CHILDREN AND FAMILIES The RAND Corporation is a nonprofit institution that

EDUCATION AND THE ARTS helps improve policy and decisionmaking through
ENERGY AND ENVIRONMENT research and analysis.
HEALTH AND HEALTH CARE
This electronic document was made available from
INFRASTRUCTURE AND www.rand.org as a public service of the RAND
TRANSPORTATION
Corporation.
INTERNATIONAL AFFAIRS

LAW AND BUSINESS

NATIONAL SECURITY Skip all front matter: Jump to Page 16

POPULATION AND AGING

PUBLIC SAFETY

SCIENCE AND TECHNOLOGY

Support RAND
TERRORISM AND
Purchase this document
HOMELAND SECURITY Browse Reports & Bookstore
Make a charitable contribution

For More Information

Visit RAND at www.rand.org
Explore the RAND National Security
Research Division
View document details

Limited Electronic Distribution Rights

This document and trademark(s) contained herein are protected by law as indicated
in a notice appearing later in this work. This electronic representation of RAND
intellectual property is provided for non-commercial use only. Unauthorized posting
of RAND electronic documents to a non-RAND website is prohibited. RAND
electronic documents are protected under copyright law. Permission is required
from RAND to reproduce, or reuse in another form, any of our research documents
for commercial use. For information on reprint and linking permissions, please see
RAND Permissions.
This report is part of the RAND Corporation research report series.
RAND reports present research findings and objective analysis that
address the challenges facing the public and private sectors. All RAND
reports undergo rigorous peer review to ensure high standards for re-
search quality and objectivity.
Countering Others’
Insurgencies
Understanding U.S. Small-Footprint
Interventions in Local Context

Stephen Watts, Jason H. Campbell, Patrick B. Johnston,

Sameer Lalwani, Sarah H. Bana

C O R P O R AT I O N
 N AT I ONAL SECUR I T Y R ESE AR C H DIV IS IO N

Countering Others’
Insurgencies
Understanding U.S. Small-Footprint
Interventions in Local Context

Stephen Watts, Jason H. Campbell, Patrick B. Johnston,

Sameer Lalwani, Sarah H. Bana

Sponsored by the Smith Richardson Foundation

The research described in this report was sponsored by the Smith Richardson
Foundation and conducted within the International Security and Defense
Policy Center of the RAND National Security Research Division.

Library of Congress Cataloging-in-Publication Data is available for this publication.

ISBN: 978-0-8330-8404-0

The RAND Corporation is a nonprofit institution that helps improve policy

and decisionmaking through research and analysis. RAND’s publications do
not necessarily reflect the opinions of its research clients and sponsors.
Support RAND—make a tax-deductible charitable contribution at
www.rand.org/giving/contribute.html

R® is a registered trademark
Cover image: Pakistani soldiers place barbwire to secure a street during a curfew
in Rawalpindi on Nov. 17, 2013 (Reuters/Faisal Mahmood).

© Copyright 2014 RAND Corporation

This document and trademark(s) contained herein are protected by law. This
representation of RAND intellectual property is provided for noncommercial use only.
Unauthorized posting of RAND documents to a non-RAND website is prohibited.
RAND documents are protected under copyright law. Permission is given to duplicate
this document for personal use only, as long as it is unaltered and complete. Permission
is required from RAND to reproduce, or reuse in another form, any of our research
documents for commercial use. For information on reprint and linking permissions,
please see the RAND permissions page (www.rand.org/pubs/permissions.html).

RAND OFFICES
SANTA MONICA, CA • WASHINGTON, DC
PITTSBURGH, PA • NEW ORLEANS, LA • JACKSON, MS • BOSTON, MA
DOHA, QA • CAMBRIDGE, UK • BRUSSELS, BE
www.rand.org
Preface

This report documents the results of a project that assessed the reasons
why counterinsurgents in the developing world so often adopt counter-
insurgency strategies and practices at odds with U.S. preferences and
how the United States can influence its potential partners’ choices in
these conflicts.
The report is written for a general audience, although it is likely
to be of particular interest to those in the foreign policy and defense
communities, both civilian and military, concerned with counterinsur-
gency and security-sector assistance.
This research was sponsored by the Smith Richardson Foundation
and conducted within the International Security and Defense Policy
Center of the RAND National Security Research Division (NSRD).
NSRD conducts research and analysis on defense and national secu-
rity topics for the U.S. and allied defense, foreign policy, homeland
security, and intelligence communities and foundations and other non-
governmental organizations that support defense and national security
analysis.
For more information on the International Security and Defense
Policy Center, see http://www.rand.org/nsrd/ndri/centers/isdp.html or
contact the director (contact information is provided on the web page).

iii
Contents

Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi
Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii

CHAPTER ONE
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

CHAPTER TWO
Counterinsurgency in Comparative Perspective.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Current American Counterinsurgency Thought. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Varieties of Counterinsurgency.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Bounded Accommodation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Discriminate Violence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Provision of Public Goods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
A Typology of Counterinsurgency Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Determinants of Counterinsurgency Strategy and Practice.. . . . . . . . . . . . . . . . . . 21
Political Inclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
State Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Military Superiority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Measures and Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

v
vi Countering Others’ Insurgencies

CHAPTER THREE
Quantitative Analysis of Counterinsurgency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Accommodation of the Reconcilable Opposition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Political Inclusivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
State Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Military Sufficiency.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Discriminate Use of Violence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Political Inclusivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
State Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Military Sufficiency.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Integrating Quantitative Findings on Counterinsurgent Strategies.. . . . . . . . . 52
Conflict Termination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Discriminate Use of Violence.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Using Quantitative and Qualitative Analysis to Reinforce One
Another. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

CHAPTER FOUR
Counterinsurgency in the Philippines.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Overview of the Counterinsurgency and Counterterrorism
Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Insurgent and Terrorist Organizations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Government Forces and Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
The Role of the United States .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Analysis of Philippine Counterinsurgency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Political Inclusion, Accommodation, and Restraint. . . . . . . . . . . . . . . . . . . . . . . . . 81
Military Superiority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
State Capacity and Counterinsurgency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Analysis of U.S. Assistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Capacity-Building and Enhanced Civilian Oversight.. . . . . . . . . . . . . . . . . . . . 104
The U.S. Role: Limitations and Risks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
 Contents vii

CHAPTER FIVE
Counterinsurgency in Pakistan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Context.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Conflict Narrative. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Analysis of Pakistan’s Counterinsurgency.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Political Inclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
State Capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Military Superiority. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Analysis of U.S. Assistance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Effects of U.S. Military Assistance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Effects of U.S. Civil Assistance.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
U.S. Efforts to Influence Pakistan and Their Consequences. . . . . . . . . . . . . 167
Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

CHAPTER SIX
Managing Troubled Partnerships. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Approaches to Managing Troubled Partnerships.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Selectivity and Avoiding “Bad” Counterinsurgents.. . . . . . . . . . . . . . . . . . . . . . . 176
Conditionality and Making Partners into “Good”
Counterinsurgents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Policy Prescriptions for Managing Troubled Partnerships. . . . . . . . . . . . . . . . . . 185
General Principles.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Bounded Accommodation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Regulated Violence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

APPENDIX
Coding Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Figures

S.1. Counterinsurgent Regimes and Conflict Outcomes. . . . . . . . . . . . xvi

S.2. Mass Killings by Counterinsurgent Regimes.. . . . . . . . . . . . . . . . . . . xvii
2.1. Typology of Counterinsurgency Models. . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.1. Relationship Between Democracy and Negotiated
Settlements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2. Relationship Between Democracy and Acceptable
Decisive Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.3. Relationship Between Democracy and Durable Peace. . . . . . . . . . . 39
3.4. Relationship Between State Reach and Acceptable
Decisive Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.5. Relationship Between Government Effectiveness and
Acceptable Decisive Outcomes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.6. Relationship Between State Reach and Indeterminate
Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.7. Relationship Between Democracy and Mass Killings of
Civilians by Government Forces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.8. Relationship Between Government Effectiveness and
Mass Killings of Civilians by Government Forces.. . . . . . . . . . . . . . . 49
3.9. Relationship Between Force-to-Force Ratio and Mass
Killings of Civilians by Government Forces.. . . . . . . . . . . . . . . . . . . . . 50
3.10. Relationship Between Force-to-Population Ratio and
Mass Killings of Civilians by Government Forces.. . . . . . . . . . . . . . . 51
3.11. Conflict Outcomes by Regime Type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.12. Durable Peace by Regime Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.1. Insurgent-Group Strength by Administration. . . . . . . . . . . . . . . . . . . . 89
4.2. Time Trend in PSF and NPA Numeric Strength and Force
Ratio, 1978–2006.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

ix
x Countering Others’ Insurgencies

4.3. ASG Estimated Strength, 1993–2003. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

4.4. Development Projects, Insurgent-Affected Barangays,
and Guerrilla Fronts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.1. Insurgent Incidents Over Time in Pakistani Regions,
2007–2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
5.2. Manpower and Operational Tempo in the Swat
Valley/Malakand Region, 2007–2012.. . . . . . . . . . . . . . . . . . . . . . . . . . . 137
5.3. The Balochistan Insurgency, 2004–2013. . . . . . . . . . . . . . . . . . . . . . . . 140
Tables

2.1. Determinants of Counterinsurgency Strategy.. . . . . . . . . . . . . . . . . . . 28

2.2. A Typology of Counterinsurgent Regimes.. . . . . . . . . . . . . . . . . . . . . . . . 31
3.1. Conflict Outcomes in the Post–Cold War Era.. . . . . . . . . . . . . . . . . . 36
3.2. Counterinsurgent Regimes and Conflict Outcomes. . . . . . . . . . . . . 53
3.3. Counterinsurgent Regimes and Mass Killings. . . . . . . . . . . . . . . . . . . . 58
4.1. Regime and Campaign Features of Philippine
Counterinsurgency, 1972–2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.2. AFP Unit Type and Conflict Deaths per Reported
Incident, 2001–2007. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.1. FATA and Malakand Conflict Deaths per Engagement,
2002–2012.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
5.2. Regime and Campaign Features of Pakistan
Counterinsurgency, 2001–2013.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
5.3. Estimates of Use of Indiscriminate Force in Different
Counterinsurgency Theaters, 2009–2012.. . . . . . . . . . . . . . . . . . . . . . . 148

xi
Summary

With the United States exhausted by more than a decade of war and
facing severe fiscal limitations, decisionmakers are striving to place
American defense policy on a more sustainable footing. Central to
this effort is a commitment to work through partner nations wherever
possible, providing support to countries with which the United States
shares interests or values while also ensuring that the primary responsi-
bility for these nations’ security remains their own. Thus the document
that currently guides U.S. defense policy states:

Building partnership capacity elsewhere in the world . . . remains

important for sharing the costs and responsibilities of global lead-
ership. Across the globe we will seek to be the security partner
of choice, pursuing new partnerships with a growing number
of nations. . . . Whenever possible, we will develop innovative,
low-cost, and small-footprint approaches to achieve our security
objectives.1

This emphasis on partnership strategies is particularly central to U.S.

efforts to manage the “security externalities” of fragile and conflict-
affected states. Such states increase the risk of spillover conflicts
throughout the region in which they occur, damage the economies of
neighboring states, contribute to flourishing transnational crime net-
works, spread pandemic disease, and foster transnational terrorism.

1 U.S. Department of Defense, Sustaining U.S. Global Leadership: Priorities for 21st Century
Defense, Washington, D.C.: Department of Defense, January 2012, p. 3.

xiii
xiv Countering Others’ Insurgencies

When pursuing partnerships with regimes that are fighting insur-

gencies, however, the likelihood of success using such “small-footprint”
approaches is inextricably bound with the local context and particu-
larly the nature of the partner government. Too often the importance
of local context is minimized in discussions of U.S. security strategy,
particularly in relation to the problems of fragile states. Proponents of
small-footprint and indirect approaches overwhelmingly cite as models
the recent U.S. operations in the Philippines and Colombia—without
providing any indication of the generalizability of these models.
This study seeks to understand the extent to which the “success
stories” of U.S. partnerships such as those with the Philippines and
Colombia can be generalized—or, phrased differently, to understand
the conditions under which the small-footprint model is likely to suc-
ceed in bringing an end to an insurgency that both the United States
and its partner seek to combat. It explores how local circumstances
shape the “art of the possible” in such partnerships and how the United
States can best maximize the potential and minimize the risks of these
often uneasy alliances. More specifically, the study asks three central
questions:

• Why do counterinsurgents adopt particular counterinsurgency

strategies and practices?
• What are the likely consequences of these strategies, in terms of
conflict outcomes and civilian casualties?
• When the United States finds a partner government’s counter-
insurgency strategy and practices problematic, what can it do to
influence its partner’s actions to improve the chances of a favor-
able outcome?

Research Findings

The answers to these questions provided in this report are derived from
a mixed-method research design incorporating both quantitative and
qualitative analysis. Simple statistical analyses are applied to a dataset
of counterinsurgencies that have terminated since the end of the Cold
 Summary xv

War (72 in all) to understand the manner in which they were fought
and how they terminated. This broad overview frames more in-depth
analyses of two important recent cases of U.S. partnerships, the Philip-
pines and Pakistan, drawing on secondary literature, a wide variety of
quantitative data sources, and interviews conducted with several dozen
government officials, military officers, and civil society actors in the
Philippines, Pakistan, and the United States.
The report finds that the counterinsurgency strategies and prac-
tices adopted by regimes fighting rebellions are strongly shaped by the
characteristics of these regimes—in particular, the degree to which
they are politically inclusive and the extent of state capacity they pos-
sess. “Success stories” like the Philippines and Colombia have occurred
in countries characterized by relatively inclusive politics and reason-
able levels of state capacity. The governments of such countries typi-
cally adopt strategies that approximate the Western model of counter-
insurgency, often (misleadingly) referred to as the “hearts-and-minds”
approach. Unfortunately, only approximately one insurgency in eight
occurs in such best-case countries. The majority of rebellions take place
in worst-case conditions—that is, in countries that lack both inclusive
politics and state capacity. Regimes in this latter category are prone to
relying on blunt applications of military force to contain or suppress
rebellion.
The quantitative analysis conducted in this study paints a stark
picture of the different trajectories that conflicts follow in these best-
case and worst-case environments. As shown in Figure S.1, only 13
percent of civil wars in the best-case environments fail to reach an out-
come that the government finds acceptable (that is, either outright mil-
itary victory or a negotiated settlement acceptable to both sides); the
failure rate is nearly five times as high (60 percent) in the worst-case
environments. Non-inclusive regimes are much more likely to suffer
outright defeat than are more-inclusive ones. Weak regimes are much
more likely to experience indeterminate ends to their conflicts, where
insurgents retain their capabilities and de facto control over parts of
the country.
xvi Countering Others’ Insurgencies

Figure S.1
Counterinsurgent Regimes and Conflict Outcomes

100

80 Insurgent victory
Outcome (percentage)

Military stalemate,
60 no settlement
Military stalemate,
40 settlement
Government
victory
20

0
Low High Low High
inclusiveness inclusiveness inclusiveness inclusiveness
Low-capacity High-capacity
government government
RAND RR513-S.1

Similarly, as shown in Figure S.2, whereas fewer than 10 percent

of all politically inclusive regimes (high- and low-capacity combined)
have resorted to indiscriminate violence as a tool of counterinsurgency,
39 percent of regimes that are less inclusive have used such tactics.
In other words, the chances of wide-scale abuse by security forces are
four times greater among less-favorable types of regimes than they are
among the more-favorable types.
Case studies of the Philippines and Pakistan broadly support
these quantitative findings. In both the Philippines and Pakistan, rela-
tively more-democratic governments were more likely to adopt a clas-
sical counterinsurgency model that sought accommodation with the
reconcilable opposition and used violence relatively discriminately.
This tendency was particularly pronounced in regions where the gov-
ernments possessed the necessary civil capacity to implement hearts-
and-minds approaches. In contrast, during periods when these coun-
tries were ruled by more-autocratic governments and in regions where
the state exercised little effective control, the governments were much
 Summary xvii

Figure S.2
Mass Killings by Counterinsurgent Regimes

100
Percentage of cases with mass
killings by government forces

80

60

40

20

0
More-democratic Less-democratic
governments governments
RAND RR513-S.2

more likely to seek to contain and suppress insurgents through raw

force.
The successes of U.S. operations in countries such as the Philip-
pines appear to have more to do with the partner nation than they do
with U.S. policies. This is not meant to deny the importance of U.S.
assistance; to the contrary, the case study of the Philippines suggests
that U.S. aid played a critical role in that country’s recent successes.
But similar U.S. policies with less-promising partner nations should
not be expected to produce anywhere near the same levels of success.
And as the quantitative analysis in this study reveals, the large majority
of potential U.S. partner nations—including many that are central to
ongoing U.S. counterterrorism efforts—are much less promising.
Of course, the fact that more than half of all insurgencies occur
in countries governed by the least-favorable type of regime does not
mean that more than half of all U.S. military interventions are likely
to occur in such countries. Indeed, the historical record suggests that
a disproportionate share of U.S. military interventions—somewhere in
the vicinity of half—occur in best-case environments. This propensity
might be explained by the fact that the odds of success are higher in
such countries, or it may be that the United States shares many values
xviii Countering Others’ Insurgencies

and interests with such states. It is sobering, however, to consider the

number of least-favorable environments in which the United States
has intervened and the levels of success it has experienced. Roughly
half of U.S. interventions have been in these worst-case environments,
and the record of U.S. interventions in Afghanistan and Iraq is hardly
promising.

Policy Implications

Counterinsurgency is perhaps the most context-dependent activity in

which militaries engage. Consequently, no universal set of policy pre-
scriptions is possible. A few broad rules of thumb can nonetheless be
discerned.
First, where U.S. and partner-nation interests fundamentally
diverge, there is little hope of a productive partnership to combat an
insurgency. The amount of U.S. leverage and information is too small
for conditionality to be used effectively to overcome the divergence.
Second, conditionality can reasonably be used to enforce “red-
lines,” where the United States would be willing to walk away from a
partner if the partner crossed certain thresholds of acceptable behav-
ior, or to press for a limited number of important but narrowly scoped
reforms, or to take advantage of specific moments in time when more
wide-ranging political change is possible. In most cases, however, it is
extremely unlikely that conditionality can be used to press for more-
fundamental transformations of partner nations in accordance with
U.S.-preferred models of counterinsurgency.
Third, given the duration of most contemporary insurgencies and
the length of time it typically takes to build state capacity or institu-
tionalize mechanisms of political inclusion, the United States should
enter into partnerships with the expectation that they will be long-term
and will have relatively low odds of success in the short-to-medium
term. Decisionmakers should carefully weigh their ability to make such
long-term commitments, particularly where potential partner nations
are problematic. With the memories of the September 11, 2001, attacks
fading and public attention no longer monopolized by the wars in Iraq
 Summary xix

and Afghanistan, decisionmakers should expect more critical scrutiny

of such light-footprint engagements.
Beyond these broad rules of thumb, a number of specific mech-
anisms may be useful to manage partnerships with difficult partner
regimes. Such mechanisms include the following:

• The United States can help partner regimes credibly commit to

political compromises with reconcilable elements of the armed
opposition through a variety of instruments, potentially includ-
ing large-scale commitments of foreign aid and, in some contexts,
international peace operations.
• Progress toward greater democracy is normally heavily con-
tested, usually occurring only when incumbents are unable to
resist demands for greater political inclusion. Periods of pro-
longed military stalemate in a civil war may provide such open-
ings. In these cases, political reforms are less about alleviating
popular grievances and winning the general population’s hearts
and minds in order to defeat insurgents. Rather, they are about
providing a framework in which reconcilable opposition leaders
come to believe they can participate with minimal fears of perse-
cution or marginalization. These processes typically do not look
like Western conceptions of “democracy,” at least for many years,
often two decades or more. But they can provide—and often have
provided—a means for ending violent conflicts. Such fragile
political systems require support in their initiation phase, and
they require buttressing to prevent collapse. The United States
and other international partner nations can help in both of these
phases.
• The United States should make the principle of civilian oversight
and other accountability mechanisms central to its security-sector
assistance. As a general rule, the United States should also stress
quality over quantity in developing partners’ security forces. This
finding highlights the importance of imparting the necessary
doctrine, leadership, discipline, and (where appropriate) technol-
ogy to manageable numbers of partner-nation forces and then
sustaining these qualitative improvements. Wherever possible,
xx Countering Others’ Insurgencies

partner-nation units receiving such assistance should be closely

paired with U.S. forces to ensure that the United States has vis-
ibility into how its assistance is being used.
• Security forces that do not include members of the same ethnic
or religious affiliation as the population in which they are operat-
ing are at particularly high risk of abusive behavior. The United
States, therefore, should work with partner regimes to improve
the representativeness of their security services. Unfortunately,
incorporating personnel from disaffected populations during the
course of intensive fighting risks subversion within these forces.
The ideal time to integrate personnel from different communities,
therefore, is before fighting erupts or, if that is not possible, as
early within a conflict as possible.
• Unfortunately, all of the above prescriptions are long-term and
uncertain. Moreover, they are all substantially more difficult to
implement during ongoing fighting than in peacetime. These
challenges suggest that security-sector reform efforts should be a
central element of U.S. “phase-zero,” or peacetime, engagement
strategy, not a peripheral concern or an issue to which significant
resources are devoted only after a crisis erupts.
Acknowledgments

The authors are grateful to Nadia Schadlow of the Smith Richardson

Foundation for her support throughout this study. The study relied
on dozens of interviews with officials, military officers, and ana-
lysts from Pakistan, the Philippines, and the United States. To guar-
antee their anonymity, we cannot thank them by name, but we are
greatly indebted to them for their insights. Special thanks are owed
to COL (Ret.) Joseph Felter for his support and subject-matter exper-
tise on both the United States and the Philippines. We thank Helge
Holtermann and Benjamin Valentino for sharing their quantitative
data with us. Emma Cutrufello provided valuable research assistance
on security issues, and Katherine Pfrommer was extraordinarily help-
ful in providing quantitative and database support. Natalie Ziegler and
Joy Merck provided expert assistance in formatting the draft document
and preparing it for publication. Janet DeLand edited the final docu-
ment and smoothed its many rough edges. We are grateful to Ambassa-
dor James Dobbins for his thoughtful suggestions and guidance in the
early stages of the study. Finally, we thank Daniel Byman, Paul Miller,
and Eric Peltz for their insightful reviews of the manuscript. The final
report was much improved by their suggestions. The authors, of course,
accept full responsibility for any remaining errors.

xxi
Abbreviations

AFP Armed Forces of the Philippines

AJK Azad Jammu and Kashmir
AMISOM African Union Mission in Somalia
ARMM Autonomous Region of Muslim Mindanao
ASG Abu Sayyaf Group
CHCD clear-hold-consolidate-develop
CPP Communist Party of the Philippines
DSO Development Support Operations
EPPA Enhanced Partnership with Pakistan Act
FATA Federally Administered Tribal Areas
FC Frontier Corps
FM field manual
GDP gross domestic product
GoP Government of Pakistan
GRP Government of the Republic of the Philippines
IDP internally displaced person
ISR intelligence, surveillance, and reconnaissance
JI Jemaah Islamiyah
JSOG Joint Special Operations Group

xxiii
xxiv Countering Others’ Insurgencies

JSOTF-P Joint Special Operations Task Force–Philippines

JTF joint task force
KP Khyber-Pakhtunkwa
LGU local government unit
MDA Mutual Defense Agreement
MILF Moro Islamic Liberation Front
MNLF Moro National Liberation Front
NADESCOM National Development Support Command
NGO non-governmental organization
NPA New People’s Army
NWA North Waziristan Agency
OEF-A Operation Enduring Freedom–Afghanistan
OEF-P Operation Enduring Freedom–Philippines
OIF Operation Iraqi Freedom
PRIO Peace Research Institute Oslo
PSF Philippine Security Forces
SLPP Sierra Leonean People’s Party
SOF Special Operations Forces
SSG Special Services Group
SWA South Waziristan Agency
TNSM Tehrik-e-Nifaz-e-Shariat-e-Mohammadi
TTP Tehrik-e-Taliban Pakistan
USAID U.S. Agency for International Development
WDI World Development Indicators
WGI world governance indicator
CHAPTER ONE

Introduction

With the United States exhausted by more than a decade of war and
facing severe fiscal limitations, decisionmakers are striving to place
American defense policy on a more sustainable footing. Central to
this effort is a commitment to work through partner nations wherever
possible, providing support to countries with which the United States
shares interests or values while also ensuring that the primary responsi-
bility for these nations’ security remains their own. Thus the document
that currently guides U.S. defense policy states:

Building partnership capacity elsewhere in the world . . . remains

important for sharing the costs and responsibilities of global lead-
ership. Across the globe we will seek to be the security partner
of choice, pursuing new partnerships with a growing number
of nations. . . . Whenever possible, we will develop innovative,
low-cost, and small-footprint approaches to achieve our security
objectives.1

This emphasis on partnership strategies is particularly central to U.S.

efforts to manage the “security externalities” of fragile and conflict-

1 U.S. Department of Defense, Sustaining U.S. Global Leadership: Priorities for 21st Century
Defense, Washington, D.C.: Department of Defense, January 2012, p. 3. Such a turn in U.S.
defense policy is nothing new. In fact, the United States adopted a remarkably similar policy
as it sought to extricate itself from the Vietnam War. In the speech in which he announced
the “Vietnamization” of the Vietnam War, President Nixon proclaimed a core principle of
what would become known as the Nixon Doctrine: “We shall furnish military and eco-
nomic assistance when requested . . . but we shall look to the nation directly threatened
to assume the primary responsibility of providing the manpower for its defense” (Richard
Nixon, “Address to the Nation on the War in Vietnam,” November 3, 1969, P-691101).

1
2 Countering Others’ Insurgencies

affected states. Such states increase the risk of spillover conflicts

throughout the region in which they occur, damage the economies
of neighboring states, contribute to flourishing transnational crime
networks, spread pandemic disease, and foster transnational terrorism.
Yet U.S. interests in such states are seldom substantial enough to
warrant large-scale interventions.
In many ways, this guidance encapsulates what has become the
dominant sentiment within the U.S. defense community. Many of
the best-known thinkers in this community have advocated “small-
footprint” approaches to stabilizing fragile states over the past several
years, and such thinking has become dominant within the Department
of Defense.2
Such an approach has many advantages. Small investments in a
country’s stability may prevent a crisis and thus the potential need for
much more costly crisis-response measures. In addition, U.S. commit-
ment to small-footprint approaches may send an important signal to
irresponsible partner nations that might be tempted to escalate a latent
conflict or fail to offer opponents reasonable concessions in anticipa-
tion of unlimited American support. Even if a conflict erupts, a com-
mitment to small-footprint approaches helps to ensure that internal
crises do not become “Americanized”—that is, the partner government
is more likely to realize that it cannot “free ride” on an open-ended
U.S. commitment, the armed opposition is less able to frame the con-
flict as one between local patriots and a foreign invader, and American

2 See, for instance, Max Boot and Richard Bennett, “Treading Softly in the Philippines,”
The Weekly Standard, Vol. 14, No. 16, January 5–12, 2009; T. X. Hammes, “Counterinsur-
gency: Not a Strategy, but a Necessary Capability,” Joint Force Quarterly, No. 65, 2nd Quar-
ter, 2012, pp. 48–52; Robert D. Kaplan, “Imperial Grunts,” The Atlantic, October 2005;
David Kilcullen, The Accidental Guerrilla: Fighting Small Wars in the Midst of a Big One, New
York: Oxford University Press, 2009, p. 283; Steven Metz, The Army’s Strategic Role, Carl-
isle Barracks, Pa.: Strategic Studies Institute, Army War College, 2009; Thomas E. Ricks,
“More Salvadors, Fewer Vietnams,” The Best Defense, February 19, 2013; David Tucker and
Christopher J. Lamb, “Restructuring Special Operations Forces for Emerging Threats,” Stra-
tegic Forum, Institute for National Strategic Studies, National Defense University, No. 219,
January 2006. On the internal debates within the Department of Defense, see Nancy A.
Youssef, “Pentagon Rethinking Value of Major Counterinsurgencies,” McClatchy News, May
12, 2010.
 Introduction 3

military personnel, diplomats, and aid workers are more likely adapt
to local circumstances rather than attempting to re-create a partner
nation in the image of the United States.
The likelihood of success using the small-footprint, or indirect,
approach in such circumstances is inextricably bound with the local
context and particularly the nature of the partner government. Where
the partner government is hopelessly corrupt, abusive toward its own
population, or incompetent and ineffective, or where the partner gov-
ernment is pursuing goals that diverge significantly from those of the
United States, the small-footprint approach can become deeply prob-
lematic.3 Generally speaking, the United States faces two risks when
working with and through partner governments fighting counterin-
surgencies and similar conflicts. First, there is the risk that the partner
government will be ineffective, thus potentially embroiling the United
States in a perpetual conflict.4 Second, there is the risk that the partner
government will be abusive, potentially subjecting the United States
to blowback from the populations that have been abused.5 Working
with even highly problematic partners may be the appropriate policy
choice, depending on the U.S. interests at stake and the goals of the
U.S. partnership. But in such cases the United States should enter into
these partnerships with eyes wide open to the likelihood of success and
the potential risks.
Too often, the importance of local context is minimized in discus-
sions of U.S. security strategy, particularly in relation to the problems
of fragile states. Proponents of small-footprint and indirect approaches,
for instance, overwhelmingly cite as models the recent U.S. operations
in the Philippines and Colombia, without ever providing any indica-

3 For recent cautionary observations about the potential of working through partner gov-
ernments, see, for instance, Daniel L. Byman, “Friends Like These: Counterinsurgency and
the War on Terrorism,” International Security, Vol. 31, No. 2, Fall 2006, pp. 79–115; Nathan
Freier, “The New Theology: Building Partner Capacity,” Small Wars Journal, May 27, 2010.
4 See especially Ariel E. Levite, Bruce W. Jentleson, and Larry Berman, eds., Foreign Mili-
tary Intervention: The Dynamics of Protracted Conflict, New York: Columbia University Press,
1992.
5 See, for instance, Kilcullen, 2009; Michael Scott Doran, “Somebody Else’s Civil War,”
Foreign Affairs, Vol. 81, No. 1, January/February 2002, pp. 22–42.
4 Countering Others’ Insurgencies

tion of the generalizability of these models.6 In fact, as this report will

demonstrate, the conditions in the Philippines and Colombia were par-
ticularly favorable for small-footprint operations. The large majority of
potential partner nations present considerably greater challenges to the
United States’ renewed faith in indirect approaches.
This study examines U.S. partnerships with governments fighting
counterinsurgencies—or, more specifically, the conditions under which
the small-footprint model is likely to succeed in bringing an acceptable
end to an insurgency that both the United States and its partner seek to
combat. For the purposes of this analysis, an insurgency is essentially
synonymous with a civil war and is defined as

an armed conflict that pits the government and national army

of an internationally recognized state against one or more armed
opposition groups able to mount effective resistance against the
state; the violence must be significant, causing more than a thou-
sand deaths in relatively continual fighting that takes place within
the country’s boundaries; and the rebels must recruit mostly
locally, controlling some part of the country’s territory.7

The study seeks to understand how local circumstances shape the “art
of the possible” in such contexts and how the United States can best
maximize the potential and minimize the risks of these often uneasy
alliances. More specifically, the study asks three central questions:

• Why do counterinsurgents adopt particular counterinsurgency

strategies and practices?

6 Examples include Boot and Bennett, 2009; Geoffrey Lambert, Larry Lewis, and
Sarah Sewall, “Operation Enduring Freedom–Philippines: Civilian Harm and the Indi-
rect Approach,” Prism, Vol. 3, No. 4, September 2012, pp. 117–135; Linda Robinson,
“The Future of Special Operations,” Foreign Affairs, Vol. 91, No. 6, November-December
2012; Gregory Wilson, “Anatomy of a Successful COIN Operation: OEF-Philippines and
the Indirect Approach,” Military Review, November-December 2006, pp. 2–12; and Paul
Wolfowitz and Michael O’Hanlon, “Plan Afghanistan,” Foreign Policy, October 28, 2011.
7 Michael W. Doyle and Nicholas Sambanis, Making War and Building Peace, Princeton,
N.J.: Princeton University Press, 2006, p. 31.
 Introduction 5

• What are the likely consequences of these strategies, in terms of

conflict outcomes and civilian casualties?
• When the United States finds a partner government’s counter-
insurgency strategy and practices problematic, what can it do to
influence the partner’s actions to improve the chances of a favor-
able outcome?

The report’s answers to these questions are derived from a mixed-

method research design incorporating both quantitative and qualita-
tive analysis. In Chapter Two, the policy and academic literature on
counterinsurgency is reviewed to determine core principles of coun-
terinsurgency and key areas in which counterinsurgency practice dif-
fers among those who practice it. From this discussion, a typology of
counterinsurgent regimes is derived to suggest how the characteristics
of potential U.S. partner nations shape the strategic choices available
to them. In Chapter Three, simple statistical analyses are applied to a
dataset of counterinsurgencies that have terminated since the end of
the Cold War (72 in all) to understand the manner in which these wars
have been fought and the likely outcomes of wars fought by different
types of counterinsurgent regimes.
This broad overview frames more in-depth analyses of two impor-
tant recent U.S. partnerships, those with the Philippines and Pakistan
(Chapters Four and Five, respectively). On the basis of the quantitative
analysis, the Philippines was selected to represent a U.S. partnership in
a promising environment. Pakistan was selected to represent a much
more challenging environment for the small-footprint approach. These
case studies combine insights from secondary literature, a wide vari-
ety of quantitative data sources, and interviews conducted with several
dozen government officials, military officers, and civil society actors in
the Philippines, Pakistan, and the United States.
In Chapter Six, the implications of the quantitative analysis and
case studies for U.S. partnership strategies are discussed. Two broad
partnership strategies are outlined, along with a number of specific
mechanisms for managing difficult partnerships. The chapter con-
cludes with a summary of the implications of the research findings for
U.S. foreign and security policy.
6 Countering Others’ Insurgencies

The report makes three central arguments, all of which represent

potentially important contributions to our understanding of counter-
insurgency and U.S. partnership strategies.
First, although there are certain universal (or near-universal) prin-
ciples that can be extracted from Western counterinsurgency doctrine,
not every counterinsurgent is equally capable of implementing these
principles. Understanding context—and in particular, the characteris-
tics of the counterinsurgent regime—is critically important to under-
standing the wide variety of counterinsurgency strategies and practices
that can be observed throughout the world.
Second, the “success stories” of U.S. light-footprint interventions
(especially in Colombia and the Philippines) represent relatively prom-
ising types of partner nations. Unfortunately, only about one in eight
counterinsurgencies in the post–Cold War period has been fought by a
regime that scores as highly on critical characteristics as do Colombia
and the Philippines. The vast majority of regimes fighting insurgen-
cies have much less-favorable characteristics, and consequently, their
chances of pursuing a counterinsurgency strategy compatible with U.S.
preferences are much lower. We should not generalize from U.S. expe-
riences in Colombia and the Philippines and assume that similar in-
direct approaches will work elsewhere; in fact, the odds are against
such successes.
Finally, except in rare circumstances, the United States is unlikely
to be able to exert sufficient leverage on partner nations to induce
the regimes to undertake transformational reforms. Recognizing the
limits of its influence, the United States should avoid committing to
certain partner nations altogether. Where U.S. security interests are
such that it cannot afford to remain disengaged, there are options the
United States can pursue to induce more modest changes in its part-
ners’ counterinsurgency practices. These alternative approaches are all
problematic, but in the conditions that characterize a substantial por-
tion of potential U.S. partners, they are the best that can realistically
be adopted.
CHAPTER TWO

Counterinsurgency in Comparative Perspective

This study examines the strategies and practices that regimes adopt
when fighting insurgencies. Other studies have examined the strategies
and practices that counterinsurgents supposedly should adopt, and still
others have focused on how external powers can help to provide them
the capabilities with which to conduct counterinsurgency. Surprisingly,
few studies have systematically examined how regimes actually do prac-
tice counterinsurgency.1 Fewer still have assessed whether the capabili-
ties the United States or other outside actors provide to these regimes
are used in support of counterinsurgency models that are likely to pro-
duce the strategic effects desired by the regimes’ sponsors.
The existing policy literature is focused overwhelmingly on the
experience of the United States and other Western nations. The ends
and means of counterinsurgency are frequently taken as a given, lead-
ing quickly to a narrow focus on operations and tactics. Such a focus
is understandable: Over the past decade, the United States has faced
the urgent task of dealing with its wars in Iraq and Afghanistan. As
the United States plans for the drawdown of its military presence in
Afghanistan and contemplates a wider range of light-footprint “shap-
ing” activities throughout the world, however, it is important for the
defense community to broaden its perspective.

1 For important exceptions, see Christopher Paul, Colin P. Clarke, Beth Grill, and Molly
Dunigan, Paths to Victory: Lessons from Modern Insurgencies, Santa Monica, Calif.: RAND
Corporation, RR-291/1-OSD, 2013; Christopher Paul, Colin P. Clarke, and Beth Grill,
Victory Has a Thousand Fathers: Sources of Success in Counterinsurgency, Santa Monica, Calif.:
RAND Corporation, MG-964-OSD, 2010.

7
8 Countering Others’ Insurgencies

In this chapter, we first briefly review the current state of think-

ing on counterinsurgency within the U.S. defense community and the
limitations of that view. Next, we examine counterinsurgency strategy
from a broader perspective, seeking to understand the many different
approaches that countries adopt when fighting internal wars. We then
propose possible explanations for the implementation by many devel-
oping countries of counterinsurgency strategies and practices that are
at odds with those preferred by the United States. The chapter con-
cludes with a typology of counterinsurgent regimes, illustrating the
wide range of potential partner nations with which the United States
must negotiate and the need for the United States to adapt its policies
to each type of regime. The historical record of counterinsurgency prac-
ticed by each of these types of regimes is reviewed in Chapter Three.

Current American Counterinsurgency Thought

The current official U.S. doctrine on counterinsurgency represents an

enormous evolution from the thinking that predominated a decade
ago.2 That thinking provided critical insights for Americans deploying
to Operation Iraqi Freedom (OIF) and Operation Enduring Freedom–
Afghanistan (OEF-A), but its evolution was deeply shaped by the
requirements of the war in Iraq.3 As a result, U.S. doctrine has faced
significant limitations in Afghanistan and even more in other theaters
where the U.S. footprint is a small fraction of the massive presence it
maintained in OIF and the later years of OEF-A.4 In particular, official

2 On the U.S. military’s doctrinal and other adaptations to irregular warfare, see especially
Joint and Coalition Operational Analysis, Decade of War: Vol. I, Enduring Lessons from the
Past Decade of Operations, Suffolk, Va.: Joint Staff J-7, June 15, 2012.
3 For an insightful discussion of the development of recent U.S. counterinsurgency doc-
trine and its limitations, see Jeffrey C. Isaac et al., “The New U.S. Army/Marine Corps Coun-
terinsurgency Field Manual as Political Science and Political Praxis,” Perspectives on Politics,
Vol. 6, No. 2, June 2008, pp. 347–360.
4 These limitations have been broadly recognized and have spurred U.S. practitioners to begin
expanding and adapting policy guidance for these conflicts. In January 2009, the U.S. Govern-
ment Counterinsurgency Guide was published, incorporating the insights of nine government
 Counterinsurgency in Comparative Perspective 9

American thinking has been limited by a failure to grapple sufficiently

with the strategic dimension of counterinsurgency and by its focus on
the Western experience of expeditionary counterinsurgency.
Counterinsurgency is often conflated with a particular way of
war—exemplified, for example, by the “hearts-and-minds” approach
so often advocated in the Western counterinsurgency community. Yet
by itself, counterinsurgency is simply a type of military operation, and
it is conducted differently by different countries.5
When we shift our view from the most frequently discussed coun-
terinsurgency campaigns—the U.S. wars in Iraq and Afghanistan and,
more distantly, the U.S. involvement in Vietnam and the British expe-
rience in Malaya—to a broader range of contingencies, the variation in
counterinsurgency approaches becomes evident. Substantial variation
can be seen even in recent U.S. interventions, including in Iraq (where
the United States dedicated substantial resources) and in Yemen and
Somalia (where it did not). If we broaden our focus to the experience
of other countries, we see practices ranging from scorched-earth tac-
tics, to forcible efforts to alter the ethnic balance of a rebellious region,
to relatively hands-off approaches designed only to contain an insur-
gency, to population-centric policies more reminiscent of the hearts-
and-minds approach ostensibly favored by countries like the United
States and the United Kingdom.
Unfortunately, most of the policy literature on counterinsurgency
offers little insight into how developing countries have fought their own
insurgencies. It generally suffers from a bias toward Western or other
external interventions, largely to the neglect of indigenous incumbents

departments and agencies. Unlike U.S. Department of the Army, Field Manual (FM) 3-24,
the Counterinsurgency Guide is more forward-looking and examines the conditions under
which the United States should engage in future counterinsurgencies abroad. Five levels of
involvement spanning the gamut from the deployment of a single advisor to a robust direct
intervention are discussed. In addition, the U.S. Army is taking the lead on drafting a revi-
sion to FM 3-24 and has invited feedback from a wide range of scholars, practitioners, and
operators. As of this writing, the estimated date of publication is December 2013.
5 See, for instance, Hammes, 2012, p. 49, and remarks by Eliot Cohen, “To COIN or Not?”
transcript from a roundtable discussion co-sponsored by Foreign Policy and the RAND Cor-
poration, March 18, 2013.
10 Countering Others’ Insurgencies

(that is, regimes fighting insurgencies within their own countries).

Such regimes constitute roughly 81 percent of the post-1945 counter-
insurgencies and 97 percent of those since 1990, yet the literature typi-
cally focuses on counterinsurgency conducted by foreign militaries.6
An examination of the policy-oriented Journal of Strategic Studies illus-
trates this point. Between the early 1990s and the writing of this report,
the journal published 82 articles on counterinsurgency and counterter-
rorism; 93 percent of them contained case studies, but only 26 percent
contained case studies of non-Western domestic incumbents. A similar
pattern is found in International Security, where only 21 percent (nine
articles) of the 43 articles since 1996 that mention counterinsurgency
and deal with the subject of state response to violent non-state actors
include a domestic-incumbent case study, while 65 percent (28 articles)
focus on campaigns by foreign interveners. Excluding Israel and the
Soviet Union/Russia, that figure falls to 12 percent (ten articles). The
result is a limited understanding of the way much of the world strat-
egizes for and fights insurgencies.
Perhaps the greatest departure between the classic-counterinsur-
gency literature and the recent track record of insurgencies in the devel-
oping world is the former’s focus on institution-building and attain-
ing a monopoly on violence. Political elites equipped with limited
resources engage in cost-benefit analysis to determine what they con-
sider to be an acceptable outcome. This calculus often results in poli-
cies aimed more at mitigating the harm caused by an insurgency than
at achieving a decisive defeat. As Paul Staniland contends, “Informal
bargains, collusive state-insurgent relationships, and shared sovereignty
are often less costly and more enduring than trying to build strong
states, an endeavor that integrally involves coercion, extraction, and
centralization.”7 If the reach of the national government is weak, it is
not uncommon to find local officials colluding with insurgent elements

6 Jason Lyall, “Are Coethnics More Effective Counterinsurgents? Evidence from the Second
Chechen War,” American Political Science Review, Vol. 104, No. 1, February 2010, pp. 1–20.
There are, of course, important exceptions. See especially Paul, Clarke, and Grill, 2010.
7 Paul Staniland, “States, Insurgents, and Wartime Political Orders,” Perspectives on Poli-
tics, Vol. 10, No. 2, June 2012, p. 244.
 Counterinsurgency in Comparative Perspective 11

to keep violence levels down or to mutually profit from illicit econo-

mies. If the insurgency is confined to a region deemed by national elites
to be of secondary importance, it may be largely ignored.8
Relatively little of counterinsurgency, in other words, is actually
practiced the way U.S. military doctrine prescribes. As the United
States shifts to a global strategy that emphasizes security partnerships
and small-footprint approaches, it will increasingly need to adapt its
policies to the realities of counterinsurgency as practiced by partners
throughout the developing world.

Varieties of Counterinsurgency

Soldiers who have actually had to fight in counterinsurgencies often

claim that all insurgencies are unique; no lessons can be universally
applied, because the circumstances of each conflict are so different.9
Although it is certainly true that counterinsurgency is highly context-
dependent, a number of patterns emerge in the ways that different
countries choose to fight counterinsurgencies and the likely conse-
quences of such choices. This section examines the range of counter-
insurgency strategies that different countries adopt, and the following
section explores why regimes adopt particular counterinsurgency strat-
egies and practices.
Three principles of counterinsurgency characterize the predomi-
nant Western approach:

• The counterinsurgent regime should seek accommodation—

including opportunities for participation in the governance of the
country—with the reconcilable armed opposition.
• The counterinsurgent regime should use violence discriminately
(i.e., should avoid harming the general civilian population) when

8 David R. Haines, “COIN in the Real World,” Parameters, Winter 2008–2009, pp. 50–51.
9 For a related argument, see Huw Bennett, “The Reluctant Pupil? Britain’s Army and
Learning in Counter-Insurgency,” London: Royal United Services Institute, 2009.
12 Countering Others’ Insurgencies

targeting the irreconcilable opposition, or “hard core,” of the

insurgency.
• The counterinsurgent regime should provide public goods (includ-
ing both security and various social services such as education and
health care) to disaffected populations in order to ensure their
loyalty and prevent their support of insurgent groups.

These three principles have been articulated with remarkable consistency

from the early “classics” of the Western counterinsurgency canon such
as those of David Galula and Sir Robert Thompson through to the
recent U.S. field manual on counterinsurgency, FM 3-24.10
As discussed in detail below, there is reason to believe that two
core principles of Western counterinsurgency doctrine—the need to
accommodate the reconcilable opposition and the need to use violence
discriminately against irreconcilable armed groups—are applicable
to the large majority of counterinsurgencies. Many counterinsurgent
regimes stray from these principles, and some have been success-
ful while adopting harsher approaches, but the past several decades
of experience suggest that regimes following those two principles are
more likely to terminate their internal wars on satisfactory terms.
The third principle—the requirement to provide public goods to
the general population—has much less universal applicability. Indeed,
Western counterinsurgency thinking has in many ways been hampered
by a focus on a state-building model overly influenced by the European
experience.

Bounded Accommodation
The need to bring the reconcilable opposition into the political process
has long been recognized, but its importance has increased in recent
decades as the dynamics of insurgencies have changed.
Insurgencies endure much longer than they once did: The aver-
age duration of civil wars in the period immediately after World War

10 David Galula, Counterinsurgency Warfare: Theory and Practice, Westport, Conn.: Praeger
Security International, 2006 [1964]; Sir Robert Thompson, Defeating Communist Insurgency:
The Lessons of Malaya and Vietnam, Saint Petersburg, Fla.: Hailer Publishing, 2005 [1966].
 Counterinsurgency in Comparative Perspective 13

II was only approximately two years, but by the post–Cold War era,
their duration had risen to about 15 years.11 This trend is rooted in
a number of factors that appear unlikely to change, including the
increasing availability of small arms and light weapons, the widespread
adoption of guerrilla tactics, the ready availability of rebel funding
through the trafficking of contraband, and the general unwillingness
of major powers to make large and indefinite commitments to allies
in the developing world.12 This steady upward trend in the duration of
insurgencies suggests that it is increasingly difficult to end these con-
flicts through outright military victory.
These conflicts are also extremely damaging to the countries in
which they are fought—so much so that they have been described
as “development in reverse.”13 Consequently, leaders with a stake in
the post-conflict future of their country may well decide that offering
at least portions of the armed opposition the opportunity to share in
power is superior to a long and damaging fight to the finish.
Moreover, insurgencies have extremely damaging consequences
beyond the borders of the country originally affected: They generally
increase the risk of spillover conflicts throughout the region in which
they occur, damage the economies of neighboring states, contribute
to flourishing transnational crime networks, spread pandemic disease,
and foster transnational terrorism.14 Unless outside powers are willing

11 James D. Fearon and David D. Laitin, “Ethnicity, Insurgency, and Civil War,” American
Political Science Review, Vol. 97, No. 1, February, 2003, p. 78. The exact durations of wars
vary according to how social scientists define what counts as a “civil war” and when the wars’
precise start and end dates are, but the general trend is broadly recognized. See also Monica
Duffy Toft, Securing the Peace: The Durable Settlement of Civil Wars, Princeton, N.J.: Prince-
ton University Press, 2010.
12 Fearon and Laitin, 2003; Toft, 2010; see also James D. Fearon, “Why Do Some Civil
Wars Last So Much Longer than Others?” Journal of Peace Research, Vol. 41, No. 3,
pp. 275–301.
13 Paul Collier, V. L. Elliott, Havard Hegre, Anke Hoeffler, Marta Reynal-Querol, and
Nicholas Sambanis, Breaking the Conflict Trap: Civil War and Policy Development, Washing-
ton, D.C.: The World Bank and Oxford University Press, 2003, Ch. 1.
14 Collier et al., 2003, Ch. 2.
14 Countering Others’ Insurgencies

to intervene decisively in these conflicts (a rare occurrence15), the inter-

national community has a vested interest in ensuring some form of
negotiated settlement and working to enforce it on the warring parties.
Given these trends, it is not surprising that a growing propor-
tion of internal conflicts end in negotiated settlements, cease-fires, or
stalemates. Whereas the vast majority of civil wars in the decades after
World War II ended in military victory for one side or the other, now
only a minority do.16 Regimes that find it difficult to accommodate
the reconcilable opposition are therefore increasingly likely to consign
themselves to lengthy, costly wars. They eventually find themselves
either facing outright defeat, yielding concessions in the end, or co-
existing indefinitely with an insurgency that exercises de facto control
over parts of the country.
Determining which segments of the opposition are reconcilable
is, of course, rarely a straightforward matter. Groups with limited,
clearly defined political objectives seldom need to resort to insurgency
in the first place against governments that are willing to accommo-
date the reconcilable opposition. In most cases, reaching accommoda-
tion with insurgents is instead achieved through an iterated process of
fighting and talking. As the costs of continued fighting become clear,
the outlines of an acceptable compromise begin to emerge—at least
for those warring parties whose minimum acceptable outcomes are
compatible with one another’s. In many conflicts, however, portions
of the opposition have minimum acceptable outcomes that will never
coincide with the government’s. Any accommodation with these par-
ties is inevitably tactical—a pause before renewed hostilities.17 Thus,

15 Stephen Watts, Caroline Baxter, Molly Dunigan, and Christopher Rizzi, The Uses and
Limits of Small-Scale Military Interventions, Santa Monica, Calif.: RAND Corporation,
2012.
16 Toft, 2010, Ch. 1.
17 These distinctions between different types of insurgents parallel those in Stephen
Stedman’s typology of “spoilers,” in which he divides parties that can undermine peace pro-
cesses into three groups: what he terms limited, greedy, and total. The reconcilable opposition
corresponds to the first two of these groups, while the irreconcilable opposition corresponds
to the last. See Stephen John Stedman, “Spoiler Problems in Peace Processes,” International
Security, Vol. 22, No. 2, Fall 1997, pp. 5–53.
 Counterinsurgency in Comparative Perspective 15

while accommodation with a significant portion of the armed oppo-

sition is increasingly necessary to end insurgencies, accommodation
cannot be open-ended. We therefore refer to this principle as bounded
accommodation—reconciliation with those portions of the opposition
that can be accommodated without sacrificing the core interests of the
government. For the remainder of the insurgency, the irreconcilable
opposition, the counterinsurgent’s only realistic alternative is the con-
tinued use of force. Even when confronting the hard core of insurgents,
however, violence must be used discriminately.

Discriminate Violence
There is strong support for the notion that government use of force usu-
ally must be discriminate to be successful.18 If the government employs
violence only against those who oppose it while leaving unharmed
those who are neutral, the population has considerable incentive to
avoid opposing the government. On the other hand, if neutrality is no
guarantee of safety (that is, if violence is indiscriminate), the popula-
tion is much more likely to seek safety or payment from rebels.19 There
is significant evidence to suggest that, much like the need to accommo-
date the reconcilable opposition, the imperative to use force discrimi-
nately is a counterinsurgency principle with broad applicability.

Provision of Public Goods

While there is a consensus of opinion on the first two elements of clas-
sical Western counterinsurgency doctrine, there is no such consensus
on the need for the state to provide public goods to its population as

18 Paul et al., 2013; Paul, Clarke, and Grill, 2010; Matthew Adam Kocher, Thomas B. Pepinsky,
and Stathis N. Kalyvas, “Aerial Bombing and Counterinsurgency in the Vietnam War,”
American Journal of Political Science, Vol. 55, No. 2, April 2011, pp. 201–218; Monica Duffy
Toft and Yuri M. Zhukov, “Denial and Punishment in the North Caucasus: Evaluating
the Effectiveness of Coercive Counterinsurgency,” Journal of Peace Research, Vol. 49, No. 6,
2012, pp. 785–800. Not all analyses agree, however. See, for instance, Jason Lyall, “Does
Indiscriminate Violence Incite Insurgent Attacks? Evidence from Chechnya,” Journal of Con-
flict Resolution, Vol. 53, No. 3, June 2009, pp. 331–362.
19 T. David Mason and Dale A. Krane, “The Political Economy of Death Squads: Toward
a Theory of the Impact of State-Sanctioned Terror,” International Studies Quarterly, Vol. 33,
No. 2, June 1989, pp. 175–198.
16 Countering Others’ Insurgencies

part of a counterinsurgency strategy. Given the slow pace at which

state-building usually progresses and the myriad ways in which many
regimes abuse their capacity, efforts to improve the state’s ability to
deliver public goods among discontented populations may in fact
undermine counterinsurgency goals.
This claim challenges much recent orthodoxy on counterinsur-
gency and “good practice” for fragile states. The U.S. Army’s doctrine
for stabilizing fragile states, for instance, charges the Army with help-
ing the host nation provide health care and education and assert con-
trol over the illicit economy.20 Similarly, in a recent guide to repairing
fragile states, the World Bank asserted that “[a] long-term focus on
state capacity . . . is critical in all fragile state contexts if these countries
are ever to find a durable exit from crisis.”21 Some research has also sup-
ported the contention that provision of public goods is key to counter-
insurgent victory, but others have reached different conclusions.22
As a long-term goal, state-building efforts are entirely reasonable.
But durable improvements in state capacity typically require decades.23
In the meantime, efforts to expand state control may be counterpro-
ductive, at least in those countries with little history of a powerful state.
In many countries, the state has historically exercised little authority in

20 U.S. Department of the Army, Stability, Army Doctrinal Publication 3-05, August 2012,
p. 10.
21 The World Bank, Fragile States: Good Practice in Country Assistance Strategies, Washing-
ton, D.C., 2005, p. 3.
22 See especially, Paul, Clarke, and Grill, 2010, 2013, for research in favor of this view. For
a contrasting perspective, see Eli Berman, Jacob N. Shapiro, and Joseph H. Felter, “Can
Hearts and Minds Be Bought? The Economics of Counterinsurgency in Iraq,” Journal of
Political Economy, Vol. 119, No. 4, August 2011, pp. 766–819. Note that Paul’s work exam-
ines specific practices rather than regimes’ ability to implement those practices. It may well
be that highly capable states are more likely to succeed in counterinsurgency campaigns if
they provide public goods, while efforts by extremely weak states to provide public goods
may have the opposite effect. The importance of understanding the character of the counter-
insurgent regime is a theme to which this report frequently returns.
23 Lant Pritchett, Michael Woolcock, and Matt Andrews, Capability Traps? The Mechanisms
of Persistent Implementation Failure, Washington, D.C.: Center for Global Development,
Working Paper 234, December 2010.
 Counterinsurgency in Comparative Perspective 17

peripheral regions remote from the major cities.24 Attempting to rap-

idly build state capacity in such regions risks alienating populations
who have traditionally mistrusted the state—often with good reason.
As long as control over resources is divided equitably between the
regime and the armed opposition, the record of civil-war termination
suggests that insurgencies can be brought to a satisfactory end without
extending the state’s reach. Indeed, many peace deals are predicated on
informal understandings and a tacit acknowledgment by state leaders
that the regime will exercise little effective control in remote regions.
In Mali, for instance, the formal democratization and decentralization
that underlay the peace that the country enjoyed from the latter 1990s
until recently was accompanied with a more informal acceptance of the
fact that the central state would exercise less control over even illegal
activities in the north of the country.
Nevertheless, it is important to be clear about the limitations
of such peace deals. They often prove less durable than more formal
arrangements, in part because they typically provide few institutional-
ized conflict-resolution mechanisms between the warring communi-
ties.25 They accept the continued existence of spaces with little formal
governance in which terrorist organizations and transnational criminal
networks may flourish.26 And they often provide a poor framework for
long-term development, which is usually thought to require the rule of
law and expectations of durable stability. Informal understandings are

24 See for instance Jeffrey Herbst, States and Power in Africa: Comparative Lessons in Author-
ity and Control, Princeton, N.J.: Princeton University Press, 2000; Christopher Clapham,
“The Global-Local Politics of State Decay,” in Robert I. Rotberg, ed., When States Fail:
Causes and Consequences, Princeton, N.J.: Princeton University Press, 2004; Pierre
Englebert and Denis M. Tull, “Postconflict Reconstruction in Africa: Flawed Ideas About
Failed States,” International Security, Vol. 32, No. 4, 2008, pp. 106–139.
25 See, for instance, Robert H. Bates, Prosperity and Violence: The Political Economy of Devel-
opment, New York: W. W. Norton, 2001.
26 Angel Rabasa, Steven Boraz, Peter Chalk, Kim Cragin, Theodore W. Karasik, Jennifer
D. P. Moroney, Kevin A. O’Brien, and John E. Peters, Ungoverned Territories: Understand-
ing and Reducing Terrorism Risks, Santa Monica, Calif.: RAND Corporation, MG-561-AF,
2007.
18 Countering Others’ Insurgencies

clearly second-best alternatives to more institutionalized frameworks,

but in many cases, they may be the best realistic alternative available.

A Typology of Counterinsurgency Models

Counterinsurgent regimes’ strategies and practices can be understood

in terms of the three principles discussed above: the political accom-
modation of the reconcilable armed opposition, the discriminate use of
violence, and the provision of public goods to disaffected communities.
As Figure 2.1 suggests, a variety of counterinsurgency models can exist
within these three dimensions.
More specifically, we can discern four different models or “ideal
types” of strategies of particular interest:27

Figure 2.1
Typology of Counterinsurgency Models

High

Political
accommodation
of reconcilable
opposition

Low
Public-goods provision
Low to disaffected
High
Low High communities
Discrimination in use of
violence against irreconcilable
opposition
RAND RR513-2.1

27 Other strategies besides these four are, of course, feasible. As the remainder of this chapter
and Chapter Three reveal, however, these four appear to be particularly common in practice.
 Counterinsurgency in Comparative Perspective 19

1. Classic counterinsurgency. The classic Western approach to

counterinsurgency emphasizes the three principles outlined
above. It is not purely a hearts-and-minds approach in that it
recognizes the importance of the well-targeted use of force,
but it places stronger emphasis on an inclusive, capable, and
accountable government than do the other strategies. This strat-
egy is most often associated with the United Kingdom and the
United States and, to a lesser degree, France—highly capable
Western democracies.
2. Strong-state repression. On the opposite corner of the front
side of the diagram in Figure 2.1 lies strong-state repression,
a strategy associated with the Soviet Union in Afghanistan,
Russia in Chechnya, and—at the lower end of the conflict
spectrum—Egypt’s repression of Islamists in the 1990s. In this
approach, the regime is unable or unwilling to offer accommo-
dations to any part of the opposition, relying instead on broad,
often indiscriminate use of force and terror to suppress orga-
nized dissent. Frequently, however, such regimes offer substan-
tial services to the general population, in part in an effort to
“buy off” discontent.
3. Informal accommodation. At the back of the three-
dimensional space in Figure 2.1, we find strategies in which
central governments are unable or unwilling to provide public
goods to discontented populations—typically those in rural
or peripheral regions, far away from the cities over which the
regime maintains stronger control. Although the state does
not provide substantial services in these regions, however, this
does not mean that the regime cannot reach an accommoda-
tion with the armed opposition. Weak states are often able to
reach peace agreements with insurgents. Such deals—which we
term informal accommodation—usually are tacit understand-
ings. Members of the armed opposition may be offered a vari-
ety of “participation rights” in the government, ranging from
leadership positions to guarantees of participation in the secu-
rity services or civil bureaucracy. Perhaps even more important,
informal accommodation normally entails tacit agreement on
20 Countering Others’ Insurgencies

the armed opposition’s control over territory and various illicit

or semi-licit economic activities such as smuggling, unlicensed
resource extraction (e.g., narcotics trafficking, illicit timber
sales, or trade in “conflict diamonds”) or the steering of foreign
aid to the insurgents’ supporters. As noted above, the Bamako
government adopted such an approach in 1996 to bring peace
to northern Mali.
4. Containment. Finally, weak states that are unable or unwill-
ing to accommodate the reconcilable opposition typically
rely on containment strategies, depicted in the bottom left-
hand corner of the back of Figure 2.1. Such strategies accept
that some level of residual violence is the inevitable result of
an inability to exercise effective control or to coopt the armed
opposition. The counterinsurgent regime uses force—usually
indiscriminately—to repress insurgent activity whenever it
becomes too threatening to the regime. As long as the insur-
gents stay within certain (usually ill-defined) bounds, the
regime typically accepts their presence and concomitant limits
on its control over portions of its population and territory. Such
containment strategies have been adopted at various points by
the ruling regimes of countries such as Myanmar and Yemen.

Each of these approaches to counterinsurgency is potentially

viable. Iraq under Saddam Hussein, for instance, was able to crush
Kurdish and Shiite insurgencies through strong-state repression. Simi-
larly, as will be discussed in greater detail in Chapter Five, Pakistan has
relied on containment strategies in the Federally Administered Tribal
Areas (FATA) for decades. However, there is substantial evidence that
the strategies in the upper right-hand corners of Figure 2.1—classic
counterinsurgency and informal accommodation—are most often
successful in the current international environment, where small arms
and light weapons abound, guerrilla tactics and illicit economies make
insurgents extremely resilient, and major powers exhibit little willing-
ness to commit to extensive interventions on behalf of client regimes.
The designation of these strategies as models or ideal types indi-
cates that no country practices counterinsurgency in a “pure” form.
Indeed, even the United Kingdom, perhaps the foremost proponent of
 Counterinsurgency in Comparative Perspective 21

the classic counterinsurgency strategy, has never truly practiced coun-

terinsurgency as its doctrine would suggest.28 Creating these concep-
tual “bins” is nonetheless useful for understanding in general terms
the major differences counterinsurgency practiced by various regimes.
Moreover, the historical record reveals that there are clear patterns in
which types of regimes adopt these different approaches.

Determinants of Counterinsurgency Strategy and Practice

Given the considerable evidence that suggests that accommodating the

reconcilable opposition and applying violence discriminately maximize
a regime’s chances of waging counterinsurgency successfully, why do
all counterinsurgents not adopt this approach? The answer appears to
be that not all regimes can apply these principles with equal success.
Both the policy and academic literatures suggest that two charac-
teristics of a regime are particularly important in determining how it
will prosecute counterinsurgency: its degree of political inclusion and
the extent of its capacity.
These characteristics, as well as the extent of the regime’s mili-
tary superiority, lie at the center of the analysis in this section and in
Chapter Three. Here, we explain why different types of counterinsur-
gent regimes are likely to adopt different counterinsurgency strategies
and practices and discuss the likely implications of these choices for
achieving negotiated settlements with the armed opposition and using
force discriminately. Chapter Three examines the record of the past
two decades to determine the extent to which the characteristics of
counterinsurgent regimes can explain outcomes in these wars. As will
be seen, those characteristics shape the ability of the United States to
positively influence regimes’ counterinsurgency practices.

Political Inclusion
The first critical characteristic of counterinsurgent regimes, politi-
cal inclusiveness, is understood here as the extent to which organized

28 Paul Dixon, “‘Hearts and Minds’? British Counter-Insurgency from Malaya to Iraq,”
Journal of Strategic Studies, Vol. 32, No. 3, 2009, pp. 353–381.
22 Countering Others’ Insurgencies

political actors are afforded opportunities to meaningfully influence

the formulation and execution of public policies.29 Relatively few of the
countries fighting insurgencies could be understood to be “democra-
cies,” as the term is typically conceived. Especially in recent decades,
however, a great many conflict-affected states have significant demo-
cratic characteristics and meaningful opportunities for opposition
groups to participate in influencing politics.
A central lesson to be learned from both the policy and academic
literatures is that non-democratic regimes find it difficult to accommo-
date elements of the armed opposition. There are at least two reasons
why more-democratic regimes find it easier to accommodate the recon-
cilable opposition:

• Breadth of coalitions. Non-democratic regimes are often based

on narrow coalitions that would be highly threatened by major
concessions to the opposition, even if top leaders wanted to
compromise.30
• Ability to make credible commitments. Non-democratic re-
gimes may be willing to make concessions, and there may be suf-
ficient areas of mutual agreement between the regime and the
armed opposition to make a peace deal possible. But the regime
may find it difficult to convince the opposition of the credibil-
ity of its commitment to compromise due to a lack of institu-
tionalized “hand-tying” mechanisms.31 The opposition, in other
words, is likely to fear that any concessions made by the gov-

29 This concept is similar to Dahl’s concept of “polyarchy.” See Robert A. Dahl, Polyarchy:
Participation and Opposition, New Haven, Conn.: Yale University Press, 1972.
30 This claim was particularly common during the Cold War. See, for instance, Douglas
S. Blaufarb, The Counterinsurgency Era: U.S. Doctrine and Performance, 1950 to the Present,
New York: The Free Press, 1977; D. Michael Shafer, Deadly Paradigms: The Failure of U.S.
Counterinsurgency Policy, Princeton, N.J.: Princeton University Press, 1988; and Mason and
Krane, 1989, pp. 175–198. However, the logic is more generalizable. Political scientists have
spelled out this logic in “selectorate theory”; see especially Bruce Bueno de Mesquita and
Alastair Smith, “Leader Survival, Revolutions, and the Nature of Government Finance,”
American Journal of Political Science, Vol. 54, No. 4, October 2010, pp. 936–950.
31 James D. Fearon, “Rationalist Explanations for War,” International Organization, Vol. 49,
No. 3, Summer 1995, pp. 379–414; Barbara F. Walter, “The Critical Barrier to Civil War
 Counterinsurgency in Comparative Perspective 23

ernment under the threat of continued insurgency will be with-

drawn as soon as the insurgents disband. The accountability
mechanisms that might prevent a regime from reneging on its
commitments—institutions such as a free press, courts to uphold
agreements, guaranteed levels of representation or veto rights in
legislatures, etc.—are typically much weaker in autocracies than
in democracies.

If more-democratic regimes are in fact better able to accommo-

date the reconcilable opposition, we should see a number of patterns
in the way wars terminate. More-democratic regimes should be better
able to broker formal peace deals with the armed opposition. Conse-
quently, unless their record of military victory is poorer than that of
less-democratic regimes, they should also be more capable of securing
a favorable outcome—that is, either a victory or a formally negotiated
settlement that meets at least the minimum demands of all parties.32
Finally, they should be able to better sustain the peace in the post-
conflict period due to the presence of institutions that promote trans-
parency and accountability, such as a free press.

Settlement,” International Organization, Vol. 51, No. 3, Summer 1997; Kenneth A. Schultz,
Democracy and Coercive Diplomacy, New York: Cambridge University Press, 2001.
32 The existing literature has not reached a consensus about the effects of democracy on coun-
terinsurgency outcomes. Paul, Clarke, and Grill, 2010, maintain that democracy is weakly
associated with counterinsurgent victory, but updated work in Paul et al., 2013, finds no such
relationship. Ben Connable and Martin C. Libicki find a very complex relationship between
democracy and victory. See Ben Connable and Martin C. Libicki, How Insurgencies End,
Santa Monica, Calif.: RAND Corporation, MG-965-MCIA, 2010, pp. 186–187. Jason Lyall
finds no relationship. See Jason Lyall, “Do Democracies Make Inferior Counterinsurgents? Re-
assessing Democracy’s Impact on War Outcomes and Duration,” International Organiza-
tion, Vol. 64, Winter 2010b, pp. 167–192. It is important to note that the research in this
report is focusing on the post–Cold War era, while these other studies examined much
longer periods of time. There are many reasons to believe that the dynamics of insurgency
and counterinsurgency have shifted in recent decades, making it more necessary for regimes
now to reach an accommodation with the reconcilable elements of the armed opposition.
We have not examined previous eras (i.e., the Cold War era or earlier) to determine how this
relationship has changed over time, although there remains much important research to be
done on this topic.
24 Countering Others’ Insurgencies

There is also reason to believe that greater political inclusion

restricts a regime’s ability to use violence indiscriminately. The relation-
ship is greatly tempered, however, by the fact that counterinsurgency
is typically a particularly brutal form of conflict—a fact that not even
the best-intentioned regimes can escape.
Numerous examples exist of public opinion restraining democ-
racies’ use of violence in expeditionary counterinsurgency.33 More-
systematic studies of domestic counterinsurgency, however, have only
recently begun to disentangle the ways in which democratic account-
ability may restrain the use of indiscriminate violence in such cam-
paigns. Two mechanisms in particular stand out:

• Transparency and public accountability. Democracies have

robust civil societies, including a free press and advocacy organi-
zations, that can expose brutal and repressive practices carried out
by the regime. They also have numerous channels through which
citizens can express their preferences and hold government leaders
accountable.
• Executive constraints. Democracies typically diffuse executive
power and more effectively constrain executive autonomy than do
autocracies, providing more opportunities for other sectors of the
government to restrain abusive security services.34

Existing evidence suggests a generally positive relationship

between democracy and the discriminate use of force. Democracies
are much less likely to engage in genocide and similar campaigns.35

33 See Gil Merom, How Democracies Lose Small Wars, New York: Cambridge University
Press, 2003.
34 For a highly nuanced discussion of the effects of democracy on the state’s use of violence,
see especially Christian Davenport, State Repression and the Domestic Democratic Peace, New
York: Cambridge University Press, 2007a. On the general propensity of more-democratic
regimes to use violence more selectively, see Kristine Eck and Lisa Hultman, “One-Sided
Violence Against Civilians in War: Insights from New Fatality Data,” Journal of Peace
Research, Vol. 44, No. 2, 2007, pp. 233–246.
35 Barbara Harff, “No Lessons Learned from the Holocaust? Assessing Risks of Genocide
and Political Mass Murder Since 1955,” American Political Science Review, Vol. 97, No. 1,
February 2003, pp. 57–73; Michael Colaresi and Sabine C. Carey, “To Kill or to Pro-
 Counterinsurgency in Comparative Perspective 25

Evidence about lesser forms of indiscriminate violence is more mixed.

Although “consolidated democracies” are generally agreed to be the
least likely form of government to conduct violent repression, what
have alternatively been labeled “intermediate,” “hybrid,” or “illiberal”
democracies may be as likely as autocracies to engage in such acts.36

State Capacity
The second critical characteristic of the counterinsurgent is state
capacity—what Samuel Huntington called the degree rather than the
type of government.37 Like political inclusiveness, state capacity is an
important determinant of a regime’s ability to reach acceptable accom-
modations with armed opposition groups. In particular, state capacity
can influence a regime’s ability to achieve negotiated outcomes in two
ways:

• Provision of public goods. Both U.S. counterinsurgency doc-

trine and academic studies suggest that governments gain the
support of their populations through the widespread and effec-
tive provision of public goods such as public sanitation and health
services, physical infrastructure (e.g., roads and bridges), and
education.38 If the government can provide few services that are

tect: Security Forces, Domestic Institutions, and Genocide,” Journal of Conflict Resolution,
Vol. 52, No. 1, February 2008, pp. 39–67; R. J. Rummel, “Democracy, Power, Genocide,
and Mass Murder,” Journal of Conflict Resolution, Vol. 39, No. 1, 1995, pp. 3–26.
36 Bruce Bueno de Mesquita, George W. Downs, Alastair Smith, and Feryal Marie Cherif,
“Thinking Inside the Box: A Closer Look at Democracy and Human Rights,” International
Studies Quarterly, Vol. 49, 2005, pp. 439–457; Davenport, 2007a; Christian Davenport,
“State Repression and Political Order,” Annual Review of Political Science, Vol. 10, 2007b,
pp. 1–23; Sabine C. Zanger, “A Global Analysis of the Effect of Political Regime Changes
on Life Integrity Violations, 1977–93,” Journal of Peace Research, Vol. 37, No. 2, 2000,
pp. 213–233.
37Samuel P. Huntington, Political Order in Changing Societies, New Haven, Conn.: Yale
University Press, 1968, p. 1.
38 See, for instance, U.S. Department of the Army, 2006; Paul et al., 2013; Paul, Clarke, and
Grill, 2010; Connable and Libicki, 2010, pp. 186–187; and Robert I. Rotberg, “The Failure
and Collapse of Nation-States: Breakdown, Prevention, and Repair,” in Robert I. Rotberg,
ed., When States Fail: Causes and Consequences, Princeton, N.J.: Princeton University Press,
2004.
26 Countering Others’ Insurgencies

desired by insurgents or the populations that support them, it

has few “carrots” with which to pursue accommodation with the
armed opposition. As a result, it will be more dependent on either
accepting de facto insurgent control over portions of territory and
population (as in the informal accommodation or containment
strategies) or winning an outright military victory.
• Support for coercive capabilities. Unfortunately, weak state
capacity also limits the state’s ability to coerce insurgents. If the
state does not penetrate relatively deeply into the social life of
ordinary people, it is difficult for the regime to acquire the infor-
mation necessary to identify insurgents.39

State capacity may also influence a regime’s ability to use force

discriminately. The propensity to engage in indiscriminate violence
may be influenced by several mechanisms:

• Sustainment and discipline. Strong, capable states are more

likely to be able to pay their forces regularly and provide them
with necessary supplies and other material support, and they are
more likely to be able to effectively monitor the performance of
their agents in the field and discipline personnel who behave in
abusive, predatory, or counterproductive ways.40 Accountabil-

39 For a formal explication of this logic, see especially Berman, Shapiro, and Felter, 2011. See
also Helge Holtermann, “Explaining the Development–Civil War Relationship,” Conflict
Management and Peace Science, Vol. 29, No. 1, 2012, p. 63.
40 Fearon and Laitin, 2003, p. 80, for instance, argues that “effective counterinsurgency
requires government forces to distinguish active rebels from noncombatants without
destroying the lives and living conditions of the latter. This is an extremely difficult political,
military, and organizational problem even for well-equipped and well-paid modern militar-
ies. . . . For less well-financed and bureaucratically competent states, the problem appears
to be nearly insoluble. Such states either cannot prevent the abuse of local powers by field
commanders or may even permit these abuses as a sort of tax farming to the military. That
is, they ‘pay’ the soldiers with the opportunity to loot and pillage, a practice that tends to
sustain rather than end insurgencies.” Stathis Kalyvas and Laia Balcells argue that govern-
ment forces have become almost indistinguishable from rebels in many conflicts, particularly
in the post–Cold War era. See Stathis N. Kalyvas and Laia Balcells, “International System
and Technologies of Rebellion: How the End of the Cold War Shaped Internal Conflict,”
American Political Science Review, Vol. 104, No. 3, August 2010, pp. 415–429. Perhaps the
 Counterinsurgency in Comparative Perspective 27

ity over and discipline among counterinsurgent forces are often

less a function of democracy than of state capacity; even well-
intentioned democratic leaders may be unable to prevent indis-
criminate violence if the state is unable to sustain its personnel
and discipline abuses of authority.
• Futility. Weak regimes may rely excessively on “sticks” because
they have too few “carrots” to offer disaffected populations. If the
regime cannot effectively provide public services to its population,
it may feel that it must resort to terror tactics to compel obedi-
ence, if not loyalty.41

Military Superiority
Counterinsurgency is, in the end, a form of warfare. The quality of
governance provided by the regime is critical, but without the req-
uisite military capabilities, governance is inadequate. Violent con-
flict emerges precisely because the parties to the conflict cannot find
a mutually acceptable solution to their disagreements within existing
political mechanisms. Even if the government improves the quality of
its governance during the course of an insurgency, a political solution
that does not sacrifice core government interests requires at a mini-
mum a military stalemate, if not the clear superiority of the governing
forces.42
For this reason, many observers have argued that the military
capabilities of a regime are a third determinant of counterinsurgency
approaches and outcomes. We argue that military capabilities play an
important role, but their effects can be understood only in terms of
the regime for which they fight. Consequently, we examine the effects
of military capabilities throughout this study and, in particular, their

archetypal example is the rise of “sobels” during Sierra Leone’s civil war—government sol-
diers by day who turned into rebels by night, preying on the general population in both of
these incarnations.
41 See for instance Jeffrey Herbst, “African Militaries and Rebellion: The Political Economy
of Threat and Combat Effectiveness,” Journal of Peace Research, Vol. 41, No. 3, May 2004,
pp. 361–362.
42U.S. Department of the Army, 2006, especially pp. 1–21; Nathan Leites and Charles
Wolf, Jr., Rebellion and Authority, Chicago: Markham Publishing Company, 1970.
28 Countering Others’ Insurgencies

ability to achieve outright military victory, but we do not believe that

they by themselves will determine a regime’s ability to reach a negoti-
ated settlement or its propensity to use violence discriminately.

Measures and Predictions

The examination of counterinsurgents’ characteristics and their conse-

quences suggests four broad generalizations about their likely behavior:

• Regimes that are more politically inclusive and have higher degrees
of state capacity are more likely to adopt the classical approach to
counterinsurgency.
• Regimes that are more politically inclusive but possess only weak
state capacity are more likely to adopt informal accommodation
as their predominant strategy.
• Regimes that are less politically inclusive and have higher degrees
of state capacity are more likely to engage in strong-state repression.
• Regimes that are less politically inclusive and possess weak state
capacity are more likely to pursue containment strategies.

These generalizations are summarized in Table 2.1.

Predictions are not particularly helpful, however, without indica-
tors of the key causal variables—we need to know how to characterize
a particular regime before we can predict what its likely behavior will

Table 2.1
Determinants of Counterinsurgency Strategy

Regime Characteristics

Political Inclusivity State Capacity Predominant Strategy

High High Classic counterinsurgency

High Low Informal accommodation

Low High Strong-state repression

Low Low Containment

 Counterinsurgency in Comparative Perspective 29

be. Fortunately, considerable efforts have been made to measure the

key characteristics of counterinsurgent regimes by the academic and
research communities. While all of the measures have been criticized,
they provide a reasonable approximation of the underlying concepts.
Moreover, from a more practical perspective, each of the measures can
predict regime behavior, as demonstrated in Chapter Three.
The first key characteristic of counterinsurgent regimes, their
degree of political inclusion, can be approximated with the Freedom
House index of democracy, using the midpoint of the Freedom House
scale as the basis for distinguishing between more- and less-inclusive
regimes.43 This measure provides an acceptable approximation of the
underlying concept, with full coverage of all the countries in our sample
for all of the years under review.
We rely on two indications of the second critical determinant of
counterinsurgency strategies and practices, state capacity. One may be
called state reach, the ability of the state to exercise influence within
society. States with poor state reach are characterized by “low state
penetration and capacity; a rural, scattered settlement structure; and
poor communications.”44 The other may be termed government effec-
tiveness, a world governance indicator (WGI) defined by researchers
at the World Bank as “the quality of public services, the quality of the

43 Freedom House, Freedom in the World, Washington, D.C., 2013. As described in greater
detail in Chapter Three and the Appendix, Freedom House scores were based on three differ-
ent time periods: the average score of the country throughout all years of conflict, the average
score in the final two years of conflict, and the average score from the five years prior to the
conflict’s onset.
44 Holtermann, 2012, p. 63. This variable is constructed as an index of three different fac-
tors: road density, proportion of the population that is urban, and telephone density. Unfor-
tunately, the data do not cover the entire period of the conflicts under review. Consequently,
we were forced to use the entire post–Cold War period as the basis for each country’s state-
reach score. Doing so clearly introduces some degree of distortion to the measure; countries
in which conflicts terminated in the early post–Cold War period, for instance, are more
likely to have a higher state-reach score than countries with conflicts that endured for much
of the post–Cold War period because the former states would have had more years of peace
(and thus opportunity to recover from wartime destruction) in the period being measured.
Despite this limitation, the state-reach variable still does a creditable job of differentiating
lower- and higher-capacity states. As with the rest of the quantitative analysis, details may be
found in the Appendix. The authors are grateful to Helge Holtermann for sharing his data.
30 Countering Others’ Insurgencies

civil service and the degree of its independence from political pressures,
the quality of policy formulation and implementation, and the cred-
ibility of the government’s commitment to such policies.”45
Finally, military superiority in counterinsurgency has tradi-
tionally been measured in terms of force-to-force ratios or force-to-
population ratios—that is, the ratio of government forces to insur-
gent forces or the ratio of government forces to the population
(typically expressed in terms of government forces for each 1,000
inhabitants of a country).46 James T. Quinlivan argued that force-to-
population ratios are the most appropriate rule of thumb for sizing
forces in counterinsurgency:

A “hearts and minds” counterinsurgency campaign places the

focus on the people, the military consequences of which are
requirements for population control measures and local security
of the population. Population control measures and local secu-
rity both demand security force numbers proportional to the
population.47

Others have argued instead that force-to-force ratios are the more
appropriate measure.48 Unfortunately, there is very little standardized

45 Daniel Kaufmann, Aart Kraay, and Massimo Mastruzzi, Governance Matters VIII: Aggre-
gate and Individual Governance Indicators, 1996–2008, Washington, D.C.: The World Bank,
Development Research Group, Policy Research Working Paper 4978, June 2009, p. 6. As
with the state-reach variable, the limited number of years for which data were available
meant that we were forced to use available data for the entire post–Cold War period rather
than data from only the period of conflict. Similarly, despite this limitation of the data, gov-
ernment effectiveness remains a useful—if rough—measure of the underlying concept.
46 U.S. Department of the Army, 2006, pp. 1–13. On force-to-population ratios specifically,
see James T. Quinlivan, “Force Requirements in Stability Operations,” Parameters: U.S.
Army War College Quarterly, Vol. 25, No. 4, Winter 1995/1996, pp. 59–69. For critical exam-
inations of both ratios, see Steven M. Goode, “A Historical Basis for Force Requirements
in Counterinsurgency,” Parameters: U.S. Army War College Quarterly, Winter 2009/2010,
pp. 45–57; Joshua Thiel, “COIN Manpower Ratios: Debunking the 10 to 1 Ratio and
Surges,” Small Wars Journal, January 15, 2011.
47 Quinlivan, 1995-1996, pp. 59–69.
48 FM 3-24 suggests the use of both measures.
 Counterinsurgency in Comparative Perspective 31

data on police forces. Consequently, we have had to rely strictly on

numbers of government military forces.49
The indicators discussed here are only crude approximations of
the underlying concepts. However, by relying on multiple measures
and examining a sizable number of cases, and by comparing the gen-
eralizations we observe in these data with the specific dynamics we
observe in our case studies of the Philippines and Pakistan, we can
have a substantial degree of confidence in the results.
Together, these variables can explain a great deal about why
counterinsurgents adopt particular counterinsurgency strategies and
practices. Table 2.2 provides a typology of counterinsurgent regimes
based on the characteristics outlined in this discussion, and it offers
predictions about the type of strategy each type of regime is most likely
to adopt. Chapter Three reviews the record of recent insurgencies to
determine the extent to which regime characteristics do, in fact, shape
counterinsurgent behavior in the ways that the findings reviewed above
would suggest. The case studies in Chapters Four and Five examine
how differences in political inclusiveness and state capacity over time
and between subregions can provide a more nuanced understanding of
the variation in counterinsurgency policies within particular countries.

Table 2.2
A Typology of Counterinsurgent Regimes

Regime Characteristics

Predicted
Political State Dominant
Inclusivity Capacity n Cases Strategy

High High 11 Colombia (1978–), Croatia (1992–1995), Classic

El Salvador (1979–1992), Israel (1987– counter-
1997), Israel (2000–), Peru (1980–1996), insurgency
Philippines (1971–), Philippines (1972–
1992), Russia (1994–1996), Turkey (1984–
1999), UK (1971–1998)

49 Data on government forces and country populations during the period of conflict are
from The World Bank’s World Development Indicators (WDI) dataset.
32 Countering Others’ Insurgencies

Table 2.2—Continued

Regime Characteristics

Predicted
Political State Dominant
Inclusivity Capacity n Cases Strategy

High Low 12 Central African Republic (1996–1997), Informal

Guinea Bissau (1998–1999), India accommo-
(1984–1993), India (1989–), India (1990–), dation
Indonesia (1999–2005), Mali (1990–1995),
Nepal (1996), Papua New Guinea (1988–
1998), Senegal (1989–1999), Sri Lanka
(1983–2002), Sri Lanka (2003–2009)

Low High 15 Algeria (1992–), Azerbaijan (1991–1994), Strong-state

Bosnia (1992–1995), Congo-Brazzaville repression
(1993–1997), Congo-Brazzaville (1998–
1999), Djibouti (1991–1994), Georgia
(1992–1994), Iran (2005–), Moldova (1991–
1992), Morocco (1975–1991), Nicaragua
(1981–1990), Russia (1999–2007), South
Africa (1976–1994), Yugoslavia (1991),
Yugoslavia (1998–1999)

Low Low 49 Afghanistan (1978–1992), Afghanistan Contain-

(1992–1996), Afghanistan (1996–2001), ment
Afghanistan (2003–), Angola (1975–1991),
Angola (1992–1994), Angola (1997–2002),
Bangladesh (1974–1997), Burundi (1991–),
Cambodia (1979–1991), Chad (1980–1994),
Chad (1994–1997), Chad (2002–), Congo-
Kinshasa (DRC) (1996–1997), Congo-
Kinshasa (DRC) (1998–), Egypt (1994–1997),
Ethiopia (1974–1991), Ethiopia (1978–1991),
Guatemala (1978–1994), Haiti (1991–1995),
Indonesia (1990–1991), Indonesia (1975–
1999), Iraq (1985–1996), Iraq (1991–1993),
Iraq (2004–), Kenya (1991–1993), Lebanon
(1975–1999), Liberia (1989–1990), Liberia
(1992–1997), Liberia (1999–2003),
Mozambique (1976–1992), Myanmar
(1960–1995), Nepal (2000–2006), Pakistan
(1994–1999), Pakistan (2007–), Rwanda
(1990–1993), Rwanda (1994), Sierra Leone
(1991–1996), Sierra Leone (1997–1998),
Sierra Leone (1998–2001), Somalia
(1988–1991), Somalia (1991–2002), Somalia
(2006–), Sudan (1983–2005), Sudan (2003),
Tajikistan (1992–1997), Uganda (1995–),
Uganda (1990–1992), Yemen (1994)
CHAPTER THREE

Quantitative Analysis of Counterinsurgency

In this chapter, we quantitatively assesses the extent of the relationship

between regime characteristics and patterns of conflict termination
and use of indiscriminate violence. More specifically, we use the simple
statistical technique of cross-tabulation to assess the historical record of
all the insurgencies that have taken place in the post–Cold War era—
a total of 89 cases, of which 17 were ongoing as of 2008 (the last year
in our dataset)1—to examine the extent to which a regime’s political
inclusivity, state capacity, and military superiority are associated with
several key outcomes:

• Decisive war termination, including both government victories

and negotiated settlements
• Durable peace
• Indiscriminate use of force by the regime (i.e., large-scale violence
directed against non-combatants).

We use as transparent a method as possible so that readers can under-

stand exactly how the results were derived, which cases were influential
in driving those results, and how they relate to the case studies pre-
sented later in this report.2

1 More specifically, if any portion of an insurgency was fought after 1989 and before 2010,
it was included in this analysis. The original dataset of insurgencies underlying this research
was taken from Doyle and Sambanis, 2006, with data updated to 2010 from Watts et al.,
2012. See the Appendix for the cases and coding notes.
2 More detailed notes on all of the concepts, data, and coding decisions underlying this
quantitative analysis are given in the Appendix. The dataset on which the analysis relied and
more detailed findings are available upon request from the authors.

33
34 Countering Others’ Insurgencies

This analysis is designed primarily to indicate the extent to which

outcomes are associated with regime characteristics. It does not pro-
vide a direct test of the predominant counterinsurgency “models”
adopted by each type of regime, described in Chapter Two. To the
extent that patterns of conflict termination and discrimination in the
use of force correspond to the various counterinsurgency models, how-
ever, the analysis also provides an indirect test of the hypotheses pre-
viously developed. More-inclusive regimes, for instance, are expected
to adopt counterinsurgency models that emphasize accommodation of
the armed opposition and discriminate uses of force. Insofar as political
inclusion is associated with negotiated settlements and the avoidance
of mass killings of non-combatants, it appears to be associated with
the use of the classical counterinsurgency model. The case studies of
the Philippines and Pakistan in Chapters Four and Five, respectively,
examine the precise linkages between the characteristics of counter-
insurgents, the counterinsurgency models they adopt, and the likely
outcomes of their counterinsurgency campaigns.
The following two sections explain why regimes’ political inclu-
siveness, state capacity, and military superiority shape their ability to
accommodate the reconcilable opposition and use violence discrimi-
nately, as well as the implications of these practices for favorable ter-
mination of insurgencies. A final section integrates the quantitative
analysis with the typology of counterinsurgent regimes introduced in
Chapter Two to help decisionmakers understand the counterinsur-
gency strategies and practices that potential partner nations are likely
to adopt.

Accommodation of the Reconcilable Opposition

It is extremely difficult to assess the general record of governments’

success in negotiating with elements of armed opposition. Many deals
exist informally. Some are made with individuals or small groups of
insurgents—side deals so minor that there is often no historical record
of them.
 Quantitative Analysis of Counterinsurgency 35

We can, however, learn a great deal about governments’ abilities

to accommodate the reconcilable opposition by looking at processes of
war termination. Wars can end in a number of ways: through military
victory by the government; military victory by the insurgents; a mili-
tary stalemate followed by a formal, comprehensive peace agreement
between the warring parties that clearly delineates the terms of the
post-conflict order; or a military stalemate followed by informal, tacit
agreements. As Table 3.1 shows, in the post–Cold War era, 17 conflicts
ended in victory for the government, 24 ended in victory for the insur-
gents, 17 ended with a military stalemate followed by a formal peace
agreement accepted by both sides, and 14 ended in a military stalemate
but without a formal settlement.3
From the U.S. perspective, either a victory by a partner govern-
ment or a formal peace agreement that satisfies the regime’s core secu-
rity interests is generally preferable. A cessation of hostilities (or reduc-
tion to an extremely low level of residual violence) can be adequate
from the U.S. perspective, and in many circumstances it may be the
best outcome that is realistically possible. But war termination without
either outright military victory or a formal negotiated settlement risks
the perpetuation of regions without any effective government control—
regions that risk becoming safe havens for terrorists, transit routes for
illegal traffickers, or other sources of regional instability.4 Such war ter-
mination also creates a greater risk that conflict will reignite.5

3 Data on formal peace agreements were taken from Doyle and Sambanis, 2006, and were
updated to 2008 by the authors of this report. See the Appendix for details of the coding
procedures.
4 See, for instance, Rabasa et al., 2007.
5 On the utility of formal peace agreements, see Virginia Page Fortna, “Scraps of Paper?
Agreements and the Durability of Peace,” International Organization, Vol. 57, No. 2, Spring
2003, pp. 337–372. Although Fortna writes specifically about interstate conflicts, the logic is
generally applicable to internal conflicts as well. On how the specific content of peace agree-
ments shapes post-conflict prospects for peace, see, for instance, Caroline Hartzell, Matthew
Hoddie, and Donald Rothchild, “Stabilizing the Peace After Civil War: An Investigation of
Some Key Variables,” International Organization, Vol. 55, No. 1, Winter 2001, pp. 183–208.
36 Countering Others’ Insurgencies

Table 3.1
Conflict Outcomes in the Post–Cold War Era

Number of
Type of Conflict Outcome Conflicts

Government victory 17
Military stalemate, formal peace agreement 17
Military stalemate, no formal peace agreement 14
Insurgent victory 24

We next examine how the characteristics of the counterinsurgent

regime—specifically, its political inclusivity, its reach and effectiveness,
and its military superiority—affect the chances of

• A formally negotiated settlement acceptable to both sides

• A decisive outcome that is acceptable from the standpoint of the
government—either a military victory for the government or a
military stalemate that leads to a formal peace agreement
• An indeterminate outcome—a military stalemate without a
formal peace agreement
• A durable peace, defined here as a peace that lasts at least five
years following the end of a civil war.

This discussion provides a baseline estimate for a regime’s likelihood

of success in counterinsurgency following the core counterinsurgency
principles discussed in Chapter Two.

Political Inclusivity
The record of all conflicts that have terminated in the post–Cold
War era strongly supports the notion that more-democratic countries
are much better able than less-democratic countries to accommo-
date armed opposition movements without sacrificing core interests. 6
Figure 3.1 shows the relationship between a government’s degree of
democracy and its likelihood of securing a formally negotiated peace
to end an internal conflict. More-democratic countries are those that

6 Conflicts that were ongoing as of 2008 were excluded from the analysis.
 Quantitative Analysis of Counterinsurgency 37

Figure 3.1
Relationship Between Democracy and Negotiated Settlements

100
Percentage of cases terminating
in negotiated settlements

80

60

40

20

0
More-democratic Less-democratic
countries countries
RAND RR513-3.1

score in the upper half of the Freedom House democracy index over
the course of a conflict, and less-democratic countries are those that
score in the bottom half of this index.7 As can be seen from Figure 3.1,
nearly half of the conflicts in more-democratic countries end in for-
mally negotiated settlements, while less than one-quarter of those in
less-democratic countries do.
Similarly, more-democratic countries are better able to achieve
an acceptable decisive outcome—that is, an end to an internal war in
which the government either wins an outright military victory or is
able to reach a formally negotiated peace agreement with the armed

7 Specifically, an average value of 4.0 on both the Political Rights and Civil Liberties
dimensions of the Freedom House index was used as the break point between more- and
less-democratic countries. Data on negotiated settlements were taken from Doyle and
Sambanis, 2006, updated to 2008 by the authors of this report. c2 = 2.7386, and the rela-
tionship is statistically significant at the 0.1 level (p = 0.098). In this calculation, all cases in
which a large military intervention occurred in the post-conflict period were dropped from
the sample on the assumption that the prospect of such an intervention would heavily condi-
tion the likelihood of a negotiated settlement. The relationship remains statistically signifi-
cant at the 0.1 level if those cases are retained. Data on foreign interventions, including the
distinction between large- and small-scale interventions, were taken from Watts et al., 2012.
38 Countering Others’ Insurgencies

opposition that guarantees each side’s core interests in the absence

of military victory for either side. Figure 3.2 shows that 41 percent
of conflicts in less-democratic countries end in an acceptable deci-
sive outcome, whereas approximately two-thirds of conflicts in more-
democratic states terminate with a favorable outcome.8 The odds of
failure, in other words, are close to twice as great for less-democratic
counterinsurgent regimes.
Finally, more-democratic countries appear better able to sustain
peace in the aftermath of civil wars. As Figure 3.3 shows, 22 percent
of more-democratic countries experience renewed conflict within five
years of the termination of a civil war, whereas almost exactly twice as

Figure 3.2
Relationship Between Democracy and Acceptable Decisive Outcomes

100
Percentage of cases terminating
in an acceptable outcome

80

60

40

20

0
More-democratic Less-democratic
countries countries
RAND RR513-3.2

8 c2 = 3.6409, and the bivariate relationship is statistically significant at the 0.1 level
(p = 0.056). If we are interested only in outright military victories, we find—perhaps
surprisingly—that democracies are somewhat more likely to win, although the difference
is not sufficient to achieve statistical significance. Relatively more-democratic states won
approximately one-quarter of the time (six cases out of 23), whereas non-democracies won
only approximately one-sixth of their counterinsurgencies (11 cases out of 65).
 Quantitative Analysis of Counterinsurgency 39

Figure 3.3
Relationship Between Democracy and Durable Peace

100
Percentage of cases returning to

80
war within five years

60

40

20

0
More-democratic Less-democratic
countries countries
RAND RR513-3.3

many—43 percent—of less-democratic countries experience a return

to war within the same period.9
These results should be sobering for those who recommend exten-
sive use of small-footprint operations. The partner countries in the suc-
cess stories of U.S. small-footprint interventions that have frequently
been cited in recent years—the Philippines, Colombia, and El Salva-
dor in the late Cold War era—all would be characterized as relatively
more democratic, using the measures adopted in this analysis.10 When
we look at all insurgencies that have taken place at least in part in the
post–Cold War era, however, only 24 out of 89 (or 27 percent) coun-

9 c2 = 2.8236 and p = 0.093. Cases in which a large foreign military force was deployed in
the post-conflict period were excluded from this analysis on the assumption that the post-
conflict trajectory of the country was too strongly influenced by the large-scale foreign inter-
vention to yield accurate information about the relationship between the domestic regime
and the conflict outcome. Data on foreign interventions, including the distinction between
large- and small-scale interventions, were taken from Watts et al., 2012.
10 Colombia had an average Freedom House score of 3.2 during the course of its conflict;
the Philippines had an average score of 3.5 throughout the period of its conflict with Muslim
separatist groups; and El Salvador had an average score of 3.8.
40 Countering Others’ Insurgencies

tries would be characterized as relatively more democratic.11 The “suc-

cess stories” of U.S. small-footprint operations, in other words, had
levels of political inclusion experienced by only approximately one-
quarter of all counterinsurgent regimes.12

State Capacity
The data from the post–Cold War period also support the proposition
that state capacity influences a counterinsurgent’s ability to achieve an
acceptable outcome to a conflict.13 As discussed previously, this analy-
sis uses two measures of state capacity: state reach, a measure of a state’s
ability to penetrate social relations and exercise meaningful effects in
the population’s day-to-day lives, and government effectiveness, a mea-
sure of the bureaucratic capacity of the state.14 Counterinsurgents with
strong state capacity (as measured using either indicator) are more

11 These 89 cases include ongoing conflicts that were dropped from the statistical analysis of
outcomes.
12 The relationship between political inclusiveness and political accommodation is even
stronger if we focus only on a counterinsurgent regime’s level of democracy at the tail end of
a conflict. It may be that insurgents are less concerned about a regime’s democratic character
at the start of a conflict than they are about its character toward the end, when it is attempt-
ing to negotiate an end to fighting. On the basis of a counterinsurgent regime’s average
democracy score in the final two years of a conflict, more-democratic governments are more
than three times as likely to achieve a formally negotiated settlement than less-democratic
regimes, securing such settlements in 55 percent and 17 percent of cases, respectively (c2 =
9.2960 and p = 0.002, an extremely strong statistical relationship). Similarly, governments
that are more democratic toward the end of a conflict are much better able to achieve accept-
able decisive outcomes (again, defined as either a military victory by the government or a
military stalemate that is formally codified in a peace deal accepted by both sides). Approxi-
mately 70 percent of more-democratic governments are able to achieve such decisive results,
whereas just over 40 percent of less-democratic governments (c2 = 4.3799) can do so, and the
relationship is statistically significant at the 0.05 level, with p = 0.036). Finally, a government
that is more democratic in the final two years of a conflict is more likely to sustain a peace.
Nearly half (48 percent) of less-democratic regimes experienced a return to conflict within
five years of the end of a civil war, whereas only one-fifth (20 percent) of more-democratic
regimes did (c2 = 4.2656 and p = 0.039).
13 As in the analysis of political inclusiveness, the analysis of the effects of state capacity on
conflict outcomes excluded all cases of conflict that were still ongoing as of 2008.
14 States with high state reach are those that score above the midpoint (0.5) of Heltermann’s
state-reach index, while states with high government effectiveness are those with scores above
 Quantitative Analysis of Counterinsurgency 41

likely to achieve negotiated settlements to conflicts, and they are more

likely to sustain peace. These relationships, however, are insufficient to
be statistically significant. Nevertheless, counterinsurgents with strong
state capacity are nearly twice as likely as weaker states to achieve an
acceptable conflict outcome (i.e., either an outright military victory
by the government or a formally negotiated peace deal in which both
sides are able to secure their core interests). As shown in Figure 3.4,
approximately three-quarters of regimes with high state reach achieve
acceptable decisive outcomes, whereas only 41 percent of regimes with
low state reach do so.15
The World Bank’s measure of government effectiveness tells a
very similar story. Regimes with high government effectiveness are
able to secure an acceptable decisive outcome two-thirds of the time,
whereas those with low government effectiveness are able to achieve
similar outcomes little more than one-third of the time (Figure 3.5).16

Figure 3.4
Relationship Between State Reach and Acceptable Decisive Outcomes

100
Percentage of cases terminating
with an acceptable outcome

80

60

40

20

0
High state reach Low state reach
RAND RR513-3.4

the median for conflict-affected countries (those that experienced civil wars in the post–Cold
War era).
15 c2 = 6.2045 and p = 0.013.
16 c2 = 4.8673 and p = 0.027.
42 Countering Others’ Insurgencies

Figure 3.5
Relationship Between Government Effectiveness and Acceptable Decisive
Outcomes

100
Percentage of cases terminating
with an acceptable outcome

80

60

40

20

0
High government Low government
effectiveness effectiveness
RAND RR513-3.5

Low-capacity states also appear substantially more likely to have

conflicts terminate in an indeterminate outcome—that is, an end to
large-scale conflict without either a military victory by either side or a
formal peace deal. In fact, all 14 cases in which a war ended without a
military victory and without a formal peace deal occurred in countries
with low state reach (Figure 3.6).17
Once again, the record of recent counterinsurgencies paints a
stark picture. Counterinsurgent regimes with higher state capacity are
able to achieve acceptable outcomes in two-thirds to three-quarters
(depending on the precise measure used) of the conflicts in which they
are involved. Those with lower state capacity achieve acceptable out-
comes in 38 and 41 percent of the conflicts—little more than half the
rate of success of stronger regimes. Just as with democracy, moreover,
the oft-cited “success stories” of U.S. small-footprint interventions—
Colombia, the Philippines, and El Salvador—score favorably on this
dimension, whether one uses a measure of state reach or government

17 c2 = 7.6469 and p = 0.006. Interestingly, the relationship between government effective-

ness and indeterminate outcomes is not nearly as strong, achieving nowhere near statistical
significance.
 Quantitative Analysis of Counterinsurgency 43

Figure 3.6
Relationship Between State Reach and Indeterminate Outcomes

100
Percentage of cases terminating
with an indeterminate outcome

80

60

40

20

0
High state reach Low state reach

RAND RR513-3.6

effectiveness. Only 30 percent of all counterinsurgents in the post–

Cold War era score above the midpoint of the state reach index, how-
ever, suggesting that the cases of Colombia, the Philippines, and El
Salvador are far from typical. Much more common are countries like
Afghanistan, Mali, Somalia, and Yemen—countries with much lower
state capacity, where the United States has had generally much less-
favorable experiences.

Military Sufficiency
Consistent with the findings in Chapter Two, data from the post–Cold
War era suggest an ambiguous relationship between a government’s
military capacity and conflict outcomes. High force-to-population
ratios and high force-to-force ratios are associated with somewhat
better odds of a government victory, but the difference between more
and less militarily-capable governments is not sufficient to be statisti-
cally significant in nearly any of the specifications we examined.18

18 We used two alternative break points for distinguishing between governments with high
and low force-to-population ratios: a 10:1,000 ratio and a 5:1,000 ratio. (We used these low
ratios rather than the doctrinally accepted 20:1,000 because, to the best of our knowledge,
there are no systematic data on police forces for all of the countries in our sample. Thus,
our calculations relied on only the military portion of a government’s security forces. Of
44 Countering Others’ Insurgencies

Similarly, governments with greater military capacity appear no

more capable of accommodating the reconcilable opposition, and gov-
ernments with higher force-to-population and force-to-force ratios
appear no more likely to achieve negotiated solutions to a conflict, nor
are they more likely to achieve a favorable decisive outcome.19 Finally,
governments with higher force ratios do not appear significantly more
capable of sustaining peace with the armed opposition—either in
instances where neither side was victorious or in cases of terminated
conflict—than governments with lower military capacity.20
What are we to make of these results? There are at least three pos-
sible explanations for the fact that greater military capacities do not
consistently yield improved counterinsurgency outcomes: (1) there may
be no relationship between military capacity and counterinsurgency

the 88 cases in our sample, 41 had force-to-population ratios of 5:1,000 or greater, so this
lower threshold is close to the median value.) We also relied on two alternative specifications
of force-to-force ratios. The first one used the doctrinally accepted ratio of 10 government
forces for every insurgent. Again, however, our data include only military forces, so such a
threshold is potentially artificially high. Consequently, we also used the median force-to-
force ratio from all of the countries in our sample as a second break point between high- and
low-capacity governments. Finally, we examined both the outcomes of all cases of insurgency
that ended in the post–Cold War era and the outcomes of only those cases in which a large
external intervention force did not deploy in the post-conflict period (on the assumption
that the prospect of such an intervention force would strongly influence the outcome of a
conflict). In all of these cases, the relationship operated in the expected direction: Govern-
ments with greater military capacity generally had somewhat better odds of winning. In only
one specification (using force-to-force ratios with the break point between high- and low-
capacity governments determined by the median value and with all cases for which we did
not have relatively precise data on insurgent strength dropped from the sample), however, did
the relationship achieve statistical significance.
19 Depending on the precise specification of the relationship used, the results can range
from no relationship whatsoever (i.e., high- and low-capacity states have almost exactly
the same odds of achieving a favorable decisive outcome) to a weak relationship (i.e., high-
capacity states tend to achieve better outcomes, but the relationship is not sufficiently strong
to be statistically significant at even the 0.1 level).
20 As with the other relationships in this analysis, higher force ratios typically lead to slightly
better odds of favorable outcomes but not sufficiently improved odds for the relationship to
be statistically significant. In some specifications, the relationship is actually reversed, and
higher force ratios are associated with slightly worse outcomes—but again, these relation-
ships achieve nowhere near statistical significance.
 Quantitative Analysis of Counterinsurgency 45

outcomes; (2) there may be a relationship, but our methods are in-
adequate to reveal it; or (3) there may be a relationship between military
capabilities and counterinsurgency outcomes, but measures of simple
numerical superiority (i.e., capacity) are inadequate to capture it.
The first possibility flies in the face of U.S. military doctrine and
the writings of many of the most respected counterinsurgency theo-
rists. The possibility should not be dismissed, however. As discussed in
the Chapter Two, counterinsurgency warfare has changed profoundly
in the past few decades due to the proliferation of small arms and light
weapons, changes in funding streams, the spread of guerrilla tactics,
and a radically altered international context, among other reasons.
Whereas military victory was the norm in the decades following World
War II, a minority of insurgencies terminate in victory today. It is pos-
sible that although they are still important, military means simply do
not exercise the same degree of influence on conflict outcomes that
they once did.
It is also possible that our methods are inadequate to capture the
relationship between military capacity and conflict outcomes. In par-
ticular, it could be that especially dangerous insurgencies cause gov-
ernments to build large security forces. In this case, our results could
be biased by adverse selection: The countries with particularly difficult
insurgencies would create large security forces, leading to an appar-
ently inverse relationship between the size of the forces and success in
battling insurgencies. Such biases may be particularly apparent when
using force-to-population ratios. Our use of the force-to-force measure,
however, should at least partially alleviate this concern. Insofar as the
number of insurgents is an indication of the threat they pose to the
government, force-to-force ratios automatically control for the possi-
bility of adverse selection. Force-to-force ratios, however, also failed to
reveal any consistent relationship between military superiority and the
ability to secure favorable outcomes.
Finally, and perhaps most plausibly, the lack of a statistically sig-
nificant relationship between military capacity and conflict outcomes
could be driven by our measure of military superiority. Force ratios
capture only the number of government forces. They say nothing about
their equipment; their tactics, techniques and procedures; or their dis-
46 Countering Others’ Insurgencies

cipline and accountability. It may well be that a more holistic measure

of military sufficiency—if one were available—would reveal a much
stronger relationship with conflict outcomes. This “quality over quan-
tity” argument is entirely plausible, and it also has significant implica-
tions for U.S. partnership policies—a point to which the discussion
will return later in this report.

Discriminate Use of Violence

The second overarching principle of counterinsurgency is that the use

of force must be discriminate; otherwise, ordinary citizens have noth-
ing to gain by not supporting the insurgency. To better understand
when regimes are likely to employ violence indiscriminately in the pros-
ecution of a counterinsurgency, we examined the relationship between
democracy, state capacity, and force sufficiency on the one hand and
intentional mass killings of civilians by government forces on the other.
Data on deaths in war zones—especially data on civilian deaths in
an insurgency environment—are notoriously elusive. Wary of unwar-
ranted precision and of missing important cases of indiscriminate vio-
lence, most studies of this issue have relied on a dichotomous coding of
mass killings (i.e., a simple yes-or-no categorization) with a low thresh-
old, so that even relatively low levels of civilian deaths are enough to
count as mass killings or indiscriminate violence. The problem with
such an approach is that it typically finds that nearly all governments
engage in some degree of indiscriminate violence during counter-
insurgency.21 Unfortunately, this finding tells us what we already know:

21 Chris Paul and his colleagues, for instance, find that in only five out of 85 cases did
government forces avoid “excessive collateral damage, disproportionate use of force, or
other illegitimate applications of force.” See Paul, Clarke, and Grill, 2010. Similarly, when
we compare Ulfelder and Valentino’s list of “mass killings” by government forces with the
list of recent civil wars, we find that governments engaged in mass killings (as defined by
Ulfelder and Valentino) in more than two-thirds of all cases of civil war or insurgency. See
Jay Ulfelder and Benjamin Valentino, Assessing Risks of State-Sponsored Mass Killing, Politi-
cal Instability Task Force, February 2008, with data updated through 2010 by the authors
of this report, and the list of civil wars taken from Doyle and Sambanis, 2006, updated in
Watts et al., 2012.
 Quantitative Analysis of Counterinsurgency 47

Counterinsurgencies are “dirty wars.” It also fails to distinguish hor-

rific campaigns of violence from much lower levels of abuse.
This report adopts a somewhat higher threshold: Intentionally
killing 2,000 or more civilians per year throughout the course of an
entire counterinsurgency is considered mass killing or truly indiscrim-
inate violence.22 Lesser totals might still be deplorable, but they are
considered lesser cases of repression rather than persistent, large-scale,
indiscriminate violence. By this standard, there were 26 instances of
mass killings in the 89 civil wars fought in the post–Cold War era, or
approximately 29 percent of all the counterinsurgencies.
Based on this standard, data from the past two decades strongly
support the contention that democratic accountability mechanisms are
critical to preventing indiscriminate violence. The relationship between
military sufficiency and the discriminate use of force is much more
complex—a finding with implications that will be discussed in the case
studies in Chapters Four and Five.

Political Inclusivity
A broad review of the historical record of the past two decades strongly
supports the claim that democratic accountability is critically impor-
tant to preventing abuses and atrocities by government forces. Only
two relatively democratic countries (fewer than 10 percent of the total)
engaged in mass killings of civilians during the course of counterinsur-
gency campaigns in this period.23 In contrast, less-democratic countries
engaged in mass killings in more than one-third of all counterinsur-
gencies.24 Figure 3.7 shows the stark difference between the two types
of regimes in their propensity to use indiscriminate violence. Demo-

22 We used the most comprehensive data on large-scale killings by government forces that
we could find—Ulfelder and Valentino’s mass-killings database, developed as part of the
CIA-funded Political Instability Task Force. We then matched these events to Doyle and
Sambanis’ list of civil wars or insurgencies. We took the average of the high-end and low-end
estimates of civilian deaths, then determined the average number of deaths per year of the
conflict.
23As before, relatively democratic is defined as scoring in the top half of Freedom House’s
democracy index (i.e., an average combined score of 4 or less during the course of a conflict).
24 c2 = 6.9282, and the results are statistically significant at the 0.01 level (p = 0.008).
48 Countering Others’ Insurgencies

Figure 3.7
Relationship Between Democracy and Mass Killings of Civilians by
Government Forces

100
Percentage of cases with mass
killings by government forces

80

60

40

20

0
More-democratic Less-democratic
governments governments
RAND RR513-3.7

cratic accountability, of course, is only one form of accountability,

and a majority of even authoritarian regimes managed to avoid such
high levels of indiscriminate violence. The risks of supporting less-
democratic regimes in their counterinsurgency campaigns, however,
should be evident.

State Capacity
As discussed in Chapter Two, state capacity might influence a regime’s
proclivity toward indiscriminate violence in two ways: Weak states may
be more likely to use indiscriminate violence because they lack suf-
ficient positive inducements to persuade discontented populations to
side with them, or they may be more likely to engage in abuses because
they lack sufficient oversight and control over their fielded forces. The
state-reach variable is probably a better choice of measures to test the
argument that weak regimes rely excessively on “sticks” because they
have too few “carrots.” The government-effectiveness variable is prob-
ably more appropriate for assessing the extent to which more-capable
governments are able sustain more-disciplined forces in the field.
 Quantitative Analysis of Counterinsurgency 49

The state-reach variable does not reveal any clear relationship

between state reach and indiscriminate violence.25 Government effec-
tiveness, however, does demonstrate a statistically significant relationship
with the likelihood of state-perpetrated mass killings.26 As Figure 3.8
shows, regimes characterized by relatively high government effective-
ness resorted to indiscriminate killings in less than one-fifth of the
counterinsurgencies they fought, whereas weaker states engaged in
such killings nearly twice as often.

Military Sufficiency
An examination of the relationship between force ratios and mass kill-
ings of civilians produces seemingly inconsistent results. Force-to-force
ratios provide apparent support for the claim that less-capable regimes
resort to indiscriminate violence in desperation, lacking the military

Figure 3.8
Relationship Between Government Effectiveness and Mass Killings of
Civilians by Government Forces

100
Percentage of cases with mass
killings by government forces

80

60

40

20

0
High state reach Low state reach

RAND RR513-3.8

25 c2 = 0.2026 and p = 0.653 for the bivariate relationship including all cases. The lack of
a clear relationship is not changed by dropping cases in which a large, external interven-
ing force was present during the conflict or by dropping cases in which conflicts had not
concluded.
26 c2 = 3.4599 and p = 0.063.
50 Countering Others’ Insurgencies

means to implement more-precise counterinsurgency campaigns.27 As

Figure 3.9 shows, governments with sufficient military forces (in this
case, a ten-to-one ratio of government to rebel forces, as prescribed in
U.S. military doctrine) appear half as likely to resort to indiscriminate
violence in their prosecution of counterinsurgencies.28
Force-to-population ratios exhibit the opposite tendency. As
shown in Figure 3.10, governments with greater military capacity
(indicated by a ratio of at least five military personnel to every 1,000

Figure 3.9
Relationship Between Force-to-Force Ratio and Mass Killings of Civilians by
Government Forces

100
Percentage of cases with mass
killings by government forces

80

60

40

20

0
High force-to-force Low force-to-force
ratio ratio
RAND RR513-3.9

27 Benjamin Valentino, Paul Huth, and Dylan Balch-Lindsay, “‘Draining the Sea’: Mass
Killing and Guerrilla Warfare,” International Organization, Vol. 58, No. 2, Spring 2004,
pp. 375–407; Alexander B. Downes, “Desperate Times, Desperate Measures: The Causes
of Civilian Victimization in War,” International Security, Vol. 30, No. 4, Spring 2006,
pp. 152–195; Reed M. Wood, Jacob D. Kathman, and Stephen E. Gent, “Armed Intervention
and Civilian Victimization in Intrastate Conflicts,” Journal of Peace Research, Vol. 49, No. 5,
2012, pp. 647–660; David Fielding and Anja Shortland, “The Dynamics of Terror During
the Peruvian Civil War,” Journal of Peace Research, Vol. 49, No. 6, 2012, pp. 847–862.
28 c2 = 3.8651 and p = 0.049. If we instead use the median force-to-force ratio among all
cases of conflict in our sample as the break point between more- and less-capable govern-
ments, the relationship remains much the same, although it is not quite as strong and nar-
rowly misses achieving statistical significance.
 Quantitative Analysis of Counterinsurgency 51

Figure 3.10
Relationship Between Force-to-Population Ratio and Mass Killings of
Civilians by Government Forces

100
Percentage of cases with mass
killings by government forces

80

60

40

20

0
High force-to-population Low force-to-population
ratio (≥5/1,000) ratio (<5/1,000)
RAND RR513-3.10

inhabitants, with police forces excluded) appear substantially more

likely to engage in indiscriminate violence.29
How can we explain these seemingly paradoxical results? One pos-
sibility is that the simple relationship between military forces and mass
killings misapprehends the causal dynamics at work. It could be that
large insurgencies cause governments to create larger security forces
and to use violence indiscriminately. In this case, both large security
sectors and mass killings are effects of powerful insurgencies, rather
than the large security being the cause of the mass killings. Or it could
be that a large security apparatus creates rebellion—either by tempting
a regime to use those capabilities for repression or by having weapons
and other coercive capabilities escape state control.30 The end result
may be that larger security forces create a rebellion that the regime
then attempts to crush through indiscriminate violence. In this case,

29 c2 = 3.0911 and p = 0.079. If we use a 10:1,000 force-to-population ratio as the break

point between counterinsurgents with military sufficiency and those without, countries with
low force-to-population ratios remain substantially less likely to use force indiscriminately,
but the relationship does not achieve statistical significance.
30 See, for instance, Clapham, 2004, p. 91.
52 Countering Others’ Insurgencies

the large security sector is the cause of both the large insurgency and
the mass killings.
Much is at stake in parsing these alternative explanations. If the
first explanation is true, the appropriate policy response may well be
to build stronger security sectors to weaken insurgents. If the second
explanation is correct, such a policy would pour oil on the fire.
Unfortunately, the simple quantitative techniques used here do
not allow us to determine which of these rival explanations is correct.
Even the most sophisticated statistical techniques would provide only
partial insight into this question.31 To establish the conditions under
which strong military capabilities are likely to yield more or less dis-
criminate use of violence, it is helpful to turn to case studies. First,
however, it will be useful to integrate these quantitative findings into
a framework for understanding different types of counterinsurgent
regimes and the strategies they are likely to adopt.

Integrating Quantitative Findings on Counterinsurgent

Strategies

So far, the quantitative analysis has proceeded by examining each of

the relevant factors in isolation. If we return to the typology of coun-
terinsurgent regimes presented in Table 3.1, however, we can see how
these factors combine to influence the ways in which counterinsur-
gency campaigns are conducted and terminate.

Conflict Termination
Table 3.2 shows the relationship between different types of counter-
insurgent regimes and the outcomes of the counterinsurgency cam-
paigns they conduct.32 As described above, acceptable outcomes are
those in which the government either wins an outright military victory
or is able to achieve a formal negotiated solution to the conflict in the

31 Experimental methods provide the strongest identification strategies, but they are diffi-
cult to devise in such circumstances, and they provide results that are highly context-specific.
32 Cases in which conflicts had not terminated as of 2010 were not included in this table.
 Quantitative Analysis of Counterinsurgency 53

Table 3.2
Counterinsurgent Regimes and Conflict Outcomes

Regime Characteristics Cases

%
Without
Accept-
able % of
Political State No Acceptable Out- All
Inclusivity Capacity Acceptable Outcome Outcome comes Cases
High High Croatia (1992–1995), El Russia (1994–1996) 13 13
Salvador (1979–1992),
Israel (1987–1997), Peru
(1980–1996), Philippines
(1972–1992), Turkey
(1984–1999), UK (1971–
1998)
High Low CAR (1996–1997), India Guinea Bissau (1998– 50 14
(1984–1993), Indonesia 1999), Nepal (1996),
(1999–2005), Mali Papua New Guinea
(1990–1995), Sri Lanka (1988–1998), Senegal
(2003–2009) (1989–1999), Sri Lanka
(1983–2002)
Low High Bosnia (1992–1995), Azerbaijan (1991– 50 17
Congo-Brazzaville 1994), Congo-
(1998–1999), Djibouti Brazzaville (1993–
(1991–1994), Morocco 1997), Georgia (1991–
(1975–1991), Nicaragua 1992), Georgia (1992–
(1981–1990), Russia 1994), Moldova (1991–
(1999–2007), South Africa 1992), Yugoslavia
(1976–1994) (1991), Yugoslavia
(1998–1999)
Low Low Angola (1997–2002), Afghanistan (1978– 60 56
Bangladesh (1974–1997), 1992), Afghanistan
Cambodia (1979–1991), (1992–1996), Afghani-
Egypt (1994–1997), stan (1996–2001),
Guatemala (1978–1994), Angola (1975–1991),
Indonesia (1990–1991), Angola (1992–1994),
Iraq (1991–1993), Chad (1980–1994),
Lebanon (1975–1999), Chad (1994–1997),
Mozambique (1976– Congo-Kinshasa
1992), Myanmar (1960– (DRC) (1996–1997),
1995), Nepal (2000– Ethiopia (1974–1991),
2006), Rwanda (1990– Ethiopia (1978–1991),
1993), Rwanda (1994), Haiti (1991–1995),
Sierra Leone (1998–2001), Indonesia (1975–1999),
Tajikistan (1992–1997), Iraq (1985–1996),
Yemen (1994) Kenya (1991–1993),
Liberia (1989–1990),
Liberia (1992–1997),
Liberia (1999–2003),
Pakistan (1994–1999),
Sierra Leone (1991–
1996), Sierra Leone
(1997–1998), Somalia
(1988–1991), Somalia
(1991–2002), Sudan
(1983–2005), Uganda
(1990–1992)
54 Countering Others’ Insurgencies

absence of military victory for either side. Because neither measure of

military sufficiency proved to have a statistically significant effect on
conflict outcomes, this variable was not included in the typology of
counterinsurgent regimes.
Grouping cases of insurgency in this way is revealing. Counterin-
surgent regimes that are both relatively more politically inclusive and
have higher levels of state capacity (as measured using the state-reach
index) are almost exactly twice as likely to reach an acceptable out-
come as regimes that are neither inclusive nor have reasonably high
levels of state capacity. The most-effective regimes—those that are both
relatively inclusive and have high levels of state capacity—achieved an
acceptable outcome in counterinsurgencies in the post–Cold War era
87 percent of the time. The least-effective regimes—those that are nei-
ther politically inclusive nor have high levels of state capacity—secured
acceptable outcomes to such conflicts 40 percent of the time, less than
half the rate of the most-effective counterinsurgents. Expressed another
way, regimes that are neither politically inclusive nor have high levels of
state capacity are nearly five times as likely as inclusive, high-capacity
regimes to fail to secure either a military victory or a formally negoti-
ated peace deal that is acceptable to all parties.
Once again, it is important to recognize that the successful U.S.
small-footprint operations cited in the 2010 Quadrennial Defense
Review and elsewhere—Colombia, El Salvador, and the Philippines—
all occurred in countries with the best possible type of counterin-
surgent regime. But the United States has also launched small-scale
interventions in many less-promising environments—for example, in
Somalia and Yemen—and it is anticipating a small-scale intervention
in Afghanistan after 2014.33 Only 13 percent of the insurgencies in our
sample were directed against the most-effective types of regimes. More
than half (56 percent) of all cases of insurgency occurred in the least-
promising environments, where regimes lacked both political inclusion

33 At the time of writing, between 8,000 and 12,000 foreign troops are anticipated to ini-
tially be in post-2014 Afghanistan. That number is expected to decline in subsequent years to
levels closer to those of Operation Enduring Freedom–Philippines (OEF-P) and similar mis-
sions, and even the initial number is still a small fraction of the 20:1,000 force-to-population
ratio specified in FM 3-24 and similar documents.
 Quantitative Analysis of Counterinsurgency 55

and state capacity.34 The United States should therefore not expect that
its successes in OEF-P and similar operations will necessarily be repli-
cated elsewhere; indeed, the odds are often against such interventions
leading to a decisive outcome.
Of course, just because more than half of all insurgencies occur in
countries governed by the least-favorable type of regime, this does not
mean that more than half of all U.S. military interventions are likely
to occur in such countries. Indeed, the historical record suggests that
a disproportionate share of U.S. military interventions—somewhere in
the vicinity of half—occur in best-case environments.35 This propen-
sity might be explained by the fact that the odds of success are higher
in such countries, or the fact that the United States shares many values
and interests with such states. It is sobering, however, to consider the
number of least-favorable environments in which the United States has
already intervened and the levels of success it has experienced. Roughly
the other half of U.S. interventions have been in these worst-case envi-
ronments, and the outcomes in countries such as Afghanistan, Iraq,
Lebanon, Somalia, and Yemen are hardly promising.
Figure 3.11 presents these differences in conflict outcomes graphi-
cally, breaking them into each of the four categories used for this analy-
sis: government victory, military stalemate followed by a formally
negotiated settlement, military stalemate without a formally negotiated
settlement, and insurgent victory. As shown, high-capacity regimes his-
torically do not experience indeterminate outcomes—that is, military
stalemates unresolved by a formal settlement acceptable to all of the
major warring parties. More-inclusive regimes are much more capable
of achieving negotiated solutions to conflicts, thereby decreasing their
risk of outright defeat.36

34 The proportion of cases in each type of country was calculated from the entire sample of
insurgencies that have taken place in the post–Cold War era, whereas the proportion of cases
ending in an acceptable outcome was calculated using only those cases in which the insur-
gency had actually ended by our data cutoff date.
35 For a list of U.S. interventions, see Watts et al., 2012. The precise proportion of interven-
tions in each type of country depends on the time period examined and the exact definition
of intervention being used.
36 Figure 3.11 includes all cases of wars that have terminated in the post-conflict era. If we
drop from our sample cases in which large-scale external military interventions occurred in
56 Countering Others’ Insurgencies

Figure 3.11
Conflict Outcomes by Regime Type

100

80 Insurgent victory
Outcome (percentage)

Military stalemate,
60 no settlement
Military stalemate,
40 settlement
Government
victory
20

0
Low High Low High
inclusiveness inclusiveness inclusiveness inclusiveness
Low-capacity High-capacity
government government
RAND RR513-3.11

More-inclusive regimes are not only better able to broker

negotiated peace deals, they are also better able to sustain a peace once
it is achieved. Of particular interest, among low-capacity regimes,
those that are more inclusive demonstrate significantly better odds
of remaining at peace for at least five years after an insurgency ends.
Indeed, as Figure 3.12 reveals, they have a more than 50 percent greater
chance of sustaining a peace than do less-inclusive regimes of similarly
low capacity.

Discriminate Use of Violence

Table 3.3 shows the relationship between regime characteristics and
mass killings of civilians by the government. Just as weak and non-
inclusive regimes are less likely to terminate insurgencies on favorable

either the conflict or the post-conflict periods, the relationship between political inclusion
and negotiated settlements becomes considerably stronger.
 Quantitative Analysis of Counterinsurgency 57

Figure 3.12
Durable Peace by Regime Type

100
Durable peace (percentage)

80
Resumption of
conflict within
60 five years
Durable peace, no
resumption of
40 conflict within
five years
20

0
Low High Low High
inclusiveness inclusiveness inclusiveness inclusiveness
Low-capacity High-capacity
government government
RAND RR513-3.12

terms, they are also less likely to use violence discriminately.37 In only
two insurgencies—less than 9 percent of the total—did relatively more
politically inclusive regimes engage in the mass killings of civilians, and
both of those regimes were only marginally democratic. In contrast,
there were 24 cases of mass killings among less politically inclusive
regimes, with the worst type of counterinsurgent regimes (those exhib-
iting low inclusion and low state capacity) resorting to mass killings 39
percent of the time. The worst type of counterinsurgents (shown in the
bottom row of Table 3.3) were more than four times as likely to use
violence indiscriminately than more-inclusive regimes such as Colom-
bia and the Philippines in recent years, and they comprise well over
half of all regimes fighting insurgencies.

37 As in Table 3.1, military superiority is not shown as a variable in Table 3.2 because of the
ambiguous quantitative findings on the relationship between military capacity and mass
killings of civilians.
58 Countering Others’ Insurgencies

Table 3.3
Counterinsurgent Regimes and Mass Killings

Regime
Characteristics Cases

Political State % with % of

Inclusiv- Capa- Mass All
ity city No Mass Killings Mass Killings Killings Cases
High High Colombia (1978–), Croatia El Salvador (1979– 18 13
(1992–1995), Israel (1987–1997), 1992), Russia (1994–
Israel (2000–), Peru (1980–1996), 1996)
Philippines (1971–), Philippines
(1972–1992), Turkey (1984–1999),
UK (1971–1998)
High Low CAR (1996–1997), Guinea Bissau 0 14
(1998–1999), India (1984–1993),
India (1989–), India (1990–),
Indonesia (1999–2005), Mali
(1990–1995), Nepal (1996),
Papua New Guinea (1988–1998),
Senegal (1989–1999), Sri Lanka
(1983–2002), Sri Lanka (2003–
2009)
Low High Azerbaijan (1991–1994), Congo– Algeria (1992–), 31 17
Brazzaville (1993–1997), Congo– Bosnia (1992–1995),
Brazzaville (1998–1999), Djibouti Russia (1999–2007),
(1991–1994), Georgia (1991– Yugoslavia (1991),
1992), Georgia (1992–1994), Iran Yugoslavia (1998–
(2005–), Moldova (1991–1992), 1999)
Morocco (1975–1991), Nicaragua
(1981–1990), South Africa (1976–
1994)
Low Low Afghanistan (1992–1996), Afghanistan 39 56
Afghanistan (1996–2001), (1978–1992), Angola
Afghanistan (2003–), (1975–1991), Angola
Bangladesh (1974–1997), (1992–1994), Angola
Cambodia (1979–1991), Chad (1997–2002), Burun-
(1980–1994), Chad (1994–1997), di (1991–), Congo-
Chad (2002–), Congo-Kinshasa Kinshasa (DRC)
(DRC) (1996–1997), Egypt (1998–), Ethiopia
(1994–1997), Guatemala (1978– (1974–1991), Ethio-
1994), Haiti (1991–1995), Iraq pia (1978–1991),
(2004–), Kenya (1991–1993), Indonesia (1975–
Indonesia (1990–1991), Lebanon 1999), Iraq (1985–
(1975–1999), Liberia (1992– 1996), Iraq (1991–
1997), Liberia (1999–2003), 1993), Liberia (1989–
Mozambique (1976–1992), 1990), Myanmar
Nepal (2000–2006), Pakistan (1960–1995),
(1994–1999), Pakistan (2007–), Rwanda (1990–
Sierra Leone (1991–1996), Sierra 1993), Rwanda
Leone (1997–1998), Sierra Leone (1994), Somalia
(1998–2001), Somalia (1991– (1988–1991), Sudan
2002), Somalia (2006–), Uganda (1983–2005), Sudan
(1990–1992), Uganda (1995–), (2003–), Tajikistan
Yemen (1994) (1992–1997)
 Quantitative Analysis of Counterinsurgency 59

Using Quantitative and Qualitative Analysis to Reinforce One

Another
This typology of counterinsurgent regimes and the quantitative analy-
sis of counterinsurgency in the post–Cold War era not only provides
a broad overview of counterinsurgency practice, it also informed our
selection of cases for in-depth study.
Drawing on well-established social science methods, we chose
cases that exhibit significant variation on the causal variables of inter-
est.38 Specifically, the Philippines represents a case of a highly promising
counterinsurgent, while Pakistan represents a much more challenging
case. The Philippines ranks among the more inclusive counterinsurgent
regimes for the post–Cold War era, and it possesses relatively greater
state capacity, regardless of whether state capacity is approximated
using state-reach or government-effectiveness measures. Pakistan, in
contrast, has consistently ranked among the less inclusive counterin-
surgent regimes in the post–Cold War era. Its degree of state capacity
is more ambiguous; it ranks in the weaker group of counterinsurgent
regimes, using the state-reach measure, but in the more-inclusive half,
using the government-effectiveness measure.39
Of course, there are many other differences between the two
countries. Ideally, all of the other factors would be as similar as pos-
sible so that their effects would be “controlled” and only the differences
in their extent of political inclusion and state capacity would be left to

38 See Gary King, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry: Scientific
Inference in Qualitative Research, Princeton, N.J.: Princeton University Press, 1994.
39 In the post–Cold War period, the Philippines have an average Freedom House score
of 2.9 (approximately two points away from complete democracy), while Pakistan has an
average score of 4.9 (two points higher than the Philippines and approximately two points
lower than complete autocracy along Freedom House’s scale from 1 to 7). The Philippines’
Freedom House ratings during both of its insurgencies are among the 19 highest-rated (most
democratic) of the 89 cases of insurgency in the post–Cold War era, while Pakistan’s ratings
from its insurgencies are both within the top 49. The Philippines’ government-effectiveness
rating places it among the top 25 cases, while Pakistan’s rating places it among the top 47.
Similarly, the Philippines’ state-reach score ranks it among the top 14 cases, while Pakistan’s
is among the top 39. Taken together, these statistics suggest that the Philippines typically
has ranked in the quartile of cases with the most-favorable characteristics (and often much
higher), whereas Pakistan has ranked roughly in the middle of all cases. Complete data are
available from the authors upon request.
60 Countering Others’ Insurgencies

explain differences in outcomes. Such a perfectly controlled compari-

son is obviously impossible. Nevertheless, the Philippines and Paki-
stan are similar in enough respects to make the comparison valuable.
Both countries have gross domestic product (GDP) per capita (adjusted
for purchasing-power parity) that ranks them between the average of
low-income and lower-middle-income states.40 Both countries have
received similar levels of net official development assistance per capita
in the post–Cold War period ($10.80 and $9.30 per person in Pakistan
and the Philippines, respectively).41 Although precise data on fatalities
in war zones are difficult to acquire, the numbers of people killed in the
conflicts in Pakistan and the Philippines are also comparable, generally
ranging in the hundreds for most years of the long-running conflicts
in these countries and only occasionally reaching the low thousands.42
In short, the two countries are similar in many relevant respects, even
though the regimes in power have been strikingly different.
It would be wrong, however, to think of the Philippines and Paki-
stan as only two cases. Perhaps even more revealing than the contrasts
between them are the differences in counterinsurgency strategy and
practice that can be observed over time and between regions within
each of them. The case-study chapters that follow therefore examine
how changes in political inclusiveness over time and differences in state
reach from one subregion to another in each country have shaped these
regimes’ policy choices.
The case studies also provide important insights into the effective-
ness of U.S. assistance to counterinsurgent regimes. If the findings of
our analyses are correct, counterinsurgency practices and outcomes are
largely a reflection of the character of counterinsurgent regimes. U.S.
assistance designed to improve the capabilities of partner nations’ secu-

40 All values and definitions were taken from The World Bank’s WDI, using post–Cold War
data.
41 Again, figures and definitions are from the WDI dataset.
42 Figures are from the Peace Research Institute Oslo (PRIO) Battle Deaths Dataset, Ver-
sion 3.0. Information on the composition of this dataset is available in Bethany Lacina and
Nils Petter Gleditsch, “Monitoring Trends in Global Combat: A New Dataset of Battle
Deaths,” European Journal of Population, Vol. 21, Nos. 2–3, 2005, pp. 145–166.
 Quantitative Analysis of Counterinsurgency 61

rity forces is thus unlikely to produce superior outcomes unless these

capabilities are harnessed to an appropriate counterinsurgency strategy.
The choice of counterinsurgency strategy, in turn, should—at least in
part—reflect a potential partner regime’s inclusivity and civil capacity.
American assistance, in other words, might play an important role in
strengthening the counterinsurgency capabilities of a regime that is
already predisposed to a classical counterinsurgency model, but it is
unlikely to be able to induce a partner government to change its basic
orientation toward counterinsurgency. Chapters Four and Five show
that these predictions have generally been fulfilled in the Philippines
and Pakistan. Chapter Six turns to the broader literature on counter-
insurgency and development to suggest that the experiences of these
two countries are generalizable, but also to explore ways in which the
United States might adapt to its limited ability to influence events on
the ground in such contexts.
CHAPTER FOUR

Counterinsurgency in the Philippines

Introduction

The Philippines is often viewed by Americans as an archetypal counter-

insurgency success story and the quintessential model for how effective
partnerships can improve partner nations’ conduct of counterinsur-
gency. In the post-9/11 Philippines, U.S. non-combat support, pri-
marily through a light-footprint advise-and-assist role, has contributed
to the severe degradation of terrorist groups such as the Abu Sayyaf
Group (ASG) and Jemaah Islamiyah (JI) and insurgent groups such as
the Moro Islamic Liberation Front (MILF).
This chapter examines counterinsurgency in the Philippines and
how the United States has influenced its course. It analyzes Philippine
approaches to counterinsurgency from the early Marcos period in
the 1970s through the present, using the analytical framework in
Chapter Two. It describes how key factors described in the analyti-
cal framework—political inclusivity, military superiority, and state
capacity—have varied in the Philippines over time and vis-à-vis par-
ticular insurgent groups and, in turn, assesses the extent to which the
presence or absence of these factors maps onto the discriminate use of
force and the will and ability to successfully accommodate reconcilable
groups or factions of groups.
It also examines how other case-specific contextual factors influ-
enced these outcomes or, alternatively, influenced the presence or
absence of the key explanatory variables. One such factor is the influ-
ence the United States has exerted over Philippine counterinsurgency
practices, particularly in the post-9/11 era, in which it deployed a light-

63
64 Countering Others’ Insurgencies

footprint military intervention that has simultaneously increased Phil-

ippine counterinsurgency capabilities and mitigated the Southeast Asia
terrorist threat to the United States.
This case study suggests the utility of looking to the post-2001
Philippines as a possible model for influencing future counterinsur-
gencies in partner nations. But given that counterinsurgencies do not
typically happen in the first place unless host-nation conditions are
extremely poor, this period in Philippine history provided a decidedly
favorable environment for a small-footprint intervention. As a result,
OEF-P should not be considered a default model for cost-effective,
indirect approaches, but rather as one among numerous possibilities.
The case study shows that practitioners should also heed the lessons
of previous Philippine counterinsurgency experiences—less-favorable
environments in which the government’s policies and practices were
only partially consistent with overarching American preferences and
in which the United States’ influence was relatively limited or con-
strained. On balance, the broader record of Philippine counterinsur-
gency suggests as many warnings as it does models for the future.
The first section of this chapter presents a historical overview of
counterinsurgency and counterterrorism in the Philippines. The next
section analyzes the sources of the Philippine government’s strategies
and practices. We then assess the ways in which the United States was
able to influence Philippine counterinsurgency strategy. The chapter
concludes with broader lessons that can be learned from the Philippine
experience.

Background
Overview of the Counterinsurgency and Counterterrorism
Environment
For most of its post-independence history, the Philippines has been a
democracy, albeit a fragile one characterized by political patronage and
coalitions dominated by relatively narrow, parochial interests. Wide-
spread poverty has hamstrung human development, and armed con-
flict has persisted since before the country’s independence in 1946. Of
 Counterinsurgency in the Philippines 65

the numerous factors that shape the environment in which insurgent

and counterinsurgent forces currently contend in the Philippines, two
in particular stand out: extreme ecological and societal diversity and a
history of intractable, low-intensity conflict.

Physical and Human Terrain

The Philippines is an archipelago of more than 7,000 islands and a
variety of ethnic, tribal, and religious groups. Slightly larger than the
state of Arizona, the Philippines has 300,000 square kilometers of ter-
ritory and shares no land borders with other countries. It is the sev-
enth most populated country in Asia and the twelfth most populated
country in the world, with a population of more than 97 million.1 The
country is both urban and rural—49 percent of Filipinos live in urban
areas, and 51 percent live in rural areas.
The human terrain of the Philippines is also defined by its diver-
sity. Tagalog is the predominant ethnic group, encompassing over
28 percent of the population. The next-largest group is the Cebuanos,
at 13 percent. Perhaps the most striking fact about Philippine ethnic
demographics is its fragmentation, as more than one-fourth of the
Philippine population does not belong to any of the country’s six larg-
est ethnic groups.2 Linguistically, Filipino and English are the Philip-
pines’ official national languages, but eight other major dialects are also
common.3 In terms of religion, recent census data show that 83 per-
cent of Filipinos consider themselves Catholic, while a small minority
(5 percent) of the population is Muslim.4

1 Republic of the Philippines, Commission on Population, “Population Statistics,” web site,

2013.
2 Philippines National Statistics Office, Census 2000: Philippines, Population and Housing
Characteristics, Manila: Republic of the Philippines, National Statistics Office, 2003. The
next largest ethnic groups, by percentage share, are the Ilocanos (9 percent), the Bisaya/
Binisaya (7.6 percent), the Hiligaynon Ilonggo (7.5 percent), and the Bikol (6 percent).
3 The official language of Filipino is based on Tagalog. The eight major dialects are Taga-
log, Cebuano, Ilocano, Hiligaynon or Ilonggo, Bicol, Waray, Pampango, and Pangasinan
(Philippines National Statistics Office, 2003).
4 The Philippines’ Muslim population is heavily concentrated in the southern part of the
country, and this geographical concentration has played an important role in the country’s
armed conflicts.
66 Countering Others’ Insurgencies

Historical Background
Conflict has played a persistent role in the Philippine Islands for cen-
turies. Dating at least to the beginning of the Philippines’ encoun-
ter with Spain in 1521, irregular warfare has endured in the southern
and other parts of the archipelago. After the United States took con-
trol of the Philippines in 1898, a nationalist resistance that had been
active against Spain continued to violently resist U.S. rule. Following
a bloody counterinsurgency campaign, the United States emerged vic-
torious in 1902, but despite tamping down the Philippine nationalist
resistance, it never fully quelled resistance throughout its new territo-
rial holding; violent resistance from the Muslim Moro population in
the southern Philippines was especially persistent.
In 1935, the Philippines was granted self-governing common-
wealth status, with a plan put in place to transition the country to
full independence in 1945. In the meantime, the Philippines became a
key World War II location in the Pacific theater. In 1942, Japan occu-
pied the islands, controlling the country until the United States, along
with Filipino resistance forces, began to reclaim control in 1944 and
regained control in 1945. Following Japan’s surrender, the Republic of
the Philippines attained independence in July 1946. A democratic gov-
ernment was elected to coincide with national independence. Manuel
Roxas was elected the country’s first president in 1946. Foreshadow-
ing later periods on which we focus, the post-independence govern-
ment was almost immediately threatened by armed insurgencies. Most
threatening was the Hukbalahap, or “Huk,” a communist insurgency
that began a rebellion against the Roxas government in 1946 and was
not defeated until 1954.5
The United States played a significant role in the Philippines fol-
lowing the country’s independence. This role was not confined to help-
ing the Government of the Republic of the Philippines (GRP) coun-
ter the Huk insurgency. The GRP, under President Roxas, passed a

5 Harvey Averch and John Koehler, The Huk Rebellion in the Philippines: Quantita-
tive Approaches, Santa Monica, Calif.: RAND Corporation, RM-6254, 1970; Benedict J.
Kerkvliet, The Huk Rebellion: A Study of Peasant Revolt in the Philippines, Lanham, Md.:
Rowman & Littlefield Publishers, 2002.
 Counterinsurgency in the Philippines 67

Military Bases Agreement in 1947 that gave the United States the
right to establish military bases and other installations, which in turn
meant that U.S. troops and materiel would have access to the Philip-
pines. The GRP-U.S. Mutual Defense Agreement (MDA) followed in
1952 to guide the overall direction for U.S.-GRP security coopera-
tion, including U.S. support to the Philippine government to counter
internal threats such as insurgencies.6 The MDA remained in place
after Philippine president Ferdinand Marcos—a close Cold War ally
of the United States—declared martial law in 1972, and it remained
an important mechanism for funneling U.S. aid to the regime until
Marcos was deposed in 1986. Following Marcos’ deposition, the Phil-
ippines transitioned to democracy. However, U.S.-GRP relations were
fragile, and when the U.S.-GRP basing agreement expired in 1991, the
Philippine Congress voted for the closure of Clark Air Base and Subic
Naval Base—two of the largest American military bases outside the
continental United States. Following the closure of the bases, U.S.-
Philippine security cooperation diminished dramatically before a Vis-
iting Forces Agreement between the two countries was signed in May
1999. This agreement provided the domestic legal basis that facilitated
the GRP’s enhanced security cooperation with the United States in the
2000s.

Insurgent and Terrorist Organizations

Today, the Philippines’ main internal threats are the New People’s
Army (NPA), a communist insurgency that is active throughout the
country; the MILF, an Islamist-separatist movement that is active in
the south; and the ASG, a thuggish Salafi jihadist organization based
in the southern provinces of Basilan and Sulu. The ASG has links to
al Qaeda and JI but is equally well known for its high-profile criminal
activities, including numerous kidnap-for-ransom operations targeting

6 Sharon Advincula Caringal, “The Impact of U.S. Military Assistance on the Communist
and Secessionist Conundrum in the Philippines,” International Journal of Arts and Sciences,
Vol. 3, No. 15, 2010, p. 437. Illustrating the rupture in U.S.-GRP relations marked by the
closure of Clark and Subic, the U.S. Ambassador to the Philippines at the time, Richard
Solomon, recalled that “dealing with the base closures was practically the only thing I did as
Ambassador” (interview with Richard Solomon, Arlington, Va., June 2013).
68 Countering Others’ Insurgencies

Westerners. None of the groups currently pose an existential threat

to the GRP. They do, however, foment widespread instability, under-
mine foreign investment, participate in rampant human trafficking,
and engage in subregional terrorism.

New People’s Army (1969–)

The NPA is the military wing of the Communist Party of the Philip-
pines (CPP). Since its foundation in 1969, it has sought to establish
a communist state in the Philippines by waging a rural-based insur-
gency according to the Maoist doctrine of protracted people’s guerrilla
war. The group has traditionally maintained a presence throughout the
country, but it is primarily active in the regions of Luzon, Visayas, and
parts of Mindanao.
The threat posed by the NPA has fluctuated over time, peaking in
the late 1980s during the unpopular regime of President Marcos, when
the NPA strength reached an estimated 25,000 armed guerrillas and a
mass base of support significantly larger than that. In the post-Marcos
era, the NPA has struggled to elicit the same degree of popular support,
and a series of brutal interfactional clashes within the organization has
weakened its organization internally and undermined its legitimacy
and credibility in the eyes of the Philippine population. However, the
threat posed by the NPA began to escalate once more following the
collapse of peace talks in 2005, and there was a sharp rise in attacks
carried out by the group.
Although the NPA has been able to maintain a steady operational
tempo following the launching of a concerted counterinsurgency cam-
paign targeting it in 2006, its capabilities have been steadily degraded,
and its areas of operation have increasingly been curtailed. Neverthe-
less, although it is no longer capable of posing the threat it once did,
and despite intelligence in early 2010 suggesting that the group was
once again suffering from internal fractionalization, the NPA remains
militarily powerful and has become a successful social and political
actor through its control of political parties and organizations. While
a resolution to the conflict remains elusive, NPA representatives met
with government negotiators in Hong Kong in early December 2010,
and peace talks resumed in Oslo in mid-February 2011. Although both
 Counterinsurgency in the Philippines 69

sides agreed to hold further talks, NPA violence continued, and dis-
agreements in early November 2011 led to the peace talks being post-
poned. The NPA insurgency continued apace into 2012.
As of 2012, the NPA asserted that it had an active presence in 70
of the Philippines’ 81 provinces, yet its presence appears to be thin:
A 2007 Armed Forces of the Philippines (AFP) estimate claimed the
NPA was highly active in only 2,121 of the Philippines’ 42,000 baran-
gays, the country’s village-level administrative divisions. In February
2010, the AFP claimed that NPA influence had been further eroded,
with the group controlling only 1,077 barangays.7 Still, the government
considers the NPA its most serious threat, and indeed, the group is
responsible for considerably more attacks annually than the MILF and
the ASG combined.

Moro National Liberation Front (1972–1996)

The primary goal of the Moro National Liberation Front (MNLF)
was the creation of an autonomous Muslim state in the southern Phil-
ippines. The MNLF sought the autonomy of Mindanao and Sulu
Archipelago, in which there are 13 provinces. There is a long history
of conflict between Christians and Muslims in the Philippines dating
back to the mid-16th century, but the proximate trigger of the MNLF
movement’s rebellion occurred in the 1950s, when the GRP encour-
aged Christian Filipinos to move to Mindanao. Christians became the
majority in Mindanao by the late 1960s, leading not only to religious
tensions but also to economic conflict over the area’s natural resources.
By the mid-1970s, the MNLF boasted more than 10,000 mem-
bers and controlled large swaths of Mindanao and Sulu. With the
country already under martial law, President Marcos responded to the
MNLF’s advance with heavy doses of force, committing 80 percent of
his available troops to the southern Philippines.
In 1976, the MNLF agreed to a 16-point accord with the GRP.
Concurrent with this accord, a referendum on autonomy for the south-
ern islands was held in 1977. But because the islands had a Christian

7 Jane’s World Insurgency and Terrorism, “JTIC Country Briefing–Philippines,” February

2012, p. 28.
70 Countering Others’ Insurgencies

majority, the referendum failed to garner a majority vote, the MNLF’s

autonomy bid failed, and the organization splintered into several
groups. In 1986, the MNLF and the GRP signed a truce in which the
parties agreed to continue discussion of full autonomy for the Min-
danao region, with a joint commission to be set up to develop the
details of a proposal for full autonomy of Mindanao, Basilan, Sulu,
Tawi-Tawi, and Palawan, subject to democratic processes. Finally, in
1996, President Fidel Ramos agreed to a peace accord with MNLF
leader Nur Misuari. The agreement created a four-province Autono-
mous Region of Muslim Mindanao, of which Nur Misuari joined the
political leadership.

Moro Islamic Liberation Front (1977–)

The MILF is an Islamist militant group based and operating in the
southern Philippine region of Mindanao, particularly in the provinces
of Lanao del Norte, Lanao del Sur, Maguindanao, and North Cota-
bato. Salamat Hashim established the MILF in 1977 as a splinter fac-
tion of the MNLF with the aim of creating an independent Islamic
state in Mindanao. However, MILF structure and policy direction
were not specifically defined until 1984.
The Philippine government never really saw the MILF as any-
thing more than a radical splinter of the MNLF, upon which it con-
tinued to focus both its military and diplomatic efforts. However, the
MILF grew steadily, particularly through its control of the mosques in
central Mindanao. The group was also a beneficiary of the 1996 peace
accord between the government and the MNLF, as 5,000 to 6,000
MNLF combatants rejected the accord and joined the MILF, doubling
its size. By 1999–2000, the MILF controlled vast swaths of central
Mindanao, where it established Islamic-based governing structures.
Peace talks resumed in Malaysia in December 2009. While
the MILF initially rejected a power-sharing offer in February 2010,
it announced the following September that it would be prepared to
accept an agreement in which a Muslim “substate” was created. The
revision of the MILF’s objectives paved the way for the resumption of
exploratory peace talks in February 2011, when the group presented
the government with a proposal to create a semi-autonomous politi-
 Counterinsurgency in the Philippines 71

cal entity in Mindanao. The future of the peace process was called
into question in October, following the MILF’s involvement in a clash
between the military and the ASG that left 19 soldiers dead.
Currently, the MILF is one of the most powerful militant groups
in Southeast Asia. It commands approximately 10,000 armed fighters
and has well-established bases from which it exerts political control
over local Muslim communities. Its principal areas of activity include
the southern provinces of Lanao del Norte, Lanao del Sur, Maguin-
danao, and North Cotabato.8 The group has extensive grassroots sup-
port, which means that the reservoir of potential fighters remains very
large. Not everyone joins exclusively for religious or ethnic purposes;
some join because of the poor economic conditions in Mindanao and a
lack of other opportunities. Senior military officials have estimated that
the MILF could easily mobilize an additional 8,000 to 10,000 fight-
ers within a short period of time. In recent years, however, the group’s
influence and threat to the GRP have declined somewhat. This decline
is in part due to government operations and the resulting deterioration
of the MILF’s conventional fighting strength, but it can also be linked
to the death of radical leader Salamat Hashim in 2003 and his replace-
ment by the more-moderate Ebrahim el-Haj Murad, who is more will-
ing to accept a negotiated settlement on expanded autonomy.9

Abu Sayyaf Group (1991–)

The ASG is a Sunni militant group based and operating in the south-
ern Philippine region of Mindanao, particularly on the island province
of Basilan and the nearby island of Jolo in Sulu province. While the
ASG’s global jihadist outlook and its ultimate desire to establish an
independent Islamic republic make it ideologically similar to al Qaeda
and JI, the group is also known for its criminal activities, including a
number of high-profile kidnappings for ransom involving Westerners.
The ASG’s links to al Qaeda and JI and its targeting of Westerners
helped ensure that the group would become a focus of U.S. counter-

8 Jane’s World Insurgency and Terrorism, 2012, p. 36.

9 Jane’s World Insurgency and Terrorism, “Moro Islamic Liberation Front,” June 7, 2013b,
p. 3.
72 Countering Others’ Insurgencies

terrorism efforts after the 9/11 terrorist attacks in the United States.
Indeed, a series of joint U.S.-Philippine military and political actions
have substantially degraded the ASG’s capabilities.
Although the ASG’s top leadership hierarchy has remained
unclear since Khadafy Janjalani was killed in 2006, numerous other
ASG commanders and subcommanders were subsequently killed
or captured, including ASG spokesman Jainal Antel Sali, alias Abu
Solaiman; Ismin Sahiron, the son of a top ASG commander; and other
key operatives, including Jundam Jamalul and Borhan Mundus. The
AFP claimed that 144 ASG rebels were killed, were captured, or sur-
rendered in 2007, bringing the total number of ASG rebels to fewer
than 400 (from a peak of between 1,000 and 1,200) for the first time
in years.
The ASG has nonetheless demonstrated that it can still inflict
significant casualties. In July 2007, ASG and MILF forces ambushed a
group of AFP Marines searching for a kidnapped Italian priest on Basi-
lan. Fourteen of the Marines were killed, ten of whom were beheaded.
In August 2007, the ASG conducted an attack that killed 15 AFP
Marines in Basilan province.10 For the foreseeable future, the group’s
propensity for kidnap-for-ransom operations ensures that it will pose a
threat to civilians, particularly foreign nationals. 11 At the same time, it
appears that the ASG is unlikely to pose more than a low-level threat to
Philippine and U.S. security forces in the south.

Government Forces and Strategies

Armed Forces of the Philippines
The AFP, established in 1935, is an all-volunteer, tri-service organiza-
tion comprising an Army, a Navy, and an Air Force. By constitutional
mandate, the AFP is under civilian control, with the President of the
Philippines as its Commander-in-Chief. All AFP branches are part of

10 Jane’s World Insurgency and Terrorism, “Abu Sayyaf Group,” June 7, 2013, p. 7.
11 Fund-raising through criminality appears to have enabled the group to replenish its
depleted ranks by offering competitive wages to recruits, many of whom do not share the
previous generation’s Salafi jihadist ideology. Rommel C. Banlaoi, “The Sources of the Abu
Sayyaf ’s Resilience in the Southern Philippines,” CTC Sentinel, Vol. 3, No. 5, May 2010,
pp. 17–19.
 Counterinsurgency in the Philippines 73

the Department of National Defense, which is headed by a Secretary of

National Defense. The AFP is organized into Joint Service Area Com-
mands, each of which is responsible for a particular region.12
Especially significant with respect to combating terrorism is
the fact that the AFP has an established special operations capability
within its Army, Navy, and Air Force, as well as a Joint Special Opera-
tions Group (JSOG). The Philippines Army Special Operations Forces
(SOF) consist of four units: the Special Forces Regiment (Airborne),
the First Scout Ranger Regiment, the Light Reaction Battalion, and
the Civil Affairs Group. The Special Forces Regiment was founded
in 1962 and has approximately 1,200 personnel trained to conduct
unconventional warfare, reconnaissance, and direct action raids. The
First Scout Ranger Regiment was first organized in the late 1940s and
today consists of approximately 1,800 personnel considered by some to
be the most experienced unconventional fighters in the AFP. The Light
Reaction Battalion consists of approximately 400 personnel taken from
the Special Forces and Scout Ranger Regiments and is organized into
three companies that are trained to conduct counterterrorist operations
under the operational control of the JSOG. The Civil Affairs Group
evolved from the Public Relations Office of the 1950s and consists of
approximately 130 personnel trained in civil-military operations. Once
a part of the Philippines Special Operations Command, the Civil
Affairs Group now falls directly under the control and supervision of
the Headquarters of the Philippines’ Army.13

Government Counterinsurgency and Counterterrorism Strategies

Throughout more than three decades of armed communist insurgency
and Muslim separatism, the GRP’s responses have been a compromise
between civilian and military interpretations of the armed threats and
how best to respond to them. This tension between civilian and military

12The commands are Northern Luzon, Southern Luzon, Visayas, Western, East Mindanao,
West Mindanao, and National Capital.
13 Stuart Farris, Joint Special Operations Task Force–Philippines, School of Advanced Mili-
tary Studies, Fort Leavenworth, Kans.: U.S. Army Command and General Staff College,
2009, pp. 22–23.
74 Countering Others’ Insurgencies

authorities is reflected in the mix of government and military responses

to the threats, as seen in the “right- and left-hand approaches” to coun-
terinsurgency since the 1950s. The balance between the right hand
(primarily the use of military force) and the left hand (socioeconomic
and political measures to address the root causes and win the hearts
and minds of the people) has tilted in favor of one side or the other,
depending on the overall political, economic, and social context, the
strategic situation on the ground, and the personality and policy prefer-
ences of political and military leaders.
Overall, a highly conventional military strategy predominated
during martial law and dictatorship, because the military was a main
partner of the civilian leader in the government, Ferdinand Marcos.
The Marcos government depended on the military for political survival,
giving the military a role in politics, a large budget, and significant
bureaucratic clout. The military’s influence in the counterinsurgency
strategy was also determined by the rise of the NPA and the MNLF
soon after the imposition of martial law and the growth of these orga-
nizations during the rest of the Marcos era, as well as the growing
opposition to his rule from the broader sociopolitical spectrum.14
After democracy was restored in 1986, the Philippine military
continued to play a significant role in politics. It had distinct prefer-
ences concerning how internal security threats should be addressed,
which involved relying heavily on coercion and conventional force.
Unlike the Marcos regime, however, each civilian democratic adminis-
tration that followed made negotiating with insurgencies a major part
of its strategy. Each also attempted to incorporate non-kinetic opera-
tions into its internal security strategies. Yet more often than not, a lack
of policy coherence—the military often did one thing while the civil-
ian government said another—or a lack of GRP capacity to implement

14 Raymond Bonner, Waltzing with a Dictator: The Marcoses and the Making of Foreign
Policy, New York: New York Times Books, 1987, pp. 204–225; William E. Kline, The Fall of
Marcos: A Problem in U.S. Foreign Policymaking, Washington, D.C.: Institute for the Study
of Diplomacy, Georgetown University School of Foreign Service, 1992, pp. 18–19.
 Counterinsurgency in the Philippines 75

these strategies effectively in the face of military opposition stymied the

success of their implementation.15

The Role of the United States

The role of the United States is inextricably intertwined in the Philip-
pine counterinsurgency experience. During the Cold War, Washing-
ton was primarily interested in the Philippines for geopolitical reasons.
In particular, the United States had a strong interest in continued use
of its military bases there.16 As a result, the United States largely stayed
out of Philippines internal affairs. Washington remained silent when
Marcos declared martial law in 1972, even though many in the U.S.
government were reportedly uncomfortable with his growing authori-
tarian bent.17 Despite the declaration of martial law, Marcos’ Philip-
pines was a significant recipient of U.S. aid: From 1972 until 1983, the
U.S. provided $2.5 billion in bilateral military and economic aid to the
Marcos regime and about $5.5 billion through multilateral institutions
such as the World Bank.18
Fourteen years after he first declared martial law, Marcos went
into exile in Hawaii amid growing pressure that culminated in the Peo-
ple’s Power demonstrations of 1986. He departed Malacanang Palace
aboard a U.S. helicopter on February 25, 1986, after U.S. officials per-
sonally informed him that he would receive no further U.S. assistance
and prepared a formal announcement that U.S. policy was to support
a “transition” to a popular government in the Philippines. Corazon
Aquino succeeded Marcos and restored democratic rule—however
imperfect—to the Philippines. In a sharp break with the policies of the
Marcos era, Aquino attempted to negotiate with the NPA, but these
talks foundered, in part due to U.S. opposition to negotiations with
communists.

15 Joel Rocamora, “Discontent in the Philippines,” World Policy Journal, Vol. 8, No. 4, Fall
1991, pp. 636–637.
16 Richard J. Kessler, “Marcos and the Americans,” Foreign Policy, Vol. 63, 1986, pp. 40–57.
17 Bonner, 1987, pp. 204–225; Kline, 1992, pp. 18–19.
18 Walden Bello, “Edging Toward the Quagmire: The United States and the Philippine
Crisis,” World Policy Journal, Vol. 3, No. 1, Winter 1985–1986, p. 31.
76 Countering Others’ Insurgencies

U.S.-Philippine military relations were largely severed in 1992,

when Clark Air Base and Subic Naval Base were closed. Philippine
demands for U.S. withdrawal were not disconnected from the United
States’ Cold War relationship with Marcos and the pressure on the
democratic government to reduce the U.S. presence in the Philip-
pines despite the major loss of revenue that was incurred.19 As U.S.-
Philippine relations deteriorated, the United States dramatically
decreased security assistance to the Philippines in the early 1990s, from
about $350 million in 1991 to $5 million in 1993. The cut in U.S. mili-
tary aid likely contributed to the degradation of Philippine counterin-
surgency capabilities and the resurgence of the NPA and the MILF, as
well as the emergence of the ASG, during the mid-1990s.
In response to the rise of militancy in the Philippines, the United
States began preparations to resume military support, and in 1998,
U.S. SOF deployed to the Philippines to help establish a counterterror-
ist unit to address the Islamist threats that were growing in the south-
ern part of the country. The unit, called the Light Reaction Company,
was modeled after a U.S. SOF Special Missions Unit.20
Kidnappings of Westerners by the ASG had already pushed the
United States closer to the GRP before the 9/11 terrorist attacks. Fol-
lowing those attacks, however, the Philippines became an important
front in the U.S. Global War on Terrorism.21 The United States desig-
nated the Philippines as a Major Non-NATO Ally in October 2003.22
U.S. military aid to the Philippines spiked from $3.8 million in 2001
to $56 million in 2002 and over $52 million in 2003 and 2004, before
declining to around $35 million in 2005 and 2006—a figure still six

19 Interview with former U.S. Ambassador to the Philippines, Arlington, Va., June 2013.
This outcome is consistent with broader trends observed in Ely Ratner, “Reaping What You
Sow: Democratic Transitions and Foreign Policy Realignment,” Journal of Conflict Resolu-
tion, Vol. 53, No. 3, June 2009, pp. 390–418.
20 Interview with former commander, AFP Special Operations Command, Manila, Philip-
pines, June 2013.
21Maria Ressa, Seeds of Terror: An Eyewitness Account of Al-Qaeda’s Newest Center of
Operations in Southeast Asia, New York: Simon and Schuster, 2004.
22 Thomas Lum, The Republic of the Philippines and U.S. Interests, Washington, D.C.: Con-
gressional Research Service, Library of Congress, April 2012, p. 14.
 Counterinsurgency in the Philippines 77

times higher than the U.S. aid received by any other southeast Asian
country.23
The United States intervened to support counterterrorism in the
Philippines as part of OEF-P. Akin to traditional foreign-internal-
defense strategies, the U.S. strategy in the Philippines later became
known as “the indirect approach.”24 Compared with U.S. operations
in other theaters, the operational method in OEF-P could aptly be
described as “indirect,” because of its relatively light footprint (approx-
imately 500–750 U.S. personnel operating under the Joint Special
Operations Task Force–Philippines [JSOTF-P]) and its focus on oper-
ating through indigenous security forces.

Analysis of Philippine Counterinsurgency

Using the theoretical framework developed in Chapter Two and the

empirical findings presented in Chapter Three, we analyzed three fac-
tors associated with Philippine counterinsurgency strategy and out-
comes: political inclusivity, military superiority, and state capacity.
Consistent with general patterns in counterinsurgency campaigns,
changes in regime type from autocracy to democracy are associated
with government adoption of counterinsurgency strategies that involve
pursuing negotiations with at least certain parts of the opposition and
an expressed desire to use force discriminately. These preferences have
been evident in multiple elected administrations’ commitment to the
minimal use of conventional firepower and avoidance of indiscrim-
inate sweep operations and a greater emphasis on reintegration pro-
grams and other civil-military operations intended to win popular sup-
port vis-à-vis the insurgents they faced.
Democratic Philippines governments’ ability to implement such
strategies have invariably been imperfect and susceptible to compro-
mise by political factors or capacity gaps; in some instances, the regimes

23 Lum, 2012.
24 Carolyn H. Briscoe, “Balikatan Exercise Spearheaded ARSOF Operations in the Philip-
pines,” Special Warfare, Vol. 17, No. 1, September 2004, p. 16.
78 Countering Others’ Insurgencies

reverted back to conventional approaches favored by the military or

other actors. Despite these inconsistencies, as of mid-2013, the GRP
is close to achieving a peace agreement with the MILF and appears to
have minimized the ASG threat and the foreign terrorist threat in its
southern islands.
Table 4.1 summarizes changes in key variables in the Philippines.
It disaggregates the case study by time period and insurgent group
to describe how state capacity and political inclusiveness have shaped
key counterinsurgency outcomes. The country underwent a significant
change from lesser to greater political inclusivity with the ouster of
the Marcos regime in 1986. Variation in state capacity is more com-
plex. In general, the Philippines took significant strides toward greater

Table 4.1
Regime and Campaign Features of Philippine Counterinsurgency,
1972–2013

Discrimi-
State Political Accom- Public nate
Group Capacity Inclusion modate Goods Violence Strategy

NPA, 1972–1986 Moder- Low Low Low Low Strong state

ate repression
MNLF, 1972–1986 Low Low Moder- Low Low Containment
ate
MILF, 1977–1986 Low Low Low Low Low Containment
NPA, 1987–2000 Moder- High Moder- Moder- Moder- Informal accom-
ate ate ate ate modation/
containment
MNLF, 1987–1996 Moder- High High Low N/A Informal accom-
ate modation
MILF, 1986–2000 Moder- High Moder- Low Low Containment
ate ate
ASG, 1991–2000 Low High Low Low Low Containment
NPA, 2001–2013 High High Moder- High High Classic counter-
ate insurgency
MILF, 2001–2013 Moder- High High Low High Informal
ate accommodation
ASG, 2001–2013 High High Low High High Classic counter-
insurgency
NOTE: Values represent the authors’ and outside experts’ judgments on relative
levels of each category over time. They should be considered heuristic summaries of
the variables rather than calculations based on precise data.
 Counterinsurgency in the Philippines 79

state capacity in the four decades covered in this study. These improve-
ments, however, did not occur uniformly across the country. Typically,
the core regions closer to Manila enjoyed substantially higher access
to public goods than did the more peripheral regions of Mindanao
and especially the Sulu Archipelago. As a result, the government was
better able to employ strong-state counterinsurgency practices against
the NPA, which operated in both the core regions and the periphery,
than it was against the insurgencies that operated in the predominantly
Muslim peripheral regions—the MNLF, the MILF, and the ASG.
The GRP counterinsurgency practices shown in Table 4.1 are
divided into three broadly defined time periods. The first time period
begins with the imposition of martial law by President Marcos from
1972 to 1981 and runs through the end of Marcos’ presidency in 1986.
During this period, political inclusivity was low, as the imposition
of martial law made Marcos a de facto dictator until he was forced
from power. His investment in coercive capabilities to crush the NPA
is consistent with strong-state repression, even though Philippine gov-
ernment capacity could be considered only moderate in Manila and
some of the areas where the NPA operated. The GRP’s capacity was
even lower in the Muslim south, where the MNLF and the MILF
had established parallel state structures amid largely sympathetic co-
ethnic populations, and the broad government strategy employed
against these groups was consistent with containment.
The second time period, roughly spanning 1986 to 2000, can be
viewed broadly as a transitional period. During this period, the coun-
try transitioned toward democracy. Policies and operational methods
changed, but the government’s capacity to implement reforms was
limited. The limitations stemmed in part from the civilian govern-
ment’s weakness relative to the military, whose most senior officers
from the Marcos period retained much of the political clout Marcos
had conferred on them. Their disagreement with the civilian govern-
ment’s policy preferences and the military’s reliance on old tactics from
the Marcos period, which often involved conventional operations in
which force was applied indiscriminately, played a key role in shap-
ing the ultimate approach that was taken in countering insurgencies.
In this period, the democratically elected regime made attempts to
80 Countering Others’ Insurgencies

forge a peace agreement with each major insurgent group. But only
one such effort—the 1996 peace deal with the MNLF—succeeded.
Despite the increased level of political inclusiveness, the government’s
lack of capacity relative to the military often resulted in counterinsur-
gency approaches that most closely resembled containment. Such an
approach was evident, for instance, after peace talks with the NPA and
the MILF repeatedly broke down and intermittent large-scale military
operations resumed. The government also was left to pursue a contain-
ment strategy against the ASG, which was never interested in a nego-
tiated peace and was thus alternately ignored and dealt with through
military operations that in many cases relied on the use of indiscrimi-
nate firepower.
The third broad period began in 2001 after Gloria Macapagal-
Arroyo succeeded Joseph Estrada as President and significantly
expanded the country’s relationship with the United States in the wake
of the 9/11 attacks. U.S. assistance, both military and civil, increased
dramatically during this period. Although the dramatic expansion in
U.S. aid was inadequate to boost state capacity throughout the coun-
try in such a short period of time, it made a dramatic difference in
the government’s capabilities in the small peripheral region of the Sulu
Archipelago, where joint operations by the U.S. military and its Phil-
ippine partners gave the government an unprecedented reach. Demo-
cratic institutions were also increasingly consolidated in this period:
Despite allegations of electoral graft in 2004 and 2010, a return to
military rule appears increasingly unlikely. The government’s approach
to counterinsurgency has followed relatively predictable patterns:
Against the NPA, it undertook the closest approximation of the classic
counterinsurgency model that it had taken to date. Against the MILF,
it engaged in an ongoing peace process and worked to restrain the
military from conducting operations that could compromise the pro-
cess, an approach consistent with the informal-accommodation model
described in Chapter Two. And against the ASG, it has worked with
U.S. partners to combine civil-military operations with discriminate
targeting of ASG members—an approach that again approximates the
classic counterinsurgency model.
 Counterinsurgency in the Philippines 81

Although the GRP’s policies over these four decades do not per-
fectly align with the behavior that we might expect based on the frame-
work presented in Chapter Two, the extent of congruence is remark-
able. In only two of the ten subcases depicted in Table 4.1 do GRP
policies clearly diverge from the behavior predicted by the framework:
the government’s approach to the MILF from 1986 to 2000 and its
approach to the ASG from 1991 to 2000. In both cases, the fledgling
democratic regime implemented containment strategies more com-
monly used by weak autocratic regimes. As discussed in more detail
below, these divergences from the behavior more typical of demo-
cratic regimes can be understood in terms of the country’s transition
to democracy. In the early years after the end of the Marcos regime,
democratic administrations were preoccupied with the critical tasks of
the transition. They did not have the necessary high-level engagement
or political capital to confront a military shaped by the Marcos era over
a conflict in a peripheral region. As the new regime consolidated its
authority, however, it was increasingly able to rely on more accommo-
dative approaches to moderate insurgents. Where the regime had the
necessary resources—specifically, against the insurgency that threat-
ened core territories (the NPA) and where the United States provided
extensive assistance focused on a relatively small peripheral region (the
campaign against the ASG)—it adopted the classical counterinsur-
gency policies common to higher-capacity, more politically inclusive
regimes.
These variations are explored in greater detail below.

Political Inclusion, Accommodation, and Restraint

The Growth of Insurgency Under Martial Law, 1972–1986
After being democratically elected in 1965, Ferdinand Marcos declared
martial law in 1972, in part because of threats posed by internal armed
groups and in part because his reelection prospects were poor. Under
martial law, Marcos subsequently greatly increased the size of the AFP,
in terms of both budget and personnel. At the same time, his govern-
ment showed little genuine interest in making concessions to recon-
cilable elements of the insurgencies it faced, which grew significantly
during this period.
82 Countering Others’ Insurgencies

Under martial law, Marcos’ main weapon against insurgency was

military force. The military budget was increased by 700 percent from
1972 to 1976; the number of Philippine Security Forces (PSF) was also
increased from 60,000 to 157,000.25 Due to the partnership between
the government and the AFP during the period of martial law, the AFP
fully supported the government’s campaigns against the CPP-NPA and
the MNLF. The AFP patterned its operations closely on the overall
policy of the martial-law regime, which sought to eliminate its oppo-
sition through crackdowns and coercion rather than inducement or
compromise. In practice, this meant that the Marcos regime imple-
mented a heavy-handed counterinsurgency strategy, unleashing both
the AFP and armed local militias that had little or no formal training.
Marcos’ gambit backfired: Civilian casualties skyrocketed, Fili-
pino public opinion increasingly turned against the regime, and the
insurgencies flourished. The NPA grew its base enough to shift from
an almost exclusively rural movement to one that also drew from urban
areas, particularly local universities whose students were prone to radi-
cal ideas. Overall, the CPP-NPA grew from a small group of about
1,100 in 1971 to an estimated 7,200–10,660 members by 1983.26 The
Moro-based Muslim separatist movements also grew under martial
law. The challenge from the MNLF became a serious threat to the Phil-
ippines in the 1970s. The fiercest fighting occurred between 1969 and
1976, when Marcos’ counterinsurgency campaign killed an estimated
60,000 people, wounded another 54,000, and displaced as many as
350,000.27

25 These numbers come from the Correlates of War Project’s National Material Capabilities
Database (version 4.0). For more background on the database, see J. David Singer,
“Reconstructing the Correlates of War Dataset on Material Capabilities of States, 1816–
1985,” International Interactions, Vol. 14, 1987, pp. 115–132.
26 For a detailed overview of the NPA’s evolution, see Paz Verdades M. Santos, “The Com-
munist Front: Protracted People’s War and Counterinsurgency in the Philippines (Over-
view),” in Soliman M. Santos, Jr., and Paz Verdades M. Santos, eds., Primed and Purposeful:
Armed Groups and Human Security Efforts in the Philippines, Geneva: Small Arms Survey,
2005, pp. 17–32.
27 Eva-Lotta E. Hedman, The Philippines: Conflict and Internal Displacement in Mindanao
and the Sulu Archipelago, Writenet Report commissioned by the United Nations High Com-
missioner for Refugees, Emergency and Technical Support Service, March 2009, pp. 2–5.
 Counterinsurgency in the Philippines 83

The Marcos government was not completely opposed to nego-

tiating with insurgents, but, consistent with the framework outlined
in Chapter Two, it was neither willing nor able to credibly commit to
a comprehensive peace accord. In 1977, the GRP entered into a tem-
porary peace deal with the MNLF, the Tripoli Agreement. The terms
of the agreement’s implementation, however, which consisted of the
MNLF’s requirements for regional autonomy, were left open to a refer-
endum that was to be held at a later time. Meanwhile, both the GRP
and the MNLF violated its initial terms, and the MNLF splintered
into multiple groups. The Autonomous Region of Muslim Mindanao
(ARMM) would not be formally created until 12 years had passed
and a democratically elected administration was in place, and even
then, the newly created entity still failed to end the Muslim separatist
insurgency.28

Transition to Democracy and Conciliatory Counterinsurgency

Strategies, 1986–2001
After Corazon Aquino assumed the presidency in 1986, she immedi-
ately announced plans to negotiate with the insurgents and to offer
concessions such as amnesties and reintegration packages. Peace talks
with the NPA were accompanied by a short cease-fire from December
1986 to January 1987, when the talks collapsed due to the NPA’s lack
of a serious commitment to peace, dissent from Marcos loyalists in the
military, and a general lack of U.S. support for any settlement ceding
substantive concessions to communists.29
Military operations also changed under Aquino, once again in
line with the discussion in Chapters Two and Three. Particularly after
the collapse of peace negotiations, the military’s counterinsurgency
strategy against the NPA also shifted in a new direction, focusing on

28 Marcos was actively supporting the resettlement of Christian Filipinos in Mindanao

during his tenure as the MNLF peace process continued, with demographic changes under-
mining the chances of a successful comprehensive agreement. See, for example, Nathan
Gilbert Quimpo, The U.S. and the Southern Philippines’ Quagmire, Murdoch, Western
Australia: Asia Research Center, July 2007, p. 4.
29 Roger W. Fontaine, “The Philippines: After Aquino,” Asian Affairs: An American Review,
Vol. 19, No. 3, Fall 1992, pp. 171–172.
84 Countering Others’ Insurgencies

conducting intelligence-driven operations that targeted the NPA’s local

infrastructure. Under this strategy, the conventional-style search-and-
destroy operations so common under Marcos received less emphasis.
Lambat Bitag, the GRP’s comprehensive counterinsurgency strat-
egy launched in the post-Marcos era, emphasized the integration of
civil-military operations, intelligence, and combat operations, and it
focused on the military tactics of clear, hold, consolidate, and develop.
This strategy was derived from a right-hand and left-hand approach,
which sought to neutralize the insurgency by reducing the emphasis on
force and instead focusing on development and popular support. The
strategy was intended to include a variety of civilian government agen-
cies and non-governmental organizations (NGOs). A key part of the
strategy involved reeducating AFP personnel about human rights and
the importance of popular support in counterinsurgency and internal-
security operations.30
More formally, this strategy focused on the clear-hold-consolidate-
develop (CHCD) methodology. It was to be executed in stages, with the
aim of dismantling insurgent political and military infrastructures in
the countryside and establishing government control in contested areas
to create opportunities for development. The strategy was designed to
be relatively flexible, and its stages could be conducted sequentially or
simultaneously, depending on the prevailing situation in an area.
The “clear” stage would involve combat, intelligence, and civil-
military operations conducted to defeat the insurgent armed groups
and to neutralize their political and military infrastructures. The aim
was to gain possession or control of strategic areas or key terrain to allow
holding forces to come in for the next stage or operation. The “hold”
stage was intended to reestablish government control and authority and
to preserve the initial gains of the clearing forces through occupation
and control of the cleared areas. It was to signal the strength of local
defense capabilities to defend the cleared areas from incursions and
reentry of rebel groups. The “consolidate” stage involved the collabora-

30 Leonardo I. Pena, Finding the Missing Link to a Successful Philippine Counterinsurgency

Strategy, master’s thesis, Monterey, Calif.: Naval Postgraduate School, 2007, pp. 43–44.
 Counterinsurgency in the Philippines 85

tive participation of the AFP, police, civil government agencies, and

the local population to strengthen government control. Consolidation
was to include assisting local government units in the delivery of basic
services to help them gain the population’s trust and confidence. The
“develop” stage was intended to foster the implementation of the gov-
ernment’s socioeconomic and political reforms, including the delivery
of basic services to address root causes of insurgency. The military and
police were to focus on the clear and hold stages and were to assume
supportive roles during the consolidate and develop stages. Other agen-
cies were intended to support the military and police in the clear and
hold stages but to take the lead role in the consolidate and develop
stages.31
Aquino’s preferred military counterinsurgency strategy achieved
only partial implementation and suffered de facto reversal by 1990.32
The preliminary results of the new strategy had looked promising, at
least according to AFP intelligence estimates and metrics. These esti-
mates suggested that NPA influence dropped from a peak of almost
8,500 NPA-influenced barangays in 1987 to fewer than 1,000 in 1993,
and that NPA strength declined from just under 24,000 in 1987 to
about 6,000 in 1995.33 But Aquino had less luck sustaining these gains
than in achieving them in the first place. After AFP troops withdrew
from an area, the NPA would typically return and reclaim control of
its villages. In a strategy oriented around the principles of CHCD,
the consolidate and develop aspects of Lambit Bitag were chronic
shortcomings of the civilian agencies tasked with reconstruction and
development.
GRP counterinsurgency operations rarely played out this way in
practice. Resourcing for civilian government agencies to perform these

31 Roy T. Devesa, An Assessment of the Philippine Counterinsurgency Operational Methodol-

ogy, master’s thesis, Fort Leavenworth, Kans.: U.S. Army Command and General Staff Col-
lege, 2005, pp. 40–41.
32 Fontaine, 1992, pp. 174–175.
33 Some of these gains undoubtedly came after Aquino left office in 1992. The annual data
required to assess the proportion of gains made during her presidency, however, are unavail-
able. See Ricardo C. Morales, “Perpetual Wars: The Philippine Insurgencies,” unpublished
master’s thesis, Monterey, Calif.: Naval Postgraduate School, 1993.
86 Countering Others’ Insurgencies

missions and their coordination with the AFP were limited. Multiple
sources suggest that the AFP was holding up largely on its own against
the NPA. In practice, the unsuccessful implementation of a whole-
of-government approach exacerbated the already strained relations
between the Aquino government and its civilian agencies and the AFP,
over which Aquino had only tenuous control.34
The civilians in Aquino’s government were poorly coordinated
and could not reach consensus on the appropriate measures to take
against insurgents. Many in the Congress viewed the insurgencies as
purely local phenomena, while others viewed them as a task for the
military. Either way, support for Aquino’s desired strategy was limited,
and implementation was often difficult.
With nowhere else to turn, Aquino, under intense pressure from
the military and with promises of increased U.S. support, reversed
course and launched a total-war policy against the NPA. Interestingly,
however, although the NPA sought to take advantage of the uptick in
AFP abuses, it was less able to do so under the democratically elected
Aquino than it had been under Marcos’ martial law, with support
less forthcoming from moderate elements of the population, perhaps
because the democratic system afforded possible opportunities for
reforms that would simply never happen if the NPA were to win.35
Negotiations between the Aquino government and the MNLF
led to the creation of the ARMM on August 1, 1989. This agreement
was hardly comprehensive, however: Of the 13 southern provinces and
nine cities that participated in the referendum held on the basis of the
Tripoli Accords, only four—Lanao del Sur, Maguindanao, Sulu, and
Tawi-Tawi—opted to be part of the ARMM. Marcos’ legacy haunted
the prospects for fuller implementation, as one of the reasons for the
failure of the referendum to garner a majority vote was Marcos’ encour-

34 Interview with GRP official, Quezon City, Philippines, June 2013; interview with AFP
officer, Makati, Philippines, June 2013; Devesa, 2005, p. 36.
35Dominique Caouette, Persevering Revolutionaries: Armed Struggle in the 21st Century,
Exploring the Revolution of the Communist Party of the Philippines, dissertation, Ithaca, N.Y.:
Cornell University, 2004, pp. 483, 488.
 Counterinsurgency in the Philippines 87

agement of Christians to move into Mindanao at the same time thou-

sands of Muslims were fleeing from government repression.
Ultimately, Aquino failed to reach a final negotiated settlement
with any of the country’s three major internal threat groups, despite
having engaged all of them in talks during her tenure.36 Aquino was
weak vis-à-vis a variety of actors with divergent preferences concerning
the peace processes, including powerful factions of the AFP, old inter-
ests from the Marcos administration, and in some cases, the United
States. This weakness likely undermined insurgents’ confidence in her
government’s ability to make good on its stated objectives of peace and
reintegration. At the same time, rebel fragmentation in both the NPA
and the southern Muslim separatists threatened to spoil any negotiated
peace.37
After Fidel Ramos came to power in 1992, he affirmed Aquino’s
initial policy of national reconciliation and general amnesty for will-
ing parties from the communist factions and the Muslim separatists.
Ramos’ efforts culminated in the only successful and lasting peace
agreement—the 1996 Final Mindanao Peace Agreement, brokered
with the MNLF—during the four decades examined in this study.
Ramos succeeded in reaching a deal that took a key part of the opposi-
tion off the battlefield.
In addition to the success of Ramos’ peace initiatives—at least
compared with those of his predecessors—his strategy hinged partly
on civil-military relations. A former Secretary of National Defense

36 The NPA did not receive significant assistance from external sources during the Cold
War, with the exception of support from China from 1969 to 1976. The group’s strength, and
its entry into negotiations, varied more with changes in the regime it faced. “New People’s
Army,” START Terrorist Organization Profiles database, University of Maryland, College
Park, Md., undated.
37 One unanticipated byproduct of democratization and attempts at reconciliation—
imperfect as they were—was that they triggered NPA fractionalization, pitting hard-
line “reaffirmationists” against “rejectionists,” who opposed continuing the protracted
war strategy. The bickering between the two factions escalated until a series of bloody
internal purges of suspected traitors had riven the NPA, nearly leading to its col-
lapse. See International Crisis Group, The Communist Insurgency in the Philippines: Tac-
tics and Talks, Crisis Group Asia Report No. 202, Islamabad/Brussels, February 2011,
pp. 5–6.
88 Countering Others’ Insurgencies

who had lost favor with key elements of the AFP, Ramos had an acute
understanding of the importance of both addressing root causes of the
insurgencies and reconciling with the military. His policy objectives
reflected these priorities. Ramos repealed R.A. No. 1700, also known
as the Anti-Subversion Law, to open the door for former subversive
organizations to pursue political and social goals through political
means rather than armed struggle. To accommodate individual dis-
sidents, the Balik-Baril program was created; it reintegrated rebels who
surrendered their weapons to government authorities in exchange for
an up-front cash payment and additional payments to be made in sub-
sequent months that would enable all those who surrendered to take
up a livelihood program of their choosing. With regard to the mili-
tary, Ramos sought to accommodate the coup plotters of the 1980s by
granting them unconditional amnesty.
Overall, Ramos’ conciliatory policies held promise for addressing
the country’s insurgencies. For example, during his tenure, the number
of communist-influenced villages declined to its lowest point.38 The
reduction was sufficient, in fact, that the AFP’s role in combating insur-
gency was scaled back and eventually transferred, in 1996, to the Phil-
ippine National Police, along with the Department of the Interior and
Local Government and local government units (LGUs). These agen-
cies remained poorly coordinated, underresourced, and lacking in the
training and doctrine for counterinsurgency that the AFP had enjoyed.
In subsequent years, the GRP lost ground in its efforts against multiple
insurgent groups, including the NPA, and in particular, the MILF.
Ramos’ attempt to enact a whole-of-government approach to counter-
insurgency never fully gained the population’s confidence. Despite the
democratic handover from Aquino, the Ramos government’s bureau-
cracy was perceived by many as weak, uncommitted to reform, un-
coordinated, corrupt, and financially strained, all of which limited its
capacity to deliver on the promises of Lambat Bitag.39 Although the
communist insurgency reached the lowest point of its history during
the Ramos administration, the MILF and the ASG began to rise up

38 Devesa, 2005, p. 5.
39 Pena, 2007, pp. 59–61.
 Counterinsurgency in the Philippines 89

again during the latter part of his term, as shown in Figure 4.1, in large
part because of the government’s failure to execute and sustain the con-
solidate and develop stages of the Lambat Bitag campaign strategy.40
Ultimately, the insurgents discovered many of the same opportunities
they exploited during the Aquino administration, playing on the peo-
ple’s grievances to increase their dissatisfaction with the government.41
This episode illustrates that the preferred policies of democratic regimes
can be undermined by inadequate capability to implement them—
a subject that will be further explored in the discussion of state capacity.
Ramos’ successor, Joseph Estrada, returned responsibility for
internal security operations to the AFP in 1998. After some initial
skirmishes, the military and the insurgent groups both escalated

Figure 4.1
Insurgent-Group Strength by Administration

Marcos Aquino Ramos Estrada Arroyo

30,000

25,000

20,000
Group strength

CPP-NPA

15,000 MILF

10,000

5,000
ASG

0
1968 1972 1976 1980 1984 1988 1992 1996 2000 2004

SOURCE: OJ3, GHQ, AFP, 2005, quoted in Pena, 2007, p. 57.

RAND RR513-4.1

40 Pena, 2007, p. 57.

41 Pena, 2007, p. 57.
90 Countering Others’ Insurgencies

operations in a tit-for-tat fashion, and the hope of a negotiated settle-

ment was largely dashed in 2000 when Estrada declared a strategy of
all-out war against the MILF, which resulted not only in a renewal of
combat, but also in high levels of civilian displacement in Mindanao.

Philippine Strategy Since 2001: Conciliation and Counterterrorism

Estrada was ousted in 2001 and replaced by Vice President Gloria
Macapagal-Arroyo. Although his ouster was dubbed “People Power 2,”
it came after the AFP withdrew its support, citing his corruption and
unpopularity as evidenced by popular protests in Manila, and shifted
to supporting Arroyo. Subsequent presidential elections in 2004 and
2010 were allegedly subject to irregularities, but whatever their flaws,
democratic processes are now institutionalized—an outcome that was
far from certain when Marcos went into exile in 1986. While not
impossible, a return to Marcos-style martial law appears increasingly
unlikely.
Since 2001, the GRP’s stated strategy for addressing its internal
security threats has been to pursue a negotiated solution with both the
MILF and the NPA through ongoing peace processes. As of July 2013,
a settlement with the MILF appeared increasingly likely, although the
government’s dialogue with the NPA has been less productive for a
variety of reasons.42
The ASG case is somewhat different. Like other Salafi jihadist
groups, the ASG rejects any negotiated compromise with the Catholic
government in its pursuit of an Islamic emirate in the Philippines.43 As
a result, the ASG remains the target of AFP counterterrorism opera-
tions in the southern Philippines that aim to cripple the group or defeat
it outright.

42 These reasons include severe NPA fragmentation and internal disagreement over the
group’s objectives; economic rents NPA leaders were able to capture via “revolutionary taxes”
and front companies; and the GRP’s decision to make the NPA peace process a secondary
priority to the MILF process. International Crisis Group, 2011, pp. 7–9.
43 Zachary Abuza, Balik-Terrorism: The Return of the Abu Sayyaf, Carlisle, Pa.: Strategic
Studies Institute, September 2005, pp. 41–42.
 Counterinsurgency in the Philippines 91

Military Superiority
A second explanatory factor associated with Philippine counterinsur-
gency strategy and outcomes is military superiority, which reflects the
balance—or imbalance—between government armed forces and the
size and capabilities of the insurgents and terrorists. Change over time
to the military balance can shape strategy and indeed did shape GRP
strategy. Contrary to the hypothesis that governments tend to be more
discriminating in their use of force when they enjoy significant military
superiority, however, the Philippines experience suggests that quantita-
tively superior government forces might use either high or low levels of
indiscriminate violence. The nature of the regime and the quality of
the counterinsurgency forces it deploys account better for the variation
over time in the Philippines.
The history of Philippines counterinsurgency since Marcos’
ouster maps onto this framework relatively well. Under Marcos, the
country was under de facto authoritarian rule for 14 years. Marcos’
military was significantly larger than the rebel forces he faced, particu-
larly after he invested hundreds of millions of dollars in the expansion
of the both the AFP and pro-government paramilitary forces to fight
armed dissident movements. Consistent with our framework, Marcos’
authoritarian regime did not bargain in good faith; rather, it know-
ingly implemented policies that would undermine the Moro Muslims’
demographic majority status in Mindanao and thus reduce the chances
that an agreed-upon referendum would force the concessions sought by
the MNLF. Moreover, the Marcos regime’s use of force became more
and more indiscriminate as the military’s relative dominance eroded
due to booms in rebel recruitment, particularly by the NPA from the
late 1970s through the first half of the 1980s.

Quantitative Military Superiority and Philippine Counterinsurgency

Strategy
As Figures 4.2 and 4.3 indicate, the escalation in the number of PSF
predated the NPA’s membership spike. The number of AFP person-
nel rose dramatically shortly after Marcos imposed martial law in
1972, rising from 62,000 in 1972 to 90,000 in 1974—an increase of
almost one-third in just two years. Marcos continued to increase the
 92 Countering Others’ Insurgencies

Figure 4.2
Time Trend in PSF and NPA Numeric Strength and Force Ratio, 1978–2006

160
PSF (thousands)

140

120

100
1980 1985 1990 1995 2000 2005
NPA (thousands)

25
20
15
10
5
0
1980 1985 1990 1995 2000 2005
60
PSF:NPA

40

20

0
1980 1985 1990 1995 2000 2005

SOURCE: Esperon, 2006; Correlates of War Project, 2010.

RAND RR513-4.2

size of the military in subsequent years. By 1977, it grew to 155,000,

where it remained fairly steady through 1982, at which time it declined
to 113,000, where it would remain fairly steady for the remainder
of Marcos’ tenure.44 Overall, in the 17 years from the beginning of
Marcos’ reign in 1965 to the regime’s peak military strength in 1982,
the number of military personnel more than tripled.
Marcos not only increased the number of military personnel in
his armed forces, he also increased spending on the military. From
1972 to 1974, Philippine military expenditures increased 142 percent,
jumping from $119,320,000 to $289,035,000. Philippine military

44 These statistics come from the Correlates of War project’s National Material Capabilities
database (version 4.0). For more on the National Material Capabilities data, see Singer, 1987,
pp. 115–132.
 Counterinsurgency in the Philippines 93

Figure 4.3
ASG Estimated Strength, 1993–2003

1,400
1,270

1,200 1,150
1,030
1,100
Estimated strength

1,000
890
800
800
650
600
580 480

400 460

200
120
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

SOURCE: Office of the Deputy Chief for Intelligence, J2, Armed Forces of the
Philippines, 2004.
RAND RR513-4.3

spending peaked in 1982 at $909,822,000—more than 14 times the

$64,838,000 spent in 1965, Marcos’ first year in power.45
Marcos clearly sought military superiority over the internal-
security threats his regime faced. But his military buildup did not repel
the NPA, the armed wing of the CPP, which was the main threat.
On the contrary, as Marcos invested more and more in his military,
the NPA also grew (see Figure 4.2). According to many accounts, the
Philippine government and armed forces’ repression of the commu-
nists was a potent tool for NPA mobilization and significantly contrib-
uted to its growth, which spiked from the mid-1970s through the late
1980s.46 Indeed, the communists had long described Marcos as their

45 Correlates of War Project’s National Material Capabilities database (version 4.0). See
Singer, 1987.
46 See, e.g., William Chapman, Inside the Philippine Revolution: The New People’s Army
and Its Struggle for Power, 1st ed., New York: W. W. Norton, 1987; Gregg R. Jones, Red
Revolution: Inside the Philippine Guerrilla Movement, Boulder, Colo.: Westview Press, 1989;
Soliman Santos, “Evolution of the Armed Conflict on the Communist Front,” in Phil-
ippine Human Development Report 2005: Peace, Human Security and Human Develop-
ment in the Philippines, 2nd ed., Quezon City, Human Development Network, 2005,
pp. 84–89.
94 Countering Others’ Insurgencies

best recruiter.47 For the period for which systematic estimates of NPA
strength are available, 1978–2006, the most striking period of growth
in NPA strength was during the Marcos era: During the period of
the most significant PSF growth in the Marcos era for which data are
available, 1978–1982, the estimated number of NPA members nearly
tripled, increasing from 2,760 to 7,750.48 Estimated NPA membership
increased to approximately 10,660 in 1983; 14,360 in 1984; 22,500
in 1985; and 24,430 in 1986.49 According to these data, the ratio of
PSF to NPA personnel was highest in the late 1970s and early 1980s
but declined steadily during this period and continued to decline until
1987.
Unclassified AFP annual estimates of ASG strength are shown
in Figure 4.3 for 1993 to 2003.50 These estimates confirm the conven-
tional wisdom that the ASG is the smallest of the three major Phil-
ippine threat groups. AFP intelligence suggests the group had only
120 members in 1993. This number increased steadily throughout
the 1990s and peaked at 1,270 in 2000, before beginning to steadily
decline in 2001 and 2002. The timing of the ASG’s decline coincides
with the deployment of U.S. SOF to the southern Philippines, first to
assist in establishing the Philippines’ premier counterterrorism unit,
the Light Reaction Company, in 2000, and then as Joint Task Force
510 (JTF-510) and then as JSOTF-P from 2002 on.51

47 Jones, 1989, p. 7.
48 Santos, 2005, p. 24.
49 Esperon, 2006, p. 5.
50 These were the only years for which unclassified annual estimates of ASG strength were
available. See Office of the Deputy Chief for Intelligence, J2, Armed Forces of the Philip-
pines, 2004, Annex IA. Quoted in Romulo C. Supapo, U.S.-Philippine Security Relations: Its
Implications for the Global War on Terrorism, Carlisle, Pa.: U.S. Army War College, March
2004, pp. 7–8.
51 Lambert, Lewis, and Sewall, 2012, pp. 120–121.
 Counterinsurgency in the Philippines 95

The Importance of Quality

The discussion of military superiority raises an important point. One
explanation for the steep decline in ASG strength in the early 2000s
involves the quality of the modest number of U.S. troops deployed to
the southern Philippines. Their training and doctrine, not to mention
the high quality of the equipment, weapons, and technical assets they
brought, were both sufficient and appropriate to support the PSF cam-
paign against the ASG and facilitated a quick, significant degradation
of the ASG organization. Moreover, the AFP has developed increas-
ingly potent SOF capable of performing unconventional small-unit
operations independently.
The relative quality of the PSF deployed to counterinsurgency
and counterterrorism operations clearly correlates with partner-nation
qualities the United States seeks. High-quality troops—that is, AFP
troops who have undergone rigorous, specialized training in counter-
insurgency and counterterrorism, and whose operational approach is
guided and disciplined by appropriate doctrine for such operations—
are more likely to be able to build rapport with the local population
and plan and execute intelligence-driven operations in a selective, dis-
criminate fashion than are soldiers who lack these qualities.52
Table 4.2 contrasts the average number of deaths per conflict
incident for AFP elite SOF units, such as Special Forces and Scout
Rangers; for conventional, or regular, units, such as infantry; and for
indigenous units, which serve as local militia, typically in areas where the
AFP’s presence is insufficient to hold territory. The statistics show that
of the three unit types, elite AFP units tend to suffer the fewest deaths
per incident, on average, illustrating their force-protection capabilities,
and also that these units inflict the most fatalities on insurgents
(nearly two times the rate of conventional units) and the fewest
civilian casualties. They thus have the most favorable loss-exchange
ratio of the three unit types—a finding that holds when the data are
disaggregated and analyzed for a variety of different threat conditions

52 Joseph H. Felter, Taking Guns to a Knife Fight: Effective Military Support to COIN, Car-
lisle, Pa.: U.S. Army War College, 2008, pp. 9–10.
96 Countering Others’ Insurgencies

Table 4.2
AFP Unit Type and Conflict Deaths per Reported Incident, 2001–2007

Deaths per Reported Incident

Government Insurgents Civilians Loss- Exchange
Unit Type Killed Killed Killed Ratio
Elite (N = 589) 0.11 0.38 0.03 0.80
Regular (N = 5,566) 0.14 0.20 0.05 0.63
Indigenous (N = 907) 0.30 0.17 0.06 0.47

SOURCE: Felter, 2008, p. 20.

NOTE: Felter’s statistics are based on unclassified data from the AFP, Office of the J2.

and other environmental factors.53 AFP elite units conduct operations

against all threat groups, typically independent of U.S. advice or assis-
tance against the GRP’s main adversary, the NPA. This trend supports
the notion that quality might be as important as quantity, if not more
important, in executing counterinsurgency operations efficiently and
with maximum discrimination in the use of force.
Although troop quality—of both U.S. SOF advisors and their
Philippine counterparts—certainly plays a role in PSF posture and
effectiveness against unconventional threat groups, it is important
not to divorce it from context. Thinking through a counterfactual
scenario in which the U.S. intervention occurred in a different time
period suggests that the post-9/11 U.S. intervention might have come
at a propitious time for the United States to be able to cultivate coun-
terinsurgency tactics within the AFP. The Philippine government was
committed to maintaining democratic institutions (even while govern-
ment corruption remained relatively high) and seeking political accom-
modation with those insurgent groups that could reconcile with the
central government. Had the United States provided the same train-

53 Loss-exchange ratio is a measure of the fraction of conflict deaths experienced by govern-

ment forces relative to deaths of rebels. This variable was coded by dividing the number of
rebels killed by the sum of both government forces and rebels killed. A value of 0.5 indicates
an equal number of government and rebel soldiers died as a result of operations in the unit
of analysis. Values closer to 1 reflect cases where the government killed more insurgents than
it lost in the incident; the opposite is the case as the loss-exchange fraction value moves from
0.5 toward zero. See Felter, 2008, pp. 31–32.
 Counterinsurgency in the Philippines 97

ing in the Marcos era, it might have fed a very different operational
strategy—one that proved counterproductive during that era.

State Capacity and Counterinsurgency

Geography, Terrain, and Logistics
Like the governments of many fledgling states in the developing world,
the GRP has struggled to generate sufficient capabilities to effectively
exercise governance throughout its territory. One result of the lack of
government authority in certain parts of the country is that even when
the GRP and a rebel group both appear to have a willingness to nego-
tiate toward a conflict settlement, the GRP has had difficulty making
a credible commitment to comply with and enforce such a settlement.
Indeed, factional politics and intragovernment instability severely
undermined peace processes with both the MNLF and the MILF, the
latter of which is still unresolved. If democracy is a factor associated
with governments that tend to be more willing to negotiate with mod-
erate elements of insurgencies, it appears that democratic states must
also have the capacity to implement a deal.
The Philippines has had limited capacity for several reasons. First,
the country’s geography makes committing to the implementation
of a peace agreement difficult for many of the same reasons it makes
fighting insurgents difficult. This can result in an “equilibrium” of pro-
tracted, low-intensity insurgency.
What exactly are these geographic conditions? The Philippines
is an archipelago of more than 7,000 islands, many of which have
rugged or jungle terrain, which provides an abundance of safe havens
where insurgents can recruit, stage operations, and avoid apprehension
by government authorities. The terrain also makes it difficult for the
government and insurgents to monitor each other and ensure trans-
parency during and after a peace process. Indeed, the combination of
geographic challenges and relative poverty in the Philippines means
that—as is the case in most countries in the developing world—the
logistics required for effective counterinsurgency, counterterrorism, or
stability operations are difficult.
These challenges differ across the various insurgencies in the Phil-
ippines. The more-concentrated, territorial nature of the MILF and
98 Countering Others’ Insurgencies

especially the ASG in the southern Philippines, along with the geo-
graphical constraints imposed by the islands where they operate, can
help government forces find insurgent camps and limits insurgents’
freedom of movement to elude apprehension when the military or
police are able to conduct timely operations. At the same time, parallel,
state-like structures built by the MNLF and the MILF make parts of
the southern Philippines akin to denied or hostile areas, in which the
AFP’s operational latitude is limited. Conversely, the more-dispersed
nature of the NPA has allowed it to better withstand strong govern-
ment offensives, although it has also made it more difficult for the
group to firmly establish its primacy.
The GRP’s limited capacity must be understood in comparative
perspective. Relative to all regimes battling insurgencies, which tend to
have very low governance and overall state capacity,54 the Philippines
has a reasonably competent governmental infrastructure that can pro-
vide some degree of public goods even in remote rural areas. Compared
with most of the countries of sub-Saharan Africa and many of the
poorer countries of Asia, it scores considerably better on both indices of
state capacity used in this study.55 The fact that it continues to struggle
to implement government policies, especially in the remote regions of
the country, is evidence of the challenges the classic counterinsurgency
model faces in most conflict-affected countries of the developing world.

The Threat of Institutional Weakness and Strained Civil-Military

Relations to Civilian Control
In addition to the geographical and logistical challenges involved in
vanquishing an insurgency in the Philippines or cooperating with
one, the institutional capabilities of the AFP dedicated to these efforts
are modest. Relatively weak civilian oversight, along with embedded
patronage networks, sometimes results in corruption and mismanage-
ment, especially within the AFP officer corps.56 This weakness threat-

54 Fearon and Laitin, 2003, pp. 75–76.

55 Holtermann, 2012; Kaufmann, Kraay, and Mastruzzi, 2009.
56 Dencio Severo Acop, “The Expanded Nontraditional Role of the AFP: A Reassessment,”
Prism, Vol. 3, No. 2, March 2012, pp. 110–112.
 Counterinsurgency in the Philippines 99

ens to undermine attempts to instill a culture of professionalism, obe-

dience to civilian rule, and restraint from civilian abuses. Reports of
AFP human rights abuses have decreased since the all-out war against
the MILF in 2000 and particularly after the establishment of U.S.
SOF JTFs in early 2002. Professionalism within the AFP officer corps
remains uneven, however, and paramilitary units—though vastly
improved in comparison with their predecessors under Marcos—
remain, in many cases, beyond the central government’s control and
are thus more prone to committing abuses at the behest of local bosses.
Ultimately, the AFP’s limited capacity to conduct effective, sustained,
and disciplined counterinsurgency operations means that those design-
ing the policies and strategies to combat insurgents often struggle to
ensure that they are implemented as planned.

State Capacity and Civil Counterinsurgency

Despite numerous post-Marcos strategic and doctrinal initiatives
to broaden counterinsurgency efforts to include civil functions in a
“holistic” approach, the AFP has in practice shouldered much of the
burden of internal-security operations. Consequently, deviations from
new counterinsurgency approaches have been relatively common, and
the GRP’s response has thus often been to revert to old, conventional
approaches that rely on heavy firepower and sweep operations, which
have failed to eradicate the insurgencies and have sometimes undone
progress made in addressing popular grievances.
Despite the incomplete reform of GRP counterinsurgency
approaches during successive post-Marcos democratic governments,
there is some correlative evidence linking development projects to
counterinsurgency successes.
As Figure 4.4 illustrates, after the onset of AFP Development
Support Operations (DSO) in 2004, which included development
project delivery, there was an upward trend in the number of projects
delivered and a downward trend in both the number of barangays
assessed as affected by insurgents and the number of guerrilla fronts,
the local organizational units of NPA insurgents. These trends were
strongest in 2007 and 2008, when the AFP activated the National
Development Support Command (NADESCOM) to institutionalize
100 Countering Others’ Insurgencies

Figure 4.4
Development Projects, Insurgent-Affected Barangays, and Guerrilla Fronts

Impact of DSO
(2001–2010)
3,000
Affected barangays
2,490 2,510
2,250 2,394
2,178 2,121
1,996 2,072 Projects
Number

1,500
1,381 1,077 1,123

Recipient barangays
750

0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Social Institutionalization of 2010 projects

preparation DSO through activation not
of NADESCOM delivered
110
105 106 106 107
100 106
83 87
Number

55 Guerilla fronts 62 52
51

28

0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

SOURCE: Interview with former commander, AFP NADESCOM, 2013.

RAND RR513-4.4

its development efforts in counterinsurgency. The trends also reveal a

minor reversal in 2010, when projects were not delivered. NADESCOM
was dissolved in 2012, and there currently appears to be somewhat
less enthusiasm for AFP-led development efforts, possibly as a result of
personnel turnover in high-level Philippine defense positions, includ-
ing the Secretary of National Defense and the AFP Chief of Staff in
2010.57 The longer-term implications of this suspension of AFP devel-
opment programs are as yet unclear.

57 Interview with former NADESCOM commander, June 2013, Manila, Philippines.

 Counterinsurgency in the Philippines 101

There are also qualitative examples that illustrate what robust

civil-military integration in counterinsurgency looks like when the
GRP has the capacity and will to implement a classic counterinsur-
gency approach in a concerted, sustained way. The counterinsurgency
effort in Bohol province, once an NPA stronghold, is one of the clearest
illustrations of successful civil-military integration. In Bohol, develop-
ment was assigned a significant role in the local campaign, and pro-
vincial counterinsurgency efforts were subsumed under the auspices
of the Bohol Poverty Reduction Management Office. The civil aspect
of the Bohol campaign was multifaceted: Governance, environmental
management, social welfare, peace and security, urban management,
education, health, rural improvement, population management, sus-
tainable livelihood, infrastructure, and economic management were
components of the overarching effort against the NPA.58 Militarily, the
AFP expanded the use of Special Operations teams in Bohol to collect
village-level intelligence and conduct other small-unit operations.59 But
military operations were just one component of a broader, multifaceted
GRP approach to counterinsurgency in Bohol.
While perhaps an exceptional case—the GRP considered Bohol
a high-priority province for operations against the NPA—Bohol indi-
cates the potential of such an approach. Indeed, the war-torn province
changed from among the poorest provinces in the Philippines in 1997
to one of the fastest-developing provinces by 2001. By the mid-2000s,
Bohol was virtually free of insurgency.60
Bohol is also notable for the fact that U.S. civil engagement has
supported Philippine counterinsurgency objectives there. In response
to reports of drug trafficking and NPA money-laundering operations,
U.S. Secretary of State Colin Powell authorized additional aid to build

58Kathline Anne Sigua Tolosa, “The Rhetoric and Practice of the Security-Development
Nexus,” OSS Digest, 1st Quarter, 2010, pp. 32–33.
59 Jonathan P. Hastings and Krishnamurti Mortela, The Strategy-Legitimacy Paradigm: Get-
ting It Right in the Philippines, master’s thesis, Monterey, Calif.: Naval Postgraduate School,
December 2008, pp. 59–72; Ernesto C. Torres, Jr., A Success Story of Philippine Counterin-
surgency: A Study of Bohol, master’s thesis, Fort Leavenworth, Kans.: U.S. Army Command
and General Staff College, November 2010, pp. 8–9, 66–68.
60 Hastings and Mortela, 2008, pp. 68–69.
102 Countering Others’ Insurgencies

and support Philippine capacity to counter these activities, which was

disbursed primarily by the U.S. Agency for International Development
(USAID).61 The U.S. military did not play a significant role in Bohol’s
post-2001 success, as the NPA insurgency in Bohol is domestic and
thus beyond the scope of its purview, which is focused primarily on
ASG and foreign elements in the south.

Analysis of U.S. Assistance

The U.S. military has played a significant role in the southern Philip-
pines in OEF-P since 2001, when the U.S.-Philippines security rela-
tionship was reestablished after security cooperation between the two
countries had diminished in the 1990s.
In the post-9/11 period, the Philippines became the Southeast
Asian front line of U.S. counterterrorism efforts in the Global War
on Terrorism. During the 1990s, there was relatively little interest in
the ASG or its links to foreign terrorist groups.62 But after the 9/11
attacks, the recognition that numerous key al Qaeda figures had used
the Philippines for various fund-raising and operational purposes and
that the Philippines served as one of its key potential staging grounds
for expanding al Qaeda’s influence in Southeast Asia, made the Philip-
pines a key interest of U.S. counterterrorism officials.63
U.S. interest in Philippine counterinsurgency, however, was
largely limited to the southern Philippines, and specifically to parts of
Mindanao, such as Basilan, Tawi-Tawi, and Sulu, where the al Qaeda–
linked terrorist network was believed to be most active. The U.S. resur-
gence in the Philippines did not—and was not intended to—address
the main insurgent threat to the GRP, the communist NPA. These
restrictions on U.S. activities have limited the influence of the United

61 Hastings and Mortela, 2008, pp. 68–69.

62Michael A. Sheehan, Crush the Cell: How to Defeat Terrorism Without Defeating Ourselves,
New York: Random House, 2009, pp. 123.
63 Author interview with former U.S. government official, Manila, Philippines, June 24,
2013.
 Counterinsurgency in the Philippines 103

States’ advisory support, both in practical operational terms and in the

eyes of some AFP officers who see themselves as confronting a very
different—and for them, more vital—insurgent adversary.64
Indeed, the United States’ primary interest in the Philippines
after 9/11 was to advance U.S. counterterrorism objectives, not to
develop PSF capacity to combat insurgencies. The latter was a byprod-
uct of Philippine domestic law, which prohibited the U.S. military
from pursuing its initial plan—engaging in direct-action counterter-
rorism operations against particular individuals from select groups—
and forced the United States to develop PSF capabilities in order to
facilitate the degradation and destruction of militant networks in the
region. The restrictions placed on U.S. forces by the Philippine gov-
ernment required a more-nuanced, integrated response involving a
wide range of U.S. capabilities, from civilian aid and U.S.-supported
civil-military operations to intelligence support from the United States’
technical platforms and logistics involving U.S. SOF aviation assets.65
The post-2001 U.S. military objective in the Philippines has
remained largely constant: to degrade terrorist threats in the region.
The U.S. objective and presence have resulted in high-quality intel-
ligence and in increased oversight of the PSF in its areas of operation,
reducing the incidence of indiscriminate violence due to shoddy infor-
mation or a lack of oversight and accountability. U.S. influence on the
PSF, however, is largely limited to those areas in the southern islands
where JSOTF-P operates. Human rights abuses and other counterin-
surgency practices that are undesirable from the United States’ stand-
point continue to occur and remain a significant concern in other parts
of the country. These bad counterinsurgency practices, however, should
be considered in relative terms. Overall, the AFP has transitioned from
brute-force approaches—a persistent characteristic of operations during

64 Author interview with former JSOTF-P officer, April 14, 2013; author interview with
former U.S. government official, Manila, Philippines, June 24, 2013; author interview with
AFP Lt. Col., Quezon City, Philippines, June 22, 2013; author interview with AFP (Ret.)
Brig. Gen., Makati City, Philippines, June 23, 2013.
65 Thomas M. Scanzillo and Edward M. Lopacienski, “Influence Warfare: How the
Joint Special Operations Task Force–Philippines Maximizes Effects to Create Long-Term
Stability,” Center for the Study of Irregular Warfare and Armed Groups, February 2011.
104 Countering Others’ Insurgencies

the period of martial law and a sporadic one of the fragile democratic
governments of the 1990s—to the use of focused tactics that seek to
minimize harm to the civilian population. This transformation is con-
sidered a key component of the success against militant groups in the
past decade.66

Capacity-Building and Enhanced Civilian Oversight

U.S. partnership assistance has sought to influence GRP counterinsur-
gency practices by providing military and other assistance intended to,
among other things, influence its political inclusivity, military supe-
riority, and overall state capacity and thereby shape its response to
the threats the United States cares about most. Over the past decade,
this assistance has improved GRP and AFP capabilities, particularly
through intelligence-sharing and tactical advice that has resulted in
highly precise military operations. It has also improved assessments of
local populations’ needs and support for the provision of goods and
services that satisfy those needs.67 Equally important, U.S. assistance
has demonstrated to the AFP the value of a classical counterinsurgency
strategy that balances discriminate, targeted uses of force with non-
kinetic approaches involving various types of assistance to the popula-
tion or acceptable concessions to reconcilable insurgent elements.68
OEF-P has played an additional stabilizing role: It has helped tip
the civil-military balance within Philippine power structures in favor
of the civilian government, primarily by providing funding and other
support in coordination with the GRP, which had long struggled to
assert and consolidate its primacy over the military. Important for
counterinsurgency, this occurred not only at the national level but
also at the local level, where it relied on an indirect approach in which
AFP units led operations but did so with critical support and oversight
from U.S. advisors. As such, U.S. military assistance to the Philippines

66 Lambert, Lewis, and Sewall, 2012, pp. 122–123.

67 Farris, 2009; Lambert, Lewis, and Sewall, 2012; Scanzillo and Lopacienski, 2011.
68 Rosalie Árcala Hall, “Boots on Unstable Ground: Democratic Governance of the Armed
Forces Under Post 9/11 U.S.-Philippine Military Relations,” Asia-Pacific Social Science
Review, Vol. 10, No. 2, 2010, pp. 30–34.
 Counterinsurgency in the Philippines 105

during OEF-P has resulted in dramatic changes not only in the tacti-
cal proficiency and operational capacity of the AFP and other PSF, but
also in the AFP’s role in and level of local engagement.69
As a result of the American presence during OEF-P, Philippine
civil-military engagement was influenced in two important ways. First,
the U.S.-PSF military partnership enabled the AFP to reach local areas
and their primary administrative bodies, LGUs, as well as civilian
populations that previously were beyond the effective control of GRP
national-level mechanisms. This occurred not only through the provi-
sion of logistics support for AFP operations, but also through intel-
ligence support that helped the AFP to conduct timely, intelligence-
driven operations.70 The AFP, with U.S. support, was thus able to
engage in civil-military operations, support development efforts, and
develop a robust messaging strategy intended to influence the popula-
tion favorably toward the GRP. Second, the AFP’s extended reach and
ability to provide basic security functions meant LGUs and private citi-
zens were less likely to rely on local paramilitaries. The mutual reliance
of LGUs and civilian populations, on the one hand, and the AFP and
other GRP agencies, on the other hand, has since 2001 enhanced state
capacity by enabling the GRP to provide better security and other ser-
vices and to conduct tactical military operations in previously denied
areas of Basilan, Sulu, and other locations in the southern Philippines
where ASG and foreign militants operate.71

The U.S. Role: Limitations and Risks

Despite their benefits, U.S. activities suffer from limitations of their
own. First, U.S. and Philippine objectives are imperfectly aligned,
which poses a risk to both GRP and U.S. interests. The United States’

69 Hall, 2010, pp. 27–28.

70 Interview with former JSOTF-P commander; multiple interviews with former JSOTF-P
officers, Arlington, Va., April and May 2013.
71 Interview with former Special Operations Command Pacific commander, Cambridge,
Mass., March 2011; interview with former JSOTF-P Psychological Operations officer,
Arlington, Va., April 2013; interview with former AFP Chief of Staff, Manila, Philippines,
June 2013; interview with former NADESCOM commander, Makati, Philippines, June
2013.
106 Countering Others’ Insurgencies

primary interest is in counterterrorism—specifically, disrupting,

degrading, and dismantling Salafi jihadist networks that operate in
Southeast Asia and have used the Philippines as a hub or for sanctu-
ary among the ASG or in MILF base camps.72 From the United States’
perspective, AFP counterinsurgency is a critical part of an indirect
approach intended to advance U.S. counterterrorism objectives. The
GRP-MILF peace process, however, has complicated the U.S. counter-
terrorism campaign. And from the GRP’s perspective, U.S. counterter-
rorism goals have complicated its efforts to reach an acceptable end-
state with the MILF. When the United States raised the possibility of
declaring the MILF a terrorist group because of its relationship with
the ASG, the GRP objected strenuously and has repeatedly sought to
limit military countermeasures against the MILF in its controlled ter-
ritory for fear of jeopardizing a political resolution with the group.73
Yet GRP protection of the MILF has at times allowed ASG and JI
operatives to use MILF camps as de facto safe-haven areas, resulting
in tensions between U.S. objectives and GRP policies. Indeed, aggres-
sive U.S.-backed AFP counterterrorist pursuit operations into MILF-
controlled areas have risked compromising the ever-fragile peace pro-
cess. This possibility remains a concern—tactical-level successes, such
as capture/kill raids or airstrikes against senior terrorist leaders, could
trigger MILF retaliation and escalation to war. Another possibility is
that such tactical strikes in the south could radicalize the MILF (or
other armed Muslim factions) and push it toward the ASG, which
would strengthen the United States’ main enemy in the country and
inadvertently make it a more credible spoiler of the MILF peace pro-
cess, hurting both U.S. and GRP interests in the process.74
Second, the U.S. and the GRP have divergent interests in another
key area for counterinsurgency, the NPA. As with the MILF, the

72 Ressa, 2004.
73 Abuza, 2005, pp. 14–16.
74 For a detailed overview of these scenarios and the risk of U.S.-backed counterterrorist
operations to the peace process, see David S. Maxwell, “Operation Enduring Freedom–
Philippines: What Would Sun Tzu Say?” Military Review, Vol. 84, No. 3, 2004, pp. 20–23;
International Crisis Group, Counterinsurgency vs. Counterterrorism in Mindanao, Asia Report
No. 152, Islamabad/Brussels, pp. 20–26; and Abuza, 2005, p. 15.
 Counterinsurgency in the Philippines 107

GRP anticipates a negotiated solution to the NPA conflict. Unlike the

MILF, however, the GRP views the NPA as its main threat, if not an
existential one. The United States has shown little interest in providing
military support to the GRP’s NPA counterinsurgency. Since 2001, the
United States has significantly increased foreign-military funding and
other economic assistance, which indirectly supports the GRP coun-
terinsurgency against the NPA,75 but the lack of a direct advise/assist
role and the provision of unique capabilities, such as intelligence, sur-
veillance, and reconnaissance (ISR) assets, leads some AFP officers to
questions the commitment of the United States to the mutual-defense
agreement that formalized the post-2001 reinitiation of significantly
deepened U.S.-Philippine security cooperation.76
Third, intelligence- and information-sharing between the United
States and the PSF has considerable transaction costs and limitations
due to the fact that U.S. personnel cannot divulge sensitive sources
or methods to their Philippine partners. These limitations could have
a significant impact on the sustainability of the AFP’s improvements
in the use of discriminate force against known militants. Filipinos are
likely to struggle to target insurgents and terrorists efficiently and dis-
criminately if U.S. intelligence support diminishes, which could hasten
a return to old, less-discriminate practices.77

Conclusions
The Philippines counterinsurgency environment has progressed in a
way that is largely consistent with the framework advanced in this
study. The Marcos regime declared martial law from 1972 to 1983 and

75 Since 2001, U.S. military assistance (the combined amount under the foreign military
financing [FMF], foreign military sales [FMS], international military education and training
[IMET], and excess defense articles), for example, grew from $10.5 million to $42.8 million
in 2007. Hall, 2010, p. 34.
76 Multiple interviews with AFP officers, Tarlac and Makati, Philippines, June 2013; multi-
ple interviews with former JSOTF-P officers, April 2013; interview with former U.S. military
attaché, Makati, Philippines, June 2013.
77 Interview with former JSOTF-P intelligence officer, Arlington, Va., May 2013.
108 Countering Others’ Insurgencies

adopted an increasingly repressive and indiscriminate strategy vis-à-vis

the opposition, even undermining chances for a sustained peace with
the MNLF separatists through its policies. Subsequent democratic
regimes, starting with Corazon Aquino’s in 1986 and followed by Fidel
Ramos’ in 1992, implemented policies that were more open to accom-
modating moderate or reconcilable elements of the opposition, while
continuing military campaigns against opposition that was less will-
ing to engage the government. These military campaigns were often
aimed at pacification—an integrated civilian and military approach
that would combine reconstruction and military operations—although
they sometimes reverted to more-conventional, military-led campaigns
that relied on heavy force against insurgent encampments. After 2001,
however, the Arroyo and Aquino governments largely reconstituted
earlier efforts to engage moderate insurgents, advance peace processes,
and develop methods for more-discriminate targeting of insurgent and
terrorist groups’ hard-line elements.
The democratically elected Estrada regime, starting in 1998 and
ending in 2001, is the main outlier to this general pattern. With U.S.
support to the Philippines during the 1990s significantly diminished,
the Estrada regime was particularly weak, and at the same time, it
was renowned for its corruption, siphoning off remaining available
resources that could have been used to build governmental capac-
ity to counter insurgents. With little remaining capacity in the civil
sector and low-level violence between the AFP and insurgents ongoing,
Estrada authorized the military to resume major combat operations
that derailed the southern peace process and caused widespread harm
to the civilian population.78 The resumption of significant U.S. support
to the Arroyo regime after 2001 was a key turning point in the return
to previous democratic regimes’ strategy of conciliation and a more bal-
anced counterinsurgency approach.
The Philippines provided a hospitable environment for the United
States to partner with to counter mutual threats, particularly during

78 Carolina G. Hernandez, “Institutional Responses to Armed Conflict: The Armed Forces

of the Philippines,” background paper, submitted to the Human Development Network
Foundation, Inc., for the Philippine Human Development Report, 2005, pp. 14–15.
 Counterinsurgency in the Philippines 109

the post-9/11 period, by which time the GRP had established con-
solidated, if imperfect, democratic institutions. Like other democrati-
cally elected governments that followed Marcos’ regime, the post-2001
Philippines has sought to negotiate an end to conflicts with groups it
considers reconcilable. These efforts have met with only partial suc-
cess: The MILF peace process holds significant promise for an agree-
ment, while NPA negotiations have never developed to the same stage
of maturity and are currently stalled.79
The Philippines has adapted the military side of its counterin-
surgencies toward a greater focus on the population, making progress
in minimizing civilian casualties and the incidence of indiscriminate
violence.80 Oversight and headquarters-level control of certain units in
certain locations remains limited, but U.S. military assistance and advi-
sors have facilitated a more sustained population-centric approach than
ever before by easing resource burdens and sharing operational meth-
ods and techniques for conducting a range of civil-military operations.
By injecting additional resources, oversight, and capabilities, the
U.S. military also appears to have helped to stabilize civil-military
relations—a chronic source of policy instability in the Philippines since
the transition to democracy in 1986. It is entirely possible, however,
that this stability will be fleeting, as there is no institutional mecha-
nism in place to build or sustain the capacity to maintain smoother
civil-military relations in the absence of the combination of qualities
and resources the United States brings to the environment.
Finally, although the United States’ partnership with the GRP
should be considered a success story, claims about that success should
be qualified by questions concerning the long-term sustainability of
the gains that have been made. Philippine capacity to counter insur-
gent and terrorist groups effectively and selectively is unquestionably
greater now than it was during any part of the pre-9/11 period. It is
unclear, however, whether the GRP will be able to deploy and sustain
certain capabilities that have enhanced its capacity and contributed

79International Crisis Group, The Philippines: Dismantling Rebel Groups, Asia Report
No. 248, Islamabad/Brussels, June 19, 2013b, pp. 1–3.
80 Lambert, Lewis, and Sewall, 2012, pp. 122–126.
110 Countering Others’ Insurgencies

to its counterinsurgency and counterterrorism successes over the past

decade. The United States, acting indirectly in support of the GRP, has
provided important intelligence support and logistics capabilities, such
as intelligence obtained from technical sources and airborne ISR assets,
robust helicopter support, and night-vision capabilities, all of which
are critical enablers of timely collection and validation of intelligence
and of precise military operations that minimize collateral damage and
civilian harm.81 The Philippine government is unlikely to be able to
harness and exploit these capabilities independently. Thus, a long-term
U.S.-Philippine partnership and continued U.S. support to the GRP,
albeit at a low level, are likely to be necessary to sustain success in
the Philippines, unless the GRP manages an enduring resolution to its
conflicts with the myriad groups and factions that contest its rule.

81 Lambert, Lewis, and Sewall, 2012, p. 127.

CHAPTER FIVE

Counterinsurgency in Pakistan

Introduction

Since its independence in 1947, Pakistan has faced numerous chal-

lenges from violent non-state actors. The number and intensity of
internal conflicts began to escalate following the U.S.-led invasion
of Afghanistan in 2001, particularly in the northwest regions of the
country. The United States began to take an active interest in Paki-
stan’s approach to counterinsurgency at that time and especially after
the Taliban resurgence beginning in approximately 2006. In a 2007
National Intelligence Estimate, U.S. officials expressed concern over
Pakistan’s counterinsurgency campaign against the Taliban and their
foreign and al Qaeda affiliates operating in the northwest and tribal
areas. According to the estimate, increased activity by these groups
threatened to erode gains made in Afghanistan and, more importantly,
to destabilize a nuclear-armed Pakistan.1 As a result, the United States
adopted a more activist approach to influencing the Pakistan govern-
ment’s counterinsurgency operations through military and economic
assistance.
This chapter examines the drivers and trajectory of Pakistan’s
various counterinsurgency campaigns since 2001. In contrast to the
Philippines, Pakistan possesses a large, professional, and highly capa-
ble military but weak state capacity and civilian institutions. Despite
frequent elections over the past decade, opportunities for meaningful

1 National Intelligence Council, The Terrorist Threat to the Homeland, National Intelli-
gence Estimate, Washington, D.C., July 2007.

111
112 Countering Others’ Insurgencies

political participation remain limited.2 Whereas the Philippines offered

an opportunity to examine a U.S. partnership under something close
to best-case conditions, Pakistan represents a much more challenging
case. This contrast between the two countries is useful for understand-
ing both the limitations and the potential of U.S. partnerships with
counterinsurgent regimes.
The Pakistan case merits careful consideration for other reasons
as well. Because the government of Pakistan has fought multiple insur-
gencies during this period, it offers the chance to explore many “mini-
cases” under different subnational conditions. Comparing the different
approaches that the government has taken at different points in time
and against different insurgent groups provides us with considerable
insight into the domestic determinants of counterinsurgency strategy
and practices.
Regime characteristics—in particular, the extent of political inclu-
sion and state capacity—have shaped Pakistan’s preferred strategies for
countering insurgency. The nuances of case-study methods allow us to
explore subtle variations in both domestic features and counterinsur-
gency approaches. By dividing Pakistan across distinct regions (Swat/
Malakand, Balochistan) and various agencies within FATA and over
different time periods (before and after civilian governments), we are
able to analyze how different levels of inclusion, state capacity, and mili-
tary superiority in eight subcases have impacted strategy and outcomes.
Across these eight subcases, we find that Pakistani governments
have usually pursued containment strategies typical of less politi-
cally inclusive regimes with low state capacity. As the government has
moved closer toward democracy in recent years, however, it has often
(although not uniformly) adopted strategies characteristic of more-
inclusive regimes, emphasizing greater efforts toward accommoda-
tion of the reconcilable opposition and more discriminate use of force.
Yet even in this period, government strategy has varied by region, as
some subcases reveal. In regions closer to the core of the Pakistani state
that possess better institutions for higher levels of political inclusion

2 Pakistan’s democracy score on the seven-point Freedom House index has improved over
the past five years, but it has still not reached the midpoint.
 Counterinsurgency in Pakistan 113

and state capacity (e.g., the Swat Valley), the government has adopted
approaches resembling classic counterinsurgency strategies. In periph-
eral regions like FATA and Balochistan, in contrast, the state has been
more prone to using force indiscriminately and denying even poten-
tially reconcilable opposition movements the opportunity for mean-
ingful political participation in state structures.
The U.S.-Pakistan security partnership is commonly regarded as
a failure, particularly since the steady decline in relations that began in
2011. Our analysis suggests that a more-nuanced view is appropriate.
The study clearly reveals the limits of U.S. leverage, even in countries to
which it provides large amounts of assistance, but it also reveals specific
circumstances under which the United States might positively engage.
Finally, it suggests the potential for even partners with unfavorable
characteristics to adopt elements of Western counterinsurgency best
practices, at least under certain conditions. To be clear, the story of the
U.S.-Pakistan relationship is not an encouraging one. It does, however,
hint at ways in which the United States might influence problematic
partners’ counterinsurgency practices on the margins.
The first section below provides background detail on the
Government of Pakistan (GoP), the insurgent groups it has fought
since 2001, and the counterinsurgency campaigns in FATA, Swat/
Malakand, and Balochistan during the pre- and post-2008 peri-
ods. The next section analyzes the influences of the three domestic
structural variables—political inclusion, state capacity, and military
superiority—across each of the eight subcases. The third section ana-
lyzes the role of the U.S. partnership assistance in shaping GoP strat-
egy. Finally, the chapter concludes with some key findings supporting
the broader claims of this report and some additional implications of
the Pakistan case.

Background

The GoP has presided over an extremely diverse and contentious ethnic
and political landscape carved out of British India in 1947. Throughout
nearly all of its independence, Pakistan has been considered a weak and
114 Countering Others’ Insurgencies

non-democratic state, but it has possessed a very large and capable pro-
fessional military. It has fought four major wars with its longtime rival,
India, as well as a number of counterinsurgency campaigns. Today, its
landscape is dotted with numerous militant organizations of ethno-
national, sectarian, and religious extremist variants, some of which the
GoP disregards, some which it directly fights, and some of which it
tacitly relies on as an internal or external national security tool. The
government has engaged in three major counterinsurgency campaigns
against the Tehrik-e-Taliban Pakistan (TTP) since 2001, its previ-
ously inchoate incarnations in FATA since 2002, the Tehrik-e-Nifaz-e-
Shariat-e-Mohammadi (TNSM) in the Swat Valley and Malakand
Division since 2007, and a range of Baloch nationalist insurgent organi-
zations in Balochistan since 2005. At certain points in time, it has pros-
ecuted all three counterinsurgency campaigns simultaneously, stretch-
ing the limits of an already weak state. Consequently, its approach to
these insurgencies has varied, following preexisting patterns of varying
geography, political topography, and state capacity across regions and
even subregions.

Context
Physical and Human Geography
The physical and economic landscape of Pakistan has been very
unevenly distributed, which has led some to describe it as an artificial
construct. Pakistan is roughly the size of France and the United King-
dom combined and home to 180 million people. While the province of
Punjab forms the bulk of Pakistan’s industrial base, population centers,
political authority, and arterial road networks, the province of Sindh
remains important as an agricultural center and home to Pakistan’s
largest city and commercial capital, Karachi. The Northwest Fron-
tier province, recently renamed Khyber-Pakhtunkwa (KP), has been
increasingly integrated into the political, economic, and social fabric
of Pakistan since the 1970s. However, the vast expanse of Balochistan
province, which makes up 47 percent of Pakistan’s territory but only
4.5 percent of its population, largely exists apart from the other three
provinces and is mainly valued for its natural-resource deposits and its
newly developed Gwadar port. FATA, Gilgit-Baltistan (previously the
 Counterinsurgency in Pakistan 115

Federally Administered Northern Areas), and Azad Jammu and Kash-

mir (AJK) are also physically remote and sparsely populated territories.
This geography underpins very uneven political topography, par-
ticularly for the non-provincial territories where political authority is
more closely akin to suzerainty than modern state governance. FATA
remains outside the Pakistani constitution and is governed by federal
political agents and tribal leaders under a set of draconian rules laid
down by the British in 1901, and the GoP has deployed segmented
and distinct approaches to different agencies even within FATA based
on tribal composition.3 Gilgit-Baltistan was also governed directly by
the federal government outside the constitution until 2009. AJK is a
self-governing state under Pakistani control. Meanwhile, 95 percent of
the territory of Balochistan (the rural areas where 52 to 75 percent of
the population lives) is administered as “B areas,” where the provincial
police have no jurisdiction and law and order is administered by tribal
levies controlled by Baloch tribal leaders (Sardars).
The human terrain of Pakistan is equally complicated, with most
major ethno-linguistic groups concentrated and generally dominant
in the provinces named after them, creating an array of centrifugal
forces contributing to tension, instability, and conflict. The dominant
ethno-linguistic group of Punjabis makes up 45 percent of the popu-
lation. The rest of the population is made up of 9 percent Saraikis
(concentrated in southern Punjab and sometimes counted as Punjabis),
15 percent Pashtuns, 14 percent Sindhis, 7.5 percent Mohajirs (Urdu-
speaking immigrants from Northern India, with the majority concen-
trated in urban Sindh), 3.5 percent Balochi, and 6 percent other. Along
with English, the official language of Pakistan is Urdu, although only
8 percent of the people are native speakers. Nearly all groups have been
involved in some violent internal challenge to the state or dominant
groups, through either ethno-nationalist movements or Islamist mili-
tant movements. Prior to 1947, most of modern-day Pakistan consti-
tuted the frontiers of the British Raj, indirectly ruled and policed due
to the harsh terrain and unruly tribes. Only Punjab province, which

3 For examples, see Akbar S. Ahmed, Resistance and Control in Pakistan, rev. ed., London:
Routledge, 2004.
116 Countering Others’ Insurgencies

was both one of the richest provinces and one of the chief contributors
to the British Indian Army,4 was directly governed.
While Islam is the official religion of Pakistan and over 96 per-
cent of the population is Muslim (the majority Sunni and about 10 to
15 percent Shia), religious friction persists. Pockets of religious minori-
ties of Christians, Hindus, Sikhs, and Ahmadis5 constitute the remain-
der, but their numbers have declined due to migration or conversion.
Sectarian tensions have frequently escalated to violence and acts of ter-
rorism, not only between Sunnis and Shias but also within the Sunni
majority (for instance, between the Barelvi and Deobandi sects).
On paper, Pakistan is the most urbanized country on the South
Asian subcontinent (excluding the Maldives), with 36 percent of the
population living in cities. In daily life, however, kinship networks
remain strong and have prevented the breakdown of old sociocultural
structures. The country’s ethno-linguistic diversity, combined with the
strength of kinship networks of collective solidarity, have left Pakistan
a weak state beset by many strong societies.6

Government
Regime type and political inclusion. Since its independence, Paki-
stan has vacillated between civilian-led and military-led governments.
For the most part, however, the politics of the country have not been
very democratic or politically inclusive, and the military has histor-
ically dominated politics even during periods of civilian rule.7 Even
though an elected civilian regime took office in 2008, a number of
scholars and practitioners do not consider Pakistan a democracy given

4 Tan Tai Yong, The Garrison State: The Military, Government and Society in Colonial
Punjab, 1849–1947, New Delhi, India: Sage Publications, 2005; Perry Anderson, “Why
Partition?” London Review of Books, Vol. 34, No. 14, July 19, 2012.
5 Ahmadis are a sect that claims to be a branch of Islam, but a Pakistani constitutional
amendment declared them a non-Muslim minority, and they are prohibited from identifying
themselves as Muslims.
6 Anatol Lieven, Pakistan: A Hard Country, London: Allen Lane, 2011, pp. 12–16.
7 Husain Haqqani, Pakistan: Between Mosque and Military, Washington, D.C.: Carnegie
Endowment, 2005.
 Counterinsurgency in Pakistan 117

that the military retains substantial control, and civil liberties such as
individual rights, freedom of expression, and rule of law have remained
poor even as political rights improved.8
Greater checks and balances within the government of Pakistan
have emerged since 2008, and the military has generally refrained from
overt interventions in politics. The passage of the 18th amendment
to the Constitution in 2010 reduced the powers of the executive and
devolved greater powers to the legislature and provinces. The legislature
has also grown more assertive on domestic politics and some aspects of
security such as counterterrorism, although the military continues to
dominate most aspects of national security. A strong and independent
judiciary has emerged since 2007, with the Supreme Court expanding
its oversight role and holding the military, the executive, and the leg-
islature to account. Additionally, the media has become an important
player on the political landscape and has grown substantially in terms
of the number of outlets and journalists.
These steps toward greater democracy were capped by a major
historical achievement in March 2013, when the democratically elected
civilian government became the first to complete a five-year term and
peacefully transition without military intervention to a newly elected
civilian government, suggesting a more robust and sustained com-
mitment to democratic governance by all institutions.9 Some charge
that the military has refrained from intervening in domestic politics
in large part as a choice, because it still controls national-security
policy. While this is true to a large extent, its choice has been in part
shaped by institutional developments that make interventions costlier,
such as a stronger media, judiciary, and parliament, and these civilian
institutions have begun to assert themselves in some national-security
debates. Thus, while Pakistan as a whole cannot be considered to have

8 Based on scores and subscores from Freedom House, 2013.

9 Declan Walsh, “Court Challenges Put Unusual Spotlight on Pakistani Spy Agency,” The
New York Times, February 7, 2012; Omar Waraich, “Two Cheers for Pakistani Democracy:
A Sobering Milestone,” Time, March 18, 2013; Saeed Shah, “In Historic First, Pakistan’s
Civilian Government Completes Its Term Without a Coup,” McClatchy News, March 16,
2013; Ishrat Saleem, What Pakistan’s Democratic Future Holds for the United States, Wash-
ington, D.C.: Middle East Institute, May 9, 2013.
118 Countering Others’ Insurgencies

been democratic during the period covered in this study, it is important

to pay close attention to subtle variations over time, between regions,
and among different institutions and policy arenas. It seems difficult
to deny the expansion of some political-inclusion mechanisms like
coalition breadth, transparency, and executive constraints since civil-
ians returned to power in 2008. Even as the central government has
become increasingly democratic and provincial assemblies like that in
KP more assertive, regions like Balochistan and FATA remain poorly
integrated into these liberalizing trends. Moreover, in many areas of
national security, the security and intelligence services remain a law
unto themselves. These variations in the extent of democratic gover-
nance play crucial roles in explaining the multiplicity of counterinsur-
gency strategies the state has adopted over the past decade. Later in this
chapter, we examine variation in the state’s approach to counterinsur-
gency through eight subcases divided by time (pre- and post-2008) and
by the regions described above.
State capacity. Pakistan is considered by most scholars to be a
relatively weak state (like all states in South Asia) because of its weak
economy, level of development, political capacity, and quality of gover-
nance.10 In Table 2.2 of Chapter Two, Pakistan is identified as a coun-
try with low state reach—a categorization confirmed by one scholar
who wrote that “the [Pakistani] state . . . does not necessarily affect
many people’s lives very much.”11 Pakistan also exhibits very low gov-
ernment effectiveness (with a World Bank composite score at the very
bottom of the third quartile, just above Nepal).12 Further indexes con-
firm this categorization. Pakistan is classified by the World Bank as a
lower-middle-income country, and most economic data sources place it
in the bottom quartile of countries for per-capita GDP (whether nom-
inal or purchasing power parity) and in the bottom quartile of the

10 T. V. Paul, South Asia’s Weak States: Understanding the Regional Insecurity Predicament,
Stanford, Calif.: Stanford University Press, 2010.
11 Lieven, 2011, p. 13.
12 The indexed measures of state capacity in Chapter Two identify the Philippines as pos-
sessing high state reach and moderate government effectiveness (safely located in the second
quartile).
 Counterinsurgency in Pakistan 119

United Nations human-development index.13 The Failed States Index

has ranked Pakistan between the 9th and 13th weakest states since
2006.14 Pakistan is one of the weakest countries in terms of extracting
taxes, with a tax-to-GDP ratio of 9.2 percent, which ranks it 113th
out of 122 countries and 23rd worst out of 176 countries in terms of
the percentage of the population paying income taxes.15 Pakistan also
ranks in the bottom quartile of countries in the 2012 Corruption Per-
ceptions Index.16 Even though it has a powerful, capable, and profes-
sional military, Pakistan’s state capacity also suffers from this in some
ways. In 2000, defense and debt servicing consumed three-quarters
of Pakistan’s federal budget.17 Like levels of political inclusion, state
capacity has varied tremendously in terms of reach and services among
Pakistan’s different regions.18
Security capabilities. Despite being a weak state, the Pakistani
government commands a very powerful modern military with strong
institutional underpinnings built by the British Indian Army from
1858 to 1947. Pakistan ranks twelfth in the global firepower index
of conventional military power, just ahead of Israel,19 and is one of
only nine countries in possession of nuclear weapons. It ranks eighth

13 Based on data from the IMF, The World Bank, CIA World Factbook, Penn World Table,
and United Nations Development Program, The Rise of the South: Human Progress in a
Diverse World, Human Development Report, New York, 2013.
14 Fund for Peace, “Failed States Index, 2005–2013,” web site.
15 WDI data on tax revenue as a percentage of GDP for 2008. Another model to mea-
sure relative political extraction finds Pakistan near the bottom. See Marina Arbetman-
Rabinowitz et al., “Pakistan Has Lowest Tax to GDP Ratio in the World,” The Nation,
February 18, 2013, and Jennie Matthew, “Report Unmasks Tax Evasion Among Pakistan
Leaders,” Agence France Presse, December 12, 2012.
16 Transparency International, “2012 Corruption Perceptions Index,” website, 2012.
17Syed Rifaat Hussain, “War Against Terrorism, Pakistani Perspective,” IPRI Journal,
Winter 2004.
18 A. F. Aisha Ghaus, Hafiz A. Pasha, and Rafia Ghaus, “Social Development Ranking of
Districts of Pakistan,” The Pakistan Development Review, Vol. 35, No. 4 (Part II), Winter
1996, pp. 593–614.
19 Global Firepower, “Countries Ranked by Military Strength (2013),” web site. The index
is based on 42 measures.
120 Countering Others’ Insurgencies

in terms of active forces, with over 900,000 regular and paramilitary

personnel. The forces that have been used for counterinsurgency are
the Army, with 550,000 active-duty personnel, and the Frontier Corps
(FC), estimated at between 65,000 and 80,000 personnel.20 Spend-
ing on the military has consistently outstripped domestic spending on
health, education, and public welfare.21
Pakistan has never lacked raw military superiority over insurgent
groups in the past decade, but it has been constrained nevertheless. Its
military power is primarily dedicated to countering its more-powerful
Indian neighbor, which it sees as an existential threat.22
Because Pakistan’s doctrine and forces are built and deployed
around countering this perceived threat, the government has had to
be selective about the degree of military power it deploys against insur-
gencies, ever mindful of preventing the military balance on its east-
ern border from falling too low to ensure defensibility in case of a
crisis. It has been estimated that Pakistan needs to retain about half
the number of brigades it has generally held on the border around
Punjab and Kashmir to maintain conventional deterrence.23 Although
some outside observers deride Pakistan’s continued focus on India, it is
impossible to ignore the fact that India is spending seven times more
on defense than Pakistan and accounts for more than 75 percent of all
defense spending in South Asia.24

20Seth G. Jones and C. Christine Fair, Counterinsurgency in Pakistan, Santa Monica, Calif.:
RAND Corporation, MG-982-RC, 2010; International Institute for Strategic Studies, The
Military Balance, London: Routledge, 2013.
21Stephen P. Cohen, The Idea of Pakistan, Washington, D.C.: Brookings Institution Press,
2004, p. 266; Hussain, 2004.
22Cohen, 2004, p. 121; Stephen Cohen, The Pakistan Army, Berkeley, Calif.: University of
California Press, 1984.
23 Sameer Lalwani, Pakistani Military Capabilities for a Counterinsurgency Campaign: A Net
Assessment, Washington, D.C.: New America Foundation, Counterterrorism Strategy Initia-
tive Publication, September 2009, pp. 41–46.
24 International Institute for Strategic Studies, 2013, p. 251.
 Counterinsurgency in Pakistan 121

Insurgents
A vast array of insurgent, terrorist, and militant organizations based in
Pakistan have challenged state forces, institutions, or authority at vari-
ous points in time. In addition to these clear challengers to the state,
there are a number of militant organizations that are either ostensibly
neutral or even allied with the GoP. We focus primarily on the stron-
gest insurgent organizations that have directly threatened the state and
have been the targets of counterinsurgency campaigns over the past
decade. These include the Taliban-affiliated TTP and TNSM, as well
as Baloch nationalist insurgents.
Tehrik-i-Taliban Pakistan.25 The TTP insurgency originated in
2001, following the ouster of the Taliban and al Qaeda from Afghan-
istan during the U.S.-led invasion, when a number of foreign fight-
ers took refuge in Pakistan’s northwest. Alongside local militant com-
manders indigenous to the tribal regions, they began to violently assert
their influence, challenging local political orders and attacking GoP
security forces. In response to Pakistan military operations targeting
al Qaeda after 2001 and more-robust operations beginning in 2004,
some local tribal militants and commanders who hosted these foreign
fighters fought alongside them against the state. At first the local com-
manders conducted operations just within FATA, but eventually their
attacks spread across the country.
The nascent Pakistani Taliban eventually formed the basis for the
TTP, which was officially declared in late 2007 to be an umbrella orga-
nization of 40 senior Taliban commanders and Islamist militant move-
ments totaling 40,000 fighters. These diverse movements were united
by a desire to fight a “defensive jihad” against the Pakistan Army’s
incursions in the tribal regions. In July 2007, the Pakistani military’s

25 This account is based on Hasan Askari Rizvi, “Understanding the Insurgency,” Daily
Times, October 5, 2008; Shuja Nawaz, FATA—A Most Dangerous Place: Meeting the Chal-
lenge of Militancy and Terror in the Federally Administered Tribal Areas of Pakistan, Washing-
ton, D.C.: Center for Strategic and International Studies Press, January 2009a; Imtiaz Gul,
The Most Dangerous Place: Pakistan’s Lawless Frontier, New York: Viking, 2011; Syed Saleem
Shahzad, Inside Al-Qaeda and the Taliban: Beyond Bin Laden and 9/11, London: Pluto Press,
2011; and Hassan Abbas, “A Profile of the Tehrik-i-Taliban,” CTC Sentinel, Vol. 1, No. 2,
January 2008, pp. 1–4.
122 Countering Others’ Insurgencies

siege and storming of Islamabad’s Red Mosque, where militants had

holed themselves up, provided a major impetus to TTP cohesion and
escalation of operations. The incident triggered a wave of attacks and
suicide bombings beyond FATA against military targets and civilian
population centers throughout the country. Baitullah Mehsud, a mili-
tant commander from the Mehsud tribal region in South Waziristan
Agency (SWA), initially presided as the head of the TTP fighters. After
his death in August 2009, he was succeeded by Hakimullah Mehsud.
Though the TTP is similar to the Afghan Taliban in its ethnic and
ideological origins, its goals and targeting are different, with TTP vio-
lence focused on the GoP.26 Although the TTP is a religious movement
with Islamist principles, given its Pashtun tribal base, some analysts
and officials fear it has nationalist liberation elements and could poten-
tially escalate to a separatist insurgency.27
To whittle down the strength of the TTP, the GoP sought to
distinguish between “good Taliban” and “bad Taliban.” Good Taliban
were those who focused their attention on Afghanistan and ceased to
directly participate in attacks on the GoP. The distinction between
good and bad Taliban was not an intrinsic one but developed over
time after deals struck with certain militant groups proved durable.
The designation of good was applied to the Afghan Taliban Quetta
Shura—a group that had not targeted the GoP and focused on
Afghanistan. It eventually also came to include Mullah Nazir and
Hafiz Gul Bahadur, who opted for cease-fire deals that have appeared
to remain durable since 2007 or 2008, as well as the Haqqani network,
based within Gul Bahadur’s territory.28

26Scott Shane, “Insurgents Share a Name, but Pursue Different Goals,” New York Times,
October 23, 2009.
27 M. Ilyas Khan, “Taleban Spread Wings in Pakistan,” BBC News, March 5, 2007; Per-
vaiz Iqbal Cheema, “Challenges Facing a Counter-Militant Campaign,” National Bureau of
Asian Research, NBR Analysis, Vol. 19, No. 3, August 2008, pp. 21–29.
28 Nazir and Gul Bahadur’s groups were considered GoP allies even though there was evi-
dence they may have provided tacit support to bad Taliban elements. See Anand Gopal,
Mansur Khan Mahsud, and Brian Fishman, “The Taliban in North Waziristan,” in Peter
Bergen and Katherine Tiedemann, eds., Talibanistan: Negotiating the Borders Between Terror,
Politics, and Religion, New York: Oxford University Press, 2013, pp. 128–163; Mansur Khan
 Counterinsurgency in Pakistan 123

The Mehsud tribe came to form the core of the TTP, although it
was allied with a number of militant outfits across FATA, foreign fight-
ers, and even some Punjabi Taliban. Estimates of TTP force strength
have been highly varied due to militant-organization defection, attri-
tion, and the flow of militants between the Afghan and Pakistan the-
aters. However, in 2009, the Pakistani military estimated Mehsud
fighters to number about 10,000, making total estimates of the TTP
at 25,000 seem plausible, with upper-end estimates of 53,000 fight-
ers. Over time, what made the TTP—and specifically the Mehsud
component—so formidable and threatening was its ability to project
power and violence outside of FATA and strike civilian and military
targets in Pakistan’s core urban centers. This was done through suicide
bombings (of which the Mehsud/TTP force was believed to be respon-
sible for about 70 percent) and complex, high-risk attacks, includ-
ing one on the Army’s headquarters.29 This successful conversion of a
“defensive war into an offensive and proactive campaign” beginning in
2007 “represented a major escalation and a serious threat to the coun-
try’s centers of power.”30
Tehrik-e-Nifaz-e-Shariat-e-Mohammadi. The TNSM is a mili-
tant group dedicated to implementing Sharia law in the region of KP
formerly known as the Malakand Division, encompassing Swat, Dir,
Malakand, Chitral, and parts of Buner and Shangla. Unlike the TTP,
the TNSM (led by Mullah Sufi Muhammad) has long-standing roots
in Pakistan dating back more than two decades and stemming from
Pakistan’s major Islamist political party, Jamaat-e-Islami. The TNSM
engaged in major confrontations with the state in 1994 in reaction to a
Supreme Court decision overturning the validity of local tribal courts
and incorporating the region into mainstream Pakistani law. Nominal

Mahsud, “The Taliban in South Waziristan,” in Peter Bergen and Katherine Tiedemann,
eds., Talibanistan: Negotiating the Borders Between Terror, Politics, and Religion, New York:
Oxford University Press, 2013, pp. 164–201.
29 Mahsud, 2013, pp. 179–180.
30 The first quotation is from Gul, 2011, p. 40; the second is from Jerry Meyerle, Unconven-
tional Warfare and Counterinsurgency in Pakistan: A Brief History, Alexandria, Va.: Center for
Naval Analyses, November 2012, p. 30.
124 Countering Others’ Insurgencies

government concessions on sharia law were never fully implemented,

and the region witnessed periodic political confrontations for the rest
of the decade.
After Muhammad was jailed for leading a large militia to fight
the United States in Afghanistan in 2001, his son-in-law Maulana
Fazlullah assumed leadership of the TNSM. Fazlullah managed to
revive it in 2006 by running an illegal radio station promoting Islamist
values and denouncing modernity, Pakistani President Musharraf, and
the United States (earning him the nickname “Mullah Radio”). Under
the leadership of a sympathetic provincial government run by a coali-
tion of religious parties, the region of KP and the Swat Valley in par-
ticular was subject to increasing radicalization and became a haven for
militant organizations as well as violent and overzealous enforcement
of sharia, the intimidation of locals, and disruption of development
measures like vaccinations and girls’ education. Fazlullah was given
free rein to continue his radio propaganda in exchange for halting the
TNSM’s intimidation campaign. After the July 2007 storming of the
Red Mosque, however, the TNSM declared jihad against the GoP and
began to attack security forces and government buildings, effectively
taking control of the region by October 2007.
The TNSM, initially estimated at between 500 and 3,000 armed
forces in 2007, grew to between 6,000 and 10,000 by 2009 by drawing
in foreign fighters as well as generating local recruits. The group turned
out to be surprisingly capable, motivated, and adept at guerrilla warfare,
retreating to the mountains surrounding the Swat Valley during major
Pakistan Army sweeps. Though religious in its aspirations and driven
in response to GoP military actions, the TNSM managed to mobilize
new recruits through a combination of appeals to local grievances and
opportunities for payment, power, and score-settling, as well as some
amount of forcible recruitment. While this mobilization and political
violence coincided with the rise of Pakistani Taliban violence in FATA
and shared similar motivations, the TNSM’s trajectory, agenda, and
targets of violence remained distinct to the Malakand region.
 Counterinsurgency in Pakistan 125

Baloch nationalist groups.31 Ever since Pakistan’s independence

from the British, the underdeveloped and indirectly governed province
of Balochistan (specifically, the eastern regions encompassing parts of
Kalat, Khuzdar, and Jhalwan) has witnessed a number of uprisings,
including one large one from 1973 to 1977. These rebellions generally
involved one or more militias of the three most powerful Baloch tribal
groups—the Marri, Mengal, and Bugti—though they were plagued
by disunity and vulnerable to GoP divide-and-rule tactics. Rebellion
against the state stemmed from an initial belief that Balochistan was
an independent territory and was later due to political exclusion and
marginalization. Despite periodic flare-ups, the conflict did not seri-
ously reemerge until 2004.
This most recent round of Baloch insurgency was a response to per-
ceived encroachment and subjugation by the Pakistani center, includ-
ing the construction of the Gwadar port and new Army cantonments,
underpayment for resources extracted, and administrative restructur-
ing under the Musharraf regime that would undercut the tribal levies
system. Violence was a form of coercive bargaining to restore a degree
of provincial autonomy and increase Balochistan’s resource share, and
it was confined to tribal militias that mostly targeted infrastructure
and some GoP security forces. For its part, the GoP saw this uprising
as a threat to the modernization and development of Balochistan and
sought to crush it.
Initial estimates of the size of Baloch forces in 2005–2006 ranged
from 1,000 to 14,000, though the upper figure likely included all
potential tribal militia members rather than actual hardened fight-
ers. While the Baloch insurgents suffered significant losses in the early
years of the insurgency, after the heavy-handed military response, the

31 Account drawn from Mansoor Akbar Kundi, “Insurgency Factors in Balochistan,” The
Nation, October 20, 2004; Frederic Grare, Balochistan: The State Versus the Nation, Wash-
ington, D.C.: Carnegie Endowment for International Peace, Carnegie Paper, April, 2013;
Robert G. Wirsing, Baloch Nationalism and the Geopolitics of Energy Resources: The Chang-
ing Context of Separatism in Pakistan, Carlisle, Pa.: Strategic Studies Institute, April 2008;
Human Rights Commission of Pakistan, Conflict in Balochistan, Report of Human Rights
Commission of Pakistan Fact-Finding Mission, December 2005–January 2006; and Alok
Bansal, “Factors Leading to Insurgency in Balochistan,” Small Wars & Insurgencies, Vol. 19,
No. 2, June 2008, pp. 182–200.
126 Countering Others’ Insurgencies

number of insurgent fighters is believed to have grown to between

3,000 and 12,000, expanding across social classes. At the same time,
the goals of the insurgency shifted from limited autonomy to separat-
ism, with growing public support for an independent state, and its tar-
gets expanded to include civilians.32
Other militant organizations. Though outside the scope of this
study, the Pakistani militant landscape is replete with organizations
with varying degrees of links to the state that operate for a variety
of purposes, including ethnic nationalism (Muttahida Qaumi Move-
ment, Sindu Desh Liberation Front), sectarianism (Sipah-e-Sahaba,
Lashkar-e-Jhangvi), and low-intensity proxy war (Lashkar-e-Taiba,
Jaish-e-Muhammad, and Harkut-ul-Mujahideen). This militant milieu
has unpredictable alignments and effects, with some groups threaten-
ing the state, others serving its interests, and some straddling the two
roles. Nevertheless, their presence and overlap with some insurgent
groups described earlier complicate any counterinsurgency strategy.33

Conflict Narrative
Since 2001, the Pakistani government has focused on fighting three
insurgencies: the TTP in FATA, the TNSM in the Swat Valley/
Malakand, and nationalist/separatist militants in Balochistan. Draw-
ing on numerous detailed accounts, we next provide brief accounts of
the conflicts and the campaigns by the GoP and its security forces to
counter them.

32 Estimates based on previous sources as well as “Pashto-Language TV Station [Khyber

TV] Interviews Pakistan’s President Musharraf,” BBC Monitoring South Asia, May 21, 2006;
Saeed Shah, “In Remote Baluchistan, Pakistan Fights Shadowy War,” McClatchy News,
March 29, 2012; Zahid Hussain, “It’s War Now,” Newsweek, January 15, 2006; “2006:
Who’s Who in Balochistan,” Dawn, May 28, 2011.
33 For more, see C. Christine Fair and Seth G. Jones, “Pakistan’s War Within,” Survival,
Vol. 51, No. 6, December 2009–January 2010, pp. 5–32; Ijaz Khan, “Challenges Facing
Development in Pakistan’s FATA,” NBR Analysis, Vol. 19, No. 3, August 2008, p. 20;
Saeed Shah, “Why Pakistan Supports the Haqqani Network: Fear for Its Own Security,”
McClatchy News, September 28, 2011; and International Crisis Group, Pakistan: The Militant
Jihadi Challenge, Asia Report No. 164, Islamabad/Brussels, March 13, 2009.
 Counterinsurgency in Pakistan 127

FATA
After the fallout of 9/11 and U.S. military operations in Afghanistan,
militants based in FATA who were formerly part of, allied with, or
sympathetic to the Taliban began to assert their dominance over local
power structures and administrative institutions.34 They also resisted
GoP counterterrorism efforts coordinated with U.S. and NATO opera-
tions on the Afghan side of the border to round up these actors, and
they harbored foreign fighters and al Qaeda operatives. In response, the
GoP launched a number of large- and small-scale operations broadly
grouped under the name Operation al Mizan.35 The scope of the cam-
paign from 2002 to 2007 was narrow, with the threat viewed as one
“to be contained, not defeated.”36 Military activity was “sporadic,”
“not sustained over time,” and “incomplete, inconclusive, and at times,
appeared insincere.”37 Following clearing operations, the Army was
“reluctant to maintain a presence” or exert “any effort to control the
area fully.”38

34 Accounts of this case are drawn from Fair and Jones, 2009–2010; Gul, 2011; Nawaz,
2009; Shahzad, 2011; Brian Cloughley, War, Coups, and Terror: Pakistan’s Army in Years
of Turmoil, New York: Skyhorse, 2008; Brian Cloughley, “Insurrection, Terrorism and the
Pakistan Army,” in Usama Butt and N. Elahi, eds., Pakistan’s Quagmire: Security, Strat-
egy, and the Future of the Islamic-Nuclear Nation, London: Continuum International, 2010,
pp. 93–122; Ahmed Rashid, Descent into Chaos, London: Penguin Books, 2008; Carey
Schofield, Inside the Pakistan Army, London: Biteback Publishing, 2011; Sameer Lalwani,
“Pakistan’s Counterinsurgency Strategy,” in Peter Bergen and Katherine Tiedemann, eds.,
Talibanistan: Negotiating the Borders Between Terror, Politics, and Religion, New York: Oxford
University Press, 2013, pp. 202–228; Daniel Markey, Securing Pakistan’s Tribal Belt, Wash-
ington, D.C.: Council on Foreign Relations, Council Special Report No. 36, August 2008;
International Crisis Group, Pakistan’s Tribal Areas: Appeasing the Militants, Islamabad/Brus-
sels, Asia Report No. 125, December 11, 2006b.
35 Schofield, 2011, p. 133; Fair and Jones, 2009–2010, pp. 46–56.
36 Marvin G. Weinbaum, “Hard Choices in Countering Insurgency and Terrorism Along
Pakistan’s North-West Frontier,” Journal of International Affairs, Vol. 63, No. 1, Fall/Winter
2009, p. 75. The term containment is echoed in Meyerle, 2012, p. 29; Ijaz Khan, 2008, p. 14;
Cloughley, 2008, p. 189.
37 Fair and Jones, 2009–2010, p. 76.
38 Haider Ali Hussein Mullick, “Lions and Jackals: Pakistan’s Emerging Counterinsurgency
Strategy,” Foreign Affairs (online), July 15, 2009.
128 Countering Others’ Insurgencies

The government deployed as many as 80,000 forces during this

period, but the number does not accurately reveal the level of effort.
First, most major operations during this campaign seemed to involve
only 10,000 or fewer troops. The real operations were led by Special
Services Group (SSG) forces, while most of the troops remained in
their bases rather than patrolling or even manning checkpoints.39
One analyst quotes a senior Pakistani officer’s description of this as
“a policy of ‘sitzkrieg’—meaning, sitting in camps without any aggres-
sive actions.”40 Most of the Army forces, including two divisions from
Peshawar’s XI Corps, were not deployed but already garrisoned in the
area,41 and 60 percent of the forces were paramilitary FC that were ill-
equipped, poorly trained, and at times unwilling to fight the nascent
Pakistani Taliban.42
When the Pakistani government did summon greater resolve and
effort, it still relied on conventional kinetic tactics with high levels of
collateral damage. About two dozen major operations were search-and-
destroy missions.43 The GoP ended up using a heavier hand than it may
have intended because the unwillingness or incapability of FC forces to
directly confront militants led to an overreliance on airpower and artil-
lery to substitute for the forces or bail them out.44 Meanwhile, despite
plans to integrate security, development, and governance, the strategy
was in fact quite narrow, without any systematic efforts to hold and

39 Interviews with U.S. and Pakistan military officials, June–July 2013.

40Shuja Nawaz, Learning by Doing: The Pakistan Army’s Experience with Counterinsurgency,
Washington, D.C.: Atlantic Council, February 2011, p. 10.
41 Fair and Jones, 2009–2010, pp. 36, 42.
42 On estimating FC proportion, see the Appendix. On FC capabilities, see Lalwani, 2009,
pp. 37–38; Fair and Jones, 2009–2010, pp. 38–39; Hassan Abbas, “Transforming Pakistan’s
Frontier Corps,” Terrorism Monitor, Vol. 5, No. 6, March 30, 2007; Tariq Mahmud Ashraf,
“The Pakistan Frontier Corps in the War on Terrorism—Part I,” Terrorism Monitor, Vol.
6, No. 15, July 25, 2008; and Arthur Keller, “Propaganda and Peace Deals: The Taliban’s
Information War in Pakistan,” CTC Sentinel, Vol. 1, No. 8, July 2008.
43Quotation from Fair and Jones, 2009–2010, pp. 168, 172; Gul, 2011, p. 26; Markey,
2008, p. 11; Jamal Hussain, “PAF in the Counterinsurgency Role in the Tribal Areas of
Pakistan, Defence Journal, November 2012.
44Jones and Fair, 2009–2010, pp. 49–50; Keller, 2008; interview with Pakistan analyst,
May 2013.
 Counterinsurgency in Pakistan 129

build in areas following the military sweeps. Due to the economization

of resources, “FATA was left to fester.”45
The GoP’s conventional “enemy-centric” approach was observed
and described by one counterinsurgency expert through a list of tacti-
cal shortfalls: (1) a near-exclusive focus on enemy targeting and high-
value targets; (2) an overdependence on large-scale, multi-unit forces
rather than small, lithe patrol units; (3) the frequent defensive deploy-
ment of forces to static garrisons, checkpoints, or asset tasks; (4) the
absence of maneuver room and a shortage of resources for flexible
responses to contingencies; (5) an overreliance on kinetic direct-action
operations; and (6) the discounting of local knowledge and assets in
planning operations.46
Facing stiffer resistance than expected, lacking sufficiently capa-
ble deployed forces, and hamstrung by public opposition to these oper-
ations, the GoP was fought to a stalemate with much higher losses
than it expected. As a result, the Army sought to sign ceasefire agree-
ments with tribal groups and armed militants in order to gain a tactical
pause, stem its losses, withdraw, and regroup. Major deals were signed
in 2004, 2005, 2006, and 2008, and a number of informal deals were
believed to exist as well. The deals were intended to “[co-opt] influ-
ential tribal elders” on the condition that they “generally commit to
provide stability in the area.” GoP officials believed they could manage
militants, use them to stem foreign fighters believed to be the source
of instability, and return the region to the pre-2001 system of indirect
governance run by tribal elders and political agents, when FATA was
believed to be “very peaceful.”47 While these deals yielded some tacti-
cal gains and turned some Taliban forces against foreign militants such

45 Quotation from Ijaz Khan, 2008, p. 14. Also see Markey, 2008, p. 12; Schofield, 2011,
p. 136; Joshua T. White and Shuja Ali Malik, Governance Reforms in Pakistan’s Tribal Areas:
The Long Road to Nowhere? Washington, D.C: United States Institute of Peace, Peace Brief
No. 135, October 15, 2012, p. 4.
46 David J. Kilcullen, Terrain, Tribes, and Terrorists: Pakistan, 2006–2008, Washington,
D.C.: Brookings Institution, Counterinsurgency and Pakistan Paper Series, No. 3, Septem-
ber 10, 2009, pp. 10–12.
47 Kamran Rasool, “Pakistan’s Perspective on the ‘War on Terror,’” Military Technology,
Vol. 32, No. 11, November 2008, p. 17; Schofield, 2011, p. 136.
130 Countering Others’ Insurgencies

as the Uzbek fighters in the area, militancy continued unabated. The

Army nonetheless stuck with this approach.48
Following the July 2007 Red Mosque incident that triggered a
wave of suicide bombings throughout urban Pakistan, the GoP, the
military, and the country as a whole began to reconsider their approach
to counterinsurgency. In 2008, new operations were launched across
FATA that were sustained with a higher operational tempo and were
even more unrestrained in the use of force. Led by higher-quality Army
units, they began to occupy (though not really build) territory in some
places. Total manpower in FATA escalated to 90,000 troops in early
2008, 120,000 troops by early 2009, and more than 140,000 troops
through 2010.49
During this time, the Army as an institution also began to adapt
in a number of ways, reducing security-force losses (see Table 5.1).
Army Chief of Staff General Kayani created a special inquiry commis-
sion to investigate the Army’s counterinsurgency failures, triggering a
series of reforms to boost training, morale, materiel, information col-

Table 5.1
FATA and Malakand Conflict Deaths per Engagement, 2002–2012

Loss per Engagement/

Operation

Engage- Security Loss-Exchange

Group ments Forces Militants Civilians Fraction
FATA, 2002–2006a 125 5.6 7.5 ?.0 0.57
FATA, 2007–2008 219 0.057 11.4 1.55 0.96
FATA, 2009–2012 1,246 0.20 8.50 0.38 0.98
Malakand, 2007–2008 163 0.56 6.70 2.30 0.94
Malakand, 2009–2012 391 0.36 7.50 0.68 0.96
SOURCE: Data from Pakistan Institute for Peace Studies.
aVery rough estimates derived from various accounts and data sources.

48 Gul, 2011, pp. 55–59; Ejaz Haider, “Intelligent Intelligence,” Daily Times, July 26, 2007;
Ismail Khan, “Militant Leader Threatens to Attack Forces, Dawn, July 30, 2005; Daud
Khattak, “Reviewing Pakistan’s Peace Deals with the Taliban,” CTC Sentinel, September
2012; interview with former U.S. government official, June 2013.
49 Rasool, 2008; Shuja Nawaz, 2009a; Shuja Nawaz, 2011.
 Counterinsurgency in Pakistan 131

lection, and synchronization and intelligence-sharing between units.

Counterinsurgency began to be incorporated into the regular military’s
curriculums at all levels, and U.S. and British SOF were brought in to
help upgrade and train the FC, while professionalism and pay were
increased.50
By 2009, the GoP also began to focus on destroying rather than
weakening the base of the Pakistani Taliban. Prior to 2007, the insur-
gency consisted of a disparate set of militant groups and commanders
with unclear organization or goals. As the Mehsud camp increasingly
became the face of the TTP insurgency in FATA, organizing under a
single banner with more clearly stated goals and demands and claiming
credit for much of the suicide bombing throughout the country, the
GoP began to more directly confront the group and its base. Govern-
ment operations were initially launched primarily in SWA, but they
were later pursued in other parts of FATA. In October 2009, the GoP
launched a major offensive in a part of SWA to destroy the TTP and
the Mehsud’s base of combat power, lines of communication, and com-
mand and control. The operation involved as many as 30,000 troops,
mostly Army, as well as massive firepower and aerial bombing (with at
least 150 targets hit in the first week). Victory was declared in Decem-
ber, and most of the forces remained to occupy SWA. Despite the
greater persistence and intensity of military operations, many observers
continued to doubt the extent of government efforts to introduce the
governmental and economic programs characteristic of the build phase
of classic counterinsurgency doctrine.51
Troops backed by airpower continued to conduct operations in
Orakzai, Mohmand, Kurram, and Khyber to pursue TTP remnants
over the next three years, although they seemed to resume a pattern

50 Shuja Nawaz, 2011; Haider Ali Hussein Mullick, “Recalibrating U.S.-Pakistan Relations,”
The Washington Quarterly, Vol. 35, No. 3, Summer 2012, pp. 93–107; Haider Ali Hussein
Mullick, “Holding Pakistan: The Second Phase of Pakistan’s Counterinsurgency Opera-
tions,” Foreign Affairs (online), March 24, 2010; Markey, 2008.
51 Shuja Nawaz, 2011; Frederick W. Kagan, Reza Jan, and Charlie Szrom, “The War in
Waziristan: Analysis of Operation Rah-e-Nijat: Phase 1 Analysis,” Critical Threats Project,
American Enterprise Institute, Washington, D.C., November 18, 2009; Reza Jan, “Trickling
Home to South Waziristan,” Critical Threats Project, American EnterpriseInstitute, Wash-
ington, D.C., December 13, 2010c; Jamal Hussain, 2012.
132 Countering Others’ Insurgencies

commonly derided as “whack-a-mole”—that is, the killing of individ-

ual insurgents without any indication of strategic effect. Most activi-
ties throughout FATA continued to be kinetic, narrowly focused on
degrading insurgent combat power, heavily reliant on airpower, and
with little attempt to hold areas or address internally displaced persons
(IDPs). This unbalanced effort may explain why the military, outside
of SWA, is continuing to fight the same fights in FATA, most recently
in the Tirah Valley in Khyber Agency.52

Malakand Region and the Swat Valley53

Between 2006 and 2007, a revived TNSM became increasingly con-
frontational. After a deal to limit TNSM activities collapsed follow-
ing the Red Mosque incident, the TNSM declared open war on the
Pakistani state, attacked government buildings and police, and seized
control of the Swat Valley and parts of Malakand.
In November 2007, the Army launched its first campaign,
Operation Rah-e-Haq (Just Path), to take back control of Swat,
deploying between 12,000 and 15,000 troops. Government forces
nonetheless“appeared hesitant to take them on.”54 While the GoP con-

52 Reza Jan, “Daily Tracker: Pakistani Military Operations in Orakzai,” Critical Threats
Project, American Enterprise Institute, Washington, D.C., May 13, 2010a; Reza Jan and Sam
Worby, “Limited Goals, Limited Gains: The Pakistan Army’s Operation in Kurram, Critical
Threats Project, American Enterprise Institute, web site, September 6, 2011; Tayyab Ali Shah,
“Pakistan’s Challenges in Orakzai Agency,” CTC Sentinel, Vol. 3, No. 7, July 2010; Daud
Khan Khattak, “‘Clearing’ Kurram,” Foreign Policy’s AfPak Channel, August 25, 2011;
“PAF Conducted 5,500 Bombing Runs in FATA Since 2008,” Express Tribune, November
14, 2011.
53 Accounts drawn from Schofield, 2011, p. 178; Gul, 2011, pp. 112–119; Lalwani, 2009:
15–17; Christine Fair, “Pakistan Loses Swat to Local Taliban,” Terrorism Focus, Vol. 4,
No. 37, November 13, 2007; Afzal Khan, “Revolt in Pakistan’s NWFP: A Profile of Maulana
Fazlullah of Swat,” Terrorism Focus, Vol. 4, No. 38, November 20, 2007; International Crisis
Group, Pakistan Countering Militancy in PATA, Asia Report No. 242, Islamabad/Brussels,
January 15, 2013; Daud Khan Khattak, The Battle for Pakistan: Militancy and Conflict in
the Swat Valley, Washington, D.C., New America Foundation, Counterterrorism Strategy
Initiative Policy Paper, April 2010; Ahmed Rashid, “Pakistan’s Continued Failure to Adopt
a Counterinsurgency Strategy,” CTC Sentinel, Vol. 2, No. 3, March 2009; Ahmed Rashid,
Pakistan on the Brink: The Future of America, Pakistan, and Afghanistan, New York: Viking,
2012; Reza Jan, “Paradise Regained: Swat One Year On,” Critical Threats Project, American
Enterprise Institute, Washington, D.C., May 25, 2010b.
54 Gul, 2011, p. 118.
 Counterinsurgency in Pakistan 133

ducted a conventional sweep of the Swat Valley through mid-Decem-

ber and forcibly shut down jihadist radio stations and transmissions,55
most TNSM forces simply retreated to the mountains, where the Army
tried to shell them with artillery. These conventional operations and
reliance on area weapons produced a moderate amount of collateral
damage and won little public support. The GoP wound down this first
phase of operations in early 2008 prior to the general elections and
soon brokered a second peace deal through the newly elected provin-
cial government.
When the deal collapsed, a second round of major operations was
required, and the military was sent back in late July 2008, this time
with a slightly larger force. The military repeated the same “blow up,
patch up, blow up” cycle of clear-and-occupy operations relying on
conventional tactics and artillery. Fearing excessive collateral damage,
however, it failed to clear militants out of the towns and abandoned
their checkpoints to retreat to the ridges of the valley. This excessive
caution and premature declarations of victory resulted in a loss of con-
trol of the valley by late 2008. As border tensions with India mounted
in the wake of the Mumbai terrorist attack in November 2008, the
Pakistani military nonetheless hesitated to commit larger numbers of
troops for more-decisive action. Peace negotiations began again in Feb-
ruary 2009, and another deal was signed in April.56
Things took a decisive turn after the collapse of this third peace
deal. An emboldened TNSM quickly broke the deal as it publicly
derided the Pakistani Constitution and stormed into the neighbor-
ing province of Buner. The Pakistani government, the Army, and the
broader public finally realized that they would have to dispense with
previous economization efforts. In April 2009, the GoP launched a
new operation titled Rah-e-Rast (Righteous Path) with a much larger

55 “Pak Troops Seize Swat Peak, Shut ‘Mullah Radio,’” Times of India, November 28, 2007;
Riaz Khan, “Pakistan Shuts Down 3 Radio Stations,” Associated Press, March 21, 2008;
Richard A. Oppel, Jr., and Pir Zubair Shah, “In Pakistan, Radio Amplifies Terror of Tali-
ban,” New York Times, January 24, 2009.
56Mullick, 2009; Schofield, 2011, pp. 178–181; Syed Saleem Shahzad, 2011, p. 171;
Rahimullah Yusufzai, “Fallen City,” Newsline Magazine, January 3, 2009; interview with
Pakistan military analyst, June 2013; Khattak, 2010, pp. 9–10; Rashid, 2012, p. 141.
134 Countering Others’ Insurgencies

force of 52,000 high-quality troops to clear out an estimated 5,000

to 10,000 militants and adopted a new “presence-oriented strategy.”57
Although this operation regained full control of the region, it proved
costly for the military, with 985 casualties endured over two years.
Following these “clearing” operations, the Army converted to a much
more defensive and sustained posture of holding the territory to build
it up and transfer it back to local control.58 These actions defied predic-
tions of skeptical experts and observers and in two years substantially
reduced insurgent incidents in the region (see Figure 5.1).59
One of the more unconventional and unprecedented methods
it used to reduce collateral damage was to rapidly evacuate 2 million
to 2.5 million people from the area, providing temporary camps and
allowances for about 25 percent of them so that they could more ably
prosecute their operations and dislodge dug-in militants using over-
whelming firepower. Though widely criticized early on as another des-
perate and poorly thought-out plan, this option had clearly been the
subject of extensive deliberations, and it ultimately proved quite suc-
cessful. About 90 percent of the residents of Swat were able to move
back to their homes within a few months. One retired officer com-
mented that many people in Swat were so desperate for GoP inter-
vention, they “offered to migrate and vacate their homes . . . [they]
have accepted their dislocation as a price for getting rid of militancy.”60
This controlled evacuation, combined with more-discriminate use of
force, rapid repatriation efforts, and compensation and reconstruction
of homes damaged from operations, indicates that this Swat campaign

57Mullick, 2009; Khattak, 2010; Mullick, 2010; Shuja Nawaz, “Pakistan’s Summer of
Chaos,” ForeignPolicy, June 17, 2009b; Cloughley, 2010, p. 108; Zahid Hussain and Matthew
Rosenberg, “Pakistani Peace Deal Gives New Clout to Taliban Rebels,” Wall Street Journal,
April 14, 2009.
58Inter-Services Public Relations (Pakistan), “COAS Keynote Address— National Seminar
Deradicalization at Mingora Swat,” ISPR No PR157/2011, July 6, 2011.
59 James Blitz, James Lamont, and Farhan Bokhari, “Swat Outlook ‘Pretty Bleak’ for Paki-
stan,” Financial Times, May 13, 2009; “Pakistan and the Taliban: A Real Offensive or a
Phony War?” The Economist, April 30, 2009.
60 Jeremy R. Hammond, “The Situation in Swat: An Interview with Shahid R. Siddiqi,”
Foreign Policy Journal, July 9, 2009.
 Counterinsurgency in Pakistan 135

Figure 5.1
Insurgent Incidents Over Time in Pakistani Regions, 2007–2013

800

700
Malakand
600 Other FATA
incidents
incidents
500
Number

400

300
NWA SWA
200 incidents incidents
100
0
2007 2008 2009 2010 2011 2012 2013

SOURCE: Data from Pakistan Institute for Peace studies.

NOTE: Figures for 2013 through June and estimated/doubled. NWA = North
Waziristan Agency.
RAND RR513-5.1

was far better managed than previous military operations.61 One poll
in FATA found that even among the displaced, 52 percent supported
military operations.62
Following the repatriation of the population, the Army adapted
its approach to emphasize population control and protection. Army
forces maintained a sustained presence for years to prevent regrouping
of TNSM cadres, coordinated with locals and village councils, devel-
oped and utilized informant networks to monitor and prevent militant
reinfiltration, rebuilt a 3,700-strong professional police force, armed
and coordinated about 30,000 lashkars in defense of the area, facili-
tated the efforts of reconstruction and development teams, manned

61 Cheema, 2008, p. 27, had proposed this option earlier; Mullick, 2010; Danielle
Kurtzleben, “Pakistan: U.S. to Aid Civilians Fleeing Embattled Swat Valley,” Inter Press Ser-
vice, May 19, 2009; Jan, 2010c; Tara McKelvey, “A Return to Hell in Swat,” Foreign Policy,
March 2, 2011; Omar Warraich, “Refugees Head Home After Army Scatters the Taliban,”
The Independent, August 3, 2009..
62 Naveed Ahmad Shinwari, Understanding FATA: Attitudes Towards Governance, Religion
and Society in Pakistan’s Federally Administered Tribal Areas, Vol. IV, Islamabad: Community
Appraisal and Motivation Program, 2010, p. 57.
136 Countering Others’ Insurgencies

checkpoints and patrolled throughout the region, and even developed

their own low-grade version of monitoring and regulating the popula-
tion through colored lanyards and identification cards that had to be
worn at all times.63 Despite what might have been seen as an intrusion
in other areas, operations worked in Swat because most of the inhabit-
ants wanted the Army to remain in place to provide security and “felt
they were equally involved in the operations.”64
The Army also engaged in an unprecedented set of non-kinetic
tactics during this holding phase, including information operations,
reconstruction, and beginning to address some grievances over local
justice systems. The GoP engaged in its own rebuilding of infrastruc-
ture such as roads, bridges, schools, and vocational centers, as well as in
providing water, electricity, and sanitation services. All told, the GoP
may have spent as much as much as 2 percent of its budget on rehabili-
tating IDPs and another 0.05 percent on reconstruction projects. The
Army’s sustained presence long after offensive operations concluded
(see Figure 5.2) and checkpoints also provided a safe environment for
the revitalization of the cosmetic, agriculture, and tourism industries
and the reestablishment of local marketplaces; the rebuilding of some
physical and social infrastructure by NGOs; and the rebuilding and
reopening of girls’ schools. One NGO leader described the continued
Army presence as “a blessing.”65
In order to transfer security gains to local control, the GoP rebuilt
the police force, almost doubling its size in Swat and KP and improving

63 Jan, 2010b; Mullick, 2010; Mullick, 2009; Inter-Services Public Relations (Pakistan),
2011; “Pakistan Army Strategy in Question After Attacks,” Dawn.com, October 22, 2012;
interview with Pakistan analyst, 2012.
64 McKelvey, 2011; interviews with Pakistan analysts confirm this, 2012, 2013.
65 Mullick, 2012; McKelvey, 2011; Jan, 2011; “Pakistani Authorities Close Three Radio Sta-
tions for Pro-Taliban Broadcasts,” Associated Press, November 1, 2009; Ajaz Maher, “Paki-
stan’s Army Steps Up Radio Wars,” BBC News, August 14, 2012; “Pakistan Army Strategy
in Question After Attacks,” October 22, 2012; Rina Saeed Khan, “Swat Valley NGO Finds
a Solution to the Assault on Education,” The Guardian, May 10, 2011; Delwar Jan, “Mal-
akand Division on Road to Civilian Control,” The News International, August 8, 2012, p. 98;
Ahmad Hassan, “Operation in Malakand, Waziristan Is No Solution,” Dawn, November
8, 2009; Emmanuel Duparcq, “Swat Rebuilds Year After Pakistan Floods,” Agence France
Presse, July 24, 2011.
 Counterinsurgency in Pakistan 137

Figure 5.2
Manpower and Operational Tempo in the Swat Valley/Malakand Region,
2007–2012

60 60
Manpower
Defensive engagements
50 50
Offensive operations

Manpower (thousands)
Operations/month

40 40

30 30

20 20

10 10

0 0
Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul Jan Jul
2007 2007 2008 2008 2009 2009 2010 2010 2011 2011 2012 2012
SOURCES: Data from Pakistan Institute for Peace studies data; numerous sources
(for manpower).
RAND RR513-5.2

pay, equipment, training, and recruitment while supporting local

levies and peace jirga militias. Additionally, the democratically elected
provincial government began to implement much-promised judicial
reforms to improve the speed and equity of the justice system, which
had been the rallying cry for the TNSM. It reduced the backlog of
cases by about 87 percent by increasing the number of judges, estab-
lishing new circuit and mobile courts for easier access, imposing dura-
tion limits on cases, and embracing some alternative appellate courts
run on the principles of Sharia and dispute-settlement mechanisms like
reconciliation committees.66

66 Reza Sayah, “Can the Return of Justice Halt the Taliban?” CNN, November 14, 2009;
“Judicial Reform in Swat,” Dawn, March 25, 2010; Rania Abouzeid, “Taliban Gone, Paki-
stan Area Still Wants Islamic Justice,” Time, April 25, 2010; Fazal Khaliq, “Speedy Justice:
Mobile Courts for Swat Announced,” Express Tribune, November 26, 2011; Fazal Khaliq,
“Nizam-e-Adl Regulation: Top Sharia Court Set Up in Swat,” Express Tribune, January 19,
2011a; Fazal Khaliq, “Terrorists Behind Swat Unrest to Face Justice,’” Express Tribune,
138 Countering Others’ Insurgencies

Finally, the GoP also began the process of rehabilitating and rein-
tegrating several thousand former TNSM members deemed recon-
cilable. “Socio-psychological” deradicalization was conducted in two
Army-run centers, Mishal and Sabaoon, to reeducate these former mil-
itants and teach them basic vocational skills. More than 1,000 of them
have thus far been reintegrated.67
As of the writing of this report, violence has been reduced sub-
stantially, and Swat is considered by many to be a victory. Despite
having strongly criticized Pakistani counterinsurgency efforts previ-
ously, Ahmed Rashid wrote of these operations:

The war in Swat was the first, and so far the only, time the Paki-
stan Army successfully completed a counterinsurgency cam-
paign according to the book: the militants were killed, captured,
or driven out, the area was secured, the displaced population
returned, their homes were rebuilt, and the civic administration
was revived. The army had finally learned the principles of “clear,
hold, build, and transfer” . . . and could carry them out when it
had the will to do so.68

Balochistan
After nearly 30 years without major conflict, Baloch insurgents
reemerged in late 2004.69 The rebellion was fueled by the construc-
tion of the Gwadar port and the establishment of new Army garrisons,

November 27, 2011c; Fazal Khaliq, “Timely Justice: Community Body Gives Quick Solu-
tions for Small Disputes,” Express Tribune, March 13, 2013.
67 ISPR, 2011; Muhammad Amir Rana, “Swat Deradicalization Model: Prospects for
Rehabilitating Militants,” Conflict and Peace Studies, Vol. 4, No. 2, April-June 2011; Dina
Temple-Raston, “Pakistan’s Ambitious Program to Re-educate Militants,” NPR, April 1,
2013; Taha Siddiqui, “Deradicalizing Boys in Pakistan,” Christian Science Monitor, May 25,
2012; Kristen Seymour, “De-radicalisation: Psychologists’ War Against Militants,” Express
Tribune, July 17, 2011; Rohit Gandhi, “Deradicalization: A Tall Order,” Asia Pacific Defense
Forum, July 1, 2012.
68 Rashid, 2012, pp. 143–144.
69 Kundi, 2004; Grare, 2013; Wirsing, 2008; Bansal, 2008; Malik Siraj Akbar, Refined
Dimensions of Baloch Nationalist Movement, Bloomington, Ind.: Xlibris Corporation, 2011;
International Crisis Group, Pakistan: The Worsening Conflict in Balochstan, Asia Report
 Counterinsurgency in Pakistan 139

both believed to be preludes to ushering in major demographic changes

that would further weaken the ethnic Baloch grip on the province, as
well as by perceptions that the Baloch had been undercompensated for
resources extracted from their territory. The insurgents began to attack
infrastructure in the region, particularly around the Sui gas fields and
pipelines in late 2004 and early 2005. President Musharraf, then head
of the GoP military regime, believed the tribal groups had been black-
mailing Pakistan for decades and threatened to strike them so hard
“they won’t know what hit them.”70 The GoP’s strategy combined
efforts to harass, intimidate, and eliminate some nationalist and tribal
leaders, while politically dividing and co-opting others. This consis-
tently harsh response by the GoP has been attributed in part to suspi-
cions that Indian intelligence agencies like the Research and Analysis
Wing have been covertly supporting Baloch insurgents.
In 2005, the GoP deployed tens of thousands of troops supported
by tanks and helicopter gunships. Operations involved indiscriminate
bombing and shelling, with substantial loss of civilian life. This first
round of highly firepower-intensive military operations culminated in
the assassination of the Bugti and Marri tribal leaders (in late 2006 and
2007, respectively) and the arrest of the Mengal leader.71
Though casualty estimates are hard to come by, the military
appears to have relied heavily on the indiscriminate application of
firepower between 2005 and 2007. The use of indiscriminate shell-
ing caused as many as 84,000 IDPs to flee the area, including at least
85 percent of the town of Dera Bugti, without any government coor-
dination, relief, or repatriation efforts. Unlike the efforts conducted

No. 119, Islamabad/Brussels, September 14, 2006a; Center for Research and Security Stud-
ies, Balochistan’s Maze of Violence, Islamabad, Pakistan, 2011.
70 Naveed Ahmad, “Trouble in Pakistan’s Energy-Rich Balochistan,” Zurich, Switzerland:
International Relations and Security Network, January 30, 2006; Khurram Shahzad, “What
Ails Balochistan?” The Nation, February 14, 2005.
71 Force numbers range from 4,500 to 70,000. See Bansal, 2008; Human Rights
Commission of Pakistan, 2005–2006; Akbar, 2011, p. 54; Tim McGirk, “Code of the
Frontier,” Time, January 30, 2005.
140 Countering Others’ Insurgencies

during the Swat campaign, in fact, the GoP is believed to have taken
steps to block NGO relief aid to Baloch IDPs.72
By 2007, some GoP officials believed they had crushed the
insurgency. These actions, however, appear to have further inflamed
it and increased insurgent intensity and activity (see Figure 5.3).
The GoP acknowledged in 2007 that insurgent activity and attacks
had increased, likely attributable to a martyrdom effect.73 Analysts
assessed that the majority of Baloch activists previously in support of
autonomy had been radicalized into the nationalist separatist camp.74

Figure 5.3
The Balochistan Insurgency, 2004–2013

800 1.2
Lethality
700
1.0
600 Attacks
0.8
500

Lethality
Attacks

400 0.6

300 Fatalities 0.4

200
0.2
100

0 0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

SOURCES: Data from Pakistan Institute for Peace studies and Worldwide Incidents
Tracking System.
NOTE: Figures for 2013 through June and estimated/doubled.
RAND RR513-5.3

72 Quotation from Wirsing, 2008, pp. 29–30; Human Rights Commission of Pakistan,
2005–2006, documents 22 killed in Jabbar Pakal, another 43 in Dera Bugti, and 41 from
the Jamhoori Watan Party, as well as 14 summary executions and 16 people missing, totaling
136; International Crisis Group, Pakistan: The Forgotten Conflict in Balochistan, Asia Briefing
No. 69, Islamabad/Brussels, October 22, 2007, p. 6.
73 Wirsing, 2008, pp. 39–40; International Crisis Group, 2006a.
74 Grare, 2013; Akbar, 2011, pp. 54–55.
 Counterinsurgency in Pakistan 141

The insurgency spread across the province, encompassing areas

that had previously remained at peace, including southwest Balochistan
and urban centers such as Quetta. It also spread beyond its rural and
tribal roots, gaining the support of other sectors of society, including
the middle class and student organizations. Ultimately, the uprising
took a more radical turn, spawning organizations that were willing to
target civilians and opposed to compromise with the government.75
Following the end of the military regime, the departure of
Musharraf, and the emergence of a democratic government in early
2008, the GoP reduced the military footprint by about 7,000 soldiers as
a confidence-building measure. Repressive operations nevertheless con-
tinued in a different guise, described as a “dirty war.” GoP force levels
rose as high as 50,000, but these were predominantly poorly trained
paramilitary and FC, as most Army troops appear to have been with-
drawn. The GoP forces were backed by about 19,000 police assigned
to urban “A areas” and about 13,500 levies assigned to the “B areas.”76
This dirty-war strategy, privately acknowledged by some in the
Pakistani security establishment and reported to be waged by irregu-
lar assets and intelligence agents, is believed to have involved arbitrary
arrests, abductions, enforced disappearances, and a “kill-and-dump”
policy, with mutilated bodies discovered on the roadside. Details and
actual figures on the strategy are difficult to ascertain, but by 2008,
5,000 Baloch had been arrested and the government officially acknowl-
edged that at least 1,100 were missing, many suspected to have been
the victims of extrajudicial killings.77 Additionally, within a 16-month
span between 2011 and 2012, 300 corpses were recovered.78 The Baloch

75 Grare, 2013; Mavish Ahmad, “Balochistan: Middle Class Rebellion,” Dawn, June 5,
2012.
76 M. Ilyas Khan, “Balochistan ‘Troops Numbers Cut,’” BBC News, May 14, 2008; Saeed
Shah, 2012; Siraj Ahmed, “Police Reform in Balochistan,” in Hassan Abbas, ed., Stabilizing
Pakistan Through Police Reform, Independent Commission on Pakistan Police Reform, New
York: Asia Society, July 2012, pp. 113–119.
77 International Crisis Group, 2007, p. 5; Grare, 2013, p. 13.
78 Matthew Green, “Hidden War Embodies Pakistan’s Struggle,” Financial Times, May 25,
2012.
142 Countering Others’ Insurgencies

Student Organization, a nationalist but non-violent political organiza-

tion, had been extensively targeted, with one-third of all kill-and-dump
victims believed to have been its members.79 Outside analysts and the
Pakistan Supreme Court suspect the security and intelligence services
were also making use of death squads to target nationalist political
leaders.80 This policy is believed to have contributed to the breakdown
of social order and to have unleashed a wave of ethnic violence across
the province.81
This form of counterinsurgency “on the cheap” results from a lack
of state reach. Limited penetration of society yields poor information
to target or apprehend insurgents involved in attacks.82 Meanwhile,
weak and corrupt provincial government institutions are ineffective at
channeling allocated resources for public goods and services to substi-
tute for force or coercion.
Despite these repressive measures, sympathy for the movement
has increased substantially since 2006. A Gallup survey in July 2012
found that 37 percent of the Baloch population favored independence
outright—a position not quantified in previous years but thought by
analysts to be much lower—while 67 percent supported greater auton-
omy, suggesting that state policies were increasing rather than depress-
ing nationalist sentiments.83
The GoP has made some attempts at negotiations or accommo-
dations (in 2004, 2008, 2009, 2010, and 2012), the more-frequent
and sincere efforts occurring under civilian rule, but these have been
undermined by continued repressive security measures. Early on, “gen-
uine accommodation of the Baloch minority [did] not seem to com-

79 Declan Walsh, “Pakistan’s Secret Dirty War,” The Guardian, March 29, 2011.
80Sohail Khan, “Disband Agencies ‘Death Squads,’ Stop [Military] Action in Balochistan:
SC,” The News, September 12, 2012.
81 Grare, 2013, pp. 15–19. See “We Only Receive Back the Bodies,” The Economist, April 7,
2012.
82 Interview with Pakistani journalist, Washington, D.C., May 2013.
83Ansar Abbasi, “37 Percent Baloch Favour Independence: UK Survey,” The News,
August 13, 2012.
 Counterinsurgency in Pakistan 143

mand the [military-led government’s] sincere and sustained interest.”84

Later, efforts under the civilian regime failed to halt the dirty war, offer
increased political autonomy, or provide real development rather than
increased provincial resources siphoned off by the center’s handpicked
elites. 85

Analysis of Pakistan’s Counterinsurgency

Even these abbreviated accounts of the counterinsurgency campaigns

between 2001 and 2013 clearly reveal the relationship between Paki-
stan’s general level of political inclusion and state capacity and its selec-
tion of counterinsurgency strategies and practices. For most of this
period, the government relied on heavy-handed but unsustained coun-
terinsurgency operations approximating the containment model dis-
cussed in Chapter Two. A closer look at these operations, however,
reveals important variation over time and between regions. As the
regime became more democratic, it increasingly turned to a classical
counterinsurgency model in the core region of Malakand and the Swat
Valley. Due in part to low state penetration into the hinterlands of
FATA and Balochistan, on the other hand, the regime continued to
rely on other models in these regions.
The approaches adopted by the Islamabad government cannot,
however, be fully explained by the regime’s degree of political inclu-
sivity and state capacity, nor is military superiority a determining
factor—in fact, the regime only selectively applied its tremendous mil-
itary capabilities, depending on local contexts and threat perceptions.
A closer examination of these cases reveals the importance of other,
albeit secondary, factors, including ethnic representation within the
security forces.
The accounts from the previous section suggests that the domi-
nant GoP counterinsurgency strategy was containment—on average,

84 Wirsing, 2008, p. 39.

85 Ahmad, 2012; Grare, 2013; Malik Siraj Akbar, “The Futility of Senate Debate,”
The Baloch Hal, August 5, 2012.
144 Countering Others’ Insurgencies

more repressive and indiscriminate, involving few meaningful political

accommodations to redress grievances, and offering even fewer public
goods to stabilize the region. This approach appears to be consistent
with our analytic framework and quantitative findings.
When making comparisons with other countries, general mea-
sures of political inclusion and state capacity are useful, but they cap-
ture only an average for what is often highly uneven political terrain.
Moving away from this average treatment of Pakistan to its subnational
variation, one finds further evidence for the expected consequences of
these structural variables. As described earlier, some subregions of Pak-
istan were more politically inclusive and had greater state capacity than
others (core regions), and this variation appeared to shape the GoP’s
diverse counterinsurgency approaches, as Table 5.2 shows.
During the period of military rule, the government employed
a containment strategy in the peripheral regions of Balochistan and
FATA—an approach typical of less-inclusive regimes with weak state

Table 5.2
Regime and Campaign Features of Pakistan Counterinsurgency, 2001–2013

Military State Political Discrimi-

Superi- Capa- Inclu- Accom- Public nate
ority city sion modate Goods Violence Strategy
Balochistan, High low Low Low Low Low Containment
2004–2007
Balochistan, Moder- Low Low/ Low/ Low Low Containment
2008–2012 ate medium moder-
ate
Swat, Low High Medium Moder- Low Low/ Containment/
2007–2008 ate moder- informal accom-
ate modation
Swat, High High High High High High Classic counter-
2009–2012 insurgency
FATA, Low Low Low Moder- Low Low Containment
2002–2007 ate
South High Low Low Low High Low Strong state
Waziristan, repression
2008–2012
North High Low Low High Low High Informal
Waziristan, accom-
2008–2012 modation
Rest of FATA, Low Low Low Low Low Low Containment
2008–2012
 Counterinsurgency in Pakistan 145

capacity. After the transition to civilian rule and increasing (albeit

highly imperfect) democratization from 2008 on, the pattern becomes
more complicated. In the region of Malakand, where political inclu-
sion and state capacity were higher than in the periphery, the state was
eventually able to concentrate much higher levels of resources. Con-
sequently, the government responded to the direct threat to the core
regions of the country posed by the growing insurgency in the Swat
Valley with a classic counterinsurgency campaign. In the more periph-
eral regions of FATA and Balochistan, the regime continued to use
counterinsurgency models more in keeping with a low-capacity state.
Interestingly, the state also appeared to rely on more-autocratic
approaches in the periphery despite its gradual transition toward higher
levels of democracy. Specifically, it continued to employ containment
strategies in Balochistan and much of FATA, as predicted in the dis-
cussion in Chapter Two. Only the Waziristans depart from these pre-
dictions and warrant further explanation. In SWA, the government
eventually drew considerable resources from the center and undertook
counterinsurgency operations most similar to the strong-state repres-
sion model because the tremendous violence inflicted from insurgents
of this region on Pakistan’s core was no longer containable. In NWA,
an overstretched government’s recognition of militant political neutral-
ity led to informal accommodation typical of more-inclusive but low-
capacity regimes. This variation is explained in greater detail below,
including a discussion of factors outside of the analytic framework pre-
sented in Chapter Two that account for some of the Pakistan govern-
ment’s observed behavior.

Political Inclusion
The case of Pakistan generally supports predictions of how political
inclusivity and regime type shape counterinsurgency strategy. Given
that Pakistan was a non-democracy for much of this period, we would
expect that on average, it would rely on more-repressive, indiscriminate
military operations with few meaningful measures of political accom-
modation and reconciliation—in other words, strong-state repression
or containment policies, depending on the resources the government
could muster. This prediction generally holds true for the campaigns in
146 Countering Others’ Insurgencies

Balochistan and most of the campaigns in FATA and Swat. The GoP
struggled to make credible commitments, relied on a narrow political
constituency, was less accountable to the broader public, and allowed
little civilian input on national-security matters, generally resulting in
the use of indiscriminate force with high civilian casualties and low
prospects for reconciliation. The deals and cease-fires the military
regime did make in FATA prior to the return of civilian rule in 2008
were merely tactical, shallow (involving a narrow subset of militants),
and neither trusted by insurgents nor durable.
Indiscriminate force continued in Balochistan throughout the
insurgency. After the 2008 elections, however, repressive tactics became
less visible. The irrelevance of the Baloch constituency, “a small, politi-
cally weak, recalcitrant, and untrustworthy ethno-tribal minority,”
allowed the GoP to continue with this approach.86 Even when the GoP
did make attempts at political negotiations or reconciliation, its cred-
ibility was undermined by a shortage of dialogue and quick resorts
to force.87 Eventually, even good-faith efforts routinely failed, because
Baloch opposition leaders do not trust the government. Ethnic Baloch
leaders have little role in central-government coalitions or bureaucra-
cies, and the civilian government cannot credibly commit to greater
Baloch autonomy when it cannot even withdraw federal security forces
or hold them accountable for indiscriminate violence.
Civilian casualties in FATA were not particularly high in the
early stages of the conflict, not because of any commitment to the dis-
criminate use of force by the military government, but rather due to
limited state capacity to support coercion and sustained operations.
When the military did conduct major operations, it relied on large-
unit actions, intensive firepower, and “indiscriminate and excessive
force.”88 A number of missteps may have resulted from not involving
civilians—both politicians and the civil bureaucracy—in decisionmak-
ing processes with the potential to disrupt a military’s organizational

86 Wirsing, 2008, p. 38.

87 Wirsing, 2008, pp. 38–39
88 International Crisis Group, 2006b; also see Weinbaum, 2009, and Kilcullen, 2009.
 Counterinsurgency in Pakistan 147

pathologies, particularly in counterinsurgency.89 As the GoP escalated

operations in FATA, it continued with the indiscriminate approach
of continuous raids, resulting in higher civilian casualties. This pat-
tern even persisted in the core region of Swat for some time. Only in
early 2009 did the civilian government switch tactics to a population-
oriented approach in Swat. Even then, it retained a generally enemy-
centric approach to the peripheral regions of FATA and Balochistan
(see Table 5.3).

Variation in Political Inclusion over Time

Pakistan’s gradual transition toward more-democratic politics follow-
ing the 2008 elections led to significant changes in its counterinsur-
gency campaigns. The most obvious change in strategy took place in
Swat, with the government transitioning from a containment strategy
to a classic counterinsurgency strategy focused on securing the popula-
tion, instituting reforms, and redressing grievances. Scholars and prac-
titioners alike credit the civilian government for enabling this change
in strategy by generating resolve, public support, and consensus, which
bought time, legitimacy for a sustained military campaign, and the
military’s eventual transition from an enemy-centric to a population-
centric approach.90
Though public perceptions of insurgent threat levels and appro-
priate government actions were shifting toward greater resolve,91 a nar-
rower regime coalition might have remained uninterested in the long-
term resolution of the conflict through both force and reforms and
rehabilitation of insurgents. A less transparent and legitimate govern-

89 Barry R. Posen, The Sources of Military Doctrine: France, Britain, and Germany Between
the World Wars, Ithaca, N.Y.: Cornell University Press, 1984; Colin Jackson, “Defeat in Vic-
tory: Organizational Learning Dysfunction in Counterinsurgency,” unpublished disserta-
tion, Cambridge, Mass.: Massachusetts Institute of Technology, June 2008.
90 Istiaq Ahmad, “Promising Partnership? American Af-Pak Strategy and Pakistan’s
Counterterrorism Response,” in Usama Butt and N. Elahi, eds., Pakistan’s Quagmire: Security,
Strategy, and the Future of the Islamic-Nuclear Nation, London: Continuum International,
2010, pp. 66–67. Interview with Pakistan military official, July 2013.
91 Based on survey data released by Pew Global Attitudes, Gallup, International Republican
Institute, Terror Free Tomorrow, Gilani Research, and World Public Opinion.
148 Countering Others’ Insurgencies

Table 5.3
Estimates of Use of Indiscriminate Force in Different Counterinsurgency
Theaters, 2009–2012

Malakand FATA Balochistan

Civilians killed in offensive 252 358 370

operations (2009–2012)
Civilians killed per 1,000 people 0.041 0.089 0.057
Population size in area of 6.1 million 4 million 6.5 million
operation
Offensive clashes/defensive clashes 0.707 1.832 N/A
Airpower
Average sorties/month (during 83.3 131 N/A
major operations)
Bombs dropped 1,700 10,600 N/A
SOURCE: Pakistan Institute for Peace Studies; Zahid Gishkori, “Violence Is on the Rise
in Balochistan: Report,” Express Tribune, January 3, 2013; Jamal Hussain, 2012; “PAF
Conducted 5,500 Bombing Runs in FATA Since 2008,” 2011.

ment might also have been hampered and unable to win public support
for costly actions like the mass evacuation of roughly 2 million people
or the follow-on expenditures to repatriate IDPs, maintain security
forces, and rebuild the region. Moreover, under President Musharraf’s
military government, it may have been difficult to distinguish neces-
sary military operations from power grabs, since the military could not
credibly commit to not overreaching—a limitation that may explain
the popular hostility to military operations prior to 2009.92
The elected KP provincial government, now far more accountable
to the general public, took costly actions to accommodate the insur-
gent support base in May 2008 and April 2009. Even though these
deals were insufficient and further military action was required, a GoP
sensitive to its political constituency not only paid lip service to politi-
cal accommodations on the core issues motivating the insurgency—
economic disenfranchisement and the sluggish justice system—it also
implemented judicial-sector reforms and established greater opportuni-
ties for education and vocational training.

92 Ijaz Gilani, Voice of the People: Public Opinion in Pakistan, 2007–2008, Karachi, Paki-
stan: Oxford University Press, 2010.
 Counterinsurgency in Pakistan 149

Some change in strategy after 2008 was also seen, though to a

much lesser degree, in Balochistan and FATA. Attempts at accom-
modation to redress Baloch grievances increased under civilian rule,
with four major attempts launched in the past few years, two of which
involved concrete steps such as the passage of legislation, implemen-
tation of some political reforms, and redistribution of a greater share
of economic resources. A fifth attempt may soon be launched by the
newly elected Sharif government.93 Even in FATA, most informal deals
with militant commanders in Waziristan (the informal accommoda-
tion model of Chapter Two) that repeatedly collapsed prior to 2008
proved more durable in keeping a relative lid on violence without major
military operations after civilian rule returned, in part because there
are more counterbalancing institutions to constrain the military’s free-
dom of action. And in the Mehsud tribal area still occupied by the mil-
itary, some dim signs of a determination to build the region emerged
in 2013.94 It seems no coincidence that this change corresponds with
recent efforts by the legislature and the judiciary to more actively exer-
cise oversight to hold the military accountable for its conduct of coun-
terinsurgency operations.95

Variation in Political Inclusion Across Regions

In addition to shifts in the level of democracy over time, different levels
of accountability in political institutions and quality of governance at
the provincial level may explain some of the uneven shifts to a classic

93 Mohammad Zafar, “Stabilising Balochistan: Government to Talk to Insurgent Groups

After Eid, Says CM,” Express Tribune, July 30, 2013.
94 Rebecca Santana, “Pakistan Army Tries to Win Over Local Population in War-Torn
Tribal Region,” Associated Press, April 10, 2013; Mehreen Zahra-Malik, “Pakistan Army
Battles Legacy of Mistrust in Taliban Heartland,” Reuters, February 3, 2013; Dion Nissen-
baum, “In Former Taliban Sanctuary, an Eerie Silence Takes Over,” Wall Street Journal, Jan-
uary 26, 2013; Amna Nawaz and Waj S. Khan, “A Rare Glimpse Inside Pakistan’s Ground
Zero for Terrorists,” NBC News, March 4, 2013.
95 Mohammad Zafar, “Balochistan Conundrum: Hearings Spotlight ‘Crumbling’ Khuzdar
Situation,” Express Tribune, October 11, 2012; “Performance of the Parliamentary Commit-
tee on National Security, March 16, 2012–March 15, 2013,” Pakistan Institute of Legislative
Development and Transparency, March 2013, p. 13.
150 Countering Others’ Insurgencies

counterinsurgency model after 2008. The fact that the strategy shifted
to a more discriminate model in the Swat Valley but not in FATA
and Balochistan may have been the result of higher levels of political
enfranchisement and more direct accountability of the government in
KP. Even after the elections in 2008, FATA had no elected representa-
tives and was primarily controlled by the military with little input from
civilians. Moreover, government officials had greater confidence in the
efficacy of investments in public goods and institutional reforms in KP,
where the social and governance infrastructure existed to support and
legitimize them. By contrast, government officials saw little incentive
to invest effort or resources in FATA or Balochistan, where many per-
ceived that “it wouldn’t take.”96 Though improved over the pre-2008
period, political inclusivity in Balochistan is still very low. For the most
part, the now broader central government coalition remains aloof to its
Baloch constituency, and the events in the province remain immune to
transparency by the press, accountability by the judiciary, and civilian
constraints on the security forces.
The Pakistani government’s approach to NWA provides another
departure from its typical policy of containment, albeit of a different
kind. In NWA, short-term cease-fires graduated to informal accommo-
dations as the government came to believe that militant groups in the
region did not directly threaten the Pakistani state or core, at least for
the time being. Given their geographic location on the border, Wazir
tribes were a key political constituency cultivated since the Afghan
jihad of the 1980s, represented well in the military, and potentially
useful as leverage in a future Afghanistan political settlement.97 Larger
military engagements that other counterinsurgency models demanded
might overstretch a military already committed to Swat and SWA,
break this “live-and-let-live” bargain, and invite even greater violence
on the mainland than that suffered between 2007 and 2010. For all
of these reasons, informal accommodation appeared to offer the best
approach.

96 Interview with former U.S. official, June 2013; International Crisis Group, 2006b, pp. 3,
9, 27.
97 Gul, 2011, p. 36; interviews with Pakistan analysts, May–June, 2013.
 Counterinsurgency in Pakistan 151

State Capacity
Pakistan’s weak state capacity throughout the past decade has led it to
generally pursue less-resource-intensive strategies, particularly contain-
ment and, on occasion, informal accommodation, both of which are
more prone to indecisive outcomes and conflict resumption. Even when
the GoP attempted to complement military operations with develop-
ment in FATA and Balochistan, it produced very little return on the
ground due to scarce resources, a lack of focus, and weak institutions
for implementation.98
In six of the eight cases in our Pakistan case study, the GoP pur-
sued a strategy consistent with the predictions of a weak state (see Table
5.1 above). Scarcity led to selective strategies that would mitigate but
not eliminate insurgency. As long as violence remained below a cer-
tain threshold and did not threaten survival, the state could tolerate it.
This approach is consistent with the rationale expressed in Pakistani
Army doctrinal writings. In the 2002 Pakistan Army Green Book, one
author writes:

The type of aggression encountered in [low-intensity conflict] is

not as blatant as that in war. Subversion, sabotage, assassination
and guerilla operations . . . may pose a threat to national inter-
ests, but the threat to national survival may be neither immi-
nent nor obvious. . . . The inevitable ambiguity of the proper
employment of force demands that weight be given to other con-
siderations. Limited national interests, the presumption against
intervention, and lack of feasibility help explain apparent toler-
ance of some undesirable situations. On the other hand, non-
involvement accepts the piecemeal degradation of security inter-
ests and tolerates unnecessary human suffering, both of which
might be prevented or alleviated by a more active, if necessarily
selective, approach.99

98 Most development funds in Balochistan get siphoned off by provincial officials. See
Lieven, 2011, p. 364; International Crisis Group, 2006b, p. 10.
99 Ata Ur Rehman, “Responsibilities of Pakistan Army in LIC Environment,” in Pakistan
Army Green Book 2002: Low Intensity Conflict, Rawalpindi, Pakistan: Pakistan Army, 2002,
p. 203.
152 Countering Others’ Insurgencies

In two cases, the state did not adopt a counterinsurgency model

typical of low-capacity counterinsurgents: the second round of cam-
paigns in Swat and SWA. In both cases, the GoP’s threat perceptions
were heightened, although for different reasons in each case. The exis-
tential threat in SWA stemmed from the TTP core that had initiated
a rash of suicide bombings across the country. By contrast, the height-
ened threat in the Swat Valley derived from the nature of the territory
threatened—what Pakistani officials termed “settled areas” rather than
tribal or unsettled areas.
In the case of SWA, the increased and direct threat to the Paki-
stani mainland forced the GoP to mobilize its scarce resources for a
more-comprehensive and decisive campaign following the strong-state
repression model. Between 2007 and 2009, the insurgents in FATA,
and specifically the TTP insurgents based primarily in the Mehsud
region of SWA, were able to project violence outside the tribal areas
and into urban centers of Pakistan, including Lahore, Karachi, and
Islamabad, yielding more lethal attacks in denser areas and much
higher levels of civilian casualties than in previous years. The TTP also
became an existential threat to Pakistan, deliberately targeting not only
military outposts in FATA but military bases, barracks, command and
intelligence headquarters, and even places of worship frequented by
military personnel throughout core Pakistan, particularly in Punjab.
This direct threat to the Pakistani mainland led the GoP to embark
on a strong but repressive counterinsurgency campaign in the eastern
part of SWA to dismantle the Mehsud command center. Though SWA
state capacity was weak prior to 2009, the acute threat compelled Paki-
stan to make investments in rapidly developing it, beginning with a
large and sustained military presence following two months of major
ground operations to extend state reach and collect better information
for intelligence-driven operations. Military occupation continues today
with the Army’s 9th and the 40th infantry divisions remaining in place
to hold the area and consolidate their grip through some limited public
goods such as the construction of roads and basic infrastructure. The
treatment of the civilian population, however, has generally been con-
 Counterinsurgency in Pakistan 153

sidered harsh, with little attempt to repatriate and accommodate tribes-

men, 90 percent of whom remain IDPs.100
The post-2009 counterinsurgency campaign in Swat also diverges
from what we would expect in a low-capacity state, but a closer exami-
nation validates the broader theory. Swat and KP province had much
higher levels of integration and state reach (measured in GDP per
capita, human development, infrastructure, or representative provin-
cial government), consistent with what was earlier described as the
unevenness and subnational variation of state capacity in Pakistan. By
2009, the Swat Valley–based TNSM insurgents, not satisfied with pre-
vious deals, had escalated their reach, ambitions, and blatant challenge
of federal authority in settled areas. While the TNSM’s proximity to
Islamabad was not unimportant, of greater concern was its breakout
into other parts of KP that could threaten core lines of communica-
tion or urban populations that were more closely tied into the national
economy and government. TNSM escalation was ultimately met with
a more resource-intensive classical counterinsurgency strategy, in part
because in this settled area, the GoP had more to lose and greater state
capacity and institutional legitimacy upon which to draw for coun-
terinsurgency than it ever had in the traditionally indirectly governed
tribal areas.101
The campaign in KP and Swat was more comprehensive because
the insurgency threatened more-valuable real estate, which in turn jus-
tified greater resource allocations from the center. Moreover, existing
state capacity and expectations were higher, and physical and social
infrastructure needed only to be rebuilt rather than built from scratch.
Preexisting social infrastructure enabled better penetration of society
for information and effective coercion. Physical infrastructure allowed
easier sustainment, resupply, and monitoring of security operations.
Resources, existing and allocated, enabled a greater and wider variety

100 Interview with development official, August 2013; Muhammad Zubair, “Army Operation
in South Waziristan: The TTP and IDPs,” Daily Times, June 19, 2012.
101 The
distinctiveness of the settled area and its impact on counterinsurgency was described
in numerous conversations with retired Pakistani military officials, including an interview
(December 2009) and a private briefing (April 2011).
154 Countering Others’ Insurgencies

of public goods and services, and the state’s prior presence and legiti-
macy made these viable substitutes for futile, excess violence.
FATA’s and Balochistan’s historically weak capacity and isolation
had acquired a path-dependence due to increasingly prohibitive costs
of development and a set of vested interests in the semiautonomous
nature of these ungoverned spaces.102 Despite its failure to curtail the
Baloch insurgency with negotiation offers, the central government
appears content with containment, “confident it can digest the cur-
rent amount of violence that is taking place.”103 With the exceptions
of NWA and SWA, explained above, this was true for much of FATA
as well.

Military Superiority
While military superiority and manpower has been a focal point for
the counterinsurgency literature, this case study of Pakistan’s recent
counterinsurgency campaigns finds mixed outcomes. Only one case
found a correlation between military superiority (and quality) and the
type of strategy adopted.
Military superiority is no guarantor of classic counterinsur-
gency approaches. In the Swat Valley, a more than threefold increase
in troop levels resulted in an almost prototypical classic counterinsur-
gency strategy involving restraint, population protection, public goods,
accommodation, and a decisive outcome. But when Pakistan’s enor-
mous military capabilities were deployed to the tribal regions of Balo-
chistan from 2005 to 2007 and in SWA in 2009–2010, none of these
processes or outcomes resulted. Though perhaps obvious, the utility of
military superiority is constrained by what a state or military chooses
to actually do with the forces deployed. The case of NWA is most
telling. The Pakistan military is believed to have deployed 40,000
troops in the agency since 2010, but it has generally avoided military
engagements—both major offensives and defensive encounters.

102 Ijaz
Khan, 2008, p. 20; Amina Khan, “FATA: Voice of the Unheard—Path-Dependency
and Why History Matters,” Strategic Studies (Islamabad), Vol. 31, Nos. 1 and 2, Spring 2011,
pp. 40–74.
103 Malik Siraj Akbar, “The Ifs and Buts of Negotiating with Insurgents in Balochistan,” The
Baloch Hal, August 1, 2013.
 Counterinsurgency in Pakistan 155

Meanwhile, higher force quality might be associated with im-

proved fighting effectiveness, but it does not appear to be related to
a particular model of counterinsurgency. The deployment of higher-
quality forces (more regular military from the eastern border and SSG
forces) only seems to correlate with more-sustained operations, not
with a particular model of counterinsurgency. In Swat, higher-quality
forces were used for a more classic counterinsurgency approach, while
in SWA, they were used for a more repressive, conventional campaign.
Elsewhere, higher proportions of FC have been effectively utilized in
support of a “second-best” strategy of informal accommodation, as
appears to be the case in NWA at present.104
It is possible that certain governance and institutional contexts
condition the effects of military superiority on strategy. In areas with
high political inclusiveness and state capacity, like Swat, military supe-
riority may enable a more classical counterinsurgency strategy, whereas
in regions where these conditions are absent, military superiority may
have indeterminate effects (as in FATA) or even harmful effects (as in
Balochistan).
Perhaps more important than either the quantity or quality of
security forces is their composition—that is, the extent to which they
are representative of all the important subpopulations within a coun-
try, in terms of both overall numbers and representation in the officer
corps and general staff. Particularly in a state like Pakistan, in which
the military has played a dominant role, the composition and represen-
tativeness of the military appears to exert a major influence on counter-
insurgency strategy toward different subnational groups.
In Pakistan, Pashtuns emerged as a powerful ethnic constitu-
ency in the 1970s. They have been highly represented—perhaps
overrepresented—in the GoP security forces, with a greater share in the
military (15 to 20 percent) than their population share (14 percent), an
artifact dating back to the recruitment policies of the British Indian

104 FC are estimated to constitute between 50 and 75 percent of the 40,000 troops in NWA
given the number of regular army divisions based there (Nawaz and Khan, 2013, and inter-
views with Pakistani analysts, August 2013). On NWA strategy, see Syed Talat Hussain,
“Profiling Our North Waziristan Policy,” Express Tribune, October 29, 2012.
156 Countering Others’ Insurgencies

Army. Moreover, they comprise nearly all of the FC. By contrast, the
Baloch are very poorly represented in both forces.105 They comprise
1 percent or less of the military, and even in the FC wings specifi-
cally tasked to the province of Balochistan (a force of some 30,000 to
50,000), ethnic Baloch are believed to constitute only 5 percent.106
Unsurprisingly, the Pakistani state was far more sensitive to puni-
tive measures against Pashtuns, because of ethnic and tribal ties and
concern that excessive force against them might threaten the cohe-
sion of a military of which they constitute a high proportion. In the
early years of the Taliban insurgency in FATA, the Pakistani military
faced a strategic risk due to problems of defection, desertion, refusals
to fight, and court-martials among some of its Pashtun officers and
soldiers who refused to conduct military operations against their kin.
Fears arose that continued operations “could split the army, which was
clearly unhappy at the prospect of fighting their own people.”107 With
Pashtuns in its ranks, the GoP also believed it had a better grasp of
Pashtun tribal motives and behavior and a method of managing the
insurgency without risking blowback. This may help to explain the
GoP’s persistence in pursuing peace deals over pure coercion, which
numerous observers described as appeasement.108

105 Ian Talbot, Pakistan: A Modern History, New York: St. Martin’s Press, 1998, p. 252;
Eliot Cohen, 2004, pp. 98, 218; Shuja Nawaz, 2011, p. 44; Cloughley, 2008, p. 187; Feroz
Ahmed, Ethnicity and Politics in Pakistan, Karachi, Pakistan: Oxford University Press, 1998,
pp. 253–254; Ayaz Ahmed Khan, “Balochistan Situation,” The Nation, April 10, 2005;
Shuja Nawaz, Crossed Swords: Pakistan, Its Army, and the Wars Within, Karachi, Pakistan:
Oxford University Press, 2008, p. 570; Tahir Amin, Ethno-National Movements of Pakistan,
Islamabad, Pakistan: Institute of Policy Studies, 1988, p. 174; Sher Baz Khan, “Punjab’s
Dominance in Army Being Reduced: ISPR,” Dawn, September 14, 2007; Hassan Abbas,
“Musharraf Contends with the Pashtun Element in the Pakistani Army,” Terrorism Focus,
Vol. 3, No. 42, November 1, 2006.
106 Author interview with Pakistan analyst, May 2013.
107 ZahidHussain, The Scorpion’s Tail: The Relentless Rise of Islamic Militants in Pakistan—
and How It Threatens America, New York: Free Press, 2010, p. 90; Abbas, 2006; Keller, 2008;
Cloughley, 2008, p. 187.
108 Based on interviews in Islamabad, Pakistan, October 2011; Sohail Habib Tajik, “Analysis
of Peace Agreements with Militants and Lessons for the Future,” Conflict and Peace Studies,
Vol. 4, No. 1, January–March 2011.
 Counterinsurgency in Pakistan 157

By contrast, the Pakistani state seemed unconcerned at the harsh

measures taken against the Baloch insurgency from 2005 onward. If
President Musharraf had had Baloch advisors or generals cautioning
that indiscriminate force could fan the flames of a nationalist insur-
gency or affect army morale and cohesion, the way he had from Pash-
tun advisors on FATA, the government might have exercised a more
cautious and comprehensive approach early on in its counterinsurgency
operations in Balochistan.109

Analysis of U.S. Assistance

Although the United States engaged in a strategic partnership with
Pakistan throughout the period under review in this study, the part-
nership was primarily motivated by the United States’ need to obtain
Pakistani cooperation on counterterrorism and counterinsurgency
operations in Afghanistan. The United States did not make serious
efforts to influence GoP counterinsurgency until 2007.110 Even then,
however, U.S. assistance was unable to significantly influence Pakistani
strategy directly, nor was it able to induce reforms in the regime fea-
tures most determinative of counterinsurgency strategy. It was able to
provide real but tactical contributions to the GoP’s military superiority,
some of which proved helpful. Efforts to foster state capacity and politi-
cal inclusion began only much later in the partnership, but these efforts
encountered significant resistance and often hostility.
The U.S. security partnership with Pakistan had three goals:
(1) buying logistics support for U.S. operations in Afghanistan; (2)
securing counterterrorism cooperation against al Qaeda and some
Afghan militant remnants in Pakistan; and (3) bolstering Pakistani

109 Interviews with Pakistani political and national security analysts, May 2013.
110 Daniel S. Markey, Securing Pakistan’s Tribal Belt, Washington, D.C.: Council on Foreign
Relations, Council Special Report No. 36, August 2008, p. 117. Funding to bolster the FC
did not begin until 2007. See Markey, 2008, p. 21. The Pakistan counterinsurgency fund was
not initiated until 2009. See Susan B. Epstein and K. Alan Kronstadt, Pakistan: U.S. Foreign
Assistance, Congressional Research Service, R41856, October 4, 2012, p. 20.
158 Countering Others’ Insurgencies

capabilities and incentives to combat militant threats on its own soil.

The United States did not consistently prioritize any of these goals,
however, and at times the goals even worked at cross-purposes. From
2001 to 2005, the United States primarily focused on buying support
for its Afghan and al Qaeda missions, particularly access to bases, over-
flight rights, border coordination, and the capture of al Qaeda oper-
atives who fled to Pakistan. For much of the period through 2005,
Pakistan’s indifference to Afghan Taliban regroupings was ignored and
security assistance was not intended to be transformative or to influ-
ence Pakistan’s approach to its own internal security. The types and
destination of assistance reflected this emphasis, as the majority of the
funds reimbursed Pakistan’s military for operations in support of U.S.
and allied operations in Afghanistan, and most assistance was provided
as direct budgetary support to the GoP.111
Beginning in roughly 2005, U.S. goals shifted to focus on the
threats within Pakistan, and consequently, assistance expanded to bol-
ster and better enable Pakistani military operations against al Qaeda,
foreign militants, and tribes providing support to them (the nascent
Pakistani Taliban). But not until 2009 did the United States try to
leverage this aid to press for the adoption of sound counterinsurgency
capabilities and tactics, sustained operations, and the expansion of tar-
gets. Additionally, it was only around this time that U.S. economic
assistance sought to promote structural reforms—specifically, support-
ing projects to bolster economic development and employment, pro-
viding services to enhance perceptions of state legitimacy, and encour-
aging civilian governance.
While assistance led to some improvements in GoP military
capabilities, particularly tactical and operational ones, it seems to
have had little success in fostering improvements to state capacity or
political inclusion. Upgrading military superiority was, for the most
part, straightforward and undoubtedly in the GoP’s interest, allow-
ing military assistance to translate into real gains. Assistance to alter

111 Interviewwith former U.S. official, July 2013; also see Markey, 2013, pp. 113, 127, and
Vali Nasr, The Dispensable Nation: American Foreign Policy in Retreat, New York: Doubleday,
2013, p. 69.
 Counterinsurgency in Pakistan 159

regime structures, however, targeted more complex and contested

institutions—a much more challenging goal. Moreover, continued
U.S. reliance on Pakistan for logistics and other support made it dif-
ficult for U.S. policymakers to compel the GoP to do anything it did
not perceive to be unequivocally in its own interest.112

Effects of U.S. Military Assistance

U.S. military assistance to Pakistan sought to augment Pakistan’s abil-
ity to combat foreign and al Qaeda militants amid a nascent Pakistani
Taliban insurgency. In addition to helping finance Pakistan deploy-
ments and operations through coalition support funds, military assis-
tance provided tactical inputs such as equipment, intelligence support,
and training. U.S. assistance, however, did relatively little to induce
Pakistan to adopt U.S. models of counterinsurgency, nor did it ensure
a close alignment of Pakistani actions with U.S. objectives in Afghani-
stan. As a general rule, military assistance could not induce the GoP
to take actions it did not want to take. Where both countries were
committed to a campaign of mutual interest (as occurred in 2009),
however, U.S. military assistance improved GoP capabilities, rendering
its military campaigns more tactically successful—even if they were
in service of what the United States considered to be a problematic
strategy.

Equipment
Equipment from the United States was perhaps the most coveted form
of military assistance, from the Pakistani perspective. While a number
of conventional weapons systems were sold to the GoP for leverage
or access, U.S. provision and sales of lift capability, airpower, ground
mobility, ordinance disposal, and night vision equipment made the
Pakistan military a more competent force for fighting a counterin-
surgency campaign. Beginning in 2007, deliveries of 26 combat and
transport helicopters seemed to correspond with the increasing aggres-
siveness and operational tempo of Pakistani counterinsurgency opera-

112 Paul D. Miller, “How to Exercise U.S. Leverage Over Pakistan,” Washington Quarterly,
Vol. 35, No. 4, Fall 2012.
160 Countering Others’ Insurgencies

tions. These assets proved critical for a number of close-air-support mis-

sions and the SSG deployments in mountainous terrain to retake the
Swat Valley in May 2009.113 Other equipment also enabled Pakistan to
step up its operational tempo. The sales of F-16s and joint direct-attack
munitions, largely seen as a quid pro quo for Pakistani cooperation
rather than a counterinsurgency tool, were in fact heavily relied upon
by GoP operations in FATA and Swat between 2008 and 2012, though
perhaps to the disappointment of some advisors.114 The continually
requested night vision equipment, initially subject to strict accounting
measures that hampered its effectiveness, eventually improved tacti-
cal capabilities and enabled round-the-clock air operations and ground
patrols. Additionally, with improvised-explosive-device attacks becom-
ing the leading cause of Pakistani casualties by 2010 and triggering
security-force overreactions, 20 U.S.-provided explosive-ordinance-
disposal vehicles proved valuable for route clearance.115

Intelligence
As early as 2004 and perhaps earlier, a small SOF presence in Waziristan
was deployed to provide support to the Pakistani military, but with-
out mutual interests in operations in FATA, GoP commanders made a
deliberate choice to underutilize these assets. Once the GoP internal-
ized counterinsurgency operations in Swat and FATA as its own and
committed to serious campaigns in 2009, however, the United States
deployed ISR units to support the 11th Corps prior to and during
its operations in SWA and Bajaur, as well as intelligence fusion cells
embedded with the SSG and FC. Much of this support was politically

113 Kilcullen,2009, p. 12; Shuja Nawaz, 2011, pp. 7, 14; SIPRI Arms Transfer Database,
accessed April 1, 2013.
114 Eric Schmitt and Jane Perlez, “U.S. Unit Secretly in Pakistan Lends Ally Support,” New
York Times, February 23, 2009; “PAF Conducted 5,500 Bombing Runs,” 2011; Jamal Hus-
sain, 2012.
115 ShujaNawaz, 2011, pp. 7, 10, 14, 21–23; Kilcullen, 2009, pp. 12–13; Kagan, Jan, and
Szrom, 2009; SIPRI Arms Transfer Database, web site, 2013; Riaz Anwar Bashir et al., “Pat-
tern of Combat Casualties in War Against Terror Among Soldiers Wearing Body Armor at
CMH Peshawar,” Pakistan Armed Forces Medical Journal, No. 2, June 2012.
 Counterinsurgency in Pakistan 161

feasible as long as it was kept largely secret from the broader public,
though this eventually proved untenable as relations deteriorated.116

Training
For years, the FC was continually criticized for deficiencies such as
unskilled personnel, poor training, inconsistent leadership, lack of
mobility, outdated weapons, shortage of equipment, a proclivity for
desertion, and compromised loyalties.117 In 2007, the United States
stepped up efforts to improve and expand these forces through finan-
cial assistance for new infrastructure, equipment (including 450 vehi-
cles), and a train-the-trainers program to improve fighting capacity and
morale specifically for counterinsurgency.118
As part of an advise-and-assist mission, U.S. SOF began training
Pakistani trainers in October 2008. The impact of this training, which
a few thousand FC underwent, appears to have been positive, although
it is difficult to estimate with certainty.119 Alongside the training, pay
for the FC was quadrupled, and better officers were introduced by a
more determined leader, Maj. Gen. Tariq Khan, making it difficult to
disentangle the effects of the training from these other factors.120 U.S.
officials involved in the training program argued that it had an imme-
diate effect on the skills, discipline, cohesion, and morale of the FC and
that competition by battalion commanders to have their units trained
offered evidence of its success.121 A former official not involved in the

116 Interview
with former U.S. official, June 2013; Asad Hashim, “Pakistan and the U.S.:
A Too-Close Embrace?” Al Jazeera, June 3, 2011.
117 Abbas, 2007; Ashraf, 2008; “U.S. Begins Training Pakistan Frontier Corps,” Jane’s
Country Risk Daily Report, October 27, 2008.
118 Markey,
2008, p. 21; Anwar Iqbal, “No Operation in North Waziristan, U.S. Told,”
Dawn, March 31, 2010.
119 Estimate based on rate of training described in various accounts, including interviews and
Jeremy Page, “Britain to Train Pakistan’s Frontier Corps Troops in Baluchistan,” The Times,
October 9, 2009.
120 Eric Schmitt and Thom Shankar, Counterstrike: The Untold Story of America’s Secret Cam-
paign Against Al Qaeda, New York: Henry Holt, 2011, pp. 198–200.
121 Interview with former U.S. advisor to the FC, April 2013.
162 Countering Others’ Insurgencies

training but able to observe the FC noticed a significant change in

its forces’ quality and confidence between 2007 and 2010.122 Another
account claims the training produced immediate returns, with a
400-man commando unit developed within the FC responsible for kill-
ing or capturing 60 militants, including five high-ranking command-
ers, within months of the training program.123 At the very least, FC
survival rates improved substantially after the onset of the training,124
but the training ended abruptly in May 2011.
Pakistani officials and analysts agreed that there were some
improvements in FC skill sets, but they contended that these gains
were a result of training on newly provided U.S. equipment rather than
in basic skills. Organizational tensions between the FC and the regu-
lar Army and suspicions that U.S. SOF were used as a cover to spy on
Pakistan, particularly following the Raymond Davis episode,125 may
have led some to discount the training’s effectiveness. Nevertheless, FC
training and investment seemed to expand the pool of capable counter-
insurgency forces just as the need arose by 2009.126

Effects of U.S. Civil Assistance

Economic assistance provided to the GoP from the beginning of the
U.S. partnership in 2001 until 2008 essentially aimed to buy coopera-
tion. Assistance was primarily focused on sustaining the government,
with an initial $1 billion in debt forgiveness, deferral of debt payments,
and direct budget support, largely in the form of cash transfers without
any conditions or accountability measures. Most funds prior to 2008
were intended to reward Pakistan but lacked any transformative goals,
with very little (about 9 percent of all assistance) dedicated to develop-

122 Interview with former U.S. official, June 2013.

123 Schmitt and Perlez, 2009.
124 Survivalrates were calculated using Pakistan Institute for Peace data; see the Appendix
for more details.
125 Raymond Davis is an American who was arrested in Pakistan in 2011 after shooting two
men at a crowded traffic stop. He was part of a covert, CIA.-led team collecting intelligence
and conducting surveillance on militant groups deep inside the country.
126 Interviews with American and Pakistani military officials and analysts, summer 2013.
 Counterinsurgency in Pakistan 163

ment. Effectively, this assistance substituted for state capacity rather

than engaging in the lengthy process of building it.127
By 2008, however, U.S. assistance was more directly targeted
at transforming the Pakistani economy and governance. This change
began with a $750 million package for development in FATA. It
was soon followed by the 2009 Kerry-Lugar-Berman Act, titled the
Enhanced Partnership with Pakistan Act (EPPA), also referred to as
the Kelly-Lugar bill, which offered to triple the level of non-military
assistance to $1.5 billion per year for five years by overhauling spend-
ing on social, economic, and political development projects. The goals
of this effort were threefold: (1) to generate a diplomatic and political
victory by signaling a sustained commitment to the people of Pakistan,
(2) to rebalance the U.S. partnership from one that was predominantly
military-centric to one that bolstered civilians, and (3) to provide a
meaningful level of assistance to improve Pakistani development and
governance for continued stability. Though well-intentioned, this civil-
assistance overhaul faced numerous challenges.
First, assistance to improve state capacity suffered from signifi-
cant implementation problems. The first major civil-assistance effort,
the 2008 Livelihood Development Program for FATA, revealed the
difficulties of development in conflict-affected areas. An audit of this
program found massive shortfalls between targets and actual results on
the ground and highly inefficient use of funds, with very high overhead
and operating costs, due to security issues, a lack of local expertise, and
some contractor fraud. Subsequent efforts reinforced a belief that civil
assistance did not work well in highly insecure environments, nor did
it easily complement anemic counterinsurgency operations.128 Imple-
mentation problems persisted under the EPPA. In the first year, less

127 Markey, 2008, pp. 19–20; Markey, 2013, p. 112; Craig Cohen and Derek Chollet, “When
$10 Billion Is Not Enough: Rethinking U.S. Strategy Toward Pakistan,” Washington Quar-
terly, Vol. 30, No. 2, 2007, p. 12.
128 Nancy Birdsall, “Risking the Dog for the Tail,” Newsweek, February 28, 2011; Hassan
Abbas and Shehzad H. Qazi, “Rebellion, Development and Security in Pakistan’s Tribal
Areas,” CTC Sentinel, Vol. 6, No. 6, June 2013; Jane Perlez, “U.S. Aid Plan for Pakistan
Is Foundering,” New York Times, May 1, 2011; interview with former Pakistani official,
July 2013.
164 Countering Others’ Insurgencies

than $180 million was disbursed. Despite improvements, a $400–$500

million shortfall persisted in subsequent years. Concern for adverse
selection of development partners and limited knowledge of the local
partner landscape led to an overreliance on foreign contractors, with
between 30 and 50 percent (though some alleged 90 percent) of the
assistance making its way back to the pockets of U.S. firms.129 When
disbursement rates remained slow in an effort to minimize waste,
fraud, and abuse, Pakistani disappointment with a perceived pattern of
U.S. overpromising and underdelivering further undermined U.S. dip-
lomatic objectives. Ultimately, conflicting priorities—the diplomatic
goal of disbursing a “flood of cash” quickly to secure Pakistani support
and the development goal of spending this money through account-
able Pakistani channels to generate sustainable capacity—proved to be
a problem.130
Second, the conditions attached to the EPPA to bolster civilian
power—motivated in part by U.S. domestic politics—backfired and
resulted in a “diplomatic disaster” that further strained relations and
generated resentment.131 The explicit conditioning of this assistance on
domestic politics within Pakistan, specifically on civil-military rela-
tions, was seen by some as attempted micromanagement of Pakistani
politics and an infringement of Pakistan’s sovereignty. These condi-
tions generated a firestorm of controversy and significant resentment
from the military, some civilian leaders, and even the general public.
At the time of its passage, only 15 percent of the Pakistani public sup-
ported the Kerry-Lugar bill, while 60 percent believed it would not
improve development and 70 percent believed it would not make a
difference in the lives of ordinary Pakistanis. While the United States
quickly diluted many of its conditions and routinely waived them out
of strategic necessity, by the spring of 2013, perceptions of U.S. aid had

129 Mosharraf Zaidi, “Demystifying Foreign Aid,” The News, December 15, 2010; Nasr,
2013, p. 80; International Crisis Group, Aid and Conflict in Pakistan, Asia Report No. 227,
Brussels/Islamabad, June 27, 2012; Markey, 2013, p. 149; Nancy Birdsall, Wren Elhai, and
Molly Kinder, Beyond Bullets and Bombs: Fixing the U.S. Approach to Development in Paki-
stan, Washington, D.C.: Center for Global Development, June 2011, p. 14.
130 Markey, 2013, pp. 145–150; interview with development official, August 2013.
131 Markey, 2013, p. 144.
 Counterinsurgency in Pakistan 165

not improved, with only 8 percent of the population believing it had a

positive impact and the majority claiming it had no impact or a nega-
tive impact.132
Third, much of the civil assistance to Pakistan suffered from con-
tradictions between the U.S. aims of fostering both long-term develop-
ment and diplomatic gains, with the end result that it yielded neither
in the short term. Initially, assistance had a very narrow focus and
a short time horizon to maximize diplomatic impact, but it was not
designed to have lasting development effects. For the tripling of assis-
tance to have its intended symbolic effect, it needed to be implemented
quickly. Most of the funds were allocated to quick-impact projects of
questionable value in order to influence public opinion and the Pak-
istan elite’s strategic calculations regarding U.S. policy. During this
time, civil assistance was routinely criticized for not investing in any
capacity-building projects that might have made a gradual but mean-
ingful impact on development and state legitimacy in the long run. The
result was a slowdown in actual disbursements and the accumulation
of unmet expectations that may have worsened Pakistani public per-
ceptions of the United States. Though U.S. development officials have
worked to correct some of these problems and refocus some stabiliza-
tion efforts to specifically complement Pakistani military operations
and improve government legitimacy, these remain on a small scale.
A long-term relationship and tremendous patience are necessary to
realize a return on these investments.133
Finally, the United States missed opportunities to provide the
GoP with long-sought alternatives to aid, such as support for for-
eign investment and access to protected U.S. markets (most notably

132 Gilani Research Foundation, Perceptions on Kerry-Lugar Bill: Gilani Poll/Gallup

Pakistan, Islamabad, October, 14, 2009; Ben Arnoldy, “Why Pakistanis Would Reject
$7.5 Billion in U.S. Aid,” Christian Science Monitor, October 14, 2009; Anwar Iqbal, “U.S.
Note Dilutes Some Conditions in Kerry-Lugar Bill,” Dawn, October 14, 2009; Dan Murphy,
“A Quiet Waiver for Pakistan from the Obama Administration,” Christian Science Monitor,
October 15, 2012; Pew Research Center, “On Eve of Elections, a Dismal Public Mood in
Pakistan,” Pew Global Attitudes Project, Washington, D.C., May 7, 2013.
133 Interviews with development officials, July/August 2013; Birdsall, Elhai, and Kinder,
2011.
166 Countering Others’ Insurgencies

the U.S. textile market). The GoP had long pursued trade access—
recognizing some of the inflationary risks and limited downstream
effects of direct aid—to spur economic growth and job creation and
create more stakeholders in U.S.-Pakistan relations.134
Though bolstering political inclusion and state capacity were
daunting tasks, assistance to complement specific counterinsurgency
efforts proved effective. U.S. humanitarian and stabilization aid consis-
tently remained popular among Pakistanis. Such aid not only cushioned
the impact of natural disasters, it also helped to ameliorate the civilian
impact of military operations that produced millions of IDPs, such as
those in Swat. By spring 2013, about one-third of the EPPA’s disbursed
$3.2 billion had been spent on emergency relief. These funds tended to
be dispersed quickly to address immediate needs and encountered few
bureaucratic obstacles.135 Moreover, the Swat experience later revealed
that more-robust and sustained military operations could produce fer-
tile ground for development and governance assistance to be effective.
Assistance provided after major combat operations in Swat seemed to
have a greater economic impact than aid deployed to conflict-affected
regions, and it was often matched by local entrepreneurs, confident in
the future, making capital investments. Recently, U.S. efforts in more-
stable parts of the northwest to facilitate the provision of some public
goods on a local level have seemed to be positively influencing local
attitudes and confidence in government.136

134 Birdsall,
Elhai, and Kinder, 2011; Nicolas Brulliard, “Pakistan Textile Exports: Call for
Wider Lifting of U.S. Tariffs Intensifies,” Washington Post, December 24, 2010; Kalbe Ali,
“U.S. Withdraws ROZs Offer,” Dawn, November 10, 2012.
135 Kurtzleben,2009; interview with former Pakistani official, August 2013; Alan Gelb and
Caroline Decker, “Direct Payments to Pakistan’s Flood Victims: A Smart Option for U.S.
Assistance,” The Hill, November 8, 2010; Umer Nangiana, “NA Question Hour: Pakistan
Received $3.2B Under Kerry-Lugar-Berman Act,” Express Tribune, March 7, 2013; Pew
Research Center, 2013; William S. Murphy, Review of USAID’s Internally Displaced Persons
Program in Pakistan, Washington, D.C.: Office of Inspector General, Review Report No.
5-391-10-001-S, June 28, 2010.
136 “GrowthProjections: Washington to Help Revive Tourism in Swat, Olson,” Express Tri-
bune, March 15, 2013; interview with development official citing private survey data, August
2013.
 Counterinsurgency in Pakistan 167

U.S. Efforts to Influence Pakistan and Their Consequences

Critics of the U.S.-Pakistani partnership point to its failures and seek
to explain why the United States did not get more out of it, in terms
of both support for its Afghan policy and reforms within Pakistan
itself.137 The simplest explanation is one of interests and leverage. Paki-
stan disagreed with U.S. strategy in Afghanistan and held a different
theory of stability in NWA, while the United States did not have the
leverage to compel, nor did it offer sufficient funds to buy, a shift in
Pakistani interests, behavior, and cooperation.
First, the U.S. and Pakistan simply had different and strongly
held strategic interests. For obvious reasons, Pakistan placed its domes-
tic stability above U.S. goals in Afghanistan. It feared another costly
and bloody escalation of violence, such as the one that devastated
urban Pakistan between 2007 and 2009. Instead of large-scale opera-
tions aimed at a decisive outcome, the GoP adopted a counterinsur-
gency strategy of informal accommodation in NWA involving peace
agreements (and perhaps bribes) that appeared to be keeping a lid on
violence. Additionally, some of the militant organizations in NWA
and SWA were useful allies of the government, balancing or manag-
ing other, more-threatening groups.138 Moreover, many of these groups
were reliable assets for ensuring Pakistani influence in any future
Afghan settlement. GoP officials objected to what they saw as the dis-
empowerment and exclusion of Afghan Pashtuns, and they fundamen-
tally disagreed about the political reconcilability of Taliban actors on
both sides of the Afghan border. They believed most Taliban groups
representing aggrieved Pashtuns would cease to threaten the Afghan
and Pakistani states following a U.S./NATO withdrawal and greater
inclusion of Pashtuns in the Afghan government, particularly the secu-
rity forces.139 Because of these disagreements, Pakistan was unwilling

137 Cohen and Chollet, 2007; Miller, 2012; Stephen Krasner, “Talking Tough to Pakistan,”
Foreign Affairs, Vol. 91, No. 1, January–February 2012, pp. 87–96.
138 SyedTalat Hussain, 2012; “Not Allies: Eight Dead as TTP, Haqqani Network Clash
in North Waziristan,” Express Tribune, April 19, 2012; Tahir Khan, “Rival Groups Launch
Operations Against Pakistan Taliban in Mohmand,” Express Tribune, June 26, 2013.
139 Interviews and discussions with numerous retired Pakistani officials and analysts.
168 Countering Others’ Insurgencies

to easily accede to U.S. demands—such as abandoning cease-fires and

engaging in military operations in NWA that would be harmful to its
internal security and foreign-policy interests—for the sake of the alli-
ance. 140
Second, the United States did not have the leverage to compel
this cooperation. Even though the United States occasionally tried to
use coercive diplomacy through public criticisms or demands,141 U.S.
dependence on Pakistan for its logistical tail and supply lines under-
mined its leverage and made it incapable of credibly threatening the
cutoff of support to influence Pakistani strategic behavior. Some ana-
lysts even suggest that whatever leverage the United States did possess
remained unexercised.142
In part, this leverage problem was caused by a lack of informa-
tion. Through 2007, the United States possessed poor information on
Pakistan’s strategic calculations and actions, in part due to its attention
and assets being focused on Iraq. It believed that during this period,
the GoP was “clearly committed to this challenge [the war in Iraq] and
willing to use all resources”—an assessment that was optimistic in the
extreme.143 Even during the turnaround in Pakistani policy from 2008
through early 2011, U.S. officials may not have been able to discern
Pakistan’s selective approach to counterinsurgency. Many continued to
praise Pakistan for “very significant strides,” and “unprecedented coop-
eration [and] coordination,” and they even displayed high expectations

140 JavedHussain, “The Hornet’s Nest,” The News International, January 19, 2011; Shem-
rez Nauman Afzal, “North Waziristan: The Death Trap,” Lahore, Pakistan: Spearhead
Research, December 29, 2010; interviews in Pakistan, October 2011.
141 These included numerous public-pressure campaigns by leading U.S. officials, including
threats to lead independent incursions into Pakistan in 2007, after elected civilian leaders
contemplated negotiations in the spring of 2008 and 2009, surrounding visits by U.S. offi-
cials in fall 2009, after the failed Times Square bombing in May 2010, or through Adm.
Mullen’s September 2011 testimony linking Pakistan’s Inter-Services Intelligence to the
Haqqani network.
142 Jane Perlez and Helene Cooper, “Signaling Tension, Pakistan Shuts NATO Route,” New
York Times, September 30, 2010; Miller, 2012.
143 Interview with former U.S. official, June 2013; Markey, 2013, pp. 105–135. “Impressions
of Waziristan, January 25, 2005,” National Security Archive, declassified cable, Electronic
Briefing Book No 325, September 13, 2010.
 Counterinsurgency in Pakistan 169

for new offensives in NWA.144 Ultimately, ambiguity and uncertainty

over Pakistani strategic calculations mitigated what little leverage the
United States did possess.
Third, large amounts of money intended to buy Pakistani cooper-
ation were never sufficient to fully shift Pakistani interests and behav-
ior.. The GoP saw coalition support funds, the bulk of the funds trans-
ferred early on and 40 percent of the overall funding, as something it
was owed.145 Additionally, while Pakistan received one of the largest
shares of U.S. aid, the amount was not particularly large on a per-capita
basis (given Pakistan’s extremely large population), nor did Pakistani
government officials perceive it as adequate to offset the high material,
human, and opportunity costs that had been suffered since Pakistan
allied with the United States in 2001.146 Given what was perceived as
inadequate compensation for the losses the GoP believed it had suf-
fered between 2007 and 2010, the Pakistani government was much
more reticent to fully align its strategy with U.S. interests in Afghani-
stan and embark on future operations such as an invasion of NWA.
A fundamental shift in GoP strategy was too much to expect,
despite the substantial amount of assistance the United States pro-
vided. While the United States struggled to influence GoP strategic
behavior, it was better positioned to improve certain military capa-
bilities and enable some successful tactical operations with injections

144 Deputy to the Special Representative for Afghanistan and Pakistan Daniel Feldman,
“Testimony Before South Asia Subcommittee, House Foreign Affairs Committee,” transcript,
April 5, 2011; Gen. David Petraeus, “Hearing to Receive Testimony on the Situation in
Afghanistan,” testimony before Senate Armed Services Committee, March 15, 2011; Adm.
Mike Mullen, transcript of speech delivered to Washington Institute for Near East Policy,
January 7, 2010.
145 Nasr, 2013, p. 94; Epstein and Kronstadt, 2012; interview with former U.S. national
security official, June 13, 2013; interview with Pakistani military official, July 2013.
146 Sartaj Aziz, “The Economic Cost of Extremism,” in Usama Butt and N. Elahi, eds.,
Pakistan’s Quagmire: Security, Strategy, and the Future of the Islamic-Nuclear Nation, London:
Continuum International, 2010, pp. 75–92. Some U.S. officials acknowledged this problem.
In reference to the $5 billion he helped raise through the Friends of Democratic Pakistan
donor community, Richard Holbrooke stated that “Pakistan needs $50 billion, not $5 bil-
lion.” General Petraeus also recognized this and said, “You get what you pay for. We have
not paid for much of anything in Pakistan.” See Nasr, 2013, p. 80; interview with Pakistani
military official, July 2013.
170 Countering Others’ Insurgencies

of material and economic support. These injections of tactical assis-

tance would have an impact on Pakistani counterinsurgency practices,
however, only when the government was already motivated to fight
and shared the same threat perceptions of certain militant groups. This
confluence of threat perceptions existed in Swat and SWA, but it has
been absent in other parts of FATA and in Balochistan, which has been
utterly neglected by U.S. policy.

Conclusion

The case study of Pakistan offers further support for the theory and
empirical analysis presented in Chapters Two and Three. There is broad
support for the contention that political inclusion and state capacity
influence a state’s general choices of counterinsurgency strategy in
terms of the level of discriminate violence, provision of public goods,
and openness to reconciliation. At a more nuanced level, when coun-
terinsurgency campaigns are broken down over time and space, the
predictions still hold: Regions with higher levels of political inclusion,
institutional integration, and state capacity prior to conflict are more
likely to exhibit counterinsurgency campaigns that look like the clas-
sic counterinsurgency model. Subnational conditions, in other words,
may explain much of the variation in counterinsurgency approaches
that cannot be explained by the properties of the state at the macro
level.
Since these subnational conditions are also important determi-
nants of conflict onset, counterinsurgency will take place more fre-
quently in these weak areas, leading states to more frequently adopt
strategies of containment or informal accommodation. The type of
competitive state-building the classic counterinsurgency model envi-
sions assumes that a state can impose its writ on these “ungoverned”
spaces. Generating the local architecture and tools necessary to imple-
ment this approach in a sustained manner may be infeasible, however,
due to local resistance and excessive costs, as has been the case for the
Pakistani government in FATA. In these areas, governments may have
to settle for second-best strategies like informal accommodation and
 Counterinsurgency in Pakistan 171

may need to foster a gradual, decades-long approach to integration and

state-building.
The Pakistani case offers no support for the claim that a country’s
ex ante possession of military superiority dictates the manner of its appli-
cation or the type of counterinsurgency campaign the state conducts.
States in a high-threat environment with other internal or external con-
cerns may choose to economize their force deployment and hold most
in reserve for other purposes. Pakistan’s overall military superiority has
never seriously changed since 2001, but Pakistan has employed differ-
ent types of strategies depending on local contexts (what was threat-
ened) and threat perceptions (who was threatening). The government
chose not to deploy its military superiority in the early FATA and Swat
campaigns but did so in Balochistan, because it judged the insurgency
there to be more threatening to the state at the time. Likewise, how a
state utilizes its military capabilities may vary. The government used its
military superiority for strong-state repression in Balochistan, classic
counterinsurgency operations in Swat, and containment in FATA—
even after learning the lessons of population security in Swat. Ethnic
representation in the security forces, however, can have some impact on
a state’s use of force, its willingness to make accommodations, and its
ability to adapt. In the case of the GoP encounters with Pashtun-based
militant groups, there is some evidence that empathy, information, and
concerns for cohesion motivated the security forces that have a signifi-
cant Pashtun presence to sign peace deals, minimize their use of force,
and eventually—when compelled to take action—adapt a population-
centric counterinsurgency model. In contrast, in Balochistan, the GoP
felt unencumbered and routinely defaulted to punitive measures, with
little concern for the moral or strategic consequences of indiscriminate
force or half-hearted accommodation efforts.
While political inclusion and state capacity seem to have been
quite important in shaping counterinsurgency strategy in Pakistan,
these features have been very difficult for third parties like the United
States to manipulate through assistance policies, particularly in a near-
to-medium time frame. Overtly tinkering with GoP domestic poli-
tics has proved to be not only difficult but also counterproductive for
relations with Pakistan. The United States found it easier to bolster
172 Countering Others’ Insurgencies

military superiority and enhance certain capabilities, with no certainty

that these would be employed for counterinsurgency practices consis-
tent with U.S. interests or values. U.S. material support for Pakistan
worked best in areas of greatest mutual interest—particularly in Swat
and, to a lesser extent, in SWA. There is always a risk, however, that
some military assistance—particularly equipment and forces trained—
might be redeployed for purposes other than the one that motivated
U.S. support. In the case of Pakistan, U.S. military assistance could
easily be used to prosecute highly repressive counterinsurgency opera-
tions in Balochistan, where the United States has both limited interest
and little influence.
The GoP’s employment of second-best strategies like informal
accommodation in parts of FATA (most notably in NWA) both miti-
gated some conflict and sidelined some insurgent combat power. As a
result, Pakistan was able to concentrate its military superiority for deci-
sive actions in Swat and later in SWA for operations that are still on-
going. Though harmful to U.S. interests in Afghanistan, the GoP’s
choice of informal accommodation in NWA may have been in its best
interest in its attempt to ensure some stability and avoid the fallout
of previous operations while consolidating its hold over territories in
which it had retaken control.
Finally, very little could be done by the United States to change
Pakistan’s overall cost-benefit calculus in NWA and its strategy toward
the region, at least in the near-to-medium term. For a time, the GoP’s
counterinsurgency goals and priorities were thought to overlap with
U.S. policy goals in Afghanistan, with failures to act ascribed to a defi-
cit of capability. The U.S. failure to change the GoP’s strategy of infor-
mal accommodation in NWA to a more confrontational approach,
however, was due to different strategic interests and divergent theo-
ries of “victory.”147 Neither substantial aid nor intensive engagement
has proved adequate to change the fundamentals of the U.S.-Pakistani
partnership.

147 Rahimullah Yusufzai, “Fazlullah in Afghanistan,” The News, December 1, 2012.

CHAPTER SIX

Managing Troubled Partnerships

Successful light-footprint U.S. operations such as OEF-P in the Phil-

ippines and Plan Colombia have led many in the American defense
community to argue that the United States will be able to protect its
security interests in conflict-affected countries throughout the world
through low-cost, indirect approaches. This report suggests that while
such operations can indeed have positive effects, their prospects for
success depend critically on the local context and especially on the
characteristics of the counterinsurgent regime.
Success stories of U.S.-counterinsurgent partnerships have
occurred in countries characterized by relatively inclusive politics and
reasonable levels of state capacity. The governments of such countries
typically adopt approaches to counterinsurgency that approximate
the Western model, often (misleadingly) referred to as the hearts-and-
minds approach. Unfortunately, only approximately one insurgency
in eight occurs in such best-case countries. The majority of rebellions
take place in worst-case conditions—that is, in countries that lack both
inclusive politics and state capacity. Regimes in this latter category
overwhelmingly rely on blunt applications of military force to contain
or suppress rebellion.
The quantitative analysis presented in Chapter Three paints a
stark picture of the different trajectories that conflicts follow in these
best-case and worst-case environments. Whereas only 13 percent of
civil wars in the best-case environments fail to reach an outcome that
the government finds acceptable (that is, either outright military vic-
tory or a negotiated settlement acceptable to both sides), the failure rate

173
174 Countering Others’ Insurgencies

is nearly five times as high (60 percent) in the worst-case environments.

Similarly, whereas fewer than 10 percent of all politically inclusive
regimes have resorted to indiscriminate violence as a tool of counterin-
surgency, 39 percent of regimes that are not politically inclusive and do
not possess reasonable levels of state capacity have used such tactics. In
other words, the chances of wide-scale abuse by security forces are four
times greater among the worst-case regimes than they are among the
more-favorable regimes.
The case studies of the Philippines and Pakistan in Chapters Four
and Five broadly support these quantitative findings. The focus on spe-
cific countries also provides an understanding of how various factors
operate over time and across different subregions, helping to explain
the microdynamics of conflict in these countries. In both the Philip-
pines and Pakistan, more-democratic governments were more likely to
adopt a classical counterinsurgency model that sought accommodation
with the reconcilable opposition and used violence relatively discrimi-
nately. This tendency was particularly pronounced in regions where
the governments possessed the necessary civil capacity to implement
hearts-and-minds approaches. In contrast, during periods when these
countries were ruled by more-autocratic regimes and in regions where
the state exercised little effective control, governments were much more
likely to seek to contain and suppress insurgents through raw force.
The successes of U.S. policies in the Philippines and Colombia,
in other words, were critically contingent on the character of the part-
ner regime. This conclusion is not meant to deny the importance of
U.S. assistance; to the contrary, the case study of the Philippines sug-
gests that U.S. aid played a critical role in recent successes there. But
similar U.S. policies with less-promising partner nations should not be
expected to produce anywhere near the same levels of success. And as
our quantitative analysis revealed, the majority of potential U.S. part-
ner nations—including many that are central to ongoing U.S. counter-
terrorism efforts—are much less-promising partners.
 Managing Troubled Partnerships 175

Approaches to Managing Troubled Partnerships

These results should be sobering for those proponents of indirect or

light-footprint approaches who suggest applying them in less-promising
environments than those of the Philippines and Colombia. What,
then, should U.S. policymakers do with this information if they are
not simply to throw up their hands in exasperation?
One possibility is to be highly cautious in the choice of partner
nations, partnering only with those that have favorable characteristics.
Unfortunately, such an approach provides little guidance about what the
United States should do in cases where it has important security inter-
ests at stake in a counterinsurgency being fought by a less-promising
potential partner nation. Counterinsurgent regimes may be unattrac-
tive, but not infrequently, the insurgents they are fighting are even
worse.
A second option would be to force liberalizing reforms on its part-
ners, as the United States attempted to do in El Salvador in the 1980s.1
Such an approach has been the traditional hearts-and-minds prescrip-
tion for counterinsurgency. Certainly, the United States should work
with its partners to implement a variety of reforms. Evidence from the
past several decades, however, suggests that it is unrealistic to expect
most partner regimes to institute major changes unless they are forced
to by looming military defeat or prolonged stalemate.
A third alternative is to look for small wins—that is, lesser
reforms and informal practices that might serve as the basis for achiev-
ing at least minimally acceptable outcomes. Even among the worst
types of counterinsurgent regimes, a significant number have managed
to achieve satisfactory outcomes—often through a negotiated settle-
ment with insurgents—and to do so without the use of indiscriminate
violence. These outcomes suggest that the United States and its allies

1 On the United States’ attempts to enforce liberalizing reforms on El Salvador, see, for
instance, Watts et al., 2012; Mark Peceny, “Two Paths to the Promotion of Democracy
During U.S. Military Interventions,” International Studies Quarterly, Vol. 39, 1995, pp. 371–
401; Thomas Carothers, In the Name of Democracy: U.S. Policy Toward Latin America in the
Reagan Years, Berkeley, Calif.: University of California Press, 1993, especially Ch. 1.
176 Countering Others’ Insurgencies

should explore alternatives to the traditional approaches that constitute

the canon of Western counterinsurgency theory.

Selectivity and Avoiding “Bad” Counterinsurgents

The sobering statistics presented in Chapter Three and the lessons that
can be gleaned from the case studies should encourage decisionmak-
ers to reflect long and hard before committing to a counterinsurgent
regime that lacks reasonable levels of political inclusivity or state reach.
Regimes that score low in both of these dimensions are unlikely to
bring a counterinsurgency to an acceptable end, and they run a high
risk of resorting to indiscriminate violence while pursuing their indeci-
sive campaigns. In many or most such cases, the United States would
be well served to heed Daniel Byman’s advice and “act more like a
third party to a conflict rather than an open and strong ally of govern-
ment forces.”2 In others, the United States might insist on the partner
regime implementing key reforms before it provides military assistance.
This was the approach the United States recently adopted in Mali, for
instance, where it insisted that the country’s military coup be rolled
back and democratic elections take place before it would provide aid.
Such an approach is similar to the development community’s
recent embrace of the principle of selectivity—that is, awarding aid
only to those countries that are judged likely to use the assistance effec-
tively.3 Selectivity focuses aid on reinforcing success rather than steering
it to countries that may be most in need of assistance but are unlikely
to make productive use of it. The Millennium Challenge Corporation,
created under the administration of President George W. Bush, was
created as a mechanism to execute such an assistance strategy.
Exercising selectivity is critically important in the national-
security field as well as in development assistance, but there will inevi-
tably be cases in which compelling national-security interests are at
stake and U.S. decisionmakers do not believe they can afford to run the
risk of having even a problematic partner government defeated by even

2 Byman, 2006, p. 82.

3 Nicolas Van de Walle, Overcoming Stagnation in Aid-Dependent Countries, Washington,
D.C.: Center for Global Development, 2005.
 Managing Troubled Partnerships 177

more problematic insurgents. Here is where the analogy to the develop-

ment community breaks down. Aid agencies are much more capable of
abandoning a partner government that routinely underperforms than
are defense authorities—at least when the partner represents a critical
national-security interest. In such cases, the United States may be left
with little choice but to attempt to work with the partner nation. How,
then, should it manage such difficult partnerships?

Conditionality and Making Partners into “Good” Counterinsurgents

As the previous chapters have indicated, the traditional hearts-and-
minds approach fits poorly with the characteristics of most regimes
waging counterinsurgency. The large majority of them are auto-
cratic regimes with weak state institutions, poorly outfitted to offer
broad political participation rights and public goods. The Western
response has often been to attempt to make partner regimes into better
governments—more inclusive, more capable, and less abusive. As dis-
cussed below, although this approach is attractive, it has a discouraging
record of success. In particular circumstances and for particular pur-
poses, however, it may be appropriate.

The Leverage Dilemma

Many of the classic works in the counterinsurgency literature note a
long litany of failed efforts to impose reform on unwilling partner gov-
ernments, most obviously in South Vietnam.4 Similarly, the develop-
ment community has generally become skeptical of aid conditionality.5
Two problems in particular have frequently confounded donor efforts
to effectively condition their assistance: (1) the inability to credibly
threaten to withdraw assistance, and (2) the inadequate ability to mon-
itor the use of aid and the implementation of reforms.
In theory, the United States should have considerable leverage
while supporting a partner regime threatened by insurgency. Particu-

4 See especially Blaufarb, 1977.

5 For a recent overview of conditionality debates and the growing pessimism about such
policies in much of the development community, see especially Andrew Mold, Policy Owner-
ship and Aid Conditionality in the Light of the Financial Crisis: A Critical Review, Paris: Orga-
nization for Economic Cooperation and Development, 2009.
178 Countering Others’ Insurgencies

larly if the partner regime would be at high risk of defeat without U.S.
assistance, threats to withdraw that assistance should be highly per-
suasive. The problem is that U.S. credibility and the reputations of key
U.S. actors become intertwined with the fate of the partner regime.
Withdrawing U.S. support for a partner might lead to the partner’s
defeat, which would be seen as a blow to the United States’ global
reputation for supporting its allies and would impose domestic political
costs on those decisionmakers who had initially supported the partner
regime.6 In fact, several observers have posited what might be called
the “leverage dilemma”: the greater the U.S. commitment to a partner
regime, the less leverage it has over the partner.7 South Vietnam, Iraq,
and Afghanistan are cited as the paradigmatic examples.
Stated in such linear terms, the leverage dilemma is almost cer-
tainly wrong: Taken to its extreme, it implies that the United States
would have maximum leverage when it supplied no assistance what-
soever to the partner regime. Nonetheless, studies of development
assistance do suggest that a donor’s ability to exercise leverage over an
aid recipient is significantly weakened if the recipient is critical to the
donor’s national interests.8
Sierra Leone provides a useful example of the potential and limits
of conditionality. In 2007, the incumbent Sierra Leonean People’s Party
(SLPP) lost largely free and fair elections. The SLPP had given every

6 One of the earliest and most influential arguments to this effect was Blaufarb’s (1977).
One of the best case studies on the domestic political consequences of such failures remains
Leslie H. Gelb and Richard K. Betts, The Irony of Vietnam: The System Worked, Washington,
D.C.: Brookings Institution, 1979. On “audience costs” more generally, see James Fearon,
“Domestic Political Audiences and the Escalation of International Disputes,” American
Political Science Review, Vol. 88, No. 3, September 1994, pp. 577–592, and Michael Tomz,
“Domestic Audience Costs in International Relations: An Experimental Approach,” Interna-
tional Organization, Vol. 61, No. 4, Autumn 2007, pp. 821–840.
7 See, for instance, Shafer, 1988. For a more recent example, see Hammes, 2012.
8 See, for instance, Axel Dreher, Jan-Egbert Sturm, and James Raymond Vreeland, “Devel-
opment Aid and International Politics: Does Membership on the UN Security Council
Influence World Bank Decisions?” Journal of Development Economics, Vol. 88, No. 1, 2009,
pp. 1–18, and Axel Dreher, Jan-Egbert Sturm, and James Raymond Vreeland, “Global Horse
Trading: IMF Loans for Votes in the United Nations Security Council,” European Economic
Review, Vol. 53, No. 7, 2009, pp. 742–757.
 Managing Troubled Partnerships 179

indication of its intent to subvert these elections and maintain itself in

office despite its intense unpopularity after years of official corruption.
The regime was utterly dependent, however, on international (and espe-
cially British) financial and military support. When the United King-
dom suspended budgetary support pending the ruling party’s accep-
tance of the results of the presidential poll, the SLPP had little choice
but to accept its defeat. Here is an instance in which conditionality
yielded an unambiguous success: The alternation of executive power
in Sierra Leone was a milestone achievement in the country’s post-
conflict evolution, and it almost certainly would not have happened in
the absence of donor conditionality.9 What is much less clear is whether
the United Kingdom would have been able or willing to make similar
threats had its intervention in Sierra Leone been motivated by critical
security interests rather than purely humanitarian concerns.
Seldom is conditionality such an all-or-nothing proposition, how-
ever. Conditionality is often applied to specific programs, so in theory,
the United States should be able to fine-tune its conditionality, prod-
ding a recalcitrant partner forward by threatening to withdraw specific
aid programs if the partner does not make satisfactory progress toward
reform. Such an approach would not place the regime’s survival at risk
but would threaten to deny specific programs or funds that the partner
regime finds useful.
Yet such fine-tuning of conditionality is extremely difficult in prac-
tice, particularly in conflict environments. In general, external actors
are at an acute information disadvantage when seeking to influence the
politics of foreign countries, especially countries in which formal insti-
tutions are weak and politics is even more informal and personality-
driven than in developed countries. As a result, aid intended to influ-
ence political structures in recipient countries is often manipulated by
local actors to serve purposes opposite from the ones intended.10 The

9 Jimmy D. Kandeh, “Rogue Incumbents, Donor Assistance, and Sierra Leone’s Second
Post-Conflict Elections of 2007,” Journal of Modern African Studies, Vol. 46, No. 4, 2008.
10 See, for instance, Peter Blunt and Mark Turner, “Decentralization, Democracy, and
Development in a Post-Conflict Society: Commune Councils in Cambodia,” Public Admin-
istration and Development, Vol. 25, 2005, pp. 84–85; Ben D’Exelle, “Excluded Again:
Village Politics at the Aid Interface,” Journal of Development Studies, Vol. 45, No. 9,
180 Countering Others’ Insurgencies

challenges donor nations face in monitoring the effects of their aid are
even more pronounced in conflict environments, where their person-
nel often operate under severe mobility restrictions. This dynamic was
observed in both of the case studies in this report and can be seen
even in heavy-footprint theaters such as Afghanistan, where the United
States possessed many more transportation assets and capabilities for
personnel protection than is the case in light-footprint interventions.

Potential for Changing Key Characteristics of Partner Nations

Turning to the key characteristics of counterinsurgent regimes that are
the focus of this study, we find a number of reasons to doubt the abil-
ity of most partner regimes to implement wide-ranging reforms in the
course of fighting.
In aggregate, the countries in our sample experienced very little
change in their degree of political inclusion during the course of con-
flict. On average, their Freedom House democracy scores changed by
only 0.10 point on the seven-point scale from the first year of the war to
the last. Fewer than 9 percent of cases experienced large positive change
in their democracy scores—that is, a change of two or more points—
during the course of fighting. In general, then, it is unrealistic to expect
significant liberalization as a part of a counterinsurgency strategy. As
a peacemaking and peace-building strategy, however, greater political
participation may be much more realistic. Many regimes become more
inclusive and participatory as a part of the war termination process.
In such cases, efforts to improve political participation are much less
about winning hearts and minds by redressing grievances than they are
about providing a framework that reconcilable opposition leaders will
feel offers them realistic opportunities for political participation and
influence over the allocation of state services.
Building the state capacity of partner nations is even more daunt-
ing, at least within a time frame that is relevant to counterinsurgents.

October 2009, pp. 1453–1471; Richard Fanthorpe, “On the Limits of Liberal Peace: Chiefs
and Democratic Decentralization in Post-War Sierra Leone,” African Affairs, Vol. 105, No.
418, 2006, p. 40; and Harry G. West and Scott Kloeck-Jenson, “Betwixt and Between: ‘Tra-
ditional Authority’ and Democratic Decentralization in Post-War Mozambique,” African
Affairs, Vol. 98, No. 393, 1999, pp. 455–484.
 Managing Troubled Partnerships 181

Lant Pritchett and his colleagues estimated how long it would take the
worst-rated countries of the world to reach a level of state effectiveness
equivalent to that of middle-ranked developing countries—countries
such as Algeria, Tanzania, and Guatemala—on a variety of governance
indicators. The results are sobering. If the 20 worst-rated countries con-
tinued to reform at their current pace, most of them would never reach
the current quality of governance of middle-ranked developing coun-
tries, because their recent performance has been negative. Even if the
worst-ranked countries experienced a miraculous change and started
to reform as rapidly as the 20 fastest-reforming countries in the world,
most would still require approximately 15 years to reach the middle
tier of developing-country state effectiveness, and some would require
three decades. If the bar were raised so that the goal were to become
as effective as the 75th percentile of developing countries, most of the
20 worst-rated countries would require approximately 25 years to reach
the goal, even if they consistently implemented reforms as rapidly as
the top 20 fastest-reforming countries in the world.11
The results are somewhat more encouraging if we look more nar-
rowly at building the military capabilities of partner nations. U.S.
advisors from among conventional forces, SOF, and contractors have
experienced considerable success in improving partner nations’ tacti-
cal capabilities in the short term. The case studies in this report pro-
vide evidence of such improvements in the Philippines and Pakistan,
and similar gains have been observed among the troop-contributing
countries to the African Union Mission in Somalia (AMISOM) and
elsewhere.12
Even here, however, there are important limits to the degree of
change that can be effected. While tactical skills such as marksman-
ship can be imparted relatively easily, operational-level skills are much
more challenging. Developing-country militaries frequently struggle
with logistics and sustainment, intelligence fusion and analysis, and

11 Pritchett, Woolcock, and Andrews, 2010.

12On capability improvements in AMISOM as a result of training by the U.S.-contracted
Bancroft Group, see Jeffrey Gettleman, Mark Mazzetti, and Eric Schmitt, “U.S. Relies on
Contractors in Somalia Conflict,” New York Times, August 10, 2011, p. A1.
182 Countering Others’ Insurgencies

other capabilities for which large, complex organizations are required.

As our case study of the Philippines suggested, improvements in these
fields made as a result of U.S. assistance can quickly be reversed if the
United States does not continue to provide critical inputs.
More problematic are efforts to improve leadership and person-
nel systems. Many militaries in developing countries are built around
patronage networks, where the military functions to provide important
resources and/or employment to regime loyalists. Moreover, regimes
frequently structure their militaries to minimize the risk of coups
rather than to maximize the military effectiveness of the organization.
Consequently, the quality of leadership is often highly degraded, and
foreign military advisors have considerable difficulty changing these
patterns.13
Organizational and political cultures also play a role. Many
developing-country security forces have been constructed as servants
of the regime, not the population. Alternatively, many militaries refuse
to accept the principle of civilian oversight, believing that they should
be given a free hand in dealing with security threats—a dynamic that
has repeatedly troubled both Pakistan and the Philippines. Whether
the security forces are heavily politicized in the service of a specific
regime or refuse to accept legitimate civilian oversight, the outcome
is often much the same. In either case, the notion that the security
forces should serve the population—including by using violence dis-
criminately, offering protection to vulnerable populations, and pro-
viding civil assistance in humanitarian crises or disasters—is typically
alien. Such attitudes inhibit the adoption of key counterinsurgency
principles.14

13 Byman, 2006, pp. 79–115; Benjamin C. Schwarz, American Counterinsurgency Doc-

trine and El Salvador: The Frustrations of Reform and the Illusions of Nation Building, Santa
Monica, Calif.: RAND Corporation, R-4042-USDP, 1991.
14 El Salvador in the 1980s provides another instructive example. See A. J. Bacevitch, James
D. Hallums, Richard H. White, and Thomas F. Young, American Military Policy in Small
Wars: The Case of El Salvador, Institute for Foreign Policy Analysis, Washington, D.C.:
Pergamon-Brassey’s, 1988; Byman, 2006; Schwarz, 1991.
 Managing Troubled Partnerships 183

For these and similar reasons, tactical-level gains are often

realized among partner-nation forces, but durable operational- and
strategic-level improvements prove elusive. Widespread partnering
between U.S. advisors and partner forces can help to alleviate these
problems, as the case study of the Philippines suggests. But a major
overhaul of an unprofessional force is typically a long-term, resource-
intensive commitment.

Policy Implications
Is it possible to make “bad” counterinsurgents better? The evidence
assembled here and in our case studies suggests that it is extremely
difficult to transform partner nations during the course of a conflict.
Change typically comes slowly, and attempting to use external aid to
leverage reform efforts has a poor overall record of success. Despite
these findings, the literature on development assistance and interna-
tional diplomacy suggests at least three ways in which conditionality
policies are appropriate and, indeed, critical components of an overall
partnership strategy.
First, aid conditionality is necessary to “sterilize” the negative
effects of foreign assistance. Many observers warn that foreign assis-
tance provides regimes with an independent source of income and
patronage, making them less reliant on—and thus less responsive to—
their own populations.15 In this sense, foreign aid is potentially much
like oil or other natural-resource rents that have been shown to have
highly corrosive effects on developing countries’ governance. Unlike
oil, however, aid is typically administered with a number of condi-
tions intended to limit the ability of recipient regimes to divert these
resources for private gain.16 Even if conditionality cannot be used to
reform the recipient regime as a whole, it may at least be used to reduce
the negative effects of the rents that can be derived from aid.

15 See, for instance, Nicolas van de Walle, African Economies and the Politics of Permanent
Crisis, 1979–1999, New York: Cambridge University Press, 2001.
16 Paul Collier, “Is Aid Oil? An Analysis of Whether Africa Can Absorb More Aid,” World
Development, Vol. 34, No. 9, 2006, pp. 1482–1497.
184 Countering Others’ Insurgencies

Second, it should be possible to establish credible redlines for

partner-regime behavior—that is, specified behaviors that would
automatically result in the termination of foreign assistance. As discussed
above, sponsoring states have difficulty convincing counterinsurgent
regimes of the credibility of their threats to terminate aid and
potentially permit the defeat of the partner. If the counterinsurgent
regime’s behavior is so egregious as to be counterproductive, however, a
sponsoring state such as the United States has no incentive to continue
providing aid. Under such conditions, the threat to terminate aid
should be entirely credible. The British government’s threat to terminate
assistance to Sierra Leone had the SLPP abrogated the results of the
2007 elections is a perfect example.
While possible in theory, terminating assistance to a regime that
has long received it is more difficult in practice. The central problem
is that actors in the sponsoring state become personally invested in
the success of the partner nation and the policies adopted toward it,
making it difficult for them to even recognize failed policies, much
less be willing to abandon them. High-level decisionmakers are very
reluctant to cut their losses and withdraw even after overwhelming
evidence of their strategy’s failure has accrued, because of personal
incentives to gamble on even low-probability efforts to rescue a failing
intervention.17 At lower levels of the decisionmaking hierarchy, it can
be similarly difficult to turn off aid spigots due to vested bureaucratic
interests in the perpetuation of assistance programs or simply due to
bureaucratic inertia.18
How, then, can U.S. decisionmakers establish credible redlines
in practice? There is no mechanism that can guarantee success. Incor-
porating explicit and robust discussions of redlines into preinterven-
tion planning, however, is a reasonable step in this direction. When
such redlines have been incorporated into planning documents and are

17 George W. Downs, “The Lessons of Disengagement,” in Ariel E. Levite, Bruce W.

Jentleson, and Larry Berman, eds., Foreign Military Intervention: The Dynamics of Protracted
Conflict, New York: Columbia University Press, 1992.
18On related incentives operating in the development assistance community, see van de
Walle, 2005, especially Chaps. 3 and 4.
 Managing Troubled Partnerships 185

repeatedly assessed through regularly collected metrics, the burden of

proof for extending support to a regime that violates those redlines is
clearly shifted to those who advocate in its favor.19
Finally, under certain circumstances, conditionality can and
should be used to create incentives for the implementation of specific,
often narrowly focused aid programs. The final section of this chap-
ter explores such circumstances. In these cases, conditionality is a tool
used either in extreme situations (when redlines are at risk of being
broken) or as an exception in certain narrowly defined circumstances.

Policy Prescriptions for Managing Troubled Partnerships

The research presented here has indicated that the United States can
easily partner with relatively inclusive, higher-capacity regimes such
as those in the Philippines and Colombia. Particularly where those
regimes lack needed military capabilities, the United States can offer
critical assistance. With regimes that are inclusive but lack capacity,
the United States still has substantial options—albeit at reduced odds
of success. In such cases, the United States should strongly consider
whether informal accommodation would meet minimal U.S. security
goals.
The more-challenging circumstances are those in which the
regime is not inclusive, especially when it lacks both political inclusion
and state capacity. The odds of success are not favorable in such situa-
tions, and the risks of abusive behavior by the regimes are substantial.
If the United States still believes it must partner with such regimes

19 The recently released Presidential Decision Directive on Security Sector Assistance issues
similar guidance for routine security assistance: “[The Department of] State will report on
the development of a set of notional triggers to prompt the United States Government to
consider initiating, restructuring, pausing, terminating, or reinstituting SSA programming
for use by policymakers and planners. Circumstances warranting such an interagency review
may include . . . SSA partners that are pursuing activities counter to U.S. policy, security
interests or United States Government SSA goals; partner change of government or other
landmark changes in political circumstances; or programs that fail to meet short-term aims
(in 2–3 years).” See “Implementation of Security Sector Assistance (SSA) Presidential Deci-
sion Directive (PDD),” Washington, D.C.: The White House, undated, pp. 5–6.
186 Countering Others’ Insurgencies

despite the challenges, there are a number of rules of thumb that can
be used to guide policy.

General Principles
Counterinsurgency is perhaps the most context-dependent activity in
which militaries engage. Consequently, no universal set of policy pre-
scriptions is possible. At least three principles, however, can be dis-
cerned from the case studies in this report and the broader literature on
counterinsurgency and development to guide U.S. partnerships with
problematic partners.
Alignment of interests. Where U.S. and partner-nation interests
fundamentally diverge, there is little hope of a productive partnership
to combat an insurgency within the partner nation. The United States’
leverage and information are insufficient to effectively use conditional-
ity to overcome such gulfs. Examples of fundamental divergences in
interests include cases in which the partner nation does not perceive
an insurgent group as a threat (and may even regard it as a partner
for fighting proxy conflicts, as has sometimes been the case in Paki-
stan) and cases in which the United States demands the elimination of
ungoverned spaces over which the partner nation lacks the resources
to assert effective control. Less fundamental differences in goals can
be managed, as the Philippines case demonstrated. Even so, such rela-
tionships entail risk—particularly the risk that the partner nation will
use military capabilities developed with U.S. assistance for purposes
that the United States opposes. In such cases, it is important that the
United States make clear and stand by its redlines.
Pushing on open doors. Conditionality can reasonably be used
to enforce redlines, to press for a limited number of important but nar-
rowly scoped reforms, or to take advantage of specific moments in time
when more wide-ranging political change is possible. In most cases,
U.S. efforts to reform the partner nation should focus on finding areas
of agreement, potentially helping to convene the networks of actors
in the partner nation that can implement changes, and then provid-
ing the necessary resources and technical expertise. U.S. support to
AFP operations in the Sulu Archipelago, as described in Chapter Four,
is one such example of a narrowly targeted, intensive collaboration.
 Managing Troubled Partnerships 187

U.S. assistance helped to transform the GRP strategy from contain-

ment to classic counterinsurgency by providing resources and know-
how, but it was successful largely because the governments in power at
the time were predisposed to such policies. This approach accords with
concepts such as good-enough governance and problem-driven adap-
tation, which are currently popular in the development community.20
At the same time, U.S. decisionmakers should be attentive to oppor-
tunities to press for and support more-extensive change, particularly
in political inclusivity. While state capacity takes decades to improve
dramatically, political inclusivity can change much more rapidly. Cases
of rapid change in political inclusivity are rare; fewer than one-tenth
of all regimes fighting insurgencies experienced large shifts toward
greater political inclusivity (defined here as two or more points on the
Freedom House scale). They do, however, occur, often in response to
prolonged periods of military stalemate in which the regime is look-
ing for a way out. In such circumstances, the United States might play
a significant role in bolstering local actors who are pressing for greater
political inclusivity.
Long duration. Given the lengthy duration of most contempo-
rary insurgencies and the length of time it takes to build state capacity
or institutionalize mechanisms of political inclusion, the United States
should enter into problematic partnerships with the expectation that
they will be long-term relationships, typically with low odds of success
in the short-to-medium term. This point was highlighted by the diver-
gent results of long-term U.S. counternarcotics assistance in Pakistan
and counterinsurgency aid focused on immediate results. Similarly,
the more-successful recent U.S. light-footprint interventions—those
in the Philippines and Colombia—have already lasted more than a

20 See, for instance, Merilee S. Grindle, “Good Enough Governance: Poverty Reduction and
Reform in Developing Countries,” Governance, Vol. 17, No. 4, October 2004, pp. 525–548;
Verena Fritz, Kai Kaiser, and Brian Levy, Problem-Driven Governance and Political Economy
Analysis, Washington, D.C.: World Bank, 2009; Matt Andrews, The Limits of Institutional
Reform in Development: Changing Rules for Realistic Solutions, New York: Cambridge Univer-
sity Press, 2013; and Thomas Carothers and Diane de Gramont, Development Aid Confronts
Politics: The Almost Revolution, Washington, D.C.: Carnegie Endowment for International
Peace, 2013.
188 Countering Others’ Insurgencies

dozen years, and the Philippines case study in this report highlighted
the critical role the United States is likely to play in sustaining AFP
capabilities for years to come. Decisionmakers should carefully weigh
their ability to make such long-term commitments, particularly where
partner nations are problematic. With the memories of the 9/11 attacks
fading and public attention no longer monopolized by the wars in Iraq
and Afghanistan, decisionmakers should expect more critical scrutiny
of these light-footprint engagements. The example of El Salvador in the
1980s is instructive: The U.S. Congress became increasingly critical of
U.S. support to the Salvadoran regime, drastically cutting the Reagan
administration’s requests for aid to its ally despite the fact that the com-
munist insurgency had battled the regime to a draw. Had the Cold
War not ended, the level at which the United States could have main-
tained its support is unclear. Particularly where potential U.S. partners
prove abusive or use violence indiscriminately, sustaining partnerships
long enough to see significant changes in partner counterinsurgency
behavior may be untenable.

Bounded Accommodation
Because only a minority of conflicts now end in outright military vic-
tory and the proportion decided by defeat on the battlefield has been
steadily declining since World War II, the need to reach political
accommodation with at least that portion of the opposition that can be
accommodated without compromising core interests of the counterin-
surgent appears to be increasing. Unfortunately, such accommodation
is difficult for regimes characterized by low levels of political inclusion.
What can the United States do to facilitate such bounded accom-
modation? Depending on the precise circumstances, a number of
instruments are available.
Bolstering the credibility of regime commitments. Democra-
cies sometimes seek to implement accommodative solutions to insur-
gencies but lack the resources necessary to follow through on all of
their promises. In such cases, foreign assistance can be critical in help-
ing the counterinsurgent regime make good on its commitments, as
the case of the Philippines illustrated. In other cases, even autocratic
governments are willing to cut deals that would give rebels important
 Managing Troubled Partnerships 189

political rights and resources. Autocracies, however, find it difficult to

credibly commit to such deals—that is, to convince insurgents that
they will stand by the terms of the agreement even after the rebels dis-
band. The United States and other external actors can play a critical
role in making such commitments more credible. Peacekeeping opera-
tions are the best-known mechanism in such circumstances.21 If the
United States has been an active participant in a counterinsurgency,
it may be unable to play the role of a peacekeeper, but other coun-
tries may well be able to do so. There is also a wide variety of mecha-
nisms besides peacekeeping by which external powers can help to make
regime commitments credible. In Mozambique, for instance, the inter-
national community committed to providing substantial development
assistance, of which the rebels were guaranteed a proportionate share,
but only if the parties to the conflict adhered to the terms of the agreed
peace deal.22 Although this arrangement has had its critics, peace has
endured for two decades in Mozambique.
Promoting local solutions. Counterinsurgency theorists have long
advocated more-inclusive, participatory politics at the local level as a
key element of counterinsurgency.23 Despite support for decentraliza-
tion in much of the development community, there is little evidence
that it reduces conflict or improves the quality of governance as a gen-
eral rule.24 In appropriate circumstances, however, it may be an impor-
tant part of efforts to promote the inclusion of the reconcilable opposi-
tion.25 USAID and other development agencies can play an important

21 See, for instance, Barbara F. Walter, Committing to Peace: The Successful Settlement of
Civil Wars, Princeton, N.J.: Princeton University Press, 2001, and Virginia Page Fortna, Does
Peacekeeping Work? Shaping Belligerents’ Choices After Civil War, Princeton, N.J.: Princeton
University Press, 2008.
22 Philippe le Billon, “Buying Peace or Fueling War: The Role of Corruption in Armed Con-
flicts,” Journal of International Development, Vol. 15, 2003, p. 420.
23 Galula, 2006 [1964], especially pp. 89–92.
24See, for instance, Paul Jackson and Zoe Scott, Local Government in Post-Conflict Environ-
ments, Oslo: United Nations Development Program, 2007.
25 Identifying the precise circumstances in which local governance is likely to exert positive
effects is a complicated exercise. One such attempt is Kristin M. Bakke and Erik Wibbels,
190 Countering Others’ Insurgencies

role in improving the capacity of local governments to offer meaning-

ful services to the local population, as the case study of the Philippines
suggests. Such assistance, however, must be carefully tailored to local
realities. Even within Mindanao, for instance, the appropriate sequenc-
ing of assistance varied from one locality to the next.
Seizing rare opportunities for more-dramatic change. Prog-
ress toward greater democracy is normally heavily contested, occur-
ring only when incumbents are unable to resist demands for greater
political inclusion. A prolonged military stalemate in a civil war may
provide such an opening. In this situation, political reforms are less
about alleviating popular grievances and winning the general popula-
tion’s hearts and minds in order to defeat insurgents.26 Rather, they are
about providing a framework in which reconcilable opposition leaders
come to believe they can participate with minimal fear of persecution
or marginalization. Such processes typically do not look like Western
conceptions of democracy—they may go on for at least many years,
often two decades or more.27 But they can provide a means for ending
violent conflicts, and they have often done so. Such fragile political
systems require support in their initiation phase, and they require sub-
sequent buttressing if they are not to collapse. The United States and
other international partner nations can help in both of these phases.

Regulated Violence
Accepting that the United States usually has only limited influence in
light-footprint partnerships and that success is a long-term prospect
implies an important corollary: The United States must undertake pol-

“Diversity, Disparity, and Civil Conflict in Federal States,” World Politics, Vol. 59, October
2006, pp. 1–50.
26 For a review of the social science literature on this subject, see Stephen Watts, “Political
Dilemmas of Stabilization and Reconstruction,” in Paul K. Davis, ed., Dilemmas of Interven-
tion: Social Science for Stabilization and Reconstruction, Santa Monica, Calif.: RAND Corpo-
ration, MG-1119-OSD, 2011.
27 As pointed out in Chapter Four, for example, there were many limitations on the Philip-
pines’ democratic transition. Democratically elected administrations did not consistently
adopt the counterinsurgency practices expected of more politically inclusive regimes for
nearly two decades after the country’s initial democratic transition.
 Managing Troubled Partnerships 191

icies to mitigate the risk that its partners will use violence indiscrimi-
nately or otherwise engage in large-scale abuses while receiving U.S.
military aid. Clearly such a policy is in line with American values, but
it is also in line with U.S. interests. As discussed in Chapter Two, the
discriminate use of violence is an important predictor of counterinsur-
gent success. It is also often necessary in order for the United States
and other Western democracies to sustain their assistance to a counter-
insurgent regime and sometimes to developing-nation partners more
generally. The example of increasing Congressional opposition to aid
for El Salvador in the 1980s has already been mentioned. Even more
starkly, French popular reaction against the French government’s sup-
port of the Habyarimana regime in Rwanda after the 1994 genocide
was a significant factor in France’s retrenchment from its military com-
mitments in sub-Saharan Africa.28
How can the United States mitigate the risk of abuses by partner
nations’ security forces? The communication of clear and credible red-
lines, as discussed above, is probably the most important thing that the
United States can do. Beyond such threats to terminate aid, the United
States can also undertake a number of actions in the course of its secu-
rity assistance.
Improving civil-military relations. As the case of the Philip-
pines shows, even relatively promising regimes can have their counter-
insurgency strategies undone by poor civil-military relations. Strong
democratic oversight is typically associated with more-professional,
higher-quality militaries, while many autocratic regimes suffer from
problems of leadership and capability traceable to a heavy reliance on
patronage networks, the need to prevent coups, and similar patholo-
gies.29 Militaries also often prove unwilling to accommodate even the
reconcilable opposition, insisting on outright military defeat of rebels,

28 Xavier Renou, “A New French Policy for Africa?” Journal of Contemporary African Studies,
Vol. 20, No. 1, 2002, pp. 5–27.
29 See, for instance, Byman, 2006; Herbert M. Howe, Ambiguous Order: Military Forces in
African States, Boulder, Colo.: Lynne Rienner Publishers, 2001.
192 Countering Others’ Insurgencies

no matter how distant a prospect such a victory might be.30 Although

the United States is committed to the principle of civilian oversight
and incorporates this commitment into much of its security assistance,
senior U.S. defense officials have recognized that the United States
can and should do a better job of reinforcing it.31 Military capacity-
building in the absence of appropriate civil-military relations often has
consequences the opposite of those intended, as both the Marcos-era
Philippines and many examples from Pakistan have demonstrated.
Quality over quantity. There have been frequent debates over
the relative priority of the quantity and quality of security services in a
counterinsurgency context, with force ratios that measure the quantita-
tive adequacy of security forces having recently come under consider-
able criticism.32 Both the quantitative evidence and the case studies in
this report suggest the primacy of quality, at least in the vast majority
of cases. In both the Marcos-era Philippines and Pakistani operations
in FATA and Balochistan, larger numbers of troops were more fre-
quently associated with more indiscriminate violence and ultimately
counterproductive outcomes. This finding highlights the importance
of imparting the necessary doctrine, leadership, discipline, and (where
appropriate) technology to manageable numbers of partner-nation
forces and then sustaining these qualitative improvements. Wherever
possible, partner-nation units receiving such assistance should be
closely paired with U.S. forces to ensure that the United States has
visibility into how its assistance is being used. In both Pakistan and
the Philippines, partner units with the necessary capabilities for
counterinsurgency—often elite SOF—were more likely to engage in
carefully targeted operations than were other units. At the same time,
the Pakistan case suggests that even with such qualitative improve-
ments, abuses are common without robust democratic oversight.

30 See, for instance, David D. Laitin and Drew A. Harker, “Military Rule and National
Secession: Nigeria and Ethiopia,” in Morris Janowitz, ed., Civil-Military Relations: Regional
Perspectives, Beverly Hills, Calif.: Sage Publications, 1981.
31See, for instance, Gen. Carter Ham, testimony before the U.S. Senate Armed Services
Committee, March 7, 2013.
32 For critical examinations, see Goode, 2009–2010, pp. 45–57; Thiel, 2011.
 Managing Troubled Partnerships 193

Intercommunal integration. As Pakistani operations in Balo-

chistan in particular demonstrated, security forces that do not include
members of the same ethnic or religious affiliation as the population in
which they are operating are at particularly high risk of abusive behav-
ior.33 The United States, therefore, should work with partner regimes to
improve the representativeness of their security services. Incorporating
personnel from disaffected populations during the course of intensive
fighting, however, risks subversion within the government’s security
forces, as witnessed in the recent “green-on-blue” violence in Afghani-
stan. The ideal time to integrate personnel from different communities,
therefore, is before fighting erupts, when there is still opportunity to
develop unit cohesion and esprit de corps within the security services.
This point, in turn, leads to our final recommendation: an emphasis on
“phase-zero” reform activities.
Phase zero. Unfortunately, all of the prescriptions offered in this
section are long-term propositions. Moreover, they are all substantially
more difficult to implement during ongoing fighting than in peace-
time. Consequently, when the United States partners with more auto-
cratic regimes, it is accepting a high level of risk that its assistance
will be abused by partner-nation forces. These warnings suggest that
security-sector reform efforts such as those discussed above should be a
central element of U.S. phase-zero, or peacetime, engagement strategy,
not a peripheral concern or an issue to which significant resources are
devoted only once a crisis erupts.34 Although the United States is com-
mitted to such reforms in principle, in practice its procedures for iden-
tifying and mitigating the risks are still nascent.35 Building capability

33On this point, see also Howe, 2001; Andre Le Sage, “Africa’s Irregular Security Threats:
Challenges for U.S. Engagement,” Strategic Forum, No. 255, May 2010, p. 8.
34 See also Herbst, 2004.
35 The United States’ commitment to security-sector reform is expressed in the U.S. govern-
ment’s policy document, “Security Sector Reform,” February 2009. Some elements of risk-
identification and risk-mitigation strategy are embedded in the recent “Presidential Decision
Directive on Security Sector Assistance” (undated), but the details remain to be worked out.
194 Countering Others’ Insurgencies

without building effective oversight and disciplining mechanisms risks

failure at the strategic level, even if partner forces become much more
effective at operational and tactical levels.
APPENDIX

Coding Notes

Dependent Variables
Acceptable Decisive Outcome
An acceptable decisive outcome is one in which the government either
wins an outright military victory or is able to negotiate a formal peace
agreement following a military stalemate. Our coding of military vic-
tory was derived from a version of the dataset in Connable and Libicki
(2010), updated with data from Watts et al. (2012). Coding of nego-
tiated settlements was derived from the data in Doyle and Sambanis
(2006), updated through 2008 primarily using histories from the
Europa World Plus online database, supplemented with additional
source material where necessary (e.g., Watts et al., 2012).

Mass Killings
Our indicator of indiscriminate violence was derived from data on
intentional government killings of civilians assembled by Ulfelder and
Valentino (2008).1
Ulfelder and Valentino’s instances of mass killings do not match
up perfectly with the civil-wars list defined by Doyle and Sambanis
that we used for our sample of cases. Consequently, we had to choose

1 Data from Ulfelder and Valentino were updated through 2010 through email correspon-
dence with the authors. Specifically, two new cases of mass killings were added: Sri Lanka
(from January 15, 2009, to May 19, 2009), with between 3,000 and 20,000 intentional
civilian deaths, and Côte d’Ivoire (from December 1, 2010, to April 11, 2011), with between
1,500 and 3,000 deaths.

195
196 Countering Others’ Insurgencies

between Ulfelder and Valentino’s year ranges for mass killings and
Doyle and Sambanis’ civil-war lengths to calculate an average value for
mass killings per year. We chose Ulfelder and Valentino’s year ranges
because mass killings associated with a civil war could have taken place
prior to the start date coded using Doyle and Sambanis’ procedures.
We then matched the civil wars in Doyle and Sambanis and the mass-
killing incidents in Ulfelder and Valentino based on approximate year
and event type. In cases where a mass-killings record spanned multiple
civil wars, we assumed that the mass killings were even across the civil
wars. As discussed in Chapter Two, Ulfelder and Valentino also rely
on a lower threshold of violence to code a case as an instance of mass
killings. We selected as our break point an average of 2,000 or more
intentional killings of civilians for each year of the civil war, with the
number of such killings determined by the average of Ulfelder and
Valentino’s high-end and low-end estimates. Ulfelder and Valentino’s
data were collected for the period from 1945 to 2006; if a civil war
was either ongoing or continued after 2006, we used 2006 as the latest
year in the mass-killings calculation.2 In cases where a mass-killings
incident was not reported for a country during the period of a civil war
(as determined using start and end dates from Doyle and Sambanis’
dataset), the civil war was coded as a case of no indiscriminate violence
by government forces.

Independent Variables
Government Effectiveness
The government effectiveness variable from the World Bank’s WGI
provided one of our two indicators of state capacity. According to the
WGI source document, “Government Effectiveness captures percep-
tions of the quality of public services, the quality of the civil service
and the degree of its independence from political pressures, the quality
of policy formulation and implementation, and the credibility of the

2 The exception is Sri Lanka, 2003–2009, for which we had updated data.
 Coding Notes 197

government’s commitment to such policies.”3 The indicator is reported

by country for 1996, 1998, 2000, and 2002–2011. Unfortunately, with
so many years of missing data, we had to use the average of the existing
years of data to score each country’s government effectiveness rather
than using only those years in which the country was affected by insur-
gency. The break point between high and low government effectiveness
was the median value from all cases in our sample.

State Reach
Our second indicator of state capacity is Holtermann’s state-reach vari-
able.4 Data for this variable are available from 1989 to 2006 and are
on a scale of 0 to 1. As with government effectiveness, because of miss-
ing data, we had to use each country’s average values for all years from
1989 to 2006 rather than using only those years in which the country
was affected by insurgency. Countries with average state-reach scores
greater than or equal to the index’s midpoint of 0.5 are considered
to have high state reach, and countries with average state-reach scores
lower than 0.5 are considered to have low state reach.

Freedom House Scores

We derived our indicator of political inclusion from Freedom House’s
Freedom in the World dataset. Freedom House reports two scores per
year: The first is a political rights score, on a scale of 1 through 7,
with 1 being the most free and 7 being the least free. The second is
a civil liberties score, also on a scale of 1 through 7, with 1 being the
most free and 7 being the least free. We took the average of these two
scores for each year, then averaged a country’s combined score for all
the years of a conflict.5 If a Freedom House score was not reported for
a country year, that year did not influence the average. We selected as

3 Kaufmann, Kraay, and Mastruzzi, 2009, p. 6.

4 Holtermann, 2012.
5 Some past Freedom in the World reports cover unconventional date ranges, particularly
in the 1980s. In these cases, we used the year most closely associated with the year covered.
For example, Freedom House data for the period from August 2, 1982, to November 1983
were used to calculate a country’s score for 1983.
198 Countering Others’ Insurgencies

our break point the midpoint of the Freedom House scale; thus, if the
average Freedom House score over the course of a civil war was less
than or equal to 4, the country was considered to possess high political
inclusivity; if the average Freedom House score was greater than 4, the
country was considered to possess low political inclusivity.

Force-to-Force Ratios
Force-to-force ratios were calculated as the ratio of government to insur-
gent forces. For data on insurgent forces, we used the Peace Research
Institute Oslo (PRIO) Non-State Actors dataset, Version 3.3.6 We cal-
culated the number of rebels per year using the dataset’s best estimate.
If there were multiple insurgent groups in a given country correspond-
ing to the same civil war in the Doyle and Sambanis dataset, we used
the sum of all insurgent groups. Since the PRIO dataset uses conflict
years as its unit of analysis, whereas our study examines conflicts as a
whole, we used the average number of insurgents over the length of
the conflict. Data on government forces were derived from the World
Bank’s WDI. This variable is entitled Armed Forces Personnel, Total
and consists of active duty military personnel, including paramilitary
forces if the training, organization, equipment, and control suggest
they may be used to support or replace regular military forces. The
original source of the data is the International Institute for Strategic
Studies’ The Military Balance. The indicator begins in 1989. If the indi-
cator was not available for a certain country year, it was omitted from
the average. To determine the extent of government military superior-
ity, we took the ratio of the government troop estimate over this period
of time to the rebel estimate, with a 10:1 ratio used as the break point
between countries scoring high and low.

Force-to-Population Ratios
Force-to-population ratios were calculated as the ratio of a country’s
government military forces to its inhabitants. As with force-to-force

6 For details, see David E. Cunningham, Kristian Skrede Gleditsch, and Idean Salehyan,
“Codebook for the Non–State Actor Data,” Peace Research Institute Oslo Centre for the
Study of Civil War, version 3.3, January 24, 2012.
 Coding Notes 199

ratios, data on government forces were derived from the World Bank’s
WDI. This variable is entitled Armed Forces Personnel, Total and con-
sists of active duty military personnel, including paramilitary forces if
the training, organization, equipment, and control suggest they may be
used to support or replace regular military forces. The original source of
the data is the International Institute for Strategic Studies’ The Military
Balance. The indicator begins in 1989. If the indicator was not avail-
able for a certain country year, it was omitted from the average. Data
on countries’ total populations were also found in the WDI. Using
this ratio, we created two binary variables based on two different break
points: average force-to-population ratios of 5:1,000 and 10:1,000. We
used both break points in separate cross-tabulations of the relationship
between military superiority and the outcomes of interest.
Bibliography

“2006: Who’s Who in Balochistan,” Dawn, May 28, 2011.

Abbas, Hassan, “Musharraf Contends with the Pashtun Element in the Pakistani
Army,” Terrorism Focus, Vol. 3, No. 42, November 1, 2006.
———, “Transforming Pakistan’s Frontier Corps,” Terrorism Monitor, Vol. 5,
No. 6, March 30, 2007.
———, “A Profile of the Tehrik-i-Taliban,” CTC Sentinel, Vol. 1, No. 2, January
2008, pp. 1–4.
Abbas, Hassan, and Shehzad H. Qazi, “Rebellion, Development and Security in
Pakistan’s Tribal Areas,” CTC Sentinel, Vol. 6, No. 6, June 2013.
Abbasi, Ansar, “37 Percent Baloch Favor Independence: UK Survey,” The News
International, August 13, 2012. As of December 9, 2013:
http://www.thenews.com.pk/
Todays-News-13-16755-37pc-Baloch-favour-independence-UK-survey
Abouzeid, Rania, “Taliban Gone, Pakistan Area Still Wants Islamic Justice,” Time,
April 25, 2010.
Abuza, Zachary, Balik-Terrorism: The Return of the Abu Sayyaf, Carlisle, Pa.:
Strategic Studies Institute, September 2005.
Acop, Dencio Severo, “The Expanded Nontraditional Role of the AFP:
A Reassessment,” Prism, Vol. 3, No. 2, March 2012, pp. 110–112.
Afzal, Shemrez Nauman, “North Waziristan: The Death Trap,” Lahore, Pakistan:
Spearhead Research, December 29, 2010. As of August 21, 2013:
http://www.spearheadresearch.org/pages/documents/North_Waziristan_-_The_
Death_Trap.pdf
Ahmad, Istiaq, “Promising Partnership? American Af-Pak Strategy and Pakistan’s
Counterterrorism Response,” in Usama Butt and N. Elahi, eds., Pakistan’s
Quagmire: Security, Strategy, and the Future of the Islamic-Nuclear Nation, London:
Continuum International, 2010, pp. 66–67.

201
202 Countering Others’ Insurgencies

Ahmad, Mavish, “Balochistan: Middle Class Rebellion,” Dawn, June 5, 2012.

Ahmad, Naveed, “Trouble in Pakistan’s Energy-Rich Balochistan,” Zurich,
Switzerland: International Relations and Security Network, January 30, 2006. As
of December 9, 2013:
http://www.isn.ethz.ch/Digital-Library/Articles/Detail/?id=107976
Ahmed, Akbar S., Resistance and Control in Pakistan, rev. ed., London: Routledge,
2004.
Ahmed, Feroz, Ethnicity and Politics in Pakistan, Karachi, Pakistan: Oxford
University Press, 1998, pp. 253–254.
Ahmed, Siraj, “Police Reform in Balochistan,” in Hassan Abbas, ed., Stabilizing
Pakistan Through Police Reform, Report by the Independent Commission on
Pakistan Police Reform, New York: Asia Society, July 2012, pp. 113–119.
Akbar, Malik Siraj, Refined Dimensions of Baloch Nationalist Movement,
Bloomington, Ind.: Xlibris Corporation, 2011.
———, “The Futility of Senate Debate,” The Baloch Hal, August 5, 2012.
———, “The Ifs and Buts of Negotiating with Insurgents in Balochistan,” The
Baloch Hal, July 31, 2013. As of August 21, 2013:
http://thebalochhal.com/2013/08/01/
editorial-the-ifs-and-buts-of-negotiating-with-insurgents-in-balochistan/
Albrecht, Peter, and Lars Buur, “An Uneasy Marriage: Non-State Actors and Police
Reform,” Policing and Society, Vol. 19, No. 4, 2009, pp. 390–405.
Alemika, Etannibi E. O., and Innocent C. Chukwuma, “The Poor and Informal
Policing in Nigeria,” Lagos, Nigeria: Center for Law Enforcement Education
(CLEEN), 2004.
Ali, Kalbe, “U.S. Withdraws ROZs Offer,” Dawn, November 10, 2012.
Amin, Tahir, Ethno-National Movements of Pakistan, Islamabad, Pakistan:
Institute of Policy Studies, 1988, p. 174.
Anderson, Perry, “Why Partition?” London Review of Books, Vol. 34, No. 14,
July 19, 2012.
Andrews, Matt, The Limits of Institutional Reform in Development: Changing Rules
for Realistic Solutions, New York: Cambridge University Press, 2013.
Arbetman-Rabinowitz, Marina, et al., “Pakistan Has Lowest Tax to GDP Ratio in
the World,” The Nation, February 18, 2013.
Arnoldy, Ben, “Why Pakistanis Would Reject $7.5 Billion in U.S. Aid,” Christian
Science Monitor, October 14, 2009.
Ashraf, Tariq Mahmud, “The Pakistan Frontier Corps in the War on Terrorism—
Part I,” Terrorism Monitor, Vol. 6, No. 15, July 25, 2008.
 Bibliography 203

Averch, Harvey, and John Koehler, The Huk Rebellion in the Philippines:
Quantitative Approaches, Santa Monica, Calif.: RAND Corporation, RM-6254,
1970. As of December 9, 2013:
http://www.rand.org/pubs/research_memoranda/RM6254.html
Aziz, Sartaj, “The Economic Cost of Extremism,” in Usama Butt and N. Elahi,
eds., Pakistan’s Quagmire: Security, Strategy, and the Future of the Islamic-Nuclear
Nation, London: Continuum International, 2010, pp. 75–92.
Bacevitch, A. J., James D. Hallums, Richard H. White, and Thomas F. Young,
American Military Policy in Small Wars: The Case of El Salvador, Institute for
Foreign Policy Analysis, Washington, D.C.: Pergamon-Brassey’s, 1988.
Baker, Bruce, “Policing Post-Conflict Societies: Helping Out the State,” Policing
and Society, Vol. 19, No. 4, 2009, pp. 329–332.
Bakke, Kristin M., and Erik Wibbels, “Diversity, Disparity, and Civil Conflict in
Federal States,” World Politics, Vol. 59, October 2006, pp. 1–50.
Banlaoi, Rommel C., “The Sources of the Abu Sayyaf’s Resilience in the Southern
Philippines,” CTC Sentinel, Vol. 3, No. 5, May 2010, pp. 17–19.
Bansal, Alok, “Factors Leading to Insurgency in Balochistan,” Small Wars &
Insurgencies, Vol. 19, No. 2, June 2008, pp. 182–200.
Bashir, Riaz Anwar, Shahzad Ahmad Qasmi, Muhammad Yasin, and Sahibzada
Nasir Mansoor, “Pattern of Combat Casualties in War Against Terror Among
Soldiers Wearing Body Armor at CMH Peshawar,” Pakistan Armed Forces Medical
Journal, No. 2, June 2012.
Bates, Robert H., Prosperity and Violence: The Political Economy of Development,
New York: W. W. Norton, 2001.
Bello, Walden, “Edging Toward the Quagmire: The United States and the
Philippine Crisis,” World Policy Journal, Vol. 3, No. 1, Winter 1985–1986, p. 31.
Bennett, Huw, “The Reluctant Pupil? Britain’s Army and Learning in Counter-
Insurgency,” Royal United Services Institute, 2009. As of November 5, 2009:
http://www.rusi.org/analysis/commentary/ref%3AC4AD22F8DF284C/#.
UulGg2RDtOE
Berman, Eli, Jacob N. Shapiro, and Joseph H. Felter, “Can Hearts and Minds
Be Bought? The Economics of Counterinsurgency in Iraq,” Journal of Political
Economy, Vol. 119, No. 4, August 2011, pp. 766–819.
Birdsall, Nancy, “Risking the Dog for the Tail,” Newsweek, February 28, 2011.
Birdsall, Nancy, Wren Elhai, and Molly Kinder, Beyond Bullets and Bombs: Fixing
the U.S. Approach to Development in Pakistan, Washington, D.C.: Center for
Global Development, June 2011, p. 14.
Blaufarb, Douglas S., The Counterinsurgency Era: U.S. Doctrine and Performance,
1950 to the Present, New York: The Free Press, 1977.
204 Countering Others’ Insurgencies

Blitz, James, James Lamont, and Farhan Bokhari, “Swat Outlook ‘Pretty Bleak’
for Pakistan,” Financial Times, May 13, 2009.
Blunt, Peter, and Mark Turner, “Decentralization, Democracy, and Development
in a Post-Conflict Society: Commune Councils in Cambodia,” Public
Administration and Development, Vol. 25, 2005, pp. 84–85.
Bonner, Raymond, Waltzing with a Dictator: The Marcoses and the Making of
Foreign Policy, New York: New York Times Books, 1987, pp. 204–225.
Boot, Max, and Richard Bennett, “Treading Softly in the Philippines,” The Weekly
Standard, Vol. 14, No. 16, January 5–12, 2009.
Briscoe, Carolyn H., “Balikatan Exercise Spearheaded ARSOF Operations in the
Philippines,” Special Warfare, Vol. 17, No. 1, September 2004, p. 16.
Brulliard, Nicolas, “Pakistan Textile Exports: Call for Wider Lifting of U.S.
Tariffs Intensifies,” Washington Post, December 24, 2010.
Byman, Daniel L., “Friends Like These: Counterinsurgency and the War on
Terrorism,” International Security, Vol. 31, No. 2, Fall 2006, pp. 79–115.
Caouette, Dominique, Persevering Revolutionaries: Armed Struggle in the 21st
Century, Exploring the Revolution of the Communist Party of the Philippines,
dissertation, Ithaca, N.Y.: Cornell University, 2004, pp. 483, 488.
Caringal, Sharon Advincula, “The Impact of U.S. Military Assistance on the
Communist and Secessionist Conundrum in the Philippines,” International
Journal of Arts and Sciences, Vol. 3, No. 15, 2010, p. 437.
Carothers, Thomas, In the Name of Democracy: U.S. Policy Toward Latin America
in the Reagan Years, Berkeley, Calif.: University of California Press, 1993, esp.
Chap. 1.
Carothers, Thomas, and Diane de Gramont, Development Aid Confronts Politics:
The Almost Revolution, Washington, D.C.: Carnegie Endowment for International
Peace, 2013.
Center for Research and Security Studies, Balochistan’s Maze of Violence,
Islamabad, Pakistan, 2011.
Chapman, William, Inside the Philippine Revolution: The New People’s Army and Its
Struggle for Power, 1st ed., New York: W. W. Norton, 1987.
Cheema, Pervaiz Iqbal, “Challenges Facing a Counter-Militant Campaign in
Pakistan’s FATA,” National Bureau of Asian Research, NBR Analysis, Vol. 19,
No. 3, August 2008, pp. 21–29.
Clapham, Christopher, “The Global-Local Politics of State Decay,” in Robert I.
Rotberg, ed., When States Fail: Causes and Consequences, Princeton, N.J.: Princeton
University Press, 2004.
 Bibliography 205

Cloughley, Brian, War, Coups, and Terror: Pakistan’s Army in Years of Turmoil,
New York: Skyhorse, 2008.
———, “Insurrection, Terrorism and the Pakistan Army,” in Usama Butt and
N. Elahi, eds., Pakistan’s Quagmire: Security, Strategy, and the Future of the Islamic-
Nuclear Nation, London: Continuum International, 2010, pp. 93–122.
Cohen, Stephen P., The Idea of Pakistan, Washington, D.C.: Brookings Institution
Press, 2004.
Cohen, Craig, and Derek Chollet, “When $10 Billion Is Not Enough: Rethinking
U.S. Strategy Toward Pakistan,” Washington Quarterly, Vol. 30, No. 2, 2007,
p. 12.
Cohen, Eliot, “To COIN or Not?” transcript from a roundtable discussion
co-sponsored by Foreign Policy and the RAND Corporation, March 18, 2013.
Cohen, Stephen, The Pakistan Army, Berkeley, Calif.: University of California
Press, 1984.
Colaresi, Michael, and Sabine C. Carey, “To Kill or to Protect: Security Forces,
Domestic Institutions, and Genocide,” Journal of Conflict Resolution, Vol. 52,
No. 1, February 2008, pp. 39–67.
Collier, Paul, “Is Aid Oil? An Analysis of Whether Africa Can Absorb More Aid,”
World Development, Vol. 34, No. 9, 2006, pp. 1482–1497.
Collier, Paul, V. L. Elliott, Havard Hegre, Anke Hoeffler, Marta Reynal-Querol,
and Nicholas Sambanis, Breaking the Conflict Trap: Civil War and Policy
Development, Washington, D.C.: The World Bank and Oxford University Press,
2003, Chap. 1.
Connable, Ben, and Martin C. Libicki, How Insurgencies End, Santa Monica,
Calif.: RAND Corporation, MG-965-MCIA, 2010, pp. 186–187. As of December
9, 2013:
http://www.rand.org/pubs/monographs/MG965.html
Correlates of War Project, “National Material Capabilities Data (v. 4.0),” June
2010. As of July 1, 2013:
http://www.correlatesofwar.org/COW2%20Data/Capabilities/NMC_v4_0.csv
Cunningham, David E., Kristian Skrede Gleditsch, and Idean Salehyan,
“Codebook for the Non–State Actor Data,” Peace Research Institute Oslo Centre
for the Study of Civil War, version 3.3, January 24, 2012. As of August 14, 2013:
http://privatewww.essex.ac.uk/~ksg/data/NSAEX_codebook.pdf
Dahl, Robert A., Polyarchy: Participation and Opposition, New Haven, Conn.: Yale
University Press, 1972.
Davenport, Christian, State Repression and the Domestic Democratic Peace, New
York: Cambridge University Press, 2007a.
206 Countering Others’ Insurgencies

———, “State Repression and Political Order,” Annual Review of Political Science,
Vol. 10, 2007b, pp. 1–23.
De Mesquita, Bruce Bueno, George W. Downs, Alastair Smith, and Feryal Marie
Cherif, “Thinking Inside the Box: A Closer Look at Democracy and Human
Rights,” International Studies Quarterly, Vol. 49, 2005, pp. 439–457.
De Mesquita, Bruce Bueno, and Alastair Smith, “Leader Survival, Revolutions,
and the Nature of Government Finance,” American Journal of Political Science,
Vol. 54, No. 4, October 2010, pp. 936–950.
Devesa, Roy T., An Assessment of the Philippine Counterinsurgency Operational
Methodology, master’s thesis, Fort Leavenworth, Kans.: U.S. Army Command and
General Staff College, 2005, pp. 40–41.
D’Exelle, Ben, “Excluded Again: Village Politics at the Aid Interface,” Journal of
Development Studies, Vol. 45, No. 9, October 2009, pp. 1453–1471.
Dixon, Paul, “‘Hearts and Minds’? British Counter-Insurgency from Malaya to
Iraq,” Journal of Strategic Studies, Vol. 32, No. 3, 2009, pp. 353–381.
Doran, Michael Scott, “Somebody Else’s Civil War,” Foreign Affairs, Vol. 81,
No. 1, January-February 2002, pp. 22–42.
Downes, Alexander B., “Desperate Times, Desperate Measures: The Causes of
Civilian Victimization in War,” International Security, Vol. 30, No. 4, Spring
2006, pp. 152–195.
Downs, George W., “The Lessons of Disengagement,” in Ariel E. Levite,
Bruce W. Jentleson, and Larry Berman, eds., Foreign Military Intervention: The
Dynamics of Protracted Conflict, New York: Columbia University Press, 1992.
Doyle, Michael W., and Nicholas Sambanis, Making War and Building Peace,
Princeton, N.J.: Princeton University Press, 2006.
Dreher, Axel, Jan-Egbert Sturm, and James Raymond Vreeland, “Development
Aid and International Politics: Does Membership on the UN Security Council
Influence World Bank Decisions?” Journal of Development Economics, Vol. 88,
No. 1, 2009, pp. 1–18.
———, “Global Horse Trading: IMF Loans for Votes in the United Nations
Security Council,” European Economic Review, Vol. 53, No. 7, 2009, pp. 742–757.
Duparcq, Emmanuel, “Swat Rebuilds Year After Pakistan Floods,” Agence France
Presse, July 24, 2011.
Eck, Kristine, and Lisa Hultman, “One-Sided Violence Against Civilians in War:
Insights from New Fatality Data,” Journal of Peace Research, Vol. 44, No. 2, 2007,
pp. 233–246.
 Bibliography 207

Englebert, Pierre, and James Ron, “Primary Commodities and War: Congo-
Brazzaville’s Ambivalent Resource Curse,” Comparative Politics, Vol. 37, No. 1,
October 2004, pp. 61–81.
Englebert, Pierre, and Denis M. Tull, “Postconflict Reconstruction in Africa:
Flawed Ideas About Failed States,” International Security, Vol. 32, No. 4, 2008,
pp. 106–139.
Epstein, Susan B., and K. Alan Kronstadt, Pakistan: U.S. Foreign Assistance,
Washington, D.C.: Congressional Research Service, R41856, October 4, 2012.
Esperon, Hermogenes C., Jr., “Perspective from the Military,” Digest: A Forum for
Security and Defense Issues, Office of Strategic and Special Studies, Armed Forces of
the Philippines, 2006, p. 5.
Fair, Christine, “Pakistan Loses Swat to Local Taliban,” Terrorism Focus, Vol. 4,
No. 37, November 13, 2007.
Fair, C. Christine, and Seth G. Jones, “Pakistan’s War Within,” Survival, Vol. 51,
No. 6, December 2009–January 2010.
Fanthorpe, Richard, “On the Limits of Liberal Peace: Chiefs and Democratic
Decentralization in Post-War Sierra Leone,” African Affairs, Vol. 105, No. 418,
2006, p. 40.
Farris, Stuart, Joint Special Operations Task Force–Philippines, Fort Leavenworth,
Kans.: U.S. Army Command and General Staff College, School of Advanced
Military Studies, 2009, pp. 22–23.
Fearon, James D., “Domestic Political Audiences and the Escalation of
International Disputes,” American Political Science Review, Vol. 88, No. 3,
September 1994, pp. 577–592.
———, “Rationalist Explanations for War,” International Organization, Vol. 49,
No. 3, Summer 1995, pp. 379–414.
———, “Why Do Some Civil Wars Last So Much Longer than Others?” Journal
of Peace Research, Vol. 41, No. 3, 2004, pp. 275–301.
Fearon, James D., and David D. Laitin, “Ethnicity, Insurgency, and Civil War,”
American Political Science Review, Vol. 97, No. 1, February 2003.
Feldman, Daniel F., “Testimony Before the House Foreign Affairs Committee,
Subcommittee on the Middle East and South Asia,” transcript, April 5, 2011. As
of August 21, 2013:
http://archives.republicans.foreignaffairs.house.gov/112/fel040511.pdf
Felter, Joseph H., Taking Guns to a Knife Fight: Effective Military Support to COIN,
Carlisle, Pa.: U.S. Army War College, 2008, pp. 9–10.
Fielding, David, and Anja Shortland, “The Dynamics of Terror During the
Peruvian Civil War,” Journal of Peace Research, Vol. 49, No. 6, 2012, pp. 847–862.
208 Countering Others’ Insurgencies

Fontaine, Roger W., “The Philippines: After Aquino,” Asian Affairs: An American
Review, Vol. 19, No. 3, Fall 1992, pp. 171–172.
Fortna, Virginia Page, “Scraps of Paper? Agreements and the Durability of Peace,”
International Organization, Vol. 57, No. 2, Spring 2003, pp. 337–372.
———, Does Peacekeeping Work? Shaping Belligerents’ Choices After Civil War,
Princeton, N.J.: Princeton University Press, 2008.
Freedom House, Freedom in the World, Washington, D.C., 2013. As of June 25,
2013:
http://www.freedomhouse.org/report-types/freedom-world
Freier, Nathan, “The New Theology: Building Partner Capacity,” Small Wars
Journal, May 27, 2010.
Fritz, Verena Fritz, Kai Kaiser, and Brian Levy, Problem-Driven Governance and
Political Economy Analysis, Washington, D.C.: World Bank, 2009.
Fund for Peace, “Failed States Index, 2005–2013,” web site. As of August 8, 2013:
http://www.global.fundforpeace.org/index.php
Galula, David, Counterinsurgency Warfare: Theory and Practice, Westport, Conn.:
Praeger Security International, 2006 [1964].
Gandhi, Rohit, “Deradicalization: A Tall Order,” Asia Pacific Defense Forum,
July 1, 2012.
Gelb, Alan, and Caroline Decker, “Direct Payments to Pakistan’s Flood Victims:
A Smart Option for U.S. Assistance,” The Hill, November 8, 2010.
Gelb, Leslie H., and Richard K. Betts, The Irony of Vietnam: The System Worked,
Washington, D.C.: Brookings Institution, 1979.
Gettleman, Jeffrey, Mark Mazzetti, and Eric Schmitt, “U.S. Relies on Contractors
in Somalia Conflict,” New York Times, August 10, 2011, p. A1.
Ghaus, A. F. Aisha, Hafiz A. Pasha, and Rafia Ghaus, “Social Development
Ranking of Districts of Pakistan,” The Pakistan Development Review,” Vol. 35,
No. 4 (Part II), Winter 1996, pp. 593–614.
Gilani, Ijaz, Voice of the People: Public Opinion in Pakistan, 2007–2008, Karachi,
Pakistan: Oxford University Press, 2010.
Gilani Research Foundation, Perceptions on Kerry-Lugar Bill: Gilani Poll/Gallup
Pakistan, Islamabad, October, 14, 2009.
Gishkori, Zahid, “Violence Is on the Rise in Balochistan: Report,” Express Tribune,
January 3, 2013.
Global Firepower, “Countries Ranked by Military Strength (2013),” GFP Power
Index, web site, 2006–2013. As of August 8, 2013:
http://www.globalfirepower.com/countries-listing.asp
 Bibliography 209

Goode, Steven M., “A Historical Basis for Force Requirements in

Counterinsurgency,” Parameters: U.S. Army War College Quarterly, Winter
2009–2010, pp. 45–57.
Gopal, Anand, Mansur Khan Mahsud, and Brian Fishman, “The Taliban in
North Waziristan,” in Peter Bergen and Katherine Tiedemann, eds., Talibanistan:
Negotiating the Borders Between Terror, Politics, and Religion, New York: Oxford
University Press, 2013, pp. 128–163.
Grare, Frederic, Balochistan: The State Versus the Nation, Washington, D.C.:
Carnegie Endowment for International Peace, Carnegie Paper, April 11, 2013.
Green, Matthew, “Hidden War Embodies Pakistan’s Struggle,” Financial Times,
May 25, 2012.
Grindle, Merilee S., “Good Enough Governance: Poverty Reduction and
Reform in Developing Countries,” Governance, Vol. 17, No. 4, October 2004,
pp. 525–548.
“Growth Projections: Washington to Help Revive Tourism in Swat, Olson,”
Express Tribune, March 15, 2013.
Gul, Imtiaz, The Most Dangerous Place: Pakistan’s Lawless Frontier, New York:
Viking, 2011.
Haider, Ejaz, “Intelligent Intelligence,” Daily Times, July 26, 2007.
Haines, David R., “COIN in the Real World,” Parameters, Winter 2008–2009,
pp. 50–51.
Hall, Rosalie Árcala, “Boots on Unstable Ground: Democratic Governance of the
Armed Forces Under Post 9/11 US–Philippine Military Relations,” Asia-Pacific
Social Science Review, Vol. 10, No. 2, 2010, pp. 30–34.
Ham, General Carter, testimony before the U.S. Senate Armed Services
Committee, March 7, 2013. As of August 9, 2013:
http://www.armed-services.senate.gov/hearings/
oversight-us-africa-command-and-us-transportation-command
Hammes, T. X., “Counterinsurgency: Not a Strategy, but a Necessary Capability,”
Joint Force Quarterly, No. 65, 2nd Quarter, 2012, pp. 48–52.
Hammond, Jeremy R. “The Situation in Swat: An Interview with Shahid R.
Siddiqi,” Foreign Policy Journal, July 9, 2009.
Haqqani, Husain, Pakistan: Between Mosque and Military, Washington, D.C.:
Carnegie Endowment for International Peace, 2005.
Harff, Barbara, “No Lessons Learned from the Holocaust? Assessing Risks of
Genocide and Political Mass Murder Since 1955,” American Political Science
Review, Vol. 97, No. 1, February 2003, pp. 57–73.
210 Countering Others’ Insurgencies

Hartzell, Caroline, Matthew Hoddie, and Donald Rothchild, “Stabilizing the

Peace After Civil War: An Investigation of Some Key Variables,” International
Organization, Vol. 55, No. 1, Winter 2001, pp. 183–208.
Hashim, Asad, “Pakistan and the U.S.: A Too-Close Embrace?” Al Jazeera.com,
June 3, 2011.
Hassan, Ahmad, “Operation in Malakand, Waziristan Is No Solution,”
Dawn.com, November 8, 2009. As of January 29, 2014:
http://www.dawn.com/news/501577/
operation-in-malakand-waziristan-is-no-solution
Hastings, Jonathan P., and Krishnamurti Mortela, The Strategy-Legitimacy
Paradigm: Getting It Right in the Philippines, master’s thesis, Monterey, Calif.:
Naval Postgraduate School, December 2008, pp. 59–72.
Hedman, Eva-Lotta E., The Philippines: Conflict and Internal Displacement in
Mindanao and the Sulu Archipelago, Writenet Report commissioned by the United
Nations High Commissioner for Refugees, Emergency and Technical Support
Service, March 2009, pp. 2–5. As of August 21, 2013:
http://www.observatori.org/paises/pais_63/documentos/UNHCR.pdf
Herbst, Jeffrey, States and Power in Africa: Comparative Lessons in Authority and
Control, Princeton, N.J.: Princeton University Press, 2000.
———, “African Militaries and Rebellion: The Political Economy of Threat and
Combat Effectiveness,” Journal of Peace Research, Vol. 41, No. 3, May 2004,
pp. 361–362.
Hernandez, Carolina G., “Institutional Responses to Armed Conflict: The
Armed Forces of the Philippines,” background paper, submitted to the Human
Development Network Foundation, Inc., for the Philippine Human Development
Report, 2005, pp. 14–15.
Holtermann, Helge, “Explaining the Development–Civil War Relationship,”
Conflict Management and Peace Science, Vol. 29, No. 1, 2012, p. 63.
Howe, Herbert M., Ambiguous Order: Military Forces in African States, Boulder,
Colo.: Lynne Rienner Publishers, 2001.
Human Rights Commission of Pakistan, Conflict in Balochistan, Report of
Human Rights Commission of Pakistan Fact-Finding Mission, December 2005–
January 2006.
Huntington, Samuel P., Political Order in Changing Societies, New Haven, Conn.:
Yale University Press, 1968.
Hussain, Jamal, “PAF in the Counterinsurgency Role in the Tribal Areas of
Pakistan, Defence Journal, November 2012.
Hussain, Javed, “The Hornet’s Nest,” The News International, January 19, 2011.
 Bibliography 211

Hussain, Syed Rifaat, “War Against Terrorism, Pakistani Perspective,” Islamabad

Policy Research Institute, IPRI Journal, Winter 2004. As of August 21, 2013:
http://ipripak.org/journal/winter2004/waragainst.shtml
Hussain, Syed Talat, “Profiling Our North Waziristan Policy,” Express Tribune,
October 29, 2012.
Hussain, Zahid, “It’s War Now,” Newsweek, January 15, 2006.
———, The Scorpion’s Tail: The Relentless Rise of Islamic Militants in Pakistan—
and How It Threatens America, New York: Free Press, 2010, p. 90.
Hussain, Zahid, and Matthew Rosenberg, “Pakistani Peace Deal Gives New Clout
to Taliban Rebels,” Wall Street Journal, April 14, 2009.
ICG—See International Crisis Group.
“Implementation of Security Sector Assistance (SSA) Presidential Decision
Directive (PDD),” Washington, D.C.: The White House, undated, pp. 5–6.
“Impressions of Waziristan, January 25, 2005,” declassified cable, National
Security Archive, Electronic Briefing Book No. 325, September 13, 2010.
International Crisis Group, Pakistan: The Worsening Conflict in Balochstan, Asia
Report No. 119, Islamabad/Brussels, September 14, 2006a.
———, Pakistan’s Tribal Areas: Appeasing the Militants, Asia Report No. 125,
Islamabad/Brussels, December 11, 2006b.
———, Pakistan: The Forgotten Conflict in Balochistan, Asia Briefing No. 69,
Islamabad/Brussels, October 22, 2007, p. 6.
———, Counterinsurgency vs. Counterterrorism in Mindanao, Asia Report
No. 152, Islamabad/Brussels, May 14, 2008, pp. 20–26.
———, Pakistan: The Militant Jihadi Challenge, Asia Report No. 164, Islamabad/
Brussels, March 13, 2009.
———, The Communist Insurgency in the Philippines: Tactics and Talks, Asia
Report No. 202, Islamabad/Brussels, February 2011.
———, Aid and Conflict in Pakistan, Asia Report No. 227, Islamabad/Brussels,
June 27, 2012.
———, Pakistan Countering Militancy in PATA, Asia Report No. 242,
Islamabad/Brussels, January 15, 2013a.
———, The Philippines: Dismantling Rebel Groups, Asia Report No. 248,
Islamabad/Brussels, June 19, 2013b, pp. 1–3.
International Dialogue on Peacebuilding and Statebuilding, “New Deal for
Engagement in Fragile States.” As of August 9, 2013:
http://www.pbsbdialogue.org/documentupload/49151944.pdf
212 Countering Others’ Insurgencies

International Institute for Strategic Studies, The Military Balance, London:

Routledge, 2013.
Inter-Services Public Relations (Pakistan), “COAS Keynote Address—National
Seminar Deradicalization at Mingora Swat,” ISPR No. PR157/2011, July 6, 2011.
Iqbal, Anwar, “U.S. Note Dilutes Some Conditions in Kerry-Lugar Bill,” Dawn,
October 14, 2009.
———, “No Operation in North Waziristan, U.S. Told,” Dawn, March 31, 2010.
Isaac, Jeffrey C., et al., “The New U.S. Army/Marine Corps Counterinsurgency
Field Manual as Political Science and Political Praxis,” Perspectives on Politics,
Vol. 6, No. 2, June 2008, pp. 347–360.
ISPR—See Inter-Services Public Relations (Pakistan).
Jackson, Colin, Defeat in Victory: Organizational Learning Dysfunction in
Counterinsurgency, unpublished dissertation, Cambridge, Mass.: Massachusetts
Institute of Technology, June 2008.
Jackson, Paul, and Zoe Scott, Local Government in Post-Conflict Environments,
Oslo: United Nations Development Program, 2007.
Jan, Delwar, “Malakand Division on Road to Civilian Control,” The News
International, August 8, 2012, p. 98.
Jan, Reza, “Daily Tracker: Pakistani Military Operations in Orakzai,” Critical
Threats Project, American Enterprise Institute, Washington, D.C., May 13, 2010a.
As of August 21, 2013:
http://www.criticalthreats.org/pakistan/
daily-tracker-pakistani-military-operations-orakzai-mar-30-2010
———, “Paradise Regained: Swat One Year On,” Critical Threats Project,
American Enterprise Institute, Washington, D.C., May 25, 2010b. As of
August 21, 2013:
http://www.criticalthreats.org/pakistan/
paradise-regained-swat-one-year-may-25-2010
———, “Trickling Home to South Waziristan,” Critical Threats Project,
American Enterprise Institute, Washington, D.C., December 13, 2010c. As of
August 21, 2013:
http://www.criticalthreats.org/pakistan/
trickling-home-south-waziristan-December-10-2010
Jan, Reza, and Sam Worby, “Limited Goals, Limited Gains: The Pakistan Army’s
Operation in Kurram,” American Enterprise Institute Critical Threats Project, web
site, September 6, 2011.
Jane’s World Insurgency and Terrorism, “JTIC Country Briefing—Philippines,”
February 2012.
 Bibliography 213

———, “Abu Sayyaf Group,” June 7, 2013a, p. 7.

———, “Moro Islamic Liberation Front,” June 7, 2013b, p. 3.
Joint and Coalition Operational Analysis, Decade of War, Vol. I: Enduring Lessons
from the Past Decade of Operations, Suffolk, Va.: Joint Staff J-7, June 15, 2012.
Jones, Gregg R., Red Revolution: Inside the Philippine Guerrilla Movement, Boulder,
Colo.: Westview Press, 1989.
Jones, Seth G., and C. Christine Fair, Counterinsurgency in Pakistan, Santa
Monica, Calif.: RAND Corporation, MG-982-RC, 2010. As of August 21, 2013:
http://www.rand.org/pubs/monographs/MG982.html
“Judicial Reform in Swat,” Dawn, March 25, 2010.
Kagan, Frederick W., Reza Jan, and Charlie Szrom, “The War in Waziristan:
Analysis of Operation Rah-e-Nijat—Phase 1 Analysis,” Critical Threats Project,
American Enterprise Institute, November 18, 2009. As of August 21, 2013:
http://www.criticalthreats.org/pakistan/
war-waziristan-operation-rah-e-nijat-phase-1-analysis
Kalyvas, Stathis N., and Laia Balcells, “International System and Technologies
of Rebellion: How the End of the Cold War Shaped Internal Conflict,” American
Political Science Review, Vol. 104, No. 3, August 2010, pp. 415–429.
Kandeh, Jimmy D., “Rogue Incumbents, Donor Assistance, and Sierra Leone’s
Second Post-Conflict Elections of 2007,” Journal of Modern African Studies,
Vol. 46, No. 4, 2008.
Kaplan, Robert D., “Imperial Grunts,” The Atlantic, October 2005.
Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi, Governance Matters VIII:
Aggregate and Individual Governance Indicators, 1996–2008, Washington, D.C.:
World Bank, Development Research Group, Policy Research Working Paper 4978,
June 2009, p. 6.
Keller, Arthur, “Propaganda and Peace Deals: The Taliban’s Information War in
Pakistan,” CTC Sentinel, Vol. 1, No. 8, July 2008.
Kerkvliet, Benedict J., The Huk Rebellion: A Study of Peasant Revolt in the
Philippines, 2nd ed., Lanham, Md.: Rowman & Littlefield Publishers, 2002.
Kessler, Richard J., “Marcos and the Americans,” Foreign Policy, Vol. 63, 1986,
pp. 40–57.
Khaliq, Fazal, “Nizam-e-Adl Regulation: Top Sharia Court Set Up in Swat,”
Express Tribune, January 19, 2011a.
———, “Speedy Justice: Mobile Courts for Swat Announced,” Express Tribune,
November 26, 2011b.
214 Countering Others’ Insurgencies

———, “Terrorists Behind Swat Unrest to Face Justice,” Express Tribune,

November 27, 2011c.
———, “Timely Justice: Community Body Gives Quick Solutions for Small
Disputes,” Express Tribune, March 13, 2013.
Khan, Afzal, “Revolt in Pakistan’s NWFP: A Profile of Maulana Fazlullah of
Swat,” Terrorism Focus, Vol. 4, No. 38, November 20, 2007.
Khan, Amina, “FATA: Voice of the Unheard—Path-Dependency and Why
History Matters,” Strategic Studies (Islamabad), Vol. 31, Nos. 1 and 2, Spring
2011, pp. 40–74.
Khan, Ayaz Ahmed, “Balochistan Situation,” The Nation, April 10, 2005.
Khan, Ijaz, “Challenges Facing Development in Pakistan’s FATA,” NBR Analysis,
Vol. 19, No. 3, August 2008, p. 20.
Khan, Ismail, “Militant Leader Threatens to Attack Forces,” Dawn, July 30, 2005.
Khan, M. Ilyas, “Taleban Spread Wings in Pakistan,” BBC News, March 5, 2007.
As of August 21, 2013:
http://news.bbc.co.uk/2/hi/south_asia/6409089.stm
———, “Balochistan ‘Troops Numbers Cut,’” BBC News, May 14, 2008. As of
August 21, 2013:
http://news.bbc.co.uk/2/hi/south_asia/7400790.stm
Khan, Riaz, “Pakistan Shuts Down Three Radio Stations,” Associated Press,
March 21, 2008.
Khan, Rina Saeed, “Swat Valley NGO Finds a Solution to the Assault on
Education,” The Guardian, May 10, 2011.
Khan, Sher Baz, “Punjab’s Dominance in Army Being Reduced: ISPR,” Dawn,
September 14, 2007.
Khan, Sohail, “Disband Agencies ‘Death Squads,’ Stop [Military] Action in
Balochistan: SC,” The International News, September 12, 2012.
Khan, Tahir, “Rival Groups Launch Operations Against Pakistan Taliban in
Mohmand,” Express Tribune, June 26, 2013.
Khattak, Daud Khan, The Battle for Pakistan: Militancy and Conflict in the Swat
Valley, Washington, D.C.: New America Foundation, Counterterrorism Strategy
Initiative Policy Paper, April 2010.
———, “‘Clearing’ Kurram,” Foreign Policy’s AfPak Channel, August 25, 2011.
———, “Reviewing Pakistan’s Peace Deals with the Taliban,” CTC Sentinel,
September 2012.
Khurram Shahzad, “What Ails Balochistan?” The Nation, February 14, 2005.
 Bibliography 215

Kilcullen, David J., The Accidental Guerrilla: Fighting Small Wars in the Midst of a
Big One, New York: Oxford University Press, 2009.
———, Terrain, Tribes, and Terrorists: Pakistan, 2006–2008, Washington,
D.C.: Brookings Institution, Counterinsurgency and Pakistan Paper Series No. 3,
September 10, 2009, pp. 10–12.
King, Gary, Robert O. Keohane, and Sidney Verba, Designing Social Inquiry:
Scientific Inference in Qualitative Research, Princeton, N.J.: Princeton University
Press, 1994.
Kline, William E., The Fall of Marcos: A Problem in U.S. Foreign Policymaking,
Washington, D.C.: Institute for the Study of Diplomacy, Georgetown University
School of Foreign Service, 1992, pp. 18–19.
Kocher, Matthew Adam, Thomas B. Pepinsky, and Stathis N. Kalyvas, “Aerial
Bombing and Counterinsurgency in the Vietnam War,” American Journal of
Political Science, Vol. 55, No. 2, April 2011, pp. 201–218.
Krasner, Stephen, “Talking Tough to Pakistan,” Foreign Affairs, Vol. 91, No. 1,
January–February 2012, pp. 87–96.
Kundi, Mansoor Akbar, “Insurgency Factors in Balochistan,” The Nation,
October 20, 2004.
Kurtzleben, Danielle, “Pakistan: U.S. to Aid Civilians Fleeing Embattled Swat
Valley,” Inter Press Service, May 19, 2009.
Lacina, Bethany, and Nils Petter Gleditsch, “Monitoring Trends in Global
Combat: A New Dataset of Battle Deaths,” European Journal of Population,
Vol. 21, Nos. 2–3, 2005, pp. 145–166.
Laitin, David D., and Drew A. Harker, “Military Rule and National Secession:
Nigeria and Ethiopia,” in Morris Janowitz, ed., Civil-Military Relations: Regional
Perspectives, Beverly Hills, Calif.: Sage Publications, 1981.
Lalwani, Sameer, Pakistani Military Capabilities for a Counterinsurgency Campaign:
A Net Assessment, Washington, D.C.: New America Foundation, Counterterrorism
Strategy Initiative Publication, September 2009, pp. 41–46.
———, “Pakistan’s Counterinsurgency Strategy,” in Peter Bergen and Katherine
Tiedemann, eds., Talibanistan: Negotiating the Borders Between Terror, Politics, and
Religion, New York: Oxford University Press, 2013, pp. 202–228.
Lambert, Geoffrey, Larry Lewis, and Sarah Sewall, “Operation Enduring
Freedom–Philippines: Civilian Harm and the Indirect Approach,” Prism,
Vol. 3, No. 4, September 2012, pp. 117–135.
Le Billon, Philippe, “Buying Peace or Fueling War: The Role of Corruption in
Armed Conflicts,” Journal of International Development, Vol. 15, 2003, p. 420.
216 Countering Others’ Insurgencies

Leites, Nathan, and Charles Wolf, Jr., Rebellion and Authority, Chicago: Markham
Publishing Company, 1970.
Le Sage, Andre, “Africa’s Irregular Security Threats: Challenges for U.S.
Engagement,” Strategic Forum, No. 255, May 2010, p. 8.
Levi, Margaret, Of Rule and Revenue, Berkeley, Calif.: University of California
Press, 1988.
Levite, Ariel E., Bruce W. Jentleson, and Larry Berman, eds., Foreign Military
Intervention: The Dynamics of Protracted Conflict, New York: Columbia University
Press, 1992.
Lieven, Anatol, Pakistan: A Hard Country, London: Allen Lane, 2011.
Lum, Thomas, The Republic of the Philippines and U.S. Interests, Washington, D.C.:
Congressional Research Service, Library of Congress, April 2012, p. 14.
Lyall, Jason, “Does Indiscriminate Violence Incite Insurgent Attacks? Evidence
from Chechnya,” Journal of Conflict Resolution, Vol. 53, No. 3, June 2009,
pp. 331–362.
———, “Are Coethnics More Effective Counterinsurgents? Evidence from
the Second Chechen War,” American Political Science Review, Vol. 104, No. 1,
February 2010a, pp. 1–20.
———, “Do Democracies Make Inferior Counterinsurgents? Reassessing
Democracy’s Impact on War Outcomes and Duration,” International
Organization, Vol. 64, Winter 2010b, pp. 167–92.
Maher, Aijaz, “Pakistan’s Army Steps Up Radio Wars,” BBC News, August 14,
2012. As of August 22, 2013:
http://www.bbc.co.uk/news/world-asia-18766713
Mahsud, Mansur Khan, “The Taliban in South Waziristan,” in Peter Bergen and
Katherine Tiedemann, eds., Talibanistan: Negotiating the Borders Between Terror,
Politics, and Religion, New York: Oxford University Press, 2013, pp. 164–201.
Markey, Daniel S., Securing Pakistan’s Tribal Belt, Washington, D.C.: Council on
Foreign Relations, Council Special Report No. 36, August 2008.
———, No Exit from Pakistan: America’s Tortured Relationship with Islamabad,
New York: Cambridge University Press, 2013, pp. 113, 127.
Mason, T. David, and Dale A. Krane, “The Political Economy of Death Squads:
Toward a Theory of the Impact of State-Sanctioned Terror,” International Studies
Quarterly, Vol. 33, No. 2, June 1989, pp. 175–198.
Matthew, Jennie, “Report Unmasks Tax Evasion Among Pakistan Leaders,”
Agence France Presse, December 12, 2012. As of August 22, 2013:
http://www.google.com/hostednews/afp/article/ALeqM5hrQ4phi5cnJlb9g0_0pQ
YLGiDZwQ?docId=CNG.1f5230b87c34eb7ef95150ba96d4cfb3.131
 Bibliography 217

Maxwell, David S., “Operation Enduring Freedom–Philippines: What Would Sun

Tzu Say?” Military Review, Vol. 84, No. 3, 2004, pp. 20–23.
McGirk, Tim, “Code of the Frontier,” Time, January 30, 2005.
McKelvey, Tara, “A Return to Hell in Swat,” Foreign Policy, March 2, 2011.
Merom, Gil, How Democracies Lose Small Wars, New York: Cambridge University
Press, 2003.
Metz, Steven, The Army’s Strategic Role, Carlisle Barracks, Pa.: Strategic Studies
Institute, Army War College, 2009.
Meyerle, Jerry, Unconventional Warfare and Counterinsurgency in Pakistan: A Brief
History, Alexandria, Va.: Center for Naval Analyses, November 2012, p. 30.
Miller, Paul D., “How to Exercise U.S. Leverage Over Pakistan,” Washington
Quarterly, Vol. 35, No. 4, Fall 2012.
Mold, Andrew, Policy Ownership and Aid Conditionality in the Light of the
Financial Crisis: A Critical Review, Paris: Organization for Economic Cooperation
and Development, 2009.
Morales, Ricardo C., “Perpetual Wars: The Philippine Insurgencies,” unpublished
master’s thesis, Monterey, Calif.: Naval Postgraduate School, 1993.
Mullen, Admiral Mike, transcript of speech Delivered to Washington Institute for
Near East Policy, January 7, 2010.
Mullick, Haider Ali Hussein, “Lions and Jackals: Pakistan’s Emerging
Counterinsurgency Strategy,” Foreign Affairs (online), July 15, 2009.
———, “Holding Pakistan: The Second Phase of Pakistan’s Counterinsurgency
Operations,” Foreign Affairs (online), March 24, 2010.
———, “Recalibrating U.S.-Pakistan Relations,” The Washington Quarterly,
Vol. 35, No. 3, Summer 2012, pp. 93–107.
Murphy, Dan, “A Quiet Waiver for Pakistan from the Obama Administration,”
Christian Science Monitor, October 15, 2012.
Murphy, William S., Review of USAID’s Internally Displaced Persons Program in
Pakistan, Washington, D.C.: Office of Inspector General, Review Report
No. 5-391-10-001-S, June 28, 2010.
Nangiana, Umer, “NA Question Hour: Pakistan Received $3.2B Under Kerry-
Lugar-Berman Act,” Express Tribune, March 7, 2013.
Nasr, Vali, The Dispensable Nation: American Foreign Policy in Retreat, New York:
Doubleday, 2013, p. 69.
National Intelligence Council, The Terrorist Threat to the Homeland, National
Intelligence Estimate, Washington, D.C., July 2007.
218 Countering Others’ Insurgencies

Nawaz, Amna, and Waj S. Khan, “A Rare Glimpse Inside Pakistan’s Ground Zero
for Terrorists,” NBC News, March 4, 2013. As of August 22, 2013:
http://worldnews.nbcnews.com/_news/2013/03/04/17177391-a-rare-glimpse-
inside-pakistans-ground-zero-for-terrorists?lite
Nawaz, Shuja, Crossed Swords: Pakistan, Its Army, and the Wars Within, Karachi,
Pakistan: Oxford University Press, 2008, p. 570.
———, FATA—A Most Dangerous Place: Meeting the Challenge of Militancy and
Terror in the Federally Administered Tribal Areas of Pakistan, Washington, D.C.:
Center for Strategic and International Studies Press, January 2009a.
———, “Pakistan’s Summer of Chaos,” Foreign Policy, June 17, 2009b. As of
August 22, 2013:
http://www.foreignpolicy.com/articles/2009/06/16/pakistans_summer_of_chaos
———, Learning by Doing: The Pakistan Army’s Experience with
Counterinsurgency, Washington, D.C.: Atlantic Council, February 2011.
“New People’s Army,” START Terrorist Organization Profiles database, web site,
University of Maryland, College Park, Md., undated. As of July 15, 2013:
http://www.start.umd.edu/start/data_collections/tops/terrorist_organization_
profile.asp?id=203
Nissenbaum, Dion, “In Former Taliban Sanctuary, an Eerie Silence Takes Over,”
Wall Street Journal, January 26, 2013.
Nixon, Richard, “Address to the Nation on the War in Vietnam,” November 3,
1969, P-691101.
“Not Allies: Eight Dead as TTP, Haqqani Network Clash in North Waziristan,”
Express Tribune, April 19, 2012.
OECD Development Assistance Committee, Handbook on Security Sector Reform,
Paris: OECD, 2007.
Office of the Deputy Chief for Intelligence, J2, Armed Forces of the Philippines,
AFP-Philippine National Police Year-End Order of Battle Validation Report, Quezon
City, Philippines: AFP General Headquarters, January 2004, Annex IA.
Oppel, Richard A., Jr., and Pir Zubair Shah, “In Pakistan, Radio Amplifies Terror
of Taliban,” New York Times, January 24, 2009.
“PAF Conducted 5,500 Bombing Runs in FATA Since 2008,” Express Tribune,
November 14, 2011.
Page, Jeremy, “Britain to Train Pakistan’s Frontier Corps Troops in Baluchistan,”
The Times, October 9, 2009.
“Pak Troops Seize Swat Peak, Shut ‘Mullah Radio,’” Times of India, November 28,
2007.
 Bibliography 219

“Pakistan and the Taliban: A Real Offensive or a Phony War?” The Economist,
April 30, 2009.
“Pakistan Army Strategy in Question After Attacks,” Dawn.com, October 22,
2012.
“Pakistan Has Lowest Tax to GDP Ratio in the World,” The Nation, February 18,
2013. As of August 21, 2013:
http://www.nation.com.pk/pakistan-news-newspaper-daily-english-online/
business/18-Feb-2013/pakistan-has-lowest-tax-to-gdp-ratio-in-world
“Pakistani Authorities Close Three Radio Stations for Pro-Taliban Broadcasts,”
Associated Press, November 1, 2009.
“Pashto-Language TV Station [Khyber TV] Interviews Pakistan’s President
Musharraf,” BBC Monitoring South Asia, May 21, 2006.
Paul, Christopher, Colin P. Clarke, and Beth Grill, Victory Has a Thousand Fathers:
Sources of Success in Counterinsurgency, Santa Monica, Calif: RAND Corporation,
MG-964-OSD, 2010. As of December 9, 2013:
http://www.rand.org/pubs/monographs/MG964.html
Paul, Christopher, Colin P. Clarke, Beth Grill, and Molly Dunigan, Paths
to Victory: Lessons from Modern Insurgencies, Santa Monica, Calif.: RAND
Corporation, RR-291/1-OSD, 2013. As of December 9, 2013:
http://www.rand.org/pubs/research_reports/RR291z1.html
Paul, T. V., ed., South Asia’s Weak States: Understanding the Regional Insecurity
Predicament, Stanford, Calif.: Stanford University Press, 2010.
Peceny, Mark, “Two Paths to the Promotion of Democracy During U.S. Military
Interventions,” International Studies Quarterly, Vol. 39, 1995, pp. 371–401.
Pena, Leonardo I., Finding the Missing Link to a Successful Philippine
Counterinsurgency Strategy, master’s thesis, Monterey, Calif.: Naval Postgraduate
School, 2007, pp. 43–44.
“Performance of the Parliamentary Committee on National Security, March 16,
2012 – March 15, 2013,” Pakistan Institute of Legislative Development and
Transparency, March 2013, p. 13.
Perlez, Jane, “U.S. Aid Plan for Pakistan Is Foundering,” New York Times,
May 1, 2011.
Perlez, Jane, and Helene Cooper, “Signaling Tension, Pakistan Shuts NATO
Route,” New York Times, September 30, 2010.
Petraeus, David H., “Hearing to Receive Testimony on the Situation in
Afghanistan,” transcript, testimony before the Senate Armed Services Committee,
Washington, D.C., March 15, 2011.
220 Countering Others’ Insurgencies

Pew Research Center, “On Eve of Elections, a Dismal Public Mood in Pakistan,”
Pew Global Attitudes Project, Washington, D.C., web site, May 7,
2013. As of August 22, 2013: http://www.pewglobal.org/2013/05/07/
on-eve-of-elections-a-dismal-public-mood-in-pakistan/
Philippines National Statistics Office, Census 2000: Philippines, Population and
Housing Characteristics, Manila: Republic of the Philippines, National Statistics
Office, 2003.
Posen, Barry R., The Sources of Military Doctrine: France, Britain, and Germany
Between the World Wars, Ithaca, N.Y.: Cornell University Press, 1984.
Pritchett, Lant, Michael Woolcock, and Matt Andrews, Capability Traps? The
Mechanisms of Persistent Implementation Failure, Washington, D.C.: Center for
Global Development, Working Paper 234, December 2010.
Quimpo, Nathan Gilbert, The U.S. and the Southern Philippines’ Quagmire,
Murdoch, Western Australia: Asia Research Centre, Murdoch University, July
2007, p. 4.
Quinlivan, James T., “Force Requirements in Stability Operations,” Parameters:
U.S. Army War College Quarterly, Vol. 25, No. 4, Winter 1995-1996, pp. 59–69.
Rabasa, Angel, Steven Boraz, Peter Chalk, Kim Cragin, Theodore W. Karasik,
Jennifer D. P. Moroney, Kevin A. O’Brien, and John E. Peters, Ungoverned
Territories: Understanding and Reducing Terrorism Risks, Santa Monica, Calif:
RAND Corporation, MG-561-AF, 2007. As of December 9, 2013:
http://www.rand.org/pubs/monographs/MG561.html
Rana, Muhammad Amir, “Swat Deradicalization Model: Prospects for
Rehabilitating Militants,” Conflict and Peace Studies, Vol. 4, No. 2, April–June
2011.
Rashid, Ahmed, Descent into Chaos, London: Penguin Books, 2008.
———, “Pakistan’s Continued Failure to Adopt a Counterinsurgency Strategy,”
CTC Sentinel, Vol. 2, No. 3, March 2009.
———, Pakistan on the Brink: The Future of America, Pakistan, and Afghanistan,
New York: Viking, 2012.
Rasool, Kamran, “Pakistan’s Perspective on the ‘War on Terror,’” Military
Technology, Vol. 32, No. 11, November 2008, p. 17.
Ratner, Ely, “Reaping What You Sow: Democratic Transitions and Foreign Policy
Realignment,” Journal of Conflict Resolution, Vol. 53, No. 3, June 2009,
pp. 390–418.
Rehman, Ata Ur, “Responsibilities of Pakistan Army in LIC Environment,” in
Pakistan Army Green Book 2002: Low Intensity Conflict, Rawalpindi, Pakistan:
Pakistan Army, 2002, p. 203.
 Bibliography 221

Renou, Xavier, “A New French Policy for Africa?” Journal of Contemporary African
Studies, Vol. 20, No. 1, 2002, pp. 5–27.
Republic of the Philippines Commission on Population, “Population Statistics,”
web site, 2013. As of July 5, 2013:
http://www.popcom.gov.ph/popstat.php
Ressa, Maria, Seeds of Terror: An Eyewitness Account of Al-Qaeda’s Newest Center of
Operations in Southeast Asia, New York: Simon and Schuster, 2004.
Ricks, Thomas E., “More Salvadors, Fewer Vietnams,” The Best Defense, February
19, 2013.
Rizvi, Hasan Askari, “Understanding the Insurgency,” Daily Times, October 5,
2008.
Robinson, Linda, “The Future of Special Operations,” Foreign Affairs, Vol. 91,
No. 6, November-December 2012.
Rocamora, Joel, “Discontent in the Philippines,” World Policy Journal, Vol. 8,
No. 4, Fall 1991, pp. 636–637.
Rotberg, Robert I., “The Failure and Collapse of Nation-States: Breakdown,
Prevention, and Repair,” in Robert I. Rotberg, ed., When States Fail: Causes and
Consequences, Princeton, N.J.: Princeton University Press, 2004.
Rummel, R. J., “Democracy, Power, Genocide, and Mass Murder,” Journal of
Conflict Resolution, Vol. 39, No. 1, March 1995, pp. 3–26.
Saleem, Ishrat, What Pakistan’s Democratic Future Holds for the United States,
Washington, D.C.: Middle East Institute, May 9, 2013. As of August 8, 2013:
http://www.mei.edu/content/
what-pakistan%E2%80%99s-democratic-future-holds-united-states
Santana, Rebecca, “Pakistan Army Tries to Win Over Local Population in War-
Torn Tribal Region,” Associated Press, April 10, 2013.
Santos, Paz Verdades M., “The Communist Front: Protracted People’s War and
Counter-insurgency in the Philippines (Overview),” in Soliman M. Santos, Jr., and
Paz Verdades M. Santos, eds., Primed and Purposeful: Armed Groups and Human
Security Efforts in the Philippines, Geneva: Small Arms Survey, 2005,
pp. 17–32.
Santos, Soliman, “Evolution of the Armed Conflict on the Communist Front,” in
Philippine Human Development Report 2005: Peace, Human Security and Human
Development in the Philippines, 2nd ed., Quezon City: Human Development
Network Foundation, 2005, pp. 84–89. As of August 21, 2013:
http://www.hdn.org.ph/wp-content/uploads/2005_PHDR/2005%20Evolution_
Communist_Conflict.pdf
222 Countering Others’ Insurgencies

Sayah, Reza, “Can the Return of Justice Halt the Taliban?” CNN, November 14,
2009. As of August 22, 2013:
http://www.cnn.com/2009/WORLD/asiapcf/11/14/pakistan.swat.justice/
Scanzillo, Thomas M., and Edward M. Lopacienski, “Influence Warfare: How
the Joint Special Operations Task Force–Philippines Maximizes Effects to Create
Long-Term Stability,” Center for the Study of Irregular Warfare and Armed
Groups, February 2011.
Schmitt, Eric, and Jane Perlez, “U.S. Unit Secretly in Pakistan Lends Ally
Support,” New York Times, February 23, 2009.
Schmitt, Eric, and Thom Shankar, Counterstrike: The Untold Story of America’s
Secret Campaign Against Al Qaeda, New York: Henry Holt, 2011, pp. 198–200.
Schofield, Carey, Inside the Pakistan Army, London: Biteback Publishing, 2011.
Schultz, Kenneth A., Democracy and Coercive Diplomacy, New York: Cambridge
University Press, 2001.
Schwarz, Benjamin C., American Counterinsurgency Doctrine and El Salvador:
The Frustrations of Reform and the Illusions of Nation Building, Santa Monica,
Calif.: RAND Corporation, R-4042-USDP, 1991. As of December 8, 2013:
http://www.rand.org/pubs/reports/R4042.html
Seymour, Kristen, “De-radicalisation: Psychologists’ War Against Militants,”
Express Tribune, July 17, 2011.
Shafer, D. Michael, Deadly Paradigms: The Failure of U.S. Counterinsurgency Policy,
Princeton, N.J.: Princeton University Press, 1988.
Shah, Saeed, “Why Pakistan Supports the Haqqani Network: Fear for Its Own
Security,” McClatchy News, September 28, 2011.
———, “In Remote Baluchistan, Pakistan Fights Shadowy War,” McClatchy
News, March 29, 2012.
———, “In Historic First, Pakistan’s Civilian Government Completes Its Term
Without a Coup,” McClatchy News, March 16, 2013.
Shah, Tayyab Ali, “Pakistan’s Challenges in Orakzai Agency,” CTC Sentinel,
Vol. 3, No. 7, July 2010.
Shahzad, Khurram, “What Ails Balochistan?” The Nation, February 14, 2005.
Shahzad, Syed Saleem, Inside Al-Qaeda and the Taliban: Beyond Bin Laden and
9/11, London: Pluto Press, 2011.
Shane, Scott, “Insurgents Share a Name, but Pursue Different Goals,” The New
York Times, October 23, 2009.
Sheehan, Michael A., Crush the Cell: How to Defeat Terrorism Without Defeating
Ourselves, New York: Random House, 2009, pp. 123.
 Bibliography 223

Shinwari, Naveed Ahmad, Understanding FATA: Attitudes Towards Governance,

Religion & Society in Pakistan’s Federally Administered Tribal Areas, Vol. IV,
Islamabad: Community Appraisal and Motivation Program, 2010, p. 57.
Siddiqui, Taha, “Deradicalizing Boys in Pakistan,” Christian Science Monitor,
May 25, 2012.
Singer, J. David, “Reconstructing the Correlates of War Dataset on Material
Capabilities of States, 1816–1985,” International Interactions, Vol. 14, 1987,
pp. 115–132.
SIPRI Arms Transfer Database, web site, 2013. As of April 1, 2013:
http://www.sipri.org/databases/armstransfers
Staniland, Paul, “States, Insurgents, and Wartime Political Orders,” Perspectives on
Politics, Vol. 10, No. 2, June 2012, p. 244.
Stedman, Stephen John, “Spoiler Problems in Peace Processes,” International
Security, Vol. 22, No. 2, Fall 1997, pp. 5–53.
Supapo, Romulo C., U.S.-Philippine Security Relations: Its Implications for the
Global War on Terrorism, Carlisle, Pa.: U.S. Army War College, March 2004,
pp. 7–8.
Talbot, Ian, Pakistan: A Modern History, New York: St. Martin’s Press, 1998.
Tajik, Sohail Habib, “Analysis of Peace Agreements with Militants and Lessons
for the Future,” Conflict and Peace Studies, Vol. 4, No. 1, January–March 2011.
Temple-Raston, Dina, “Pakistan’s Ambitious Program to Re-educate Militants,”
NPR, April 1, 2013.
Thiel, Joshua, “COIN Manpower Ratios: Debunking the 10 to 1 Ratio and
Surges,” Small Wars Journal, January 15, 2011.
Thompson, Sir Robert, Defeating Communist Insurgency: The Lessons of Malaya and
Vietnam, Saint Petersburg, Fla.: Hailer Publishing, 2005 [1966].
Toft, Monica Duffy, Securing the Peace: The Durable Settlement of Civil Wars,
Princeton, N.J.: Princeton University Press, 2010.
Toft, Monica Duffy, and Yuri M. Zhukov, “Denial and Punishment in the North
Caucasus: Evaluating the Effectiveness of Coercive Counterinsurgency,” Journal of
Peace Research, Vol. 49, No. 6, 2012, pp. 785–800.
Tolosa, Kathline Anne Sigua, “The Rhetoric and Practice of the Security-
Development Nexus,” OSS Digest, 1st Quarter, 2010, pp. 32–33.
Tomz, Michael, “Domestic Audience Costs in International Relations: An
Experimental Approach,” International Organization, Vol. 61, No. 4, Autumn
2007, pp. 821–840.
224 Countering Others’ Insurgencies

Torres, Ernesto C., Jr., A Success Story of Philippine Counterinsurgency: A Study

of Bohol, master’s thesis, Fort Leavenworth, Kans.: U.S. Army Command and
General Staff College, November 2010, pp. 8–9, 66–68.
Transparency International, “2012 Corruption Perceptions Index,” web site, 2012.
As of August 8, 2013:
http://cpi.transparency.org/cpi2012/
Tucker, David, and Christopher J. Lamb, “Restructuring Special Operations
Forces for Emerging Threats,” Strategic Forum, Institute for National Strategic
Studies, National Defense University, No. 219, January 2006.
Ulfelder, Jay, and Benjamin Valentino, Assessing Risks of State-Sponsored Mass
Killing, Arlington, Va.: Political Instability Task Force, February 2008,
UNDP—See United Nations Development Programme.
United Nations Development Programme, The Rise of the South: Human Progress
in a Diverse World, 2013 Human Development Report, New York, 2013. As of
August 21, 2013:
http://hdr.undp.org/sites/default/files/reports/14/hdr2013_en_complete.pdf
U. S. Agency for International Development, U.S. Department of Defense, and
U.S. Department of State, Security Sector Reform, Washington, D.C., February,
2009.
“U.S. Begins Training Pakistan Frontier Corps,” Jane’s Country Risk Daily Report,
October 27, 2008.
U.S. Department of Defense, Sustaining U.S. Global Leadership: Priorities for 21st
Century Defense, Washington, D.C.: Department of Defense, January 2012.
U.S. Department of the Army, Counterinsurgency, FM 3-24, Washington, D.C.:
Headquarters, Department of the Army, 2006, pp. 1–13.
U. S. Department of the Army, Headquarters, Stability, Army Doctrinal
Publication 3-05, August 2012, p. 10.
Valentino, Benjamin, Paul Huth, and Dylan Balch-Lindsay, “‘Draining the Sea’:
Mass Killing and Guerrilla Warfare,” International Organization, Vol. 58, No. 2,
Spring 2004, pp. 375–407.
Van de Walle, Nicolas, African Economies and the Politics of Permanent Crisis,
1979–1999, New York: Cambridge University Press, 2001.
———, Overcoming Stagnation in Aid-Dependent Countries, Washington, D.C.:
Center for Global Development, 2005.
Walsh, Declan, “Pakistan’s Secret Dirty War,” The Guardian, March 29, 2011.
———, “Court Challenges Put Unusual Spotlight on Pakistani Spy Agency,”
New York Times, February 7, 2012, p. A4. As of August 21, 2013:
http://www.nytimes.com/2012/02/07/world/asia/isi-in-pakistan-faces-court-cases.
html?pagewanted=all&_r=0
 Bibliography 225

Walter, Barbara F., “The Critical Barrier to Civil War Settlement,” International
Organization, Vol. 51, No. 3, Summer 1997.
———, Committing to Peace: The Successful Settlement of Civil Wars, Princeton,
N.J.: Princeton University Press, 2001.
Warraich, Omar, “Refugees Head Home After Army Scatters the Taliban,” The
Independent, August 3, 2009.
———, “Two Cheers for Pakistani Democracy: A Sobering Milestone,” Time,
March 18, 2013. As of August 21, 2013:
http://world.time.com/2013/03/18/
two-cheers-for-pakistani-democracy-a-sobering-milestone/
Watts, Stephen, “Political Dilemmas of Stabilization and Reconstruction,” in
Paul K. Davis, ed., Dilemmas of Intervention: Social Science for Stabilization and
Reconstruction, Santa Monica, Calif.: RAND Corporation, MG-1119-OSD, 2011.
As of December 9, 2013:
http://www.rand.org/pubs/monographs/MG1119.html
Watts, Stephen, Caroline Baxter, Molly Dunigan, and Christopher Rizzi, The Uses
and Limits of Small-Scale Military Interventions, Santa Monica, Calif.: RAND
Corporation, MG-1226-RC, 2012. As of December 9, 2013:
http://www.rand.org/pubs/monographs/MG1226.html
“We Only Receive Back the Bodies,” The Economist, April 7, 2012.
Weinbaum, Marvin G., “Hard Choices in Countering Insurgency and Terrorism
Along Pakistan’s North-West Frontier,” Journal of International Affairs, Vol. 63,
No. 1, Fall/Winter 2009, p. 75.
West, Harry G., and Scott Kloeck-Jenson, “Betwixt and Between: ‘Traditional
Authority’ and Democratic Decentralization in Post-War Mozambique,” African
Affairs, Vol. 98, No. 393, 1999, pp. 455–484.
White, Joshua T., and Shuja Ali Malik, Governance Reforms in Pakistan’s Tribal
Areas: The Long Road to Nowhere? Washington, D.C.: United States Institute of
Peace, Peace Brief No. 135, October 15, 2012, p. 4.
Wilson, Gregory, “Anatomy of a Successful COIN Operation: OEF-Philippines
and the Indirect Approach,” Military Review, November-December 2006,
pp. 2–12.
Wirsing, Robert G., Baloch Nationalism and the Geopolitics of Energy Resources:
The Changing Context of Separatism in Pakistan, Carlisle, Pa.: Strategic Studies
Institute, April 2008.
Wisler, Dominique, and Ihekwoaba D. Onwudiwe, “Community Policing in
Comparison,” Police Quarterly, Vol. 11, No. 4, pp. 427–446.
Wolfowitz, Paul, and Michael O’Hanlon, “Plan Afghanistan,” Foreign Policy,
October 28, 2011.
226 Countering Others’ Insurgencies

Wood, Reed M., Jacob D. Kathman, and Stephen E. Gent, “Armed Intervention
and Civilian Victimization in Intrastate Conflicts,” Journal of Peace Research,
Vol. 49, No. 5, 2012, pp. 647–660.
The World Bank, Fragile States: Good Practice in Country Assistance Strategies,
Washington, D.C., 2005, p. 3.
———, World Development Indicators (WDI) dataset. As of March 11, 2013:
http://data.worldbank.org/data-catalog/world-development-indicators
Yong, Tan Tai, The Garrison State: The Military, Government and Society in
Colonial Punjab, 1849–1947, New Delhi, India: Sage Publications, 2005.
Youssef, Nancy A., “Pentagon Rethinking Value of Major Counterinsurgencies,”
McClatchy News, May 12, 2010.
Yusufzai, Rahimullah, “Fallen City,” Newsline Magazine, January 3, 2009.
———, “Fazlullah in Afghanistan,” The News, December 1, 2012.
Zafar, Mohammad, “Balochistan Conundrum: Hearings Spotlight ‘Crumbling’
Khuzdar Situation,” Express Tribune, October 11, 2012.
———, “Stabilising Balochistan: Government to Talk to Insurgent Groups After
Eid, Says CM,” Express Tribune, July 30, 2013.
Zahra-Malik, Mehreen, “Pakistan Army Battles Legacy of Mistrust in Taliban
Heartland,” Reuters, February 3, 2013.
Zaidi, Mosharraf, “Demystifying Foreign Aid,” The International News, December
15, 2010.
Zanger, Sabine C., “A Global Analysis of the Effect of Political Regime Changes
on Life Integrity Violations, 1977–93,” Journal of Peace Research, Vol. 37, No. 2,
2000, pp. 213–233.
Zubair, Muhammad, “Army Operation in South Waziristan: The TTP and IDPs,”
Daily Times, June 19, 2012.
This study examines the counterinsurgency strategies and practices adopted
by threatened regimes and the conditions under which U.S. “small-footprint”
partnerships are likely to help these governments succeed. The report’s findings
are derived from a mixed-method research design incorporating both quantitative
and qualitative analysis. Simple statistical analyses are applied to a dataset
of counterinsurgencies that have terminated since the end of the Cold War
(72 in all), and more in-depth analyses are provided of two recent cases of
U.S. partnerships with counterinsurgent regimes, in the Philippines and Pakistan.
The quantitative analysis finds that the cases of small-footprint U.S. operations that
are commonly touted as “success stories” all occurred in countries approximating
a best-case scenario. Only one insurgency in eight occurred in such best-case
countries; the sizable majority have taken place in worst-case conditions. In these
worst-case environments, counterinsurgent regimes are typically unsuccessful in
their efforts to suppress or contain rebellion, and they often employ violence
indiscriminately. The case studies of the Philippines and Pakistan largely reinforce
the findings of the quantitative analysis. They also highlight the challenges
the United States faces in attempting to influence partner regimes to fight
counterinsurgencies in the manner that the United States would prefer. The study
concludes with policy recommendations for managing troubled partnerships.

NATIONAL SECUR I T Y R ESEA R C H DIV IS IO N

www.rand.org $37.95

RR-513-SRF