Understand Social Media, and You Won't Be Harmed by Them

The effect of social media on political outcomes is a hotly debated topic, and recent research casts doubts on whether there are significant any effects at all. Conversely, the detrimental effects of social media platforms on mental health seem to be well-established.

Luca Braghieri, Assistant Professor at Bocconi's Department of Decision Sciences, earned a €1.5mln Starting Grant from the European Research Council to measure causal effects of social media in domains where we still lack a clear picture – such as political engagement – and to design interventions that can mitigate the negative effects we are aware of. Since more than half of the world population is estimated to have a social media account, the importance of the topic can't escape anyone.

Braghieri's SOME (Social Media: Measuring Effects and Mitigating Downsides) project will use a unique blend of cutting-edge causal inference techniques from observational data and large-scale field experiments to generate insights that can benefit researchers, users, and policymakers alike.

The project will measure the effect of Facebook usage on voter turnout, party vote shares, and political donations, in the period 2008-2020, thus aiming to paint a comprehensive picture of the political effects of Facebook over time. Furthermore, it will investigate the formation of echo-chambers (a situation where individuals tend to interact with like-minded people and consume media that aligns with their pre-existing views) and filter-bubbles (where individuals' online experiences are personalized based on algorithms that predict their preferences and interests, thus preventing exposure to diverse views), and the proliferation and diffusion of low-quality and fake news. Braghieri will then evaluate the effectiveness of an intervention aimed at countering inertia in news consumption by requiring Facebook users to make an active choice about the news outlets that they follow on the platform.

Given the documented negative effects of social media on subjective well-being and mental health and the strong complementarity between social media and smartphone use, it is natural to worry about potential downsides of giving children smartphones from an early age. To shed light on the effects of smartphone use on a host of outcomes related to child development, including subjective well-being, mental health, academic performance, cognitive skills and non-cognitive skills, Braghieri will evaluate experimentally how a one-year delay in the purchase of a smartphone affects the outcomes above.



Another part of the project is aimed at mitigating the negative effects of social media on mental health without requiring drastic reductions in use. "I plan to leverage insights from the behavioral economics literature to design an in-app intervention," Braghieri said. A randomly-selected subset of Instagram users will be encouraged to follow a purpose-built Instagram page providing information on how to use social media with no detriment to one's mental health, and the outcome of the informative intervention will be evaluated.

"Causal evidence on the effects of social media will provide users with valuable information and will help identify potential areas for policy action, while testing scalable interventions to mitigate the downsides of social media will help design effective policies," said Braghieri, summarizing the benefits of his project.