Journal on Policy & Complex Systems Volume 3, Issue 2 | Page 157

Policy and Complex Systems
Table 1 . Instance variables for the migrant agents
Migrant Agent Behaviors
The migrant agents have two behaviors : ( 1 ) interact with the state agent , and ( 2 ) update their cooperation strategy to cooperate or deceive . For the interaction with the state , the migrant agent m i simply engages with the state agent and updates their interaction status x i
S and outcome x i
O variables accordingly . The logic for whether m i is allowed entry past the border is controlled by the state agent .
For the updating of the cooperation intention y i
, this behavior is executed at each tick of the model but handled differently for two different subsets of the migrant population M . Note that only migrants who have not yet interacted with the state agent will go through this process of updating their intention . The cooperation intention y i is meant to reflect the strategy that the migrant will follow if and only if they are randomly selected to interact with the state agent . An agent will keep updating their cooperation intention for each time step of the model that they are not interacting with the authorities .
We denote the set of migrants waiting to be processed as M X , where migrants
X
m i
M X have a first-order connection to the migrant agent , m 0
, who just completed their interaction with the state agent . This is , in effect , an abstraction of the individuals that an agent will relate their recent interaction experience to . We either define a first-order connection as being either a member of a Moore neighborhood or linked via the underlying preferential attachment network to the migrant m 0
.
X
For each m i
M X , the agent will count their own first-order connections and imitate the majority strategy of “ cooperate ” or
X
“ deceive .” Each agent m i will also compare themselves to m 0 and employ a user-defined similarity multiplier weight to have the cooperation strategy of m 0
X count more if m i has a similar attribute that could be perceived as a contributing factor to the m 0 being either accepted or rejected by the state . This was intended to capture an assump-
153