Qinyou Hu is a PhD student and job market candidate at Rice University.
Click here for Qinyou’s job market paper Elevator Pitch.
Adolescence can be a period full of joy and happiness, but also one of fear and vulnerability. Bullying is a major source of this vulnerability for adolescents worldwide. One-third of youth are victims of bullying, which deteriorates their academic performance, mental health conditions, and even increases suicide ideation and the risk of committing suicide. The situation can be even worse in the developing world, where bullying has yet to receive the level of public awareness and policy attention it deserves.
Bullying often takes place within small social groups. For example, an unfortunate teenager might find themselves isolated or targeted by rumors from peers, or more severely, they may face threats, both verbal and physical, from others. This is where social-emotional skills come into play, offering valuable support to both bullies and victims. Regrettably, current anti-bullying policies or interventions fall short of harnessing these crucial tools and instead emphasize the negative consequences of bullying.
My job market paper tackles this issue through a unique angle of cultivating adolescents’ empathy skills and aims to understand how empathy works to reduce bullying. Empathy is defined as the ability to detect or sense others’ emotions. Empathetic students can feel bad about bullying others as they don’t want to see others suffer, a pathway by which empathy can have an individual human capital effect on bullying reduction. Empathy may also exert a social effect through reshaping adolescents’ social circles and the subsequent influence from peers within those new circles. However, empirical evidence supporting this channel is scarce.
To fill this empirical gap, I analyze a parent-directed parental involvement program fostering empathy among middle schoolers in China. Getting parents involved to learn empathy with students is effective in improving students’ empathy, which further helps reduce bullying. To investigate the mechanisms by which empathy reduces bullying, I collect social network data from the field to further enhance our understanding of the ways empathy influences bullying.
How do students’ network structure in classes change when empathy increases?
First, I illustrate empathy’s social effect by visualizing changes in students’ friendship networks.
Figure 1 visualizes the friendship network structures of one sample class at baseline (Panel A) and follow-up (Panel B), where each circle denotes one student, and the size of the circles indicates the number of friendship nominations received by the student. Originally, students tend to associate with those who share a similar status, and both bullies and non-bullies can be popular. However, at the follow-up, the social circles of bullies break down into pieces, suggesting that bullies become less popular within the class after empathy education.
Decomposing empathy’s two channels
Next, I develop a structural model to (i) distinguish the contributions of empathy’s two channels — the human capital and social effects and (ii) account for the interdependent nature of individual bullying decisions within the friendship network, which, therefore, should be considered as an equilibrium outcome.
In my model, utility-maximizing parents choose whether to participate in the empathy-fostering program, and children make decisions regarding their interactions with peers in the classroom, including friendship formation and bullying behavior. Students’ empathy evolves according to a human capital production function, taking students’ baseline empathy skill endowment and parents’ participation status as inputs. Students make friends according to a dyadic link network formation model and decide on bullying involvement following a social interaction type model. Essentially, parents’ participation decision affects the students’ empathy level, which later leads to changes in their social circles. Finally, the bullying outcome is a joint product of the student’s new empathy and their friends’ bullying behavior.
The key feature of the model is that empathy plays a twofold role in shaping bullying: first, individual empathy levels directly influence the private utility of their own bullying efforts to capture its human capital effect, and second, empathy affects the probability of friendship links within the class, thus governing the process of social network formation to capture its social effect. The decomposition from the model estimates shows that approximately two-thirds of the empathy effect on bullying is attributed to the human capital effect, while the remaining one-third stems from the social effect.
Who should we target with empathy interventions?
What does this decomposition result imply? What insights can we glean from the model? Understanding how empathy works through social networks informs policy design, and network theory-based targeting can be particularly appealing, especially when resources are limited. Currently, the majority of existing anti-bullying programs adopt whole-school approaches and educational programs. When they do target, they often focus on the bullies themselves. While this may seem intuitive, the question remains: is it the most effective approach?
My model can help answer this question. I use the model to simulate the policy counterfactual of targeting bullies directly and compare it with the effectiveness under random targeting. As shown in Panel A of Figure 2, I find that targeting bullies directly, no matter how popular they are, will always result in smaller reductions in bullying than random targeting. Digging into the mechanisms, I find that this is mainly driven by the fact that bullies are usually associated with lower empathy skills, are less popular than the average student, and, as a result, targeting them always generates smaller social effects than random targeting.
Luckily, it is not the end of the world. Leveraging social network data helps us get further! Accounting for the social aspect of bullying shows that fostering empathy can have additional impacts through reshaping students’ social networks. Therefore, instead of targeting bullies, I try targeting bullies’ friends—especially their best friends. I find that targeting bullies’ friends always leads to more reductions in bullying than random targeting, as shown in Panel B of Figure 2. This is because those students identified as bullies’ closest friends have higher social status and are more popular in the class than the bullies themselves. So, targeting bullies’ closest friends generates a higher social effect than random targeting, and importantly, the effect will inevitably get transmitted to their bully friends. So, you are indeed who your friends are!
I also simulate an alternative policy counterfactual – targeting based on students’ popularity. The results again highlight the value of social networks: I find that targeting based on popularity is more effective than targeting students randomly, and at its maximum, we can get a 7.5% further reduction in bullies from popularity-based targeting. Nevertheless, if the budget allows, whole-school education is still preferred because the effects of random targeting and popularity-based targeting gradually converge.
The story continues …
“Being protected from bullying is a fundamental human right,” says a top UN advocate. My paper offers a new solution to address bullying by integrating evidence from the field with insights from a structural model. Policy counterfactuals suggest that directly targeting bullies can be insufficient. In contrast, cultivating adolescents’ social-emotional skills, like empathy, can be an efficient solution as it generates an additional benefit of reshaping their friendship networks, subsequently reducing their involvement in bullying. This social effect of empathy prompts more effective targeting strategies, such as targeting by popularity or the social circles of bullies.
Finally, empathy skills are closely associated with prosocial behaviors in a broader context. Improving students’ empathy via the intervention may generate positive externalities but is often under-appreciated in the universal education system. At the very least, it is worthwhile to educate empathy as an important skill, and not just for anti-bullying purposes.
Feature Image Source: Ashoka
