PEACE
Population and Environment Analysis for Counter-insurgency
Evaluation
Final Presentation
prepared by
Jason Southerland
Kevin Neary
Brian Kolstad
Steven Darcy
for
Colonel Manago
Center for Army Analysis
PEACE
Herakles and Athena
• Herakles was making his way through a narrow pass. He
saw something that looked like an apple lying on the
ground and he tried to smash it with his club. After
having been struck by the club, the thing swelled up to
twice its size. Herakles struck it again with his club, even
harder than before, and the thing then expanded to such
a size that it blocked Herakles's way. Herakles let go of
his club and stood there, amazed. Athena saw him and
said, “O Herakles, don't be so surprised! This thing that
has brought about your confusion is Aporia
(Contentiousness) and Eris (Strife). If you just leave it
alone, it stays small; but if you decide to fight it, then it
swells from its small size and grows large.” - Aesop,
Fables 534 (from Chambry 129)
PEACE
The Victorious Warrior
“A victorious warrior wins first and then goes
to war, while a defeated warrior goes to
war first and then seeks to win.” -Sun Tzu,
The Art of War
PEACE
Agenda
•
•
•
•
Motivation and Background (Mr. Kolstad)
Data (Mr. Darcy)
Analysis (Mr. Neary and Mr. Southerland)
Observations, Conclusions and Next
Steps (Mr. Neary)
PEACE
Agenda
•
•
•
•
Motivation and Background (Mr. Kolstad)
Data (Mr. Darcy)
Analysis (Mr. Neary and Mr. Southerland)
Observations, Conclusions and Next Steps
(Mr. Neary)
PEACE
Recent Events
• At any time since 1945, there has been at least
one insurgency somewhere in the world.
• In some manner, the Unites States has been
involved in many recent insurgencies.
–
–
–
–
–
Bosnia
Somalia
Afghanistan
Vietnam
Iraq II
PEACE
What Constitutes an Insurgency?
“A struggle between a non–ruling group and the ruling
authorities in which the non–ruling group consciously uses
political resources (e.g. organizational expertise,
propaganda, and demonstrations) and violence to destroy,
reformulate, or sustain the basis of legitimacy of one or
more aspects of politics.”
Insurgency and Terrorism
Bard E. O’Neill
It is the use of violence that distinguishes insurgencies
from other protest movements.
7
PEACE
Background
• Efforts to define the force size and time required to
restore and maintain order in a failed or failing state have
been sporadic and far from complete.
– The United States Army’s “Field Manual 3-24,
Counterinsurgency” specifies a 20:1000 force ratio.
– James T. Quinlivan, in his RAND essay, “Burden of
Victory: The Painful Arithmetic of Stability
Operations,” establishes a 20:1000 force ratio.
– John J. McGrath, “Boots on the Ground: Troop
Density in Contingency Operations,” recommends a
13.26:1000 force ratio.
• Defining the correct force size should result in less risk to
troops and greater chance of success.
PEACE
Project Goal
• The goals of this project are to:
– Find a relationship between Troop Density
and Violence in a counterinsurgency
– Find a method of predicting violence in an
insurgency
PEACE
Project Scope
• Provide an expansion upon the limited
scope of counterinsurgency troop density
studies
• Provide a means of framing a strategist’s
troop density decision by identifying the
key variables that define the operating
environment
PEACE
Project Evolution
• O’Neill classification of an insurgency
– “Can the O’Neill classifications be described
statistically?”
• Using Bayesian networks to tie troop levels to
the overall outcome of the insurgency
– “How many troops does it take to win?”
• Using Bayesian networks to tie troop levels to
violence
– “How many troops does it take to reduce violence?”
• Analysis of troop and violence data
– Which metrics are the best predictors of future
violence?
PEACE
Agenda
•
•
•
•
Motivation and Background (Mr. Kolstad)
Data (Mr. Darcy)
Analysis (Mr. Neary and Mr. Southerland)
Observations, Conclusions and Next
Steps (Mr. Neary)
PEACE
Description of the Dataset
•I Code
•Case Name
•Calendar Year
•O’Neill Classification
•Strategic Approach
•Primary Terrain Type
•Percent Urban Population
•Indigenous Government
•Government
•Rules of Engagement of
Intervening Force
•Degree of outside support
for insurgency
•Structure of the insurgency
•Developed Nation
•Political concept
•Total # Troops
•Total # Intervening
•Total # Indigenous
•Total # Ind Military
•Total # Police
•Total Troop Density
•Intervening Troop Density
•Indigenous Force Density
•Indigenous Military Density
•Indigenous Police Density
•Country
•Population
•Incidents
•Incidents per 1000 population
PEACE
Insurgencies Used
I Code
Case Name
I Code
Case Name
1
Aden (1963-1967)
39
Mozambique (1964-1974)
4
Algerian War (1954-1962)
41
Namibia (1966-1989)
7
Argentina & The Dirty War (1969-1983)
42
Northern Ireland (1968-1998)
8
Borneo (1963-1966)
43
Oman (1957-1959)
9
Cabanas Insurgency in Mexico (1967-1974)
46
Peacekeeping in Lebanon (1990-present)
12
Chad Civil War, 1965-1969
47
Peacekeeping in Liberia (1990-1997)
15
Colombian Civil War (1964-present)
51
Portuguese Guinea (1963-1974)
16
Contras in Nicaragua (1981-1990)
52
Puerto Rican Independence Movement (1950-1954)
18
Cyprus (1955-1959)
53
Rhodesia I (1966-1970)
20
El Salvador (1979-1992)
55
Sandinistas in Nicaragua (1967-1979)
22
First Intifada (1987-1993)
57
Second Intifada (2000-2005)
25
Guatemala (1960-1996)
59
Shining Path in Peru (1980-1999)
26
Guevara Guerilla Campaign in Bolivia (1966-1967)
64
Tupamaro Insurgency in Uruguay (1963-1973)
27
Hamas War (2006-present)
66
UK in Palestine (1944-48)
28
Hezbollah War (2006-present)
80
Vietnam I (1957-1960)
30
Indonesia (1945-1949)
81
Vietnam II (1961-1964)
33
Kashmir Insurgency (1988-present)
82
Vietnam War (1965-1973)
35
La Menos Violencia (1958-1964)
83
Yemen (1962-1970)
37
Malaya (1948-1960)
Statistical Summary of
Numeric Data
PEACE
PEACE
Agenda
•
•
•
•
Motivation and Background (Mr. Kolstad)
Data (Mr. Darcy)
Analysis (Mr. Neary and Mr. Southerland)
Observations, Conclusions and Next Steps
(Mr. Neary)
PEACE
Data Analysis
•
Initial Data Analysis began
with
– Scatterplot
– Histogram
•
Zoom–in
Not strong relationship
between troop density and
violent incidents
– Confirmed with Correlation
matrix
•
Transform Violence and Zoom–in
•
•
Re–scaled violence by taking
its logarithm
Again, not a strong
relationship evident in the
scatterplot
Expanded data set to include
categorical variables
PEACE
Violent Incidents Model
• Conducted stepwise linear
regression of the data
• Some of the variables of interest
–
–
–
–
–
Intervening Troop Density
Indigenous Police Density
Percent Urban Population
Degree of Outside Support (Some)
Primary Terrain Type (Foliated
Mountains)
– O’Neill Classification (3)
• This model provides reasonable fit to
the data, but is not necessarily
insightful to changes in violence
PEACE
MARS Analysis
• Multivariate non-linear regression
– Approximates non-linearity using piecewise linear functions
• Will a non-linear model fit our data well and still be useful
for predictive analysis?
– Find “reasonable” model
– Test model using set-aside data
• Target Variable: Natural Logarithm of Violent Incidents
per 1,000 population
– Continuous Predictor Variables: Intervening Troop Density,
Indigenous Troop Density, Percent Urban Population, Natural
Logarithm of Previous Relative Violence Rate
– Categorical Predictor Variables: O’Neill Classification, Degree of
Outside Support for Insurgency, Insurgent Strategic Approach,
Counter-Insurgent Rules of Engagement
PEACE
MARS Model
LogViolence = -7.312 + .704*B1 - .0206*B5 - .747*B6 + .854*B8
Variable in Basis Function
Functional Form
Natural Logarithm of Violence
From Previous Year (B1)
max(0, PrevLog + 8.822)
Indigenous Troops per 1,000
population (B5)
max(0, 5.754- Indigenous Troop Density)
Percent of National Population
that is Urban (B6)
max (0,Percent Urban Population - 80.86)
Foreign Support (B8)
Foreign Support = 2
PEACE
MARS Model
Values of Predictors for Which Basis Functions are Non-Zero
•Natural Logarithm of Previous Violence >= -8.822
•Indigenous Troop Density <= 5.754
•Percent Urban Population >= 80.86
•Foreign Support = 2: Support from foreign entities that falls short of
contribution of troops; may include: money, materiel, training and safe haven
Relationship of Values of Predictor Variables to Violence
•As Previous Violence increases, violence increases
•As Indigenous Troop Density increases, up to 5.754 troops per 1,000
population, violence increases
•As Percent Urban Population increases, violence decreases
•Foreign Support, short of foreign fighter involvement, increases violence
PEACE
Testing the Model
Max Error
Min Error
Mean Absolute Error
Mean Square Error
1.683491796
-2.778205633
0.646543367
0.885780614
PEACE
Analyzing the Residuals
y = 0.8719x
Q-Q Plot of Residuals
2
R = 0.8181
3
Residuals
2
1
0
-3
-2
-1
-1 0
1
2
3
-2
-3
-4
Comparison of Target Variable and Associated
Residuals
2
1
0
-8
-6
-4
-2
-1 0
-2
-3
-4
2
4
PEACE
Agenda
•
•
•
•
Motivation and Background (Mr. Kolstad)
Data (Mr. Darcy)
Analysis (Mr. Neary and Mr. Southerland)
Observations, Conclusions and Next
Steps (Mr. Neary)
PEACE
Observations
• Troop density was positively correlated to the
natural log of violent incidents.
– Does the apple really get bigger?
– More troops means more targets?
– Who is counting the incidents?
• Mean Troop Density is 39.5, much higher than
the 20 from Quinlivan or 13.26 by McGrath, but
Median Troop Density is 11.87
• Does aggregating violence over years and
countries provide sufficient fidelity?
PEACE
Conclusions
• Best predictor of Violence is previous
year’s Violence
• High Urban populations are less
susceptible to increasing violence
• Material foreign support of an insurgency
increases violence
• Whether troops inspire violence or
violence brings more troops is unresolved,
but bears further study
PEACE
Next Steps
• Expand the scope of the analysis to
include political, economic, diplomatic and
other factors
• Get more data
• Explore the individual nature of each
insurgency
• How do you achieve victory BEFORE
undertaking a counter-insurgency role?