?
Friend or Foe: A Computational Model of Identity Construction During Political Mobilization
Due to the heterogeneous composition of modern opposition movements, protesters may have two different types of identity: a narrow one associated with their political group or organization or a widespread identity related to the opposition as a whole. Accordingly, the authorities can use two types of repression: broad, directed against the entire opposition, or targeted, focused on its specific segments. How will the authorities’ use of one of two repressive tactics – targeted or broad – affect the structure of participants' identities in a heterogeneous protest movement? How will a change in the structure of identities, in turn, affect the protest turnout – the massive participation of opposition representatives in street actions? To answer these questions, we build an original agent-based model, based on micro-mechanisms of protest participation: on the perception of agents as members of an in-group or an out-group on the first place. Widespread repression can trigger a mechanism of re-categorization, when members of competing opposition parties may begin to perceive each other as ingroup peers. This is the basis for the formation of solidarity and broad opposition identity. As our computational experiments show, these micro-level mechanisms of identity formation provide greater resistance of protesters to widespread repressions than to targeted ones. While more “effective” in terms of undermining common opposition identity, targeted repression is, at the same time, characterized by greater uncertainty in the resulting protest turnout.