Questions

1
(No Transcript)
2
Questions

• Literature has contradictory findings on
  repression. Some argue repression decreases
  dissident activity, others argue that repression
  increases rebellion
• What are the effects of government activity on
  dissident movements?
• How does dissident activity affect government
  response to dissident groups?

3
Purpose

• Seek to explain combinations of various empirical
  findings using single theoretical framework
• Examine the strategic behavior between
  governments and dissident groups.
• Identify conditions under which two parties
  respond to each other both cooperatively and
  violently in a single theoretical framework.

4
Government and Dissident Action Set
5
Action costs, Opponent Costs, and Budget

• Both government and dissidents are subject to
  costs for taking action and costs imposed by
  their respective opponent.
• Both sides have finite amount of budget to pay
  the costs of action and absorb the costs imposed
  by their respective opponent.

6
Constraints on Budget

• Internal Constraints The ability of both the
  government and the dissidents to act is
  constrained by forces internal to these
  organizations.
• External Constraints By taking action against
  each other, both the government and the
  dissidents constrain each others ability to take
  action.

7
Government and Dissident Action Set
8
General Hypothesis 1 2
9
Context Hypotheses
10
Predicting Government and Dissident Behavior
11
Research Design Cases

• Venezuela 1987-1992 (IPI)
• Chile 1983-1992 (IPI)
• Israel 1979-2002 (KEDS)
• Afghanistan 1990-1999 (KEDS Central Asia)
• Cambodia 1975-2000 (TABARI currently collecting
  the data with NSF funds)

12
Case Aggregation
Observation of Events
Theoretical Specification Sequences
13
Data and Method

• Intranational Political Interactions (IPI), and
  the Kansas Event Data Systems (KEDS). Event data
  are "day-by-day coded accounts of who did what to
  whom as reported in the open press," and
  according to Goldstein (1992, 369).
• For models with autocorrelated disturbances,
  estimate ARIMA models. For models without
  evidence of autocorrelated disturbances, OLS
  estimates are used.

14
Results from Afghanistan
15
Conclusion

• Evidence suggests strategic aspect to government
  and rebel decision-making
• Future development of the project could move to
  questions of how to translate the current model
  into a game theoretic model
• Other potential directions may include complex
  system modeling.