Abstract
A growing literature in economics studies the causal effects of media exposure on individual behavior (see DellaVigna and Ferara 2015 for a narrative review). The outcomes domains studied are diverse and cover crime (Dahl and DellaVigna 2009, ) health (Vaughan et al. 2000, La Ferrara et al. 2012, Kearney and Levine 2015, Trudeau 2016, Banerjee et al. 2019, Breza et al. 2021), education, empowerment and labor-market outcomes (Zavodny 2006, Chong and La Ferrara 2009, Jensen and Oster 2009, Ravallion et al. 2015, Bjorvatn et al. 2019, Kearney and Levine 2019), and financial decision making (Baker and George 2010, Berg and Zia 2017).
In the context of financial education, evidence from South Africa suggests that edutainment is an effective way to foster individual financial decision making (see Berg and Zia 2017) with realized treatment effects being larger, on average, than what is being found in recent meta-analyses of financial education field experiments (Miller et al. 2015, Kaiser and Menkhoff 2017, Kaiser et al. 2020). Edutainment may be seen as an especially promising avenue for financial education, as marginal costs of these interventions are generally low and the interventions appear to be well suited to operate at scale. While the available empirical evidence suggests that media (especially television and social media) can impact individual field behaviors in a meaningful way, the mechanisms leading to behavior change are less well understood. Against this backdrop, we test leading hypotheses of media impact regarding three disctinct channels (a) information provision, (b) role modeling and preference change via emotional connections, and (c) increased salience and basic awareness (see La Ferrara 2016).
To this end, we conduct a randomized encouragement experiment with a representative sample of over 3800 individuals in Italy offering monetary incentives to follow elements of a national campaign designed to foster financial literacy. We randomly allocate the individuals to one of four experimental conditions: (a) C: pure control (i.e., no encouragement), (b) T1: monetary encouragement to follow a popular soap-opera with financial messaging embedded into the story line (i.e., the role-modelling channel), T2: monetary incentive to follow a popular quiz show with questions related to financial decision-making posed within the show (i.e., the information provision channel), and (c) T3: a monetary incentive to follow a low-intensity social and traditional media campaign relying on a female fictional character communicating simple but salient and intiuitive messages about financial decision making (i.e., the salience and basic awareness channel). In addition to studying the relative effectiveness of these campaign elements, we are interested in heterogenous treatment effects by gender, baseline financial knowledge / education, and socio-economic background. Additionally, we are interested in the potential role of cognitive vs. non-cognitive process outcomes in mediating the observed treatment effects.
The results of our study are expected to be contributions to the literature on effective and scalable financial education policies and to the literature on persuasion and causal effects of media exposure more generally.