Primary Outcomes (explanation)
1. Gender attitudes
We will report a social norm index that aggregates all subindices below. Also, we will report each subindex separately:
• Norms on women´s work and marriage subindex
• Norms on women´s education subindex
• Norms on the acceptability of IPV subindex
• Norms on women´s mobility
• Norms on household decision-making
We will report outcomes separately for men and women.
2. Women´s Empowerment
We will report an empowerment index that aggregates all subindices below. Also, we will
report each subindex separately:
• Mobility sub-index
• Reported decision-making sub-index
• Elicited decision-making sub-index
• Tayssir decision-making sub-index
3. Children Education
• Completed primary school (for children 13 years old or older)
• Completed collége (for children 16 years old or older)
• Completed lycée (for children 19 years old or older)
• Attending primary school (for children up to 12)
We will report outcomes separately for boys and girls.
4. Preferences for the design of social protection programs and views over potential tradeoffs
We will report answers to questions that capture support for women as the beneficiaries of cash transfers and perceived associated risks
5. IPV
We will report an IPV index that aggregates all subindices below. Additionally, we will report each subindex separately:
• Physical IPV subindex (ever experienced, experienced during Tayssir, experienced in past 12 months)
• Emotional IPV subindex (ever experienced, experienced during Tayssir, experienced in past 12 months)
As IPV victimization is a sensitive topic, there could be underreporting of its incidence. This can pose a special threat to estimating the causal impact of women being assigned as beneficiaries of Tayssir on IPV if its reporting changes by treatment status. For instance, if treatment impacts empowerment and this leads to higher reporting but not higher incidence, we could be wrongly concluding that the treatment leads to an increase in IPV.
This challenge has led researchers to adopt several different methods to measure IPV, such as list experiments and Audio Computer-Assisted Self-Interview. We piloted both alternative methods and concluded that they did not perform well in our context, where we interview illiterate people with low familiarity with managing a tablet. Therefore, we decided to adopt a different approach in order to increase the acceptability of reporting IPV to surveyors. We first ask respondents if they agree that if it only happens once, a woman should endure IPV for family stability. Our pilot indicates that a great majority of women agree with this sentence. After, we ask them if they have already endured IPV for family stability. By embedding the question within a culturally recognizable justification, it reduces the confrontational nature of directly asking about IPV victimization. To validate this method, we will report whether the acceptability of enduring IPV that happens once for family stability varies by treatment status. In addition, we will measure if there are any treatment effects on the probability of reporting IPV through a vignette.