Investigating Violent NonState Actors

1
IMPORTANCE Why This Matters
LITERATURE Why We Have Contradiction
THEORY What Were Bringing Together
HYPOTHESES The Questions Were Asking
DATA What it is and How its Structured
MODEL Do the Math
RESULTS What We Found
RESEARCH CONCLUSION What it Means
POLICY RAMIFICATIONS What to Do
Investigating Violent Non-State Actors in the
Middle East J. Bernhard Compton Claremont
Graduate University
2
IMPORTANCE
Only states matter? Individual non-state actors
now precipitate major changes in national and
foreign policy. If nations are to rationally set
policy that deals specifically with non-state
actor perpetrated violence, then policy levers
must be specifically related to necessary and
sufficient causes and motivators of non-state
violent actors.
3
LITERATURE

• Contradictory Ideas, Contradictory Findings
• Motivation and Purpose
• Alter local status quos Pape (2003)
• Alter international status quos Lind (2004),
  Thomas Casebeer (2004)
• Conflation of Definitions
• Civil Wars Benson Kugler (1998)
• Insurgencies Civil Wars Fearon Laitin
  (2002), Collier Hoeffler (2001)
• Suicide Terror Pape (2003)
• Conspiracies, Mutinies, Assassinations, Guerilla
  Wars Gurr (1970)
• Terrorism
• Morass of Findings
• Social Pressures Ross (1993)
• Overtakings Benson Kugler (1998)
• Accumulation of Protracted Conflicts Fearon
  Laitin (2002)
• Economic Variables Collier Hoeffler (2001)

4
THEORY

• Defining the Actor, Choosing the Relevant PoVs.
• Transnational Violent Non-State Actor
• Violence committed by individual or group to
  affect political change within or across
  countries.
• Picking Related Concepts
• In order to perpetrate violence, non-state
  actors must have opportunity and cause to do so.

5
THEORY II

• Defining the Relationships
• Opportunity
• In order for an actor to form a group purposed
  with committing violence, there must exist
  opportunity for such a group to arise.
• Factors of opportunity include weak governance,
  lower levels of income, and unstable
  environments.
• Cause
• Less ethnically and religiously homogenous
  societies have greater internal friction.
• Two ideas are pursued
• Local grievances such as poverty and perceived
  resource inequality cause identity cleavage.
• Combined with perception that governance lacks
  legitimacy.

6
HYPOTHESES
H1 Transnational terror groups originate from
countries with low levels of RPC, lower levels of
percapita GDP, and higher levels of instability.
H2 Countries of origin will have high levels
of ethnic and religious (but not linguistic)
homogeneity, and have higher levels of state
power and perceived unequal distribution of
resources.
7
DATA

• Data Sources
• International Terrorism Attributes of Terrorist
  Events (ITERATE)
• World Development Indicators
• Witches Brew Homogeneity Dataset
• Relative Political Capacity (RPC Dataset, as
  developed for POFED)
• Modus Operandi
• Disparate datasets brought into unified database
  and given common country codes
• SQL Queries written to data-mine specific dataset
  for this research
• Data Sets (created through data-mining process)
• All countries, unbalanced panel, entries without
  full data are dropped in estimation

8
DATA II
Non-Linearity in the Data
9
MODEL
Hypotheses are specified in the following
model NumIncidentsaß1(RPC4)ß2(Instability)ß3
(LogPCGDP)ß4(LogPCGDP2)ß5(LogPower) ß6(Oil)
ß7(EthnicHom) ß8(RelHomog) ß9(LingHomog)ß10(Gi
ni) Regression Method Zero-Inflated
Poisson Wooldridge test showed no serial
correlation problems. Robust standard errors
used in estimation process for heteroskedasticity.
Vuong test indicates ZIP best regression method.
10
RESULTS
General Model DV No. of Incidents Model 1 Model
2 Model 3 Opportunity Cause Full RPC4
-.7566802 -1.096231
(.2006164) (.4173566) Instability
.0000106 .0000007
(.0000005) (.0000005) LogPCGDP 9.149384
36.11765 (4.338657)
(11.60857) LogPCGDP2 -.5489765 -2.199297
(.2562721)
(.7056974) LogPower .3641231 .6941174
(.1005752) (.1341911) Oil -.
7397416 -1.42049
(.2916303) (.3199254) LingHomog .0122943
.0184953 (.0091414) (.0102433)
EthHomog -.0560409 -.0661368
(.010812) (.0174343) RelHomog .082011
-.1614689
(.0176442) (.0413361) Gini .0269573
-.0773689
(.025537) (.0415744) Constant
-36.79535 -3.942645 -125.1074
(18.30926) (2.013877) (43.97879)
Pseudo R2 .141 .215 .322 chi2 34.60808
47.60907 132.8119 p .00000006
.000000001 1.25e-23 N 368 368 368
plt0.10, plt0.05, plt0.01
11
(No Transcript)
12
(No Transcript)
13
RESEARCH CONCLUSION

• Quite strong support was given to the notion that
  countries of origin have lower levels of RPC and
  higher levels of power. Violent incidents rise
  with increasing percapita GDP to a point, then
  decline.
• GINI index runs counter to percapita GDP. This
  fits with the notion that industrializing nations
  initially have disparate wealth distributions.
  Middle Eastern countries are generally considered
  to be industrializing.
• In conjunction with the significance of
  homogeneity in society, less ethnically and
  religiously homogenized societies have higher
  rates of incidents. However, more linguistically
  homogeneous societies have higher incidents. To
  disagree, we need to speak the same language.

14
POLICY RAMIFICATIONS

• Because the effect of RPC increases as power
  increases, industrializing countries should seek
  to maximize their strength of governance as
  quickly as possible.
• Countries that have ethnic and religious
  fractionalization but speak a common language are
  at a higher risk.
• Instability plays little role in the frequency of
  non-state actor terror violence, which may
  indicate that crack-downs and other strong arm
  tactics may be effective.
• Terror violence occurs most often in the early to
  mid stages of rising percapita income.
  Demonstrating more equitable distribution of
  resources may dampen the number of attacks.

15
Questions and Comments are Welcomed. Thank you.