A growing body of research indicates the potential for social influence and behavioral spillovers to amplify the impact of policy interventions in areas such as public health and sustainable behavior. The intuition behind recruiting social influence, such as conformity influence, for behavioral change and social tipping is clear. Once a sufficiently large proportion of people in the population have adopted the desired behavior of the policy-makers intervention, this behavior change can spill over to those who haven’t changed yet. Several theoretical and empirical studies have shown that policymakers may be able to recruit conformity to endogenously drive change in the population in this way. But while the potential to leverage conformity-influence for population-level change has been explored, fewer studies have investigated the role of ‘anticonformity-influence’ in such a setting.

Simply put, while conformity influence leads decision-makers to adopt the behavior of the majority, anti-conformity influence shoves people to do whatever the majority doesn’t do. Especially in domains where group identities matter, decision-makers may jointly integrate both conformity- and anti-conformity cues. For example, an individual may want to align with her political ingroup’s position on whether to get vaccinated, while also making sure to differentiate themselves from whatever the political outgroup does. Currently, there is little theoretical or empirical work that jointly models both sources of social influence.

Here we propose to investigate how the joint presence of ingroup conformity and outgroup differentiation shapes behavioral change dynamics in a population after a policy intervention. We present results from an empirically informed agent-based model that simulates behavior change in a population after an external policy shock. The complete results of the model are not yet analyzed, but preliminary analyses indicate that a policymaker can maximize the desired behavioral change by strategically choosing who and how many to target in the subgroups. Furthermore, incorporating group identity affects the potential for social tipping in the population. While tipping is still possible, it is more likely to happen along group boundaries, which leads to a polarization of behaviors in the population. Overall, the distribution of psychological preferences that shape whether agents are more likely to respond to in- or outgroup social influence determines the scope of behavioral change. In cases where both ingroup conformity and outgroup differentiation suggest the same behavior, the case is clear-cut, but what about cases when the two forms of social influence point in opposing directions for behavior?

The full simulation results will be prepared until the meeting in May. We anticipate our simulation results to be a starting point for more empirical research that investigates the influence of ingroup conformity and outgroup differentiation of behavioral change, both in the lab and in the field. Group identities are consolidating across many domains, ranging from politics, and sustainability, to public health. When launching a behavior change intervention to achieve socially beneficial policy objectives, a given policy-makers thus need to understand the subtleties of how different forms of social influence affect population-level change after the intervention.