Discord
@hash/agenda-seeding
Simulation
1
0
Public

Agenda Seeding

Simulation inspired by the paper "Agenda Seeding: How 1960s Black Protests Moved Elites, Public Opinion and Voting". It implements the multi-round decision game between protestors, the police, the media, and the public that Professor Wasow describes.

Overview

Simulation inspired by the paper Agenda Seeding: How 1960s Black Protests Moved Elites, Public Opinion and Voting.

We implement the multi-round decision game between protestors, the police, the media, and the public that Wasow describes. We then experiment with adding agent level behaviors, in particular those inspired by a 2011 paper on the London protests that had protestors interact and persuade one another to adopt more or less intense positions.

Click the plots tab to view graphs that show changes in public opinion and composition over time.

Our goal was to start exploring the space of protest models and the decision to use violence or non-violence in a protest.

Parameters in this model

  • num_protestors: The number of protestors in the simulation
  • protestors parameters:
    • attitude_threshold: threshold for whether protestor gains a new attitude role.
    • viol_threshold: threshold for whether protestor votes for violence
    • role_division: an array where the first number determines percentage of "hardcore" protestors (RNG lower than this number), the second determines number of "hangeron" protestors (RNG > than this number) and the enthusiasts are the percentage in between those two thresholds
    • vote_percentage: if the number of protestors voting violence divided by the total number of protestors is greater than this, they vote violence, otherwise they vote non-violence.
    • join_rate: rate at which new protestors join based on positive media coverage.
    • leave_rate: rate at which new protestors leave based on negative media coverage
  • police parameters:
    • viol_threshold: If RNG > than this, police vote violence.

Extension ideas

  • Expand the number of police decision makers
  • Add in more complicated movement behaviors
  • Add more media agents

Agents

Protestors

  • Protestors act first, and decide if they are going to protest violently or non-violently. They individually decide by RNG modified by their attitude towards violence and the police actions in the previous round (if the police were violent, more likely the protestor will be violent). The group weighted choice is submitted and, if above a parameter threshold, will vote violence, otherwise non-violence.
    • Send a message to the Round Manager agent with the protestors choice

Police

  • Police act second - their choices depend on the protestors choice. If the protestors were violent, more likely the police will be violent.

Media

  • Media receives data about Protestors and about Police.
    • Media determines issue framing and agenda setting
    • The choices of the protestors and the police push on the agenda setting and the issue framing. If both choose violence, there is a discontinuous jump in agenda setting.

Public Opinion

  • Based on media coverage, the public changes their opinion of the protests

Round Manager

  • An agent is used to coordinate the voting and dispersal of information among the agents.