NORMS, NARRATIVES AND IDENTITIES: TESTING THE EFFECT OF NARRATIVE FORM AND SOCIAL NORM DEVELOPMENT ON INDIVIDUALS' ATTITUDES, BELIEFS, AND BEHAVIOURS
THE EFFECT OF NARRATIVE FORM PROJECT
Testing the effect of narrative form and social norm development on individuals' attitudes, beliefs and behaviours.
OUR AIMS
The research team are heading to rural Malawi to investigate whether and how narratives affect parents’ investment decisions in girls’ education. Do stories affect behaviour through conventional mechanisms alone (i.e., information about the economic costs and benefits of education) or do they operate through a richer set of social and cognitive processes (e.g., normative expectations, self-beliefs etc.)? The research team are heading to rural Malawi to conduct a lab-in-the-field experiment to better understand this question.
ABOUT THE PROJECT
There is a substantial body of evidence documenting significant impacts of narrative communication on individual behaviour (Bernard et al., 2014; DellaVigna and La Ferrara 2015; Vogt et al., 2016), however, as increasing levels of public resources are invested in the policy tool, it is important to understand the basic pathways through which they work.
This is part of a research study aimed at understanding ways to support girls’ education in rural Malawi. Research on educational performance in Malawi indicates that girls are underperforming compared to boys in school (Mzuza, Yudong & Kapute, 2014). Malawian girls achieve lower learning outcomes than boys and drop out earlier than boys. We are interested in exploring ways to help parents support their girls. This will to inform broad policy efforts to support parents in helping their girls thrive in school.
OUR RESULTS
In our analysis, we find no treatment effects on any of our primary or secondary outcome variables of interest. In general, the coefficient on the story outcome is negative, though always insignificant. The parent’s gender does affect allocations: Mothers give more to girls. However, there is no interaction effect between parental gender and either treatment condition. We intend to continue exploring the data to understand why we find no effects.