gender social norm by doing more household work than what the social norm expects; does not comply with the gender social norm by doing less household work than what the social norm expects). This approach allows for the examination of whether deviations from gender norms influence penalizations or rewards, and if these reactions differ based on the sex of the deviator.

2024-08-14

2024-09-05

Gender social norm regarding time allocation between paid work and unpaid household work.
Sanction or reward for the compliance or no compliance with the elicited social norm.

The gender social norm regarding time allocation between paid work and unpaid household work is elicited following Krupka and Weber's (2013) approach, which relies on an incentive system to detect social norms through simple coordination games. Specifically, participants must predict others' responses regarding time allocation, aiming to reveal the social norm. To ensure that participants disclose social norms rather than merely their personal preferences, the method includes economic incentives when their answers match the most common response among all participants (the modal value), thereby motivating participants to accurately respond about the assessment of shared norms.
The sanction or reward for non-compliance or compliance with the elicited social norm is measured by the difference in the amounts allocated in two different Dictator Games.

In this online experiment, participants are randomly assigned to one of three experimental conditions that differ in the compliance with the gender social norm and the sex of the virtual counterpart in a Dictator Game. Participants will first be asked to 'Guess how the majority of people close to you believe household chores should be divided between men and women in a family consisting of a couple and their two young children,' using an incentivized task. This captures the social norm for a specific situation, making the responses comparable.
Participants will then play a dictator game with a virtual counterpart from whom they know only their sex. Finally, participants will play another dictator game with a counterpart from whom they know both their sex and whether they comply with the gender social norm.
The decision problem is embedded in a standard dictator-game setup where the shared amount is our outcome variable. We have the following three experimental cells/treatment arms:
1. The participant decides their strategy considering the sex of the virtual person they are playing with (receiver) and the following information about the degree of conformity with the social norm: the virtual person complies with the gender social norm regarding the distribution of time between paid work and unpaid reproductive or caregiving work.
2. The participant decides their strategy considering the sex of the virtual person they are playing with (receiver) and the following information about the degree of conformity with the social norm: the virtual person does not comply with the gender social norm, as they allocate more hours to paid work and fewer hours to unpaid reproductive or caregiving work compared to what the norm stipulates.
3. The participant decides their strategy considering the sex of the virtual person they are playing with (receiver) and the following information about the degree of conformity with the social norm: the virtual person does not comply with the gender social norm, as they allocate fewer hours to paid work and more hours to unpaid reproductive or caregiving work compared to what the norm stipulates.

oTree, developed by Chen et al. (2016), was used for the randomization of the treatments.

Individual (participants).

No

No clusters are planned in this study.

246 participants.

41 per treatment arm (3 treatments, between-sex group design).

As is standard in experimental literature, a significance level of 0.05 was set. Similarly, a desired statistical power (1−beta) of at least 0.80 was established, representing the probability of detecting an effect if it truly exists. Following the conventional values proposed by Cohen (1969) for statistical power analysis in the behavioral sciences, the size of the minimum expected effect was set at 0.20, below the 0.25 considered a medium effect size by Cohen. The choice of 0.20 instead of 0.25 was based on the nature of the experimental design, where the allocation of tickets as a reward in the dictator game only allows integer values. In other words, a minimum detectable effect size of 20% indicates that the experiment was designed with a high probability of detecting a difference in the ticket allocation in the game equal to or greater than 20% of the total allocated, which equates to 2 tickets.