Rand Rr4293
Rand Rr4293
Rand Rr4293
Assessing Trade-Offs in
U.S. Military Intervention
Decisions
Whether, When, and with What Size Force
to Intervene
For more information on this publication, visit www.rand.org/t/RR4293.
About RAND
The RAND Corporation is a research organization that develops solutions to public policy
challenges to help make communities throughout the world safer and more secure, healthier
and more prosperous. RAND is nonprofit, nonpartisan, and committed to the public interest. To
learn more about RAND, visit www.rand.org.
Research Integrity
Our mission to help improve policy and decisionmaking through research and analysis is
enabled through our core values of quality and objectivity and our unwavering commitment
to the highest level of integrity and ethical behavior. To help ensure our research and analysis
are rigorous, objective, and nonpartisan, we subject our research publications to a robust and
exacting quality-assurance process; avoid both the appearance and reality of financial and
other conflicts of interest through staff training, project screening, and a policy of mandatory
disclosure; and pursue transparency in our research engagements through our commitment to
the open publication of our research findings and recommendations, disclosure of the source
of funding of published research, and policies to ensure intellectual independence. For more
information, visit www.rand.org/about/principles.
RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.
Published by the RAND Corporation, Santa Monica, Calif.
© 2021 RAND Corporation
is a registered trademark.
Library of Congress Cataloging-in-Publication Data is available for this publication.
ISBN: 978-1-9774-0506-7
Cover photo by Sean Taylor, U.S. Army.
Limited Print and Electronic Distribution Rights
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 this publication online is prohibited. 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 its research documents for commercial use. For
information on reprint and linking permissions, please visit www.rand.org/pubs/permissions.
Preface
iii
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii
Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix
CHAPTER ONE
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Theoretical Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Organization of This Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
CHAPTER TWO
What Do We Know? Reviewing Existing Literature . . . . . . . . . . . . . . . . . . . . . . 13
The Intervention Decision: When Do Interventions Improve
Outcomes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
The Timing Decision: Is It Better to Intervene Early or Late? . . . . . . . . . . . . . 24
The Size Decision: Should the Intervention Be Small or Large? . . . . . . . . . . . 30
Synthesis and Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
CHAPTER THREE
Quantitative Analysis: Studying Trends and Patterns . . . . . . . . . . . . . . . . . . . . 39
Data Set of Intervention and Nonintervention Cases . . . . . . . . . . . . . . . . . . . . . . . . 39
Exploring Outcomes by Intervention Timing and Size . . . . . . . . . . . . . . . . . . . . . 48
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
v
vi Assessing Trade-Offs in U.S. Military Intervention Decisions
CHAPTER FOUR
Lessons from Focused Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Assessment and Insights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Implications and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
CHAPTER FIVE
Counterfactuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Summary of the Counterfactual Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
CHAPTER SIX
Conclusion and Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
The Effect of Interventions on Outcomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
APPENDIX
A. List of Cases of Potential U.S. Intervention . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
ONLINE APPENDIX
Available for download at www.rand.org/t/RR4293
B. Counterfactual Case Studies
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Figures
vii
Tables
ix
x Assessing Trade-Offs in U.S. Military Intervention Decisions
xi
xii Assessing Trade-Offs in U.S. Military Intervention Decisions
ings in terms of lives lost, but also the enhancement of U.S. reputation
and diplomatic influence, the strengthening of other regional actors
that are then empowered to handle future crises, or even the potential
to prompt U.S. adversary intervention into what becomes, for them,
a quagmire. An improved understanding of the potential costs and
benefits of intervention and nonintervention, and the factors on which
they depend, could improve the information available to policymakers
and could lead to substantial benefits for the United States, given the
often highly consequential nature of military intervention decisions.
Objective
Figure S.1
The Intervention Decision
Decision1
2 3 4
1 Decision2
No intervention
5 6 7
intervening state (the United States), and outcomes for the host nation.
Throughout the report, we explore the relationships between interven-
tion decisions and these outcome measures, with the goal of identifying
patterns and insights that can be used by planners and decisionmakers.
Research Approach
Results
1 To clarify, we assessed that a U.S. intervention prior to the fall of Mosul could have been
more effective (and at lower cost) than the historical intervention that took place afterward.
Once Mosul had fallen, however, U.S. patience in rebuilding Iraqi security forces before
mounting a sustained counteroffensive appears to have paid dividends.
Summary xix
Recommendations
• The local balance of power: In crises and wars where the balance of
power already tilts toward the side favored by the United States,
xxii Assessing Trade-Offs in U.S. Military Intervention Decisions
• Warring parties and the balance of power: Who are the warring
parties? What is the balance of power between relevant parties?
Would the balance be changed in a way favorable to U.S. interests
with an intervention?
xxiv Assessing Trade-Offs in U.S. Military Intervention Decisions
• Motivations of the parties: Is the war or crisis one that seems ame-
nable to settlement? How long is each side willing to fight? Are
the issues at stake likely to remain intractable, or could they be
resolved through negotiation?
• External spoilers and influencers: Which external actors could
become involved in ways that counter U.S. interests? What are
their goals and capabilities? Are they likely to intervene with mili-
tary forces or to exert other types of influence?
• Local populations: Will the local population of the host nation be
supportive of a U.S. intervention? If not, what type of backlash
is likely, and how severe will it be? What are the relationships
between the other parties to the conflict and the local population?
• Adversary responses: How will adversaries respond to a U.S. inter-
vention? Are they likely to feel threatened by it? Are there impor-
tant escalation risks to consider? Do adversaries have the ability
and motivation to undermine U.S. interests elsewhere in response
to the intervention?
• International community: What will be the response of the inter-
national community to the intervention? Will the United States
face backlash from allies for intervening or not intervening?
Many people helped with the research and completion of this report.
The authors thank MG Christopher McPadden for his sponsorship
and support throughout the project. We also thank MG (ret) William
Hix for asking the question that led to this project and for support-
ing the effort. Tony Vanderbeek and Mark Calvo provided valuable
guidance and feedback throughout the research and writing of this
report. Stephen Biddle of Columbia University, Karin von Hippel of
the Royal United Services Institute, and Karl Mueller of RAND pro-
vided insightful reviews that led to substantial improvements in this
document. The authors are also grateful to Bruce Bennett, Ambas-
sador (ret) James Dobbins, Edward Geist, Benjamin Fishman, Jeffrey
Martini, Laurel Miller, and Ariane Tabatabai for their insights and
expertise on specific cases included in this report. Lynn Davis provided
essential guidance and insights for our research throughout. Finally,
Sally Sleeper and Stephen Watts offered valuable comments on earlier
drafts. All mistakes are the authors’ own.
xxvii
Abbreviations
xxix
CHAPTER ONE
Introduction
1
2 Assessing Trade-Offs in U.S. Military Intervention Decisions
1 Marc Grossman, Executive Secretary, U.S. Department of State, “Discussion Paper for
Deputies Committee on Peacekeeping Options in Rwanda,” memorandum to William H.
Itoh, Executive Secretary, National Security Council, Washington, D.C., May 16, 1994;
Susan E. Rice, National Security Council, “Updated Talking Points on Rwanda,” memoran-
dum to W. Anthony K. Lake et al., Washington, D.C., May 7, 1994; Susan E. Rice and Nick
Rasmussen, “Deputies Committee Meeting on African Peacekeeping Issues,” memorandum
to Samuel R. Berger, Washington, D.C., September 20, 1993.
Introduction 3
2 Dominic Tierney, “Syria and the Cycle of American Intervention,” The Atlantic, Octo-
ber 24, 2016.
3 Samantha Power, “Bystanders to Genocide,” The Atlantic, September 2001.
4 Richard N. Haass and Martin Indyk, “Beyond Iraq: A New U.S. Strategy for the
Middle East,” Foreign Affairs, Vol. 88, No. 1, 2009; Stacie L. Pettyjohn and Jennifer Kava-
nagh, Access Granted: Political Challenges to the U.S. Overseas Military Presence, 1945–2014,
Santa Monica, Calif.: RAND Corporation, RR-1339-AF, 2016.
4 Assessing Trade-Offs in U.S. Military Intervention Decisions
Objective
To this end, our objective with this report is to create a framework that
can be used to rigorously consider the trade-offs between interven-
ing militarily immediately following the outbreak of a war or crisis,
intervening later, and not intervening at all, as well as the trade-offs
involved with decisions regarding the size of the potential intervention
force to be employed. This framework can be used to provide a better
understanding of the relationships between intervention timing, inter-
vention size, and intervention outcomes to inform future debates about
whether and when to intervene.
To construct this framework and explore these trade-offs, we will
need to consider not only the historical U.S. experience with inter-
ventions of different timings and sizes, but also the historical record
of cases in which the United States did not intervene. In addition to
investigating the historical relationships between intervention deci-
sions and relevant outcomes, we will explore the possible effects of
alternative decisions. In cases where the historical decision was made
to intervene, we will endeavor to estimate the possible costs and bene-
fits of the alternative choice: a decision not to intervene. In cases where
the historical decision was made not to intervene, we will explore the
potential costs and benefits to the United States had the decision been
Introduction 5
Theoretical Framework
Figure 1.1
The Intervention Decision
Decision1
2 3 4
1 Decision2
No intervention
5 6 7
5 There are, of course, numerous other factors that could affect intervention outcomes
beyond these high-level decisions made at the outset of an intervention. Perhaps most notable
among these, from the perspective of policymakers, are part of a series of more-operational
decisions regarding how to actually execute the intervention, including decisions regarding
rules of engagement, targeting, and a host of other issues. Although these decisions poten-
tially are important to determining outcomes, we largely placed consideration of the effects
of such decisions outside the scope of this report to focus more directly on the three essen-
tial high-level decisions discussed above. That said, how interventions are executed has the
potential to confound our analysis if patterns we attribute to timing and size decisions turn
out to be attributable to execution decisions. Because of this possibility, when we considered
in our focused case studies and counterfactual cases (summarized in Chapters Four and Five)
whether size and timing decisions made independent contributions to the advancement of
U.S. interests, we assessed whether the execution of mission decisions was instead responsible
for the observed patterns. We did not have a similar ability to account for these issues in our
quantitative chapter (Chapter Three); this should be understood as an additional limitation
of that analysis.
8 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table 1.1
Intervention Outcomes of Interest
NOTE: This report does not focus on either domestic political costs or military
casualties, although both are discussed in Chapter Two.
6 There is a significant literature that considers the political costs and benefits of conduct-
ing military interventions at various points in the political and election cycles. We consider
this literature in Chapter Two. We do not, however, directly consider political costs in the
remainder of the analysis. The focus of this report is on how decisions about when and with
how many forces to intervene shape the outcomes of those interventions and their ability to
advance U.S. national interests. Concerns about the political costs of interventions to spe-
cific actors or parties is therefore outside the scope of the report.
Introduction 9
Approach
nation (or location of the conflict or crisis), and political and economic
outcomes for the host nation (or location of the conflict or crisis). We
use these comparisons as another perspective on how decisions about
whether to intervene, decisions about intervention timing, and choices
about intervention size might be related to different outcome measures.
Third, we conducted a set of 45 short, focused case studies. These
case studies include interventions and noninterventions, each address-
ing a consistent set of questions. Topics covered include an assessment
of why an intervention did or did not occur, why it occurred when it
did (early or late), and what the consequences of the intervention or
nonintervention choice were for the United States and the host nation
(or potential host, in the case of noninterventions). In particular, these
case studies helped us to isolate the effects of intervention size and
timing decisions on U.S. interests from selection effects that might
drive more-general correlations or apparent relationships. The case
studies address many of the same questions as the first two method-
ological approaches described, though they do not focus on the effects
on conflict intensity because these appeared to be well-considered else-
where. However, they allow for the inclusion of context and nuance as
well as cross-case comparisons and observations.
Finally, we conducted four in-depth counterfactual illustrations
focused on the 1992–1995 Bosnian War, the 1950–1953 Korean War,
the 2011 Libyan Civil War, and the 1979 Iranian Revolution. In these
counterfactuals, we first trace out what happened historically and why,
then we explore what might have happened if the United States had
made a different intervention decision, such as choosing not to inter-
vene in the Korean War. We highlight the role of key actors and their
interests and capabilities and explore how they most likely would have
acted in the alternative scenario. We explore the costs and benefits for
the United States of this alternate decision and identify which factors
would have been central to determining the counterfactual outcomes.
We use the insights from these cases to illustrate broader observations
about the trade-offs between different intervention decisions and out-
come measures.
Introduction 11
Limitations
Although our use of four methods will give us advantages over the
use of any single method, there remain limitations to the conclusions
we will be able to draw. The dynamics of war and crisis situations can
be immensely complex, and isolating the effects of U.S. interventions
from the context in which they occur is quite difficult. Our analysis
will not provide a single or simple answer. Instead, it will show how
the relative costs and benefits of different intervention timing and size
decisions depend very much on the context and that the relationship
between intervention decisions and outcomes is highly conditional. We
aim to provide decisionmakers with a framework for navigating the set
of choices surrounding whether, when, and how to intervene, includ-
ing insights into when early or late interventions are more likely to lead
to favorable outcomes and when interventions tend to have costs that
are not offset by their benefits. Finally, the purpose of this report is not
to directly evaluate any notable previous or anticipated U.S. interven-
tion decision. Such assessments certainly could be useful to learn les-
sons or inform specific decisions, but because each intervention deci-
sion and each context is unique, they might not provide, in isolation,
more-general guidance to policymakers facing these choices in unan-
ticipated contexts in the future. Instead, our objective is to consider the
broader U.S. historical experience and to identify patterns, trends, and
trade-offs to develop a generalizable framework that can help guide
U.S. intervention decisions going forward.
13
14 Assessing Trade-Offs in U.S. Military Intervention Decisions
1 Steven Watts, Patrick B. Johnston, Jennifer Kavanagh, Sean M. Zeigler, Bryan Freder-
ick, Trevor Johnston, Karl P. Mueller, Astrid Stuth Cevallos, Nathan Chandler, Meagan L.
Smith, Alexander Stephenson, and Julia A. Thompson, Limited Intervention: Evaluating
the Effectiveness of Limited Stabilization, Limited Strike, and Containment Operations, Santa
Monica, Calif.: RAND Corporation, RR-2037-A, 2017, p. 125.
2 Donald Wittman, “How a War Ends: A Rational Model Approach,” Journal of Conflict
Resolution, Vol. 23, No. 4, December 1979; Ibrahim A. Elbadawi and Nicholas Sambanis,
“External Interventions and the Duration of Civil Wars,” Washington, D.C.: World Bank,
What Do We Know? Reviewing Existing Literature 15
September 2000; Patricia L. Sullivan and Johannes Karreth, “The Conditional Impact of
Military Intervention on Internal Armed Conflict Outcomes,” Conflict Management and
Peace Science, Vol. 32, No. 3, 2015.
3 David Kilcullen, The Accidental Guerrilla: Fighting Small Wars in the Midst of a Big One,
New York: Oxford University Press, 2009; David M. Edelstein, “Occupational Hazards:
Why Military Occupations Succeed or Fail,” International Security, Vol. 29, No. 1, Summer
2004, pp. 50–51.
4 Richard K. Betts, American Force: Dangers, Delusions, and Dilemmas in National Security,
New York: Columbia University Press, 2012, p. 31.
5 Barbara F. Walter, “The Critical Barrier to Civil War Settlement,” International Organi-
zation, Vol. 51, No. 3, Summer 1997, p. 363.
6 This research has mixed findings, in part, because of differing definitions, focal points,
and time frames.
7 Niklas Karlén, “The Legacy of Foreign Patrons: External State Support and Conflict
Recurrence,” Journal of Peace Research, Vol. 54, No. 4, 2017.
16 Assessing Trade-Offs in U.S. Military Intervention Decisions
other relevant factors), they would only intervene in conflicts and crises
that they could influence to their desired outcome. In reality, how-
ever, states have much more trouble accurately assessing the balance of
power and face obstacles in the form of private and asymmetric infor-
mation. There are many cases in which U.S. policymakers might have
made a different choice about an intervention if they had full informa-
tion at the initial decision point. Because miscalculation does occur,
interventions only sometimes advance the intervener’s interests and
could bring unexpected economic and other costs.11 In this section, we
consider what previous research has suggested about the implications
of military interventions for the intervening party.
National Interests
Our understanding of which factors influence the relative success of
military interventions at achieving political objectives (used here as a
proxy for national interests) is based largely on the U.S. experience,
because few other states have a sufficient number of interventions for a
rigorous analysis. Previous RAND Arroyo Center work suggests that
the type of objectives defined for an intervention directly influences
the likelihood the intervention is able to achieve those objectives. U.S.
military interventions are more likely to achieve their political objec-
tives when those objectives are narrowly defined and are less success-
ful when objectives are ambitious or sweeping in scope (e.g., build
democracy, rebuild institutions, prevent the spread of communism).
In practice, this relationship has meant that the United States has been
less successful in interventions since the Cold War than was the case
before and during it because intervention goals have become increas-
ingly broad in scope.12 Success can vary by specific objective as well.
For instance, the U.S. record on democratization-focused interven-
tions is (as would be expected) mixed, with some research finding that
the military interventions have been successful in increasing levels of
11 Stephen Biddle, Military Power: Explaining Victory and Defeat in Modern Battle, Princ-
eton, N.J.: Princeton University Press, 2004.
12 Jennifer Kavanagh, Bryan Frederick, Alexandra Stark, Nathan Chandler, Meagan L.
Smith, Matthew Povlock, Lynn E. Davis, and Edward Geist, Characteristics of Successful U.S.
Military Interventions, Santa Monica, Calif.: RAND Corporation, RR-3062-A, 2019.
18 Assessing Trade-Offs in U.S. Military Intervention Decisions
13 James Meernik, “United States Military Intervention and the Promotion of Democ-
racy,” Journal of Peace Research, Vol. 33, No. 4, 1996, p. 400; Margaret G. Hermann and
Charles W. Kegley, Jr., “The U.S. Use of Military Intervention to Promote Democracy:
Evaluating the Record,” International Interactions, Vol. 24, No. 2, 1998, p. 108.
14 Jeffrey Pickering and Mark Peceny, “Forging Democracy at Gunpoint,” International
Studies Quarterly, Vol. 50, No. 3, September 2006, p. 555; Bruce Bueno de Mesquita and
George W. Downs, “Intervention and Democracy,” International Organization, Vol. 60,
No. 3, 2006, p. 647.
15 The United States has typically been more successful at achieving its political objectives
with combat interventions than during stabilization or deterrence missions. This success
could be because combat missions tend to be more heavily resourced, allowing for the appli-
cation of overwhelming force supported by technological superiority. Kavanagh, Frederick,
Stark, et al., 2019. Also see Patrick M. Regan, “Conditions of Successful Third-Party Inter-
vention in Intrastate Conflicts,” Journal of Conflict Resolution, Vol. 40, No. 2, June 1996; and
Patrick M. Regan and Aysegul Aydin, “Diplomacy and Other Forms of Intervention in Civil
Wars,” Journal of Conflict Resolution, Vol. 50, No. 5, October 2006.
16 Host-nation support will be a key consideration, since interventions could face many
more obstacles when the host nation is opposed. Edelstein, 2009, argues that host nations are
most likely to support a large intervening force when they share a common threat perception
or desired outcome with the intervener. An example in the U.S. context would be the high
willingness of South Korea to host U.S. forces since the end of the Korean War. David M.
Edelstein, “Foreign Militaries, Sustainable Institutions, and Postwar Statebuilding,” in
Roland Paris and Timothy D. Sisk, eds., The Dilemmas of Statebuilding: Confronting the
Contradictions of Postwar Peace Operations, New York: Routledge, 2009.
What Do We Know? Reviewing Existing Literature 19
17 Kavanagh, Frederick, Stark, et al., 2019; Matthew Adam Kocher, Adria Lawrence, and
Nuno P. Monteiro, The Rabbit in the Hat: Nationalism and Resistance to Foreign Occupa-
tion, Chicago, Ill.: American Political Science Association 2013 Annual Meeting, August 20,
2013.
18 Kavanagh, Frederick, Stark, et al., 2019.
19 Richard N. Haass, “Military Force: A User’s Guide,” Foreign Policy, No. 96, 1994, p. 21.
20 See Michael Mandelbaum, “The Reluctance to Intervene,” Foreign Policy, No. 95,
Summer 1994, pp. 6–9; and Watts, Johnston, et al., 2017, p. 108.
21 Walter, 1997, p. 361; Bob Woodward, Plan of Attack, New York: Simon and Schuster,
2004, p. 150.
20 Assessing Trade-Offs in U.S. Military Intervention Decisions
22 Among many others, see, for example, James H. Lebovic and Ashfaq Ishaq, “Military
Burden, Security Needs, and Economic Growth in the Middle East,” Journal of Conflict Res-
olution, Vol. 31, No. 1, March 1987; and Jülide Yildirim and Nadir Öcal, “Military Expen-
ditures, Economic Growth and Spatial Spillovers,” Defence and Peace Economics, Vol. 27,
No. 1, 2016, p. 89.
23 Pettyjohn and Kavanagh, 2016.
24 Eelco van der Maat, “Sleeping Hegemons: Third-Party Intervention Following Territo-
rial Integrity Transgressions,” Journal of Peace Research, Vol. 48, No. 2, 2011.
25 Barry R. Posen, Restraint: A New Foundation for U.S. Grand Strategy, Ithaca, N.Y.: Cor-
nell University Press, 2014.
26 Sarah E. Kreps, “Multilateral Military Interventions: Theory and Practice,” Political Sci-
ence Quarterly, Vol. 123, No. 4, Winter 2008/2009.
What Do We Know? Reviewing Existing Literature 21
ance for casualties.36 In the U.S. context, support for additional mili-
tary interventions has tended to decline for a time after conflicts that
involve large numbers of casualties, especially after military defeats.37
Host-Nation Trajectory
Previous work has also explored how external interventions affect the
governance, economies, and societies of host countries. We focus on
political and economic outcomes.
Political Outcomes
Work focused on political outcomes has emphasized how interventions
affect the host nation’s level of democracy and its record on politi-
cal and civil rights. The results are mixed. Several studies have found
that, in general, the existence of an intervention does not make the
host nation more likely to democratize.38 Other studies have found that
there does seem to be such a relationship but that it depends heavily
on context. The strategy of the intervener (both military and nation-
building) and host-nation political institutions could affect the out-
come.39 However, the literature is more pessimistic about the duration
36 Gartner and Segura, 1998; Eric V. Larson, Casualties and Consensus: The Historical Role
of Casualties in Domestic Support for U.S. Military Operations, Santa Monica, Calif.: RAND
Corporation, MR-726-RC, 1996, pp. 99–101; James Burk, “Public Support for Peacekeep-
ing in Lebanon and Somalia: Assessing the Casualties Hypothesis,” Political Science Quar-
terly, Vol. 114, No. 1, 1999.
37 Jennifer Kavanagh, Bryan Frederick, Matthew Povlock, Stacie L. Pettyjohn, Angela
O’Mahony, Stephen Watts, Nathan Chandler, John Speed Meyers, and Eugeniu Han, The
Past, Present, and Future of U.S. Ground Interventions: Identifying Trends, Characteristics, and
Signposts, Santa Monica, Calif.: RAND Corporation, RR-1831-A, 2017.
38 Bueno de Mesquita and Downs, 2006; Meernik, 1996; Alexander B. Downes and Jona-
than Monten, “Forced to Be Free? Why Foreign-Imposed Regime Change Rarely Leads to
Democratization,” International Security, Vol. 37, No. 4, 2013; William Easterly, Shanker
Satyanath, and Daniel Berger, Superpower Interventions and Their Consequences for Democ-
racy: An Empirical Inquiry, Cambridge, Mass.: National Bureau of Economic Research,
Working Paper No. 13992, May 2008; Jeffrey Pickering and Emizet F. Kisangani, “Politi-
cal, Economic, and Social Consequences of Foreign Military Intervention,” Political Research
Quarterly, Vol. 59, No. 3, September 2006; Mark Peceny, “Forcing Them to Be Free,” Politi-
cal Research Quarterly, Vol. 52, No. 3, September 1999.
39 Meernik, 1996; Hermann and Kegley, 1998; Downes and Monten, 2013; Peceny, 1999.
24 Assessing Trade-Offs in U.S. Military Intervention Decisions
40 Nils Petter Gleditsch, Lene Siljeholm Christiansen, and Håvard Hegre, Democratic
Jihad? Military Intervention and Democracy, Washington, D.C.: World Bank, Working Paper
No. WPS4242, June 2007.
41 Frederic S. Pearson, Scott Walker, and Stephanie Stern, “Military Intervention and Pros-
pects for Democratization,” International Journal of Peace Studies, Vol. 11, No. 2, 2006;
Taylor B. Seybolt, Humanitarian Military Intervention: The Conditions for Success and Fail-
ure, Oxford, United Kingdom: Stockholm International Peace Research Institute and
Oxford University Press, 2007, p. 270; Dursun Peksen, “Does Foreign Military Intervention
Help Human Rights?” Political Research Quarterly, Vol. 65, No. 3, 2012.
42 Marie Olson Lounsbery, Frederic Pearson, and Andrea Kathryn Talentino, “Unilateral
and Multilateral Military Intervention: Effects on Stability and Security,” Democracy and
Security, Vol. 7, No. 3, 2011; Pickering and Kisangani, 2006.
What Do We Know? Reviewing Existing Literature 25
Table 2.1
Summary of Prior Research on Military Interventions and Outcome
Measures
outcomes and for the interests of the intervener and host nations, but
the body of evidence surrounding the timing question is limited.
Conflict Outcomes
Duration and Intensity
Research on the timing of military and nonmilitary interventions sug-
gests that there are two points during which an intervention (military
or nonmilitary) might be able to shorten the duration of a conflict or
crisis (or prevent a crisis from escalating). The first occurs immediately
after the start of the conflict or crisis, before states have invested time
and money in the conflict and before battle lines are fully drawn.43
Duration of Interstate Disputes,” International Studies Quarterly, Vol. 44, No. 2, June 2000;
Christopher R. Mitchell, Cutting Losses: Reflections on Appropriate Timing, Fairfax, Va.: Insti-
tute for Conflict Analysis and Resolution, Working Paper No. 9, 1995; T. David Mason,
Joseph P. Weingarten, Jr., and Patrick J. Fett, “Win, Lose, or Draw: Predicting the Outcome
of Civil Wars,” Political Research Quarterly, Vol. 52, No. 2, June 1999, p. 264.
44 Zartman and Touval, 1985.
45 Stephen John Stedman, Peacemaking in Civil War: International Mediation in Zimbabwe,
1974–1980, Boulder, Colo.: Lynne Reinner, 1991.
46 Zartman and Touval, 1985; Stedman, 1991; Christopher R. Mitchell, The Structure of
International Conflict, London: Macmillan, 1981, Ch. 8.
What Do We Know? Reviewing Existing Literature 27
50 Kreps, 2008/2009. For example, interventions that occur before any diplomatic efforts
toward resolution could face questions of legitimacy. The United States has been faulted in
the case of Iraq for intervening without giving the processes of diplomacy sufficient oppor-
tunity to take their course. As Kreps argues, intervening as part of a coalition could help
to reduce some of the strategic and reputational costs that might come with an early inter-
vention. Also see Jon Western, “Sources of Humanitarian Intervention: Beliefs, Informa-
tion, and Advocacy in U.S. Decisions on Somalia and Bosnia,” in James M. McCormick,
ed., The Domestic Sources of American Foreign Policy: Insights and Evidence, 6th ed., New
York: Rowman and Littlefield Publishers, 2012, p. 399; and Seybolt, 2007.
What Do We Know? Reviewing Existing Literature 29
tic political costs might be higher when the domestic public does not
support the intervention. For this reason, election timing could play a
role in decisions about when to launch a military intervention.51 Early
interventions might generally be harder to sell to the public, especially
in the case of an emerging crisis or conflict, prior to the point at which
the threat is severe enough to be readily apparent. However, there could
also be political costs from intervening too late, especially if public
pressure in support of a specific intervention is high.52
Host-Nation Outcomes
There is very limited literature available on how intervention timing
affects the outcomes for the host nation, beyond what was already
described in the section on conflict outcomes (resolution, recurrence,
and intensity) that focused on how timing can affect the interests of
the intervening state. We can draw some inferences from this litera-
ture. For instance, an early intervention that is able to end a conflict
more quickly could have more-positive outcomes for the host nation.
As another example, early interventions into humanitarian crises could
be better able to prevent serious atrocities or limit the damage inflicted
on the host nation.53 Apart from these indirect observations, we know
little from previous work about whether intervention timing improves
the future political or economic trajectory of the host nation.
Table 2.2 summarizes the key insights from our assessment of the
prior literature regarding the relationship between intervention timing
51 Bruce Bueno de Mesquita, Alastair Smith, Randolph M. Siverson, and James D. Morrow,
The Logic of Political Survival, Cambridge, Mass.: MIT Press, 2005; Stephen Biddle, com-
ments to the authors, January 2019.
52 This pressure is hypothesized to come from media coverage and is known as the CNN
effect. Empirical support for the CNN effect is also mixed. See, for example, Piers Robin-
son, “The CNN Effect: Can the News Media Drive Foreign Policy?” Review of International
Studies, Vol. 25, No. 2, 1999; Babak Bahador, The CNN Effect in Action: How the News
Media Pushed the West Toward War in Kosovo, New York: Palgrave Macmillan, 2007; and
Bueno de Mesquita et al., 2005.
53 Zartman and Touval, 1985; Western and Goldstein, 2011, p. 57.
30 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table 2.2
Summary of Prior Research on Intervention Timing and Outcome Measures
54 James Dobbins, Michele A. Poole, Austin Long, and Benjamin Runkle, After the War:
Nation-Building from FDR to George W. Bush, Santa Monica, Calif.: RAND Corporation,
MG-716-CC, 2008.
55 We consider large forces to be those in excess of a division of deployed U.S. forces and
small forces to be those of brigade size or smaller.
What Do We Know? Reviewing Existing Literature 31
Conflict Outcomes
Duration and Intensity
Research that considers the ability of smaller interventions to contain,
end, or deescalate conflict is relatively mixed in its findings. For exam-
ple, several RAND studies have found that small stabilization missions
can lead to better outcomes for the state actor but do not typically
shorten conflicts.56 Findings on the ability of larger interventions to
shorten conflict duration or to deescalate a conflict or crisis are also
mixed. There is some evidence that larger forces could be better able to
guarantee the terms of any settlement agreement, making one or both
sides more willing to come to the negotiating table.57 The extent to
which a large intervention force is able to help the party it is support-
ing militarily could speed progress toward a resolution of the conflict.58
Larger interventions are also associated with lower fatality levels (expe-
rienced and caused), at least during nation-building.59 However, past
experience in Afghanistan and elsewhere has underscored the fact that
having a large force size might be necessary for conflict termination
56 In addition to small ground forces, limited interventions might also rely on airpower
through drone strikes or limited bombing campaigns. Previous work on the use of drone
strikes to fight insurgency campaigns has mixed empirical results, with some research sug-
gesting it can reduce militant violence and other research suggesting it can provoke backlash.
The difference in effect could depend on the way the intervening state organizes and uses the
strikes. Stephen Watts, 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, p. 45; Watts, Johnston, et al., 2017. See also Patrick B. Johnston and
Anoop K. Sarbahi, “The Impact of US Drone Strikes on Terrorism in Pakistan,” Interna-
tional Studies Quarterly, Vol. 60, No. 2, June 2016; and Javier Jordan, “The Effectiveness of
the Drone Campaign Against al Qaeda Central: A Case Study,” Journal of Strategic Studies,
Vol. 37, No. 1, 2014.
57 Walter, 1997; Watts, Johnston, et al., 2017.
58 Watts, Johnston, et al., 2017; Kavanagh, Frederick, Stark, et al., 2019.
59 James Dobbins, Seth G. Jones, Keith Crane, Andrew Rathmell, Brett Steele, Richard
Teltschik, and Anga R. Timilsina, The UN’s Role in Nation-Building: From the Congo to
Iraq, Santa Monica, Calif.: RAND Corporation, MG-304-RC, 2005; Stephen Biddle, Jef-
frey A. Friedman, and Jacob N. Shapiro, “Testing the Surge: Why Did Violence Decline in
Iraq in 2007?” International Security, Vol. 37, No. 1, Summer 2012; James T. Quinlivan,
“Force Requirements in Stability Operations,” Parameters: U.S. Army War College Quarterly,
Vol. 25, No. 4, Winter 1995–1996.
32 Assessing Trade-Offs in U.S. Military Intervention Decisions
Host-Nation Outcomes
As noted elsewhere, the key determinant of host-nation outcomes is
likely to be the context and characteristics of the host nation itself, such
as its level of political and economic development. As a result, interven-
tion size might have a fairly limited effect on the political, economic,
and development outcomes of the host nation, especially in the longer
term.75
However, intervention size could have an indirect effect in sev-
eral contexts, most already mentioned elsewhere in this chapter. First,
larger intervention forces could support better long-term outcomes if
they are able to facilitate a durable peace settlement and prevent con-
flict recurrence.76 Second, larger forces are likely necessary for the
nation-building tasks typically required in a postconflict environ-
ment.77 Other relevant factors include the level of political develop-
ment of the host nation and the application of nonmilitary resources.78
However, foreign intervention forces cannot stay forever, and having a
large force in a postconflict situation could build a dependency on or
trigger backlash against the intervening force within the host nation
that could interfere with the long-term stability and independence of
the host nation.79
Table 2.3 highlights the insights that emerge from a review of
research that considers the relationship between intervention size and
Table 2.3
Summary of Prior Research on Intervention Size and Outcome Measures
host nation). Our review reveals that although the existing work is
both extensive and diverse, the focus of the most-rigorous empirical
work has been on the intervention decision: specifically, the effect of
interventions on conflict outcomes and, to a lesser extent, the interests
of the intervener. The effects of intervention size have also been stud-
ied, with this work focusing primarily on the relationship with conflict
outcomes. There has been limited attention paid to the implications of
different intervention timings.
The other key observation that emerges from our review of the lit-
erature is just how conditional the relationships between interventions
(their timing and size) and relevant outcomes appear to be. In most
cases, the direction and strength of the relationship depends on other
mediating factors, such as the domestic context of the host nation, the
side in the conflict that the intervention supports, and the presence of
an international coalition, among others. Understanding the role of
these mediating variables is important to our approach in the remain-
ing chapters in this report. In the next chapter, we investigate quantita-
tive patterns in our database of intervention and potential intervention
cases.
CHAPTER THREE
39
40 Assessing Trade-Offs in U.S. Military Intervention Decisions
1 The UCDP Armed Conflict data include conflicts and wars. We include only wars
(annual battle deaths over 1,000) in our data. Therése Pettersson and Kristine Eck, “Orga-
nized Violence, 1989–2017,” Journal of Peace Research, Vol. 55, No. 4, 2018; Nils Petter
Gleditsch, Peter Wallensteen, Mikael Eriksson, Margareta Sollenberg, and Håvard Strand,
“Armed Conflict 1946–2001: A New Dataset,” Journal of Peace Research, Vol. 39, No. 5,
2002.
2 To be included in the ICB data, an interstate interaction must meet three criteria: “a
threat to one or more basic values, along with an awareness of finite time for response to the
value threat, and a heightened probability of involvement in military hostilities.” Michael
Brecher and Jonathan Wilkenfeld, A Study of Crisis, Ann Arbor, Mich.: University of Michi-
gan Press, 2000, p. 3.; Michael Brecher, Jonathan Wilkenfeld, Kyle Beardsley, Patrick James,
and David Quinn, International Crisis Behavior Data Codebook, Version 12, 2017.
Quantitative Analysis: Studying Trends and Patterns 41
Africa, and East and Southeast Asia), such as the 1991 Gulf War, and
(4) severe humanitarian crises (e.g., the 1994 Rwandan genocide).3
We made the decision to limit our data set to only those wars and
crises in which the United States had a strong national interest in order
to remove those cases in which there was simply no realistic possibil-
ity that the United States would intervene. It is possible that we elimi-
nated one or more crises or wars in which the United States intervened
or might have considered intervening in this process of narrowing the
universe of cases. However, the loss of any such case is likely balanced
out by the benefits that we get from eliminating the larger number of
cases that were not plausible intervention opportunities. We could also
have chosen a more restrictive set of inclusion criteria that would have
further narrowed the data. However, the more restrictive the inclusion
criteria, the higher the chance that we begin excluding cases in which
the United States could realistically have intervened. Furthermore, U.S.
intervention decisions are very idiosyncratic, often the product of a
confluence of events, personal interests, national interests, and timing,
so criteria that are too narrow might have resulted in a sample that was
not only missing meaningful nonintervention cases but also missing
intervention cases. Such a set of cases would then provide a misleading
and incomplete account of the trade-offs between intervening and not
intervening and the factors that influence this decision and its effects
on outcomes. After our attempt to balance these concerns, we were left
with a data set of 286 cases, including 130 crises and 156 wars.
Coding Interventions
With this data set of cases defined, the next step was to identify which
of the 286 cases had actually experienced a U.S. military interven-
tion. To do this, we relied on a database of such interventions that was
compiled and refined during a series of RAND Arroyo Center proj-
ects: the RAND U.S. Military Intervention Dataset (RUMID). The
data include entries for each U.S. military intervention since 1946 that
meets our size threshold.4 We relied on the following size thresholds to
identify military interventions for each domain:
4 The data cover the years 1898 to 2016, but we use only the post-1945 data for this analy-
sis. As a reminder, we do not include covert missions or any special forces missions because
these are typically classified.
5 For more information on this data set, see Kavanagh, Frederick, Stark, et al., 2019.
Quantitative Analysis: Studying Trends and Patterns 43
• no intervention
• early, small intervention
• early, medium intervention
• early, large intervention
• late, small intervention
• late, medium intervention
• late, large intervention
• forces already present.6
6 These include cases where the United States already had troops in a location before the
conflict or crisis in that location broke out. The precise coding of these cases is situation
specific. In order to fall into this category, the United States had to have already forward
deployed U.S. forces to a specific region or country for the specific intention of countering
or deterring a given threat or security challenge. In some cases, this means U.S. forces were
deployed to the country in question to deter against internal threats. In others, U.S. forces
were deployed to a “threatened” country to deter an adversary. Forces deployed or stationed
at home that were not specifically tasked with countering a given threat are not counted as
already present.
44 Assessing Trade-Offs in U.S. Military Intervention Decisions
7 We code the intervention as starting at the time of deployment and follow coding the
ICB and COW datasets to determine timing and phasing of conflict.
Quantitative Analysis: Studying Trends and Patterns 45
8 We do not count naval or air personnel involved unless they are also active in ground
activities.
46 Assessing Trade-Offs in U.S. Military Intervention Decisions
9 These size categories are obviously still broad and might group together interventions
that are quite different. For example, although our taxonomy will group interventions that
occur eight months after the conflict or crisis start with those that occur five years later, these
interventions could be different in important ways. Similarly, an intervention that involves
25,000 troops could be very different than one that involves 200,000. However, we decided
that a simpler categorization scheme, even with these inherently arbitrary break points, was
still preferable. In the case of size, commitments of 25,000 and 200,000 troops are signifi-
cant and meaningful commitments of forces that are more similar to each other than either
is to smaller commitments, which might represent a more-limited level of engagement or
interest by the United States. Similarly, interventions that include 100 troops and those that
include 2,990 troops are likely to be very different, but both are commitments that the Army
can more easily absorb without experiencing significant pressure on readiness. Again, both
are likely to be more similar to each other than to interventions of substantially larger sizes.
For the timing, a similar argument applies. An intervention that occurs immediately after a
crisis could be different than one that starts three months later, just as an intervention that
starts at month eight will be different from one that starts five years in. However, both sets
of cases are likely to share important similarities. The United States has undertaken very few
late interventions and has never undertaken a late intervention into a crisis. If the country is
going to intervene early, the intervention typically happens quickly. Interventions that occur
after six months are alike in that something prevented them from happening more immedi-
ately, whether it was a disagreement, an uncertainty, or simply a change in perceptions of the
importance of the cases for the United States. Of course, there will still be differences within
each category of cases that our quantitative analysis will not be able to capture. We will use
the focused and counterfactual case studies in Chapters Four and Five to help capture greater
nuance along these dimensions.
Quantitative Analysis: Studying Trends and Patterns 47
Table 3.1
Intervention Categories and Examples
Late, large — —
Table 3.2
War and Crisis Interventions, by Timing and Size
No intervention 103 69
Early, small 11 16
Early, medium 1 5
Early, large 1 4
Late, small 7 —
Late, medium 5 —
Late, large — —
the universe of cases into crises and wars. Early and small interventions
are still the most likely. There are nine interventions into crises that are
not small and none that are late. Late interventions do occur into wars
but slightly less frequently than those that occur early in the conflict.
48 Assessing Trade-Offs in U.S. Military Intervention Decisions
Outcome Measures
In Chapter Two, we outlined three types of outcomes that are of inter-
est when considering the trade-offs between different intervention
decisions. In this section, we will use these same three groups of out-
comes: effects on the conflict, effects on the interests of the intervener,
and effects on the host nation. We consider what patterns we can dis-
cern from our quantitative comparison, how they match to the results
described in Chapter Two, and what expectations they set for our qual-
itative analysis.
First, we consider several conflict outcome measures:10
10 We do not consider whether the United States or the side the United States supported
“won” in this analysis. First, in many cases, the United States did not intervene on one or the
other side but instead intervened to monitor the situation and deter the Soviets. Any assess-
ment of win/loss ratios would have to exclude those cases. This issue and the issue of which
wars and crises tend to escalate and which do not are covered in greater depth in Bryan Fred-
erick, Stephen Watts, Matthew Lane, Abby Doll, Ashley L. Rhoades, and Meagan L. Smith,
Understanding the Deterrent Impact of U.S. Overseas Forces, Santa Monica, Calif.: RAND
Corporation, RR-2533-A, 2020; and Watts, Johnston, et al., 2017.
11 Pettersson and Eck, 2018. The battle-death data is, however, affected by both missing
observations and uncertainty about included totals. Other measures of conflict intensity are
affected by similar limitations. We use this measure but note the uncertainty attached to
given estimates.
12 Brecher and Wilkenfeld, 2000; Brecher et al., 2017; Pettersson and Eck, 2018; Gleditsch
et al., 2002.
50 Assessing Trade-Offs in U.S. Military Intervention Decisions
crisis that later recurs as war, or a war that later recurs in a crisis
phase.
13 Political objectives are coded for success on a three-point scale. We only consider full
success here. The process for determining and coding both political objectives and degrees
of success can be found in Kavanagh, Frederick, Stark, et al., 2019. In that work, the authors
identified all political objectives for each intervention and then coded the degree of success.
They included multiple coders to increase robustness of the coding. Interventions could have
many objectives, and success for each was coded independently. For example, in the case of
the U.S. intervention in Afghanistan in 2001, the United States was successful in defeating
the Taliban but only partially successful at establishing stable governing institutions.
14 Compiled from U.S. Agency for International Development, US. Overseas Loans and
Grants, Obligations and Loan Authorizations, July 1, 1945–September 30, 2014, Washington,
D.C., 2015; and U.S. Agency for International Development, “The Greenbook,” webpage,
updated April 29, 2013.
Quantitative Analysis: Studying Trends and Patterns 51
15 Host nation is defined as “where the forces are located.” For naval and air deployments, we
consider the “host” nation to be the country whose airspace or nearby waters contains those
forces. Note that the Cuban Missile Crisis is unusual in that U.S. forces are coded as being
in Cuba because of their presence at Guantanamo, even though a large portion of forces were
deployed in U.S. waters or airspace.
It is also important to note that the intention of the United States regarding the host
nation varies in the cases in our data set. The United States was, for example, much more
motivated to improve the trajectory of Afghanistan after 2001 than it was to improve the tra-
jectory of Cuba after the Cuban Missile Crisis. This variation makes it difficult to interpret
the patterns in the data for host-nation trajectory. We discuss the implications of this when
we present those results. Although we are better able to account for such differences in our
focused case studies detailed in Chapter Four, this heterogeneity in U.S. goals highlights a
limitation of the quantitative analysis.
16 Monty G. Marshall, Ted Robert Gurr, and Keith Jaggers, Polity IV: “Political Regime
Characteristics and Transitions, 1800–2017,” Vienna, Va.: Center for Systemic Peace, 2018.
17 GDP per capita data from Angus Maddison, homepage, updated September 3, 2008.
Jutta Bolt, Robert Inklaar, Herman de Jong, and Jan Luiten van Zanden, “Rebasing ‘Mad-
dison’: New Income Comparisons and the Shape of Long-Run Economic Development,”
Maddison Project Database, Version 2018, Working Paper No. 10, 2018.
18 World Bank, “Mortality Rate, Infant (Per 1,000 Live Births), 1960–2017,” online data-
base, undated.
52 Assessing Trade-Offs in U.S. Military Intervention Decisions
These data come from the World Bank and start in 1960, so we
are not able to use this metric for all cases.
Comparisons
Conflict Outcomes
When considering conflict outcomes, we focus on conflict duration,
conflict intensity, and conflict recurrence. We first consider conflict
duration. Table 3.3 shows the duration of wars by intervention cat-
egory. Most wars are relatively short; as expected, the vast majority
of all wars in our data set last less than five years, with more than
half lasting less than two years. However, even in this context, U.S.
interventions appear to be associated with shorter wars. All wars with
early U.S. interventions last five years or less, though there are several
longer wars with late, small interventions. There are also several longer
wars in the “forces already present” category. However, selection effects
might be relevant to this association between U.S. interventions and
shorter wars. It is possible that the trend observed in Table 3.3 results,
for example, from an unwillingness by the United States to become
involved in wars that appear too messy, too difficult, or likely to last
more than a few years. Late interventions into wars, in contrast, are
somewhat more likely to be associated with longer wars, but that may
Quantitative Analysis: Studying Trends and Patterns 53
Table 3.3
Conflict Duration by Intervention Type for War Cases
Less than
Category 2 Years 2–5 Years 6–10 Years 10-Plus Years
No intervention 49 35 13 4
New U.S. 14 3 4 1
intervention
Early, small 10 1 — —
Early, medium 1 — — —
Early, large — 1 — —
Late, small 2 — 3 1
Late, medium 1 1 1 —
Late, large — — — —
Forces already 53 7 3 1
present
simply be because, in many cases, the war could have already been
lengthy by the time the United States became involved.
In crisis cases (shown in Table 3.4), about two-thirds of crisis
interventions occur in cases that resolve within less than one year, with
the rest occurring in crises that last roughly one full year or, in rela-
tively few cases, two years. For nonintervention cases, about 85 percent
resolve in less than one year, suggesting that new interventions into
crises do not seem to be associated with shorter crises than what occurs
in the absence of interventions. However, it is important to note that
the United States does not always withdraw forces at the end of a crisis,
even if the initial crisis resolves. In many cases, including Afghani-
stan 2001, the United States remains for a lengthy postcrisis period of
stabilization and other activities, which can last anywhere from a few
months to many years. The duration of the initial conflict or crisis,
then, does not necessarily tell us much about the cost or implications
of that intervention for the United States.
54 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table 3.4
Conflict Duration by Intervention Type for Crisis Cases
No intervention 58 10 1
New U.S. 17 6 2
intervention
Early, small 11 3 2
Early, medium 5 — —
Early, large 1 3 —
Late, small — — —
Late, large — — —
Forces already 24 11 1
present
19 All five-year postconflict comparisons in this chapter are calculated from the end year of
the crisis or conflict.
Quantitative Analysis: Studying Trends and Patterns 55
Table 3.5
Change in Average Annual Battle Deaths, by Intervention Category
Late, large — — — —
20 We do not see strong effects by other outliers in our other intervention categories.
21 Further investigation reveals that this increase occurs only in war cases, and that average
annual battle deaths falls in crisis cases where forces were already present.
Quantitative Analysis: Studying Trends and Patterns 57
bility and prevent crisis and conflict recurrence, then we would hope
to see a clear association between U.S. interventions and lower rates
of conflict recurrence.22 However, Table 3.6 suggests little difference
at the aggregate level in the recurrence rates for cases in which the
United States does or does not intervene. Wars appear to recur about
85 percent of the time, while crises recur about 60 percent of the time,
regardless of whether there was a U.S. intervention. There are a few
differences when we consider individual intervention categories, but
the small number of cases and potential selection effects make it dif-
ficult to draw any real conclusions. In our universe for both wars and
crises, recurrence seems lower for early interventions regardless of size
(though the larger categories have few interventions). However, it is
worth underscoring that recurrence rates following U.S. interventions,
Table 3.6
Conflict Recurrence, by Intervention Category
Late, large — — — —
22 As already noted, even where others disagree, interventions always respond to the percep-
tion of a threat or crisis, even if the crisis or threat later proves to be nonexistent.
58 Assessing Trade-Offs in U.S. Military Intervention Decisions
Percentage Percentage
of Objectives of Objectives
Number of Successfully Number of Successfully
Category Wars Achieved Crises Achieved
No Intervention 103 — 69 —
Late, large — — — —
rate of success uses prior RAND research that identified the politi-
cal objectives of U.S. military interventions and coded the degree to
which they were achieved, according to in-depth research.23 These data
do not contain information on the political objectives of noninterven-
tions; at this point, we are not able to compare this outcome across
the intervention decision point. We will conduct an assessment of the
extent to which the United States achieves its objectives with a smaller
set of nonintervention cases in Chapter Four.
Table 3.7 suggests that rates of success are higher, in general, for
early interventions when compared with late interventions. This result
also seems consistent with previous research described in Chapter Two
that suggests that early interventions tend to be more likely to advance
the interests of the intervener. This result could be because interveners
have the most flexibility during the conflict’s early phases, before both
sides dig in and compromise becomes harder. The effects of size appear
to be somewhat more mixed, with no consistent patterns emerging.
Selection effects could also play a role in explaining these pat-
terns. It is possible that the United States intervenes early and contrib-
utes the most resources (not just forces, but also equipment, funding,
and support) to cases in which more-important interests are at stake.
This additional commitment, beyond what we can measure here, might
explain the higher success rate as much as the timing of the interven-
tion itself. We further explore the possible dynamics driving the higher
success rates of early interventions in Chapter Four. It is also worth
emphasizing that our analysis does not allow us to assess whether the
United States achieves a higher rate of success on intervention objec-
tives when it intervenes than when it does not. We will explore this
relationship qualitatively in Chapters Four and Five.
Ideally, we would use proxy metrics in addition to the level of suc-
cess to compare the strategic and reputational costs, domestic political
costs, and economic costs and benefits of the intervention borne by the
United States across categories. Practically, however, we lack the data
sources to comprehensively conduct such an investigation. Instead, we
assess these dynamics in the short, focused case studies in Chapter Four
and the longer counterfactual case studies in Chapter Five.
However, one aspect of the costs and benefits to the intervener
for which we do have a relevant data source is the amount of U.S.
economic and military assistance spending. Increased levels of U.S.
assistance reflect, in part, an increasing economic burden, but they
might also reflect a degree of success in establishing a partner receptive
to receiving this assistance. Although they represent an economic cost,
increases in military or economic assistance could also reflect a strate-
gic benefit for the United States.
To assess changes in assistance, we first calculated the percent-
age of total military and economic assistance allocated to each country
during the first year of the conflict. We did the same for the five-year
postconflict period. Then we considered the change between the two.24
Table 3.8 shows the change in economic assistance and Table 3.9 shows
the change in military assistance. The two tables suggest a correlation
between U.S. intervention decisions and future levels of economic and
military assistance, but we believe that these patterns might not be
that meaningful as a reflection of advancing U.S. interests. Table 3.8
suggests a modest correlation between U.S. intervention (early U.S.
interventions in particular) and future increases in economic assis-
tance. Table 3.9 suggests a notably stronger correlation between U.S.
intervention (again, early U.S. interventions in particular) and future
increases in military assistance. It is also worth noting that we reran the
results of Tables 3.8 and 3.9 excluding two high-profile cases that were
known to involve very high levels of U.S. assistance (U.S. interventions
in Iraq post-2003 and Afghanistan post-2001) to see whether these
patterns were driven by outlier observations; although the increases in
military assistance correlated with U.S. interventions were somewhat
more modest, they were not dramatically different from those reported
in Tables 3.8 and 3.9.
24 We use this particular metric to deal with the dramatic year-over-year fluctuations in
military and economic assistance spending. The “noise” in these data makes it difficult to
identify meaningful trends. Furthermore, the size of the swings in certain countries and the
overall distribution of the data mean that outliers can significantly influence overall trends.
Quantitative Analysis: Studying Trends and Patterns 61
Table 3.8
Change in Economic Assistance to Host Nation, by Intervention Category
Late, large — — — —
Table 3.9
Change in Military Assistance to Host Nation, by Intervention Category
Late, large — — — —
25 It is important to note when interpreting the results in this section that the U.S. com-
mitment to helping improve the host nation varied substantially across the interventions
in our data set. Although many cases involve clear commitments to nation-building and
assistance (such as post-2001 Afghanistan and post-2003 Iraq), others might have seen much
more limited involvement (such as naval missions to deter Soviet involvement in the 1971
India-Pakistan war over Bangladesh), while in still other cases, the United States might have
actually wished the host nation ill (such as Cuba following the Cuban Missile Crisis or Iran
following the Iran Hostage Crisis). Because the United States was, in most cases, at least
generally interested in improving outcomes for the host nation, we decided to include the
analysis showing overall patterns in these metrics. However, caution should be taken when
interpreting them, and they should not be understood to reflect the limits of how much
the United States might be capable of improving host-nation outcomes, if that were to be a
priority.
26 Polity scores rate the level of democracy of a state on a scale from −10 to 10, with 10 being
a full democracy and −10 being a full autocracy. Marshall, Gurr, and Jaggers, 2018.
64 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table 3.10
Changes in Polity Score, by Intervention Category
Change in Average
Percentage Polity Score (Initial
Number of of Cases That Average Initial Year to Five Years
Category Cases Are Wars Polity Score Postconflict)
Late, large — — — —
following a bloody war between India and Pakistan that eventually led
to independence and the formation of a democratic Bangladesh. But
it would be misleading to argue that the U.S. intervention, limited to
the deployment of a carrier group to the Indian Ocean to monitor the
conflict, played any important role in this democratization.27 In the
case of El Salvador, the United States originally supported the Salva-
doran government in its fight against Leftist guerillas, despite the fact
that the government was less than democratic. It is true, however, that
the United States was also involved in the negotiation of the cease-
fire and supported the eventual holding of elections.28 Looking at the
data more generally, we see relatively few cases of significant increases
in polity score across intervention categories. Among those cases with
significant increases, such as Liberia in the 1990s, it can be hard to
trace the direct line between the military intervention and democrati-
zation.29 There are also cases in which U.S. interventions are followed
by large decreases in democracy, including in Indonesia, where U.S.
involvement in regime activities against Indonesian communists could
have contributed to the decline in democracy observed in the 1960s
in that country.30 Finally, a comparison of wars and crises suggests no
major differences between patterns in the two types of cases for this
metric. Taking a holistic view of all the intervention cases in the data
set, our analysis leads us to expect no substantial relationship between
U.S. interventions and democratization of the host nation.
Turning to economic outcomes, our analysis in Table 3.11 suggests
that, at the aggregate level, U.S. military interventions are not associ-
ated with any increase in GDP per capita in the five years after the end
27 Lorraine Boissoneault, “The Genocide the U.S. Can’t Remember, but Bangladesh Can’t
Forget,” Smithsonian, December 16, 2016.
28 William M. LeoGrande, Our Own Backyard: The United States in Central America,
1977–1992, Chapel Hill, N.C.: University of North Carolina Press, 1998.
29 Terrence Lyons, “Liberia’s Path from Anarchy to Elections,” Current History, Vol. 97,
No. 619, May 1998.
30 Vincent Bevins, “What the United States Did in Indonesia,” The Atlantic, October 20,
2017; Bradley R. Simpson, Economists with Guns: Authoritarian Development and U.S.-
Indonesian Relations, 1960–1968, Stanford, Calif.: Stanford University Press, 2008.
66 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table 3.11
Changes in GDP per Capita, by Intervention Category
Late, large — — — —
NOTE: Excludes ongoing wars and crises as well as cases that ended after 2013.
Change in Infant
Percentage of Average Infant Mortality (Initial
Number of Cases That Are Mortality (per Year to Five Years
Category Cases Wars 1,000 Live Births) Postconflict)
Late, large — — — —
NOTE: Excludes ongoing wars and crises as well as cases that ended after 2013.
68 Assessing Trade-Offs in U.S. Military Intervention Decisions
Summary
Table 3.13
Summary of Quantitative Investigations into Relationship Between U.S.
Interventions and Outcomes
Table 3.13—Continued
1 More information on the focused case studies is available from the authors on request.
73
74 Assessing Trade-Offs in U.S. Military Intervention Decisions
Crucially, these focused studies allow us to explore cases that cover the
full breadth of differing intervention timing and size combinations and
to consider several nonintervention cases. In addition, this approach
allows us to assess a wider range of in-country and international char-
acteristics and their relationship to intervention decisions and interven-
tion outcomes. As a complement to this approach, we also conducted
four much more detailed counterfactual case studies, which are sum-
marized in the next chapter.
Methodology
Selecting Cases
When selecting the cases for our focused case studies, we considered
several factors. We wanted to ensure that the cases chosen included a
sufficient number of nonintervention cases to provide a useful baseline
and covered a wide range of intervention timings and sizes. Further,
we also wanted the cases to cover a representative set of activity types,
2 As discussed in Chapter One, we did not assess the third category of outcome measures
(effects on the conflict) in these case studies. The effects of interventions on conflict dynam-
ics were among the most robustly studied in the literature. For this reason, we decided to use
our case studies to focus on the least well-studied outcome measures, including a particular
emphasis on whether the interventions advanced the interests of the intervening state.
Lessons from Focused Case Studies 75
such that our cases were not unduly concentrated in, for example, all
combat missions or all deterrence missions. Finally, we wanted to con-
duct the largest number of cases that resources permitted to ensure that
our assessments of cases with particular timing and size characteristics
were as likely to be representative as possible.
With these criteria in mind, we selected 45 cases from the list of
286 potential cases of U.S. intervention discussed in Chapter Three.
During our selection process, we drew cases using the intervention cat-
egories described in Chapter Three: whether the intervention occurred
early or late in a crisis or war; whether the intervention involved small,
medium, or large U.S. forces; and combinations of these timing and
size dimensions, as well as cases in which U.S. forces were already pres-
ent before the onset of the crisis or war and cases in which the United
States did not intervene at all—there were no historical cases that were
late and large. After ensuring, to the best extent the universe of poten-
tial cases allowed, a sufficient number of cases from each category, we
then randomly selected cases from each category while making sure
that we also included a mix of activity types, including deterrence,
combat, and stability operations.3 During this process, we intentionally
oversampled intervention cases; in particular, we oversampled medium
and large intervention cases, of which there have historically been
fewer, given their greater substantive importance for policymakers. We
therefore undersampled cases in which U.S. forces were already pres-
ent or in which there was no U.S. intervention at all, although we still
included 20 nonintervention cases. A full list of cases selected is shown
in Table 4.1.
Table 4.1
List of Cases Selected
Intervention Timing,
Case Name Years Size
Demilitarized Zone Poplar Tree Crisis (Korea) 1976 Forces already present
Table 4.1—Continued
Intervention Timing,
Case Name Years Size
Guiding Questions
To ensure that the information from each of our 45 focused case stud-
ies was consistently collected by the research team and could be more
easily compared, we developed a set of specific questions that would be
answered across all of the cases. These questions were designed to focus
on the U.S. intervention decision and the resulting effects, if any, of the
intervention on U.S. interests. The questions asked varied depending
on whether the case in question included a U.S. military intervention,
although both sets of questions investigated similar, and parallel, fac-
tors. Box 4.1 lists the information collected in our set of focused case
studies.
These questions focused on identifying the U.S. decisionmaking
process in each case and the interests that motivated U.S. intervention
(or made it plausible). They further investigated the outcomes of the
case across four broad categories of U.S. interests: economic interests,
Box 4.1
Focused Case Study Questions
Asked for every case:
• Describe the nature of the conflict or crisis.
• Describe the U.S. interests involved that made intervention plausible.
• Was the crisis or conflict one in a series?
• Discuss the size and scope of U.S. resources committed.
• Describe relevant other concurrent nonmilitary factors.
• Overall assessment of the U.S. intervention decision.
Asked for U.S. intervention cases:
• What was the timing of the intervention within the war or conflict?
• What factors most influenced the U.S. decision to intervene?
• If the intervention occurred late, was one considered earlier?
• If an earlier intervention was considered but not undertaken, why not?
• What casualties did U.S. forces incur during the intervention?
• What were the economic costs and benefits to the United States of the
intervention?
• What were the reputational costs and benefits to the United States of the
intervention?
• What were the strategic costs and benefits to the United States of the
intervention?
• What were the costs and benefits of the intervention for the host or target
state?
Asked for cases with no U.S. intervention:
• Was undertaking an intervention ever seriously considered?
• If an intervention was not seriously considered, why not?
• If an intervention was seriously considered, which factors were most
important in policymaker calculations?
• What were the economic costs and benefits to the United States of not
undertaking the intervention?
• What were the reputational costs and benefits to the United States of not
undertaking the intervention?
• What were the strategic costs and benefits to the United States of not
undertaking the intervention?
• What trajectory did the host nation take over the course of the war or crisis
and its aftermath?
There are several patterns of interest that emerge across the case stud-
ies. This section summarizes these patterns through two lenses. After
an initial discussion of the possibility that the size and timing of inter-
ventions are driven by the issues at stake in an intervention, this section
reviews overall patterns in the answers provided to the case study ques-
tions noted in Box 4.1. We then discuss our effort to isolate the effects
of U.S. intervention from other factors to identify its likely impact
and how this impact might have varied by the intervention’s timing
and size. This section concludes with a summary of results and their
implications.
8 Although ISIL was of course not fully in control of the recognized government of either
state, it did exercise effective territorial control in many areas, and U.S. efforts to reverse this
state of affairs are therefore similar to efforts to change regimes in other states.
9 U.S. National Security Council, “U.S. Policy to Nicaragua,” memorandum of conversa-
tion, Washington, D.C., September 4, 1978; Marcia Olander, “Costa Rica in 1948: Cold
War or Local War?” The Americas, Vol. 52, No. 4, April 1996; Barak Salmoni, Bryce Loidolt,
and Madeleine Wells, Regime and Periphery in Northern Yemen: The Huthi Phenomenon,
Santa Monica, Calif.: RAND Corporation, MG-962-DIA, 2010.
10 Christopher S. Chivvis, Toppling Qaddafi: Libya and the Limits of Liberal Intervention,
New York: Cambridge University Press, 2014; Maryann K. Cusimano, Operation Restore
Hope: The Bush Administration’s Decision to Intervene in Somalia, Washington, D.C.: Insti-
tute for the Study of Diplomacy, 1995.
11 Central Intelligence Agency, Office of National Estimates, “Implications of Madagas-
car’s Unfinished Revolution,” memorandum, July 27, 1972; Roger Hilsman, Director of the
Bureau of Intelligence and Research, “Basic Elements in the Ruanda-Urundi Situation,”
memorandum to Secretary of State Dean Rusk, June 8, 1962; Henry A. Kissinger, “U.S.
Options in Biafra Relief,” memorandum to President Richared Nixon, Washington, D.C.,
January 28, 1969, pp. 1–6.
Lessons from Focused Case Studies 83
16 Panama Canal Commission, Annual Report: Fiscal Year Ended September 30, 1990,
Washington, D.C.: U.S. Government Printing Office, 1990; Nicholas Cullather, Operation
PBSUCCESS: The United States and Guatemala, 1952–1954, Washington, D.C.: Central
Intelligence Agency, Center for the Study of Intelligence, 1994.
Lessons from Focused Case Studies 85
effects, such as Bosnia 1995 or Haiti 1994.17 But there do not appear
to be clear patterns in the timing or size of the interventions regarding
these effects. Interventions with clearly negative effects on U.S. eco-
nomic interests included those that were early (Iraq 2003) or late (Iraq
2014) as well as those that were small, medium, or large in size (Libya
2011, Afghanistan post-2001, and Iraq 2003, respectively).18 Noninter-
vention cases typically involved relatively low stakes for U.S. economic
interests, with little evidence of dramatic changes in either direction.
We therefore do not find evidence that U.S. decisions regarding the size
and timing of an intervention are overall likely to affect U.S. economic
interests, although of course individual circumstances can differ.
Reputational Costs and Benefits
Reputational costs and benefits appear to have been somewhat more
divergent across intervention size and timing categories. Early inter-
ventions such as Iraq 2003 and Guatemala 1953 had clearly negative
reputational effects for the United States, as did late interventions such
as El Salvador 1979.19 Clearly positive reputational effects were rare.
Those we identified (the 1990 Gulf War and the 1962 Cuban Missile
Crisis) were both early interventions, but there might not have been
17 Amy Belasco, The Cost of Iraq, Afghanistan, and Other Global War on Terror Opera-
tions Since 9/11, Washington, D.C.: Congressional Research Service, RL33110, December 8,
2014; Neta C. Crawford, United States Budgetary Costs of Post-9/11 Wars Through FY2018: A
Summary of the $5.6 Trillion in Costs for the US Wars in Iraq, Syria, Afghanistan, and Pakistan,
and Post-9/11 Veterans Care and Homeland Security, Providence, R.I.: Watson Institute of
International and Public Affairs, Brown University, November 2017; Kavanagh, Frederick,
Stark, et al., 2019, pp. 196–228; United Nations, “UN Comtrade Database,” online data-
base, undated b; U.S. Agency for International Development, 2015.
18 U.S. Department of Defense, “Combined Joint Task Force Operation Inherent Resolve,”
webpage, undated; Stefan Karlsson, “The Total Cost of the Libya Intervention,” Christian
Science Monitor, April 15, 2011.
19 Collins, 2008; David M. Barrett, “Sterilizing a ‘Red Infection’: Congress, the CIA, and
Guatemala, 1954,” Central Intelligence Agency, Center for the Study of Intelligence, May 8,
2007; Cullather, 1994; Raymond Bonner, “America’s Role in El Salvador’s Deterioration,”
The Atlantic, January 20, 2018; Paul P. Cale, “The United States Military Advisory Group
in El Salvador, 1979–1992,” Small Wars Journal, 1996; Kevin Sullivan and Mary Jordan, “In
Central America, Reagan Remains a Polarizing Figure,” Washington Post, June 10, 2004.
86 Assessing Trade-Offs in U.S. Military Intervention Decisions
20 Lawrence Freedman and Efraim Karsh, The Gulf Conflict, 1990–1991: Diplomacy and
War in the New World Order, Princeton, N.J.: Princeton University Press, 1993; Tom Mahn-
ken, “The Gulf War in Retrospect,” Foreign Policy, January 20, 2011; Central Intelligence
Agency, “Implications of Khrushchev’s Message of 28 October,” Boston, Mass.: John F. Ken-
nedy Library, October 29, 1962a; David Coleman, “JFK’s Presidential Approval Ratings,”
webpage, undated; Central Intelligence Agency, “The Crisis: USSR/Cuba,” Boston, Mass.:
John F. Kennedy Library, October 29, 1962b.
21 Lawrence A. Yates, Power Pack: U.S. Intervention in the Dominican Republic, 1965–1966,
Fort Leavenworth, Kan.: Combat Studies Institute, U.S. Army Command and General Staff
College, 1988; Lawrence M. Greenberg, United States Army Unilateral and Coalition Opera-
tions in the 1965 Dominican Republic Intervention, Washington, D.C.: U.S. Army Center
of Military History, November 1986; David Coleman, ed., “LBJ Regretted Ordering U.S.
Troops into Dominican Republic in 1965, White House Tapes Confirm,” National Security
Archive, Electronic Briefing Book No. 513, April 28, 2015; Association for Diplomatic Stud-
ies and Training, “The Dominican Civil War of 1965,” Huffington Post, updated Decem-
ber 6, 2017.
Lessons from Focused Case Studies 87
22 Simond de Galbert, Back to the Future: Europe, the United States, and Iran Sanctions after
the Nuclear Deal, Washington, D.C.: Center for Strategic and International Studies, Novem-
ber 2015; Gideon Rose, “What Obama Gets Right,” Foreign Affairs, September/October
2015; Olander, 1996.
23 Collins, 2008; Daveed Gartenstein-Ross, “Successes and Failures of the U.S. and NATO
Intervention in Libya,” testimony presented before the Committee on Oversight and Gov-
ernment Reform, Washington, D.C.: Foundation for Defense of Democracies, May 1, 2014,
p. 4; Michael Clarke, “The Making of Britain’s Libya Strategy,” in Adrian Johnson and
Saqeb Mueen, eds., Short War, Long Shadow: The Political and Military Legacies of the 2011
Libya Campaign, London: Royal United Services Institute, 2012; Jonathan Marcus, “Why
China and Russia Rebuffed the West on Syria,” BBC News, October 5, 2011; Paul Lewis,
“Putin Insists Russia Wants to Repair US Relations – but Ill Will for Obama Is Clear,” The
Guardian, April 17, 2014; Stephen Schlesinger, “Ghosts of Guatemala’s Past,” New York
Times, June 3, 2011; Cullather, 1994; Barrett, 2007; Stephen M. Streeter, “Interpreting the
1954 U.S. Intervention in Guatemala: Realist, Revisionist, and Postrevisionist Perspectives,”
The History Teacher, Vol. 34, No. 1, November 2000.
24 Yates, 1988; Greenberg, 1986; Joseph B. Treaster, “20 Years After Dominican War,
Wounds Linger,” New York Times, May 1, 1985; Julio Yao, “Legacies of the U.S. Invasion
of Panama,” North American Congress on Latin America, March 22, 2012; Sullivan and
Jordan, 2004; Erick O. Granados, Impact of U.S. Foreign Policy and Assistance in El Salvador,
Quantico, Va.: U.S. Marine Corps Command and Staff College, May 2012; Christopher M.
Blanchard and Carla E. Humud, The Islamic State and U.S. Policy, Washington, D.C.: Con-
gressional Research Service, September 25, 2018.
88 Assessing Trade-Offs in U.S. Military Intervention Decisions
25 Richard Stewart, Operation Urgent Fury: The Invasion of Grenada, October 1983,
Washington, D.C.: U.S. Army Center of Military History, 2008, pp 26–36; Ronald H.
Cole, Operation Urgent Fury: The Planning and Execution of Joint Operations in Grenada,
12 October–2 November 1983, Washington, D.C.: Joint History Office, Office of the Chair-
man of the Joint Chiefs of Staff, 1997, pp. 1–5; Daniel L. Haulman, “Crisis in Grenada:
Operation Urgent Fury,” in A. Timothy Warnock, ed., Short of War: Major USAF Contin-
gency Operations, 1947–1997, Maxwell Air Force Base, Ala.: Air University Press, 2000;
Ronald Reagan, “Remarks of the President and Prime Minister Eugenia Charles of Domi-
nica Announcing the Deployment of United States Forces in Grenada,” October 25, 1983.
Lessons from Focused Case Studies 89
26 de Galbert, 2015; Rose, 2015; U.S. Department of State, Communist Involvement in the
Colombian Riots of April 9, 1948, Washington, D.C., Report No. 4686, October 14, 1948;
U.S. National Security Council, “Current Policies of the Government of the United States of
America Relating to the National Security: Volume I, Geographical Area Policies: Part VII,
Latin America - Colombia,” Gale Group, November 1, 1952; Central Intelligence Agency,
“OCI Handbook: Colombia,” June 13, 1955; Olander, 1996; Andrew Exum, “What’s Really
at Stake for America in Yemen’s Conflict,” The Atlantic, April 14, 2017; Presidential Review
Committee, “Summary of a Presidential Review Committee (PRC) Meeting on Central
America. Issues Include Discussion on U.S. Policy and the Political Situation in the Follow-
ing Countries: Nicaragua; El Salvador; Guatemala; Honduras,” Gale Group, June 11, 1979;
“U.S. Aid to the Contras,” Washington Post, February 3, 1988.
27 In this context, by advance U.S. interests we mean increase the ability of the United States
to achieve its political objectives, enhance its security and strategic flexibility, and protect
access and assets globally.
90 Assessing Trade-Offs in U.S. Military Intervention Decisions
28 Specifically, we had two researchers review the focused case studies and independently
provide answers to these questions. Across all cases, 94 percent of the answers they provided
to these questions were the same. The remaining 6 percent of questions were discussed and
adjudicated as a group to develop a consensus set of responses.
29 When answering this question, coders were instructed to consider the responses to the
questions about the specific types of U.S. interests detailed in Table 4.1 as a whole, and then
identify whether there appeared to be clear evidence of the intervention having played an
important role in the advancement of these interests, according to the history of the case.
Lessons from Focused Case Studies 91
30 This difference is certainly not statistically significant in the relatively small number of
cases we considered.
92 Assessing Trade-Offs in U.S. Military Intervention Decisions
likely would have occurred among Kurdish refugees, and any attempts
to establish a similar safe haven in northern Iraq might have involved
fighting Iraqi ground forces.36 This is not to suggest that even in simi-
lar circumstances early interventions will be sufficient, as illustrated by
the 1991 U.S. intervention to establish a no-fly zone over southern Iraq
to provide protection to Shiite demonstrators there.37 Nor is an early
decision always necessary. In retrospect, the 2003 U.S. invasion of Iraq
likely would not have been less successful had the decision to under-
take it been delayed further, even if it had eventually still been under-
taken. But in cases where local circumstances are imminently threaten-
ing a U.S. interest and the U.S. intervention has the ability to prevent
this threat, an early intervention can be clearly preferable to a later one.
On the other hand, we only identified one case in which the
lateness of the U.S. intervention per se might have been beneficial,
although the extent to which it was beneficial is unclear. The 1995
U.S. intervention into the conflict in Bosnia occurred after several
years of fighting and ineffective U.S. efforts to terminate the conflict
through other forms of pressure, which helped increase the suscepti-
bility of the Bosnian Serbs to pressure.38 Had the U.S. intervention
occurred much earlier, it potentially still could have achieved a similar
outcome for the host nation while reducing the dramatic humanitarian
toll of the war, but an earlier intervention likely would have required
the United States to bear greater up-front costs, potentially including
higher casualties. This counterfactual scenario is explored in detail in
Appendix B and summarized in Chapter Five. It illustrates potential
circumstances under which a later intervention might be preferable,
Gordon W. Rudd, Humanitarian Intervention: Assisting the Iraqi Kurds in Operation PRO-
VIDE COMFORT, 1991, Washington, D.C.: Department of the Army, 2004, p. 226.
36 Benard, 2004, pp. 463–464; U.S. Government Accountability Office, Operation Desert
Shield/Storm: Costs and Funding Requirements, Washington, D.C., GAO/NSIAD-91-304,
September 1991, p. 3.
37 Benard, 2004, p. 464; Phillip Gibbons, “U.S. No-Fly Zones in Iraq: To What End?”
Washington, D.C: Washington Institute, July 1, 2002.
38 U.S. Department of State, The Road to Dayton: U.S. Diplomacy and the Bosnia Peace Pro-
cess, May–December 1995, Washington, D.C., May 1997; Ivo H. Daalder, Getting to Dayton
The Making of America’s Bosnia Policy, Washington D.C.: Brookings Institution Press, 2000.
94 Assessing Trade-Offs in U.S. Military Intervention Decisions
39 Cameron Glenn, “Timeline: The Rise, Spread, and Fall of the Islamic State,” Wilson
Center, July 5, 2016; Chivvis, 2014, pp. 25, 69.
40 Freedman and Karsh, 1993, p. 409.
Lessons from Focused Case Studies 95
tional coalition that could be assembled and putting at risk victory over
Iraqi forces on the battlefield.
However, our review of the focused case studies emphasized that
making an intervention force larger should not always be assumed to
improve the intervention’s likelihood of success. There appear to be at
least four circumstances under which limiting the size of U.S. forces
in an intervention might be preferable. First, larger forces could raise
political concerns for the host nation, with the potential to undercut
the success of the U.S. intervention and the effective cooperation of
the host government. As an example, the medium-sized U.S. inter-
vention into Iraq following the 2014 capture of Mosul by ISIL was
eventually successful in its objective, helping to roll back the territorial
gains of ISIL and reestablish Baghdad’s control over most of the coun-
try, although the long-term trajectory of the country remains unclear.41
A larger-scale U.S. invasion potentially could have achieved similar
successes on the battlefield against ISIL (and perhaps done so more
quickly), but it also would have risked a substantial popular backlash,
given the recent history of the United States in Iraq, that could have
undermined Iraqi stability more broadly and plunged the country and
intervening U.S. forces back into a renewed civil war. In that case,
restraint with the size of the U.S. force likely was a virtue.
Second, a more limited force could enable a more rapid U.S.
response to a crisis. During the 2001 U.S. invasion of Afghanistan,
only limited forces could be rapidly introduced into that country in the
aftermath of the 9/11 attacks.42 However, the advantages that such a
medium-sized force provided in terms of speed outweighed the poten-
tial benefits of waiting longer for a larger force to become available,
and (along with air and naval assets) still proved to be sufficient for the
initial objective of overthrowing the Taliban.43
41 Glenn, 2016.
42 Joseph J. Collins, “Initial Planning and Execution in Afghanistan and Iraq,” in Rich-
ard D. Hooker, Jr., and Joseph J. Collins, eds., Lessons Encountered: Learning from the Long
War, Washington, D.C.: National Defense University Press, September 2015.
43 Collins, 2015.
96 Assessing Trade-Offs in U.S. Military Intervention Decisions
44 Adam B. Siegel, The Use of Naval Forces in the Post-War Era: U.S. Navy and U.S. Marine
Corps Crisis Response Activity, 1946–1990, Alexandria, Va.: CNA, 1991; John M. Young,
When the Russians Blinked: The U.S. Maritime Response to the Cuban Missile Crisis, Wash-
ington, D.C.: History and Museums Division, Headquarters, U.S. Marine Corps, 1990,
p. 94; Graham T. Allison and Philip Zelikow, Essence of Decision: Explaining the Cuban Mis-
sile Crisis, 2nd ed., New York: Longman, 1999, p. 123; Edward T. Russell, “Cuban Missile
Crisis,” in A. Timothy Warnock, ed., Short of War: Major USAF Contingency Operations,
1947–1997, Maxwell Air Force Base, Ala.: Air University Press, 2000.
45 Burris, 1962; Michael Dobbs, “One Minute to Midnight: Kennedy, Khrushchev and
Castro on the Brink of Nuclear War,” National Security Archive, George Washington Uni-
versity, June 4, 2008; Kavanagh, Frederick, Stark, et. al, 2019.
46 Yates, 1988; Greenberg, 1986; Coleman, 2015.
Lessons from Focused Case Studies 97
47 John H. Kelly, “Lebanon: 1982–1984” in Jeremy R. Azrael and Emil A. Payin, eds.,
U.S. and Russian Policymaking with Respect to the Use of Force, Santa Monica, Calif.: RAND
98 Assessing Trade-Offs in U.S. Military Intervention Decisions
was broadly successful, but a moderately larger U.S. force might have
been equally successful (if not more so).48 This is not to suggest that the
conditions discussed above regarding when the political context, speed
of deployment, or escalation risks might make a smaller intervention
preferable cannot apply to the decision regarding whether to employ
a medium or small force. But we do not find clear evidence for these
conditions in the cases we considered. There do appear to have been
relevant cases regarding whether the size of the forces committed were
unnecessary. The 1983 U.S. invasion of Grenada involved a substantial
medium-sized force that met with very limited resistance from the tiny
island nation of Grenada.49 In this case, as with the 1965 intervention
in the Dominican Republic, the forces committed likely were overkill,
though they do not appear to have hampered the achievement of U.S.
objectives.
To sum up, then, we find evidence that the question of what size
force should be committed to an intervention to advance U.S. objec-
tives should be considered carefully because it can have a notable effect
on the likelihood of success in achieving U.S. objectives or advancing
U.S. interests. But with respect to whether such forces should be large,
medium, or small, we find that contextual factors are likely to be quite
important. We will discuss the implications of these findings in greater
detail later in this chapter.
U.S. Interests and Nonintervention Cases
The most-striking pattern among our nonintervention cases was how
rarely they tended to evolve to either improve or worsen U.S. interests
to any notable degree (again, absent U.S. military intervention). Over-
all, roughly three-quarters of cases saw the aftermath of the case as
broadly similar to the status quo ante bellum from the perspective of
U.S. interests. The Biafra War in Nigeria, for example, involved hor-
Corporation, CF-129-CRES, 1996; Caspar W. Weinberger, Fighting for Peace: Seven Critical
Years in the Pentagon, New York: Warner Books, 1990, pp. 144–145.
48 Eric Schmitt, “Obama Is Pressed to Open Military Front Against ISIS in Libya”, New
York Times, February 4, 2016.
49 Stewart, 2008, pp. 3–8; “Leader of Grenada Is Reported Killed by Troops,” New York
Times, October 20, 1983.
Lessons from Focused Case Studies 99
50 Brian Edward McNeil, Frontiers of Need: Humanitarianism and the American Involve-
ment in the Nigerian Civil War, 1967–1970, dissertation, University of Texas at Austin, 2014,
pp. 68–70; Lasse Heerten and A. Dirk Moses, “The Nigeria-Biafra War: Postcolonial Con-
flict and the Question of Genocide,” Journal of Genocide Research, Vol. 16, Nos. 2–3, 2014,
pp. 169–176; Kissinger, 1969, pp. 1–6.
51 Kenton Clymer, “The United States and the Guomindang (KMT) Forces in Burma,
1949–1954: A Diplomatic Disaster,” Chinese Historical Review, Vol. 21, No. 1, 2014,
pp. 41–42.
52 Salmoni et al., 2010, pp. 273–274.
100 Assessing Trade-Offs in U.S. Military Intervention Decisions
in leading to this change. Of these cases, there were more in which the
lack of intervention led to positive outcomes for the United States than
led to negative outcomes. Next, we will survey three such cases that
led to positive outcomes, as well as a fourth case in which the opposite
happened: U.S. interests declined, and a U.S. intervention might have
helped.
Following a disputed election in Costa Rica in 1948, a mixed
group of Costa Rican political exiles with communist ties and Nicara-
guans invaded Costa Rica from Nicaragua. The United States declined
to intervene in the crisis militarily to restore stability or prevent an
increase in communist influence in the region. Instead, it supported
a diplomatic effort through the nascent Organization of American
States that succeeded in resolving the crisis, led to a durable transi-
tion to democracy in Costa Rica, and strengthened the Organization
of American States, enabling it to play a constructive role in future
regional crises.53 A U.S. intervention to impose a friendly government
in Costa Rica might have been able to achieve that narrower goal but
likely would have undermined or put at risk these other positive devel-
opments, which is roughly what occurred in the later U.S. intervention
to overthrow a communist-linked government in Guatemala in 1954.54
In the 2006 crisis over Iran’s nuclear program, the United States,
deeply enmeshed in the conflict in neighboring Iraq, declined to take
military action to attempt to destroy Iran’s emerging nuclear capabili-
ties. Exercising restraint at this time gave space both for the interna-
tional community to build a broad-based sanctions regime to pressure
Iran over its nuclear program and for Iranian leaders to conclude that
a diplomatic deal with the international community might be a prefer-
able arrangement.55 These developments led to the conclusion of the
Joint Comprehensive Plan of Action (JCPOA) in 2015, which commit-
53 Olander, 1996.
54 Barrett, 2007; Cullather, 1994.
55 Peter Baker, “Bush Warns Against Nuclear-Armed Iran; U.S. Presses Allies for U.N.
Action,” Washington Post, January 14, 2006; Ivo H. Daalder and Philip H. Gordon, “We
Should Strike Iran, but Not with Bombs,” Washington Post, January 22, 2006; David E.
Sanger, “Why Not a Strike on Iran?” New York Times, January 22, 2006.
Lessons from Focused Case Studies 101
56 The Trump administration later decided to unilaterally withdraw from the JCPOA in
favor of renewed sanctions and an effort to modify Iranian regime behavior in other areas,
which has increased tensions. However, this later action does not alter the assessment that
the nonintervention in 2006 led to an improvement in U.S. interests in the 2006 to 2016
period because international sanctions coordination increased leverage over Iran’s decision
to pursue a nuclear weapons program. Whether that increased leverage was best used in the
negotiation of the 2015 JCPOA is a separate issue from an assessment of the likely effects of
having forgone a military strike against Iran in 2006. See Zachary Laub and Kali Robinson,
“What Is the Status of the Iran Nuclear Agreement?” Council on Foreign Relations, updated
January 7, 2020.
57 World Peace Foundation, “Colombia: La Violencia,” Fletcher School, Tufts University,
December 14, 2016.
58 U.S. Department of Defense, Security Cooperation Agency, “Fiscal Year Series: Foreign
Military Sales, Foreign Military Construction Sales and Other Security Cooperation His-
torical Facts,” September 30, 2016.
102 Assessing Trade-Offs in U.S. Military Intervention Decisions
munist sentiment and helped keep Colombia a key U.S. ally during the
Cold War.
Although these cases illustrate how nonintervention can improve
U.S. interests, we also identified one case in which a U.S. interven-
tion could very plausibly have improved U.S. interests. In Nicaragua in
1978, decades of harsh dictatorship had produced widespread opposi-
tion to the government and a growing communist-linked insurgency
in the Sandinistas. The United States sought to apply diplomatic and
political pressure to force President Anastasio Somoza Debayle to
either reform or leave office, the latter of which he did in July 1979.59
Although the United States had made plans for a transition to a
national reconciliation government that included the moderate opposi-
tion, it decided not to commit any military forces to ensure stability,
and Sandinista forces were able to take advantage of the power vacuum
and seize power themselves.60 This marked a clear reversal for U.S.
efforts to contain communism in Latin America, and it was a rever-
sal that the United States proved unwilling to let stand. Throughout
the 1980s, the United States provided extensive military support to
the anti-Sandinista Contras, widening and prolonging the Nicaraguan
civil war and helping to destabilize the broader region.61 This aid con-
tinued illegally even when prohibited by Congress, eventually leading
to the 1987 Iran-Contra scandal. Although an intervention was not
seriously considered by the Carter administration in 1979, a relatively
modest U.S. intervention in that year could plausibly have helped to
stave off a Sandinista seizure of power in the first place and to ensure
that the transition to the moderate opposition that the United States
had envisioned actually took place.
We highlight two key lessons that can be drawn from these
notable nonintervention cases. First, the international context can be
vital when determining whether wars or crises are likely to worsen or
improve on their own. In 1940s Latin America, the United States held
59U.S. Department of State, “Milestones in the History of U.S. Foreign Relations: Central
America, 1977–1980,” webpage, undated.
60 Presidential Review Committee, 1979.
61 “U.S. Aid to the Contras,” 1988.
Lessons from Focused Case Studies 103
tremendous sway, and outside actors such as the Soviet Union were not
yet capable of dramatically influencing events. In both Costa Rica and
Colombia, the United States could feel reasonably confident that the
diplomatic and other policy levers at its disposal could help shape events
in the absence of well-funded spoilers. Similarly, there was broad-based
international concern with the Iranian nuclear weapons program, even
from some of Iran’s erstwhile partners, such as Russia, enabling the
United States to build up international pressure to apparently prompt a
change in Iranian thinking about the utility of its nuclear weapons pro-
gram. In late-1970s Latin America, by contrast, the Soviet Union had
well-established funding networks and proxies in the region, enabling
it to fund spoilers, such as the Sandinistas, that could divert events in
ways the United States proved unable to counter absent the application
of military force. Assessing the regional context, particularly the influ-
ence and motivations of potential third-party spoilers, can therefore
be an important step when determining whether a war or crisis might
require U.S. military intervention to safeguard U.S. interests.
Second, the relative strength and influence of the different fac-
tions on the ground is of course a vitally important factor in deter-
mining the outcome of a war or crisis, even absent any external inter-
ference. In Colombia 1948, for example, the United States remained
confident in the ability of government forces to ultimately hold on to
power throughout a decade of violence, making the decision not to
intervene on the government’s behalf easier.62 In Iran 2006, most U.S.
analysts assessed that the likely result of military strikes against the
Iranian regime would not have been instability and regime change but
instead a reinvigorated popular backing for an otherwise increasingly
unpopular regime, strengthening the case for pursuing diplomatic and
economic leverage.63 In Nicaragua, the United States misjudged the
capabilities and support of moderate groups, leaving a vacuum that
could be filled quickly by the Sandinistas. Accurate intelligence and
64 Olander, 1996; Fernando F. Sanchez, Democracy After War: Causes and Consequences of
the 1948 Civil War in Costa Rica, paper presented at the European Consortium for Political
Research Joint Sessions of Workshops, Uppsala, Sweden: Uppsala University, April 13–18,
2004; Desmond A. Turner, Civil War and (Re)Building the State: Recovery in Costa Rica,
Nicaragua, and El Salvador, thesis, Coral Gables, Fla.: University of Miami, May 2017.
65 Exum, 2017; Hilsman, 1962; United Nations, “Rwanda: A Brief History of the Country,”
webpage, undated a; U.S. Department of State, “Interagency Intelligence Memorandum,”
Foreign Relations of the United States, 1977-1980, Vol. XVII: Horn of Africa, Part 1, Washing-
ton, D.C., July 31, 1979; U.S. Department of State, “Memorandum from Paul B. Henze of
the National Security Council Staff to the President’s Assistant for National Security Affairs
(Brzezinski),” Foreign Relations of the United States, 1977-1980, Vol. XVII: Horn of Africa,
Part 1, Washington, D.C., June 3, 1980; John M. Goshko, “U.S. War by Proxy at an End,”
Washington Post, February 27, 1990.
Lessons from Focused Case Studies 105
rarely improved absent a U.S. intervention, at least in the crisis and war
cases we considered; when such interventions were conducted, positive
trajectories became more frequent, particularly when the intervention
was larger.
66 To be clear, there are substantial endogeneity concerns with interpreting this observed
pattern. The United States would seem much more likely to intervene in cases in which it
assessed acute threats to its interests, such that the remaining set of nonintervention cases
would not be expected to be replete with cases of sharp deterioration in these interests (absent
dramatic failures in U.S. analyses or limitations on forces available for intervention). We
therefore do not view this finding as speaking directly to whether the United States might be
“over-intervening.” This finding is, however, notable in that it does not suggest any evidence
that the United States could instead be “under-intervening.”
108 Assessing Trade-Offs in U.S. Military Intervention Decisions
When Should the United States Use a Larger Force Rather than a
Smaller One?
The focused cases suggest that large or medium interventions might
more often advance U.S. interests than small interventions do. When
facing larger or stronger adversaries in combat or when the potential
intervention needs to affect a wide area or large population, the use of
a larger force could be essential. Even in other cases, however, a larger
force could still help achieve U.S. objectives and, at worst, might be
redundant or unnecessary.
Although the benefits of larger intervention forces are clear in
some cases, we also identified several circumstances in which more-
limited forces are likely to be preferable and to better support the
advancement of U.S. interests. These circumstances include when a
larger force might produce a popular backlash in the host nation, when
a larger force might enhance the risks of escalation with an adversary,
and when waiting to mobilize a larger force might limit the timeliness
(and therefore the effectiveness) of the U.S. intervention. Intervention
decisions should be informed by an analysis of whether these condi-
tions appear to be present in the particular circumstances being con-
sidered for an intervention.
Conclusion
In this chapter, we considered what can be learned about the costs and
benefits of different intervention decisions across a large set of shorter,
focused case studies. In the next chapter, we look in much greater
depth at a smaller number of cases, exploring not only the real world
intervention decisions the United States made but also possible coun-
terfactual scenarios. We use the counterfactual case studies to further
explore the observations made in Chapters Two through Four, with
a particular focus on identifying those key factors and characteristics
most likely to influence the relationship between intervention timing
and size and the ability of the United States to advance and protect its
national interests.
CHAPTER FIVE
Counterfactuals
109
110 Assessing Trade-Offs in U.S. Military Intervention Decisions
Methodology
Constructing Counterfactuals
Counterfactuals are imagined alternative versions of real-world events
in which key features are altered. For example, consider the trajectory
of the Vietnam War had John F. Kennedy not been assassinated or the
path of the United States over the past 15 years if the 2003 intervention
in Iraq had not occurred. Counterfactuals can be a useful tool to evalu-
ate causal theories because they allow researchers to explore how such
imagined changes might affect a real-world outcome.1 Counterfactual
logic is necessarily embedded in any research that aims to understand
causation: If a researcher finds that variable X is necessary for a cer-
tain outcome, then the absence of variable X (i.e., in the counterfac-
tual scenario) should also be sufficient to produce a different outcome.
This reasoning is built into most statistical analysis and qualitative case
study research in the social sciences.2
Political scientists and historians have identified several counter-
factual ideal types that can broadly be sorted into two baskets: sta-
tistical and computer-assisted approaches and qualitative approaches.
Statistical and computer-assisted methods are typically used to predict
outcomes that are measured numerically, such as the number of civil-
3 Alberto Abadie, Alexis Diamond, and Jens Hainmueller, “Comparative Politics and the
Synthetic Control Method,” American Journal of Political Science, Vol. 59, No. 2, April 2015.
4 Alberto Abadie and Javier Garbeazabal, “The Economic Costs of Conflict: A Case Study
of the Basque Country,” American Economic Review, Vol. 93, No. 1, March 2003.
5 Ian Lustick, Miguel Garces, and Thomas McCauley, “An Agent-Based Model of Counter-
factual Opportunities for Reducing Atrocities in Syria, 2011-2014,” Washington, D.C.: Simon-
Skjodt Center for the Prevention of Genocide, August 2017.
112 Assessing Trade-Offs in U.S. Military Intervention Decisions
6 Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the
Social Sciences, Cambridge, Mass.: MIT Press, 2005, p. 147.
7 Frank P. Harvey, “President Al Gore and the 2003 Iraq War: A Counterfactual Test of
Conventional ‘“W’”isdom,” Canadian Journal of Political Science, Vol. 45, No. 1, March
2012.
8 Mona Yacoubian, “Critical Junctures in United States Policy Toward Syria: An Assessment
of the Counterfactuals,” WashingUon, D.C.: Simon-Skjodt Center for the Prevention of Geno-
cide, August 2017.
9 Fearon, 1991, p. 193.
Counterfactuals 113
10 Robert William Fogel, Railroads and American Economic Growth: Essays in Econometric
History, Baltimore, Md.: Johns Hopkins Press, 1964, cited in Lebow, 2000, p. 582.
11 Jon Elster, Logic and Society: Contradictions and Possible Worlds, New York: John Wiley,
1978, pp. 204–208, cited in Lebow, 2000, p. 582.
12 Philip E. Tetlock and Aaron Belkin, eds., Counterfactual Thought Experiments in World
Politics: Logical, Methodological, and Psychological Perspectives, Princeton, N.J.: Princeton
University Press, 1996.
13 Geoffrey Hawthorn, Plausible Worlds: Possibility and Understanding in History and the
Social Sciences, New York: Cambridge University Press, 1991, pp. 31–60, quoted in Lebow,
2000, p. 582.
114 Assessing Trade-Offs in U.S. Military Intervention Decisions
Case Selection
In total, we conducted four in-depth counterfactual case studies.26 To
select the specific four cases we would study, we applied several criteria.
First, we wanted to select cases in which a different U.S. intervention
decision could have been broadly plausible. On the one hand, a U.S.
decision not to intervene in the Cuban Missile Crisis would have been
relatively difficult to justify, given the stakes involved for the United
States. On the other hand, U.S. intervention in the ongoing (as of this
writing) civil war in the Central African Republic also appears highly
improbable, given the lack of clear U.S. interests involved. Selecting
cases in which both intervention and nonintervention (or interven-
tion of a different size and timing) could plausibly have occurred helps
increase the utility of the case. Second, we focused on cases whose
histories are relatively better known and better documented. This was
for pragmatic purposes, to ensure that we would be able to develop the
detailed understanding of the actors and their capabilities and interests
necessary to conduct a rigorous counterfactual investigation. Third, we
wanted to explore counterfactual cases that varied by their interven-
tion category. That is, we wanted a mixture of nonintervention cases,
intervention cases, and interventions with different timings and sizes,
to the extent feasible with only four cases. Fourth, we also wanted to
select cases that varied according to the issues that they presented for
U.S. policymakers. We wanted to include some cases that focused on
more-geostrategic issues, while others included an important humani-
tarian component. Finally, to the extent possible, we wanted to achieve
variety in the region and time in which the cases took place.27
26 Although more in-depth cases studies are always preferable, the scope of this report lim-
ited us to conducting four such cases.
27 Of course, with this many dimensions under consideration and only four cases to select,
identifying cases that would allow us to directly assess the causal effects of intervention deci-
sions on outcomes across each category was not possible. However, these counterfactual cases
were not intended or designed for the testing of novel hypotheses in this manner. Instead,
they were intended as an illustrative tool. They represent a plausibility demonstration but
should not, on their own, be taken to represent clear evidence of generalizable relationships.
Counterfactuals 117
Table 5.1
List of Selected Counterfactual Cases
28 It is worth noting that although two of these cases (Libya and Bosnia) were investi-
gated as focused case studies, as discussed in Chapter Four, we did not require that cases
either were or were not investigated using both methods. The degree of overlap, therefore, is
coincidental.
118 Assessing Trade-Offs in U.S. Military Intervention Decisions
with strikes against the Libyan regime’s ground troops and command
and control centers, later continued as NATO Operation Unified Pro-
tector. With this international support, rebel forces gradually pushed
back Libyan government troops until they had captured all key cities
and killed Qaddafi by October 2011.
The intervention succeeded in averting a near-term humanitar-
ian catastrophe and eliminating Qaddafi from power, but it was also
accompanied by a variety of unintended consequences. In the after-
math of the intervention, Libyan politics remained fractured and cha-
otic, with a battered economy and persistent violence. The intervention
undermined the stability of neighboring countries, with jihadist groups
in neighboring countries benefiting from the instability in Libya and
weapons proliferation throughout the region. Further, the Russian
sense that the United States had violated an understanding regarding
the limits of force that would be used in Libya contributed to a wider
decline in relations between Russia and the West.
In the counterfactual scenario, President Barack Obama main-
tains the U.S. position of not seeking military involvement in Libya,
a country that presents limited strategic interest for the United States
and where the risks of entanglement are perceived to be high. In the
absence of U.S. support, European allies decide they are neither will-
ing nor able to conduct an operation on their own. Absent interna-
tional intervention, the Libyan rebellion ends up defeated and Qad-
dafi remains in power. A failure of the rebellion, however, does not
mean that rebel activity disappears in Libya. Instead, it morphs into a
low-intensity conflict consisting mainly of guerilla-type attacks against
Libyan government troops, infrastructure, and official buildings and
that is mostly confined to the Benghazi area, where protests started in
February 2011, and the Misrata area, which had shown particularly
strong resistance during its siege by Qaddafi’s troops from February
to May 2011. Regionally, Qaddafi’s continued hold on power results
in lower levels of conflict from Mali to Syria because Libya no longer
functions as a source of exported weaponry or fighters. U.S.-Russian
relations are improved, at least over the near term, while U.S. relations
with NATO allies are negatively affected following the U.S. refusal to
intervene alongside them in Libya.
Counterfactuals 125
Results
would likely have had reputational costs for the United States as it
sought to rally a coalition to contain communism. In Libya, the deci-
sion to intervene likely had mixed effects on the U.S. reputation. In
some circles, the decision to intervene, including with UN authori-
zation, suggested a commitment to protecting against human rights
atrocities. However, the perceived expansion of the goals of the inter-
vention beyond the short-term humanitarian imperatives to encompass
the fall of Qaddafi reinforced concerns with U.S. leadership in other
circles, most notably in Russia, that the United States was an incorrigi-
ble interventionist bent on regime change in non-allied states. In Iran, a
U.S. intervention might have had more-negative effects on reputation,
especially without international support. In Bosnia, the U.S. reputa-
tion for international leadership would likely have been improved with
an early intervention instead of the historical later approach, though we
did not assess the reputational effects that would likely have occurred
if the United States had decided not to intervene in the war altogether.
Finally, in only two of the counterfactual cases did interventions
have strategic benefits: Bosnia and Korea. In Bosnia, the U.S. inter-
vention restored stability in a region very close to key European allies,
though this benefit was realized in both the late (historical) and early
(counterfactual) interventions. In Korea, the intervention provided the
United States with an important regional partner and counterweight
to Russia and China. In Libya and Iran, in contrast, it is not clear that
an intervention, historical in the case of Libya and counterfactual in
the case of Iran, had any net strategic benefits. The Libyan intervention
helped increase regional instability and required a follow-up interven-
tion to help stabilize Libya, while the Iranian intervention failed to
save a key U.S. partner in the Middle East, forcing the United States to
rethink its strategic approach to the region.
A key insight from this discussion of U.S. interests in our coun-
terfactual cases is that those cases in which U.S. objectives were clearly
advanced tended to involve larger forces than those cases in which they
were not, though there was some overlap. Further, the large size of
the intervention in Korea and the relatively substantial application of
forces in Bosnia were an essential component of their ability to advance
U.S. interests. However, our cases also suggest the limits of larger inter-
Counterfactuals 129
of competing actors within the Iranian political system and the weak-
ness of groups willing to ally with the United States. In Libya, the U.S.
intervention was quite limited. Although it might have been feasible
for a larger U.S. and allied intervention to provide greater postconflict
security and facilitate a political settlement in the country, the Libyan
transitional government was opposed to this option; instead, the lim-
ited U.S. intervention led to instability, characterized by competing
militias.
The effects of U.S. interventions on the outcome categories that
are illustrated by our counterfactual cases are summarized in Table 5.2.
Table 5.3 summarizes the effects of intervention across the three coun-
terfactual cases that consider the intervention decisions. Table 5.4 sum-
marizes the effects of timing in the Bosnia case.
Implications
The Intervention Decision: Should the United States Intervene?
When deciding whether to intervene in a war or crisis in which impor-
tant U.S. interests are at stake, our cases again illustrate the importance
of considering the likely effect that a U.S. intervention could have on
the local balance of power and, in turn, the leverage that the U.S.
intervention would provide for the United States to alter the outcome.
When the United States is able to decisively and sustainably intervene,
Table 5.2
Summary of Effects on Outcomes Illustrated by Counterfactual Cases
Intervener Larger interventions, for the most part, provided greater leverage
interests to advance U.S. interests. Local context and balance of power are
also crucial.
Host nation Local conditions of the host nation, including the local balance
of power and the broader political and economic environment,
shape outcomes. The feasibility of sufficiently larger or persistent
interventions is dependent on international context.
132 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table 5.3
Summary of Effects of Intervention for Counterfactual Cases
Table 5.4
Summary of Effects of Timing in Bosnia Counterfactual Case
Category Outcome
Table 5.5
Summary of Implications for Intervention Decisions Illustrated by
Counterfactual Cases
With what Larger interventions are able to help advance U.S. interests in
size force to certain circumstances. This is not universal, however. In some cases,
intervene depending on the local balance of power, even large forces might
be insufficient to advance U.S. interests. In others, the presence of
larger U.S. forces would be likely to generate substantial backlash,
suggesting that smaller forces might be more advisable.
CHAPTER SIX
137
138
Table 6.1
Summary of Findings for Outcome Measures
Conflict • Interventions can help • No clear patterns with • Not assessed. • Effects on conflict are
outcomes terminate conflicts and respect to conflict dura- highly dependent on
reduce conflict dura- tion or recurrence. how U.S. intervention
tion, especially if the • U.S. interventions, affects the local bal-
intervention is larger especially when later ance of power, as well
and earlier. in a conflict, could as the behavior and
• But the intervention’s be associated with a effects of third-party
effect depends on reduction in conflict spoilers.
its influence on the intensity in comparison • For example, early
local balance of power with nonintervention interventions can
and the potential for cases. reduce conflict duration
backlash. • However, selection if they lead to a quick
effects likely play a victory or dramatically
role, particularly in the increase it if they lead
case of timing. to a protracted stale-
mate. In some cases,
the longer conflict is
required to advance
other U.S. interests.
Table 6.1—Continued
Outcome
Measures Literature Review Quantitative Analysis Focused Case Studies Counterfactuals
139
140
Table 6.1—Continued
Host- • Mixed or limited inter- • No clear differences • Although results varied, • Interventions, par-
nation ventions tend to have between intervention interventions are asso- ticularly when larger,
trajectory highly contingent and nonintervention ciated with positive have the potential to
effects on host nations. cases. host-nation trajectories positively alter the host
• Earlier and larger inter- more often than non- nation’s trajectory.
ventions are associated interventions are. • However, their ability
with higher economic • Positive trajectories are to do so depends on
growth, while later more likely in larger local conditions, includ-
interventions are asso- interventions, with ing the local balance of
ciated with greater notable exceptions. power and the broader
decreases in infant political and economic
mortality. environment.
• Patterns within inter-
vention subcategories
are driven by a very
small number of cases.
Conclusion and Implications 141
1 Further, an early intervention naturally has greater potential to affect conflict duration
than a later intervention, given that the conflict has already persisted for some time when the
later intervention takes place.
Conclusion and Implications 143
2 To clarify, we assessed that a U.S. intervention prior to the fall of Mosul could have been
more effective (and at lower cost) than the historical intervention that took place afterward.
Once Mosul had fallen, however, U.S. patience in rebuilding Iraqi security forces before
mounting a sustained counteroffensive appears to have paid dividends.
Conclusion and Implications 147
Recommendations
• The local balance of power: In crises and wars where the balance of
power already tilts toward the side favored by the United States,
intervention might be unnecessary. However, where the United
States’ interests are under threat, intervention could make more
sense, provided the U.S. intervention would be capable of turning
the tide.
• The role of spoilers: If external actors might respond to a U.S.
intervention by increasing their involvement in a crisis or war
sufficiently to counteract the introduction of U.S. forces, then
intervention could prove ineffective. U.S. adversaries might also
respond to the intervention by taking actions in other theaters
that undermine U.S. interests in a manner that a narrow analysis
of the host nation might not account for. During the Cold War,
for example, the United States had to carefully consider how the
Soviet Union would react to its overseas activities in order to pre-
vent unintentional escalation.
150 Assessing Trade-Offs in U.S. Military Intervention Decisions
• Warring parties and the balance of power: Who are the warring
parties? What is the balance of power between relevant parties?
Would the balance be changed in a way favorable to U.S. interests
with an intervention?
• Motivations of the parties: Is the war or crisis one that seems ame-
nable to settlement? How long is each side willing to fight? Are
the issues at stake likely to remain intractable, or could they be
resolved through negotiation?
• External spoilers and influencers: Which external actors could
become involved in ways that counter U.S. interests? What are
their goals and capabilities? Are they likely to intervene with mili-
tary forces or to exert other types of influence?
• Local populations: Will the local population of the host nation be
supportive of a U.S. intervention? If not, what type of backlash
is likely, and how severe will it be? What are the relationships
between the other parties to the conflict and the local population?
• Adversary responses: How will adversaries respond to a U.S. inter-
vention? Are they likely to feel threatened by it? Are there impor-
tant escalation risks to consider? Do adversaries have the ability
152 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table A.1 provides additional detail on our list of potential U.S. inter-
vention cases.
153
154 Assessing Trade-Offs in U.S. Military Intervention Decisions
Table A.1
List of Cases of Potential U.S. Intervention
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
South Korea North Korea Nuclear III 2006 2007 Forces already
present
South Korea North Korea Nuclear IV– 2009 2009 Forces already
Satellite Launch present
Table A.1—Continued
Country Intervention
Location Crisis or War Start Year End Year Type
169
170 Assessing Trade-Offs in U.S. Military Intervention Decisions
Betts, Richard K., American Force: Dangers, Delusions, and Dilemmas in National
Security, New York: Columbia University Press, 2012.
Bevins, Vincent, “What the United States Did in Indonesia,” The Atlantic,
October 20, 2017.
Biddle, Stephen, Military Power: Explaining Victory and Defeat in Modern Battle,
Princeton, N.J.: Princeton University Press, 2004.
Biddle, Stephen, Jeffrey A. Friedman, and Jacob N. Shapiro, “Testing the Surge:
Why Did Violence Decline in Iraq in 2007?” International Security, Vol. 37, No. 1,
Summer 2012, pp. 7–40.
Blair, Charles P., “War with Iran? Revisiting the Potentially Staggering Costs to
the Global Economy,” Federation of American Scientists, May 23, 2013.
Blanchard, Christopher M., and Carla E. Humud, The Islamic State and U.S.
Policy, Washington, D.C.: Congressional Research Service, R43612, September 25,
2018. As of November 12, 2018:
https://fas.org/sgp/crs/mideast/R43612.pdf
Bliesemann de Guevara, Berit, and Florian P. Kühn. “‘The International
Community Needs to Act’: Loose Use and Empty Signalling of a Hackneyed
Concept,” International Peacekeeping, Vol. 18, No. 2, 2011, pp. 135–151.
Boissoneault, Lorraine, “The Genocide the U.S. Can’t Remember, but Bangladesh
Can’t Forget,” Smithsonian, December 16, 2016.
Bolt, Jutta, Robert Inklaar, Herman de Jong, and Jan Luiten van Zanden,
“Rebasing ‘Maddison’: New Income Comparisons and the Shape of Long-Run
Economic Development,” Maddison Project Database, Version 2018, Working
Paper No. 10, 2018. As of January 29, 2020:
https://www.rug.nl/ggdc/historicaldevelopment/maddison/releases/
maddison-project-database-2018
Bonner, Raymond, “America’s Role in El Salvador’s Deterioration,” The Atlantic,
January 20, 2018.
Brandt, Patrick T., T. David Mason, Mehmet Gurses, Nicolai Petrovsky, and
Dagmar Radin, “When and How the Fighting Stops: Explaining the Duration
and Outcome of Civil Wars,” Defence and Peace Economics, Vol. 19, No. 6, 2008,
pp. 415–434.
Brecher, Michael, and Jonathan Wilkenfeld, A Study of Crisis, Ann Arbor, Mich.:
University of Michigan Press, 2000.
Brecher, Michael, Jonathan Wilkenfeld, Kyle Beardsley, Patrick James, and David
Quinn, International Crisis Behavior Data Codebook, Version 12, 2017.
Bueno de Mesquita, Bruce, and George W. Downs, “Intervention and
Democracy,” International Organization, Vol. 60, No. 3, 2006, pp. 627–649.
References 171
Bueno de Mesquita, Bruce, and Randolph M. Siverson, “War and the Survival
of Political Leaders: A Comparative Study of Regime Types and Political
Accountability,” American Political Science Review, Vol. 89, No. 4, December 1995,
pp. 841–855.
Bueno de Mesquita, Bruce, Alastair Smith, Randolph M. Siverson, and James D.
Morrow, The Logic of Political Survival, Cambridge, Mass.: MIT Press, 2005.
Bunzl, Martin, “Counterfactual History: A User’s Guide,” American Historical
Review, Vol. 109, No. 3, June 2004, pp. 845–858.
Burk, James, “Public Support for Peacekeeping in Lebanon and Somalia: Assessing
the Casualties Hypothesis,” Political Science Quarterly, Vol. 114, No. 1, 1999,
pp. 53–78.
Burris, Howard L., “Soviet Concern Over Cuba,” memorandum to Vice President
Lyndon B. Johnson, October 2, 1962.
Cale, Paul P., “The United States Military Advisory Group in El Salvador, 1979–
1992,” Small Wars Journal, 1996. As of November 11, 2018:
http://smallwarsjournal.com/documents/cale.pdf
Call, Charles T., Why Peace Fails: The Causes and Prevention of Civil War
Recurrence, Washington, D.C.: Georgetown University Press, 2012.
Central Intelligence Agency, “The Current Situation in Colombia,” Situation
Memorandum 32-49, May 31, 1949.
———, “OCI Handbook: Colombia,” June 13, 1955.
———, “Implications of Khrushchev’s Message of 28 October,” Boston, Mass.:
John F. Kennedy Library, October 29, 1962a.
———, “The Crisis: USSR/Cuba,” Boston, Mass.: John F. Kennedy Library,
October 29, 1962b.
———, Office of National Estimates, “Implications of Madagascar’s Unfinished
Revolution,” memorandum, July 27, 1972.
Chivvis, Christopher S., Toppling Qaddafi: Libya and the Limits of Liberal
Intervention, New York: Cambridge University Press, 2014.
Clarke, Michael, “The Making of Britain’s Libya Strategy,” in Adrian Johnson and
Saqeb Mueen, eds., Short War, Long Shadow: The Political and Military Legacies
of the 2011 Libya Campaign, London: Royal United Services Institute, 2012,
pp. 7–14. As of November 12, 2018:
https://rusi.org/publication/whitehall-reports/
short-war-long-shadow-political-and-military-legacies-2011-libya
Clymer, Kenton, “The United States and the Guomindang (KMT) Forces in
Burma, 1949–1954: A Diplomatic Disaster,” Chinese Historical Review, Vol. 21,
No. 1, 2014, pp. 24–44.
172 Assessing Trade-Offs in U.S. Military Intervention Decisions
Daalder, Ivo H., and Philip H. Gordon, “We Should Strike Iran, but Not with
Bombs,” Washington Post, January 22, 2006.
de Galbert, Simond, Back to the Future: Europe, the United States, and Iran
Sanctions After the Nuclear Deal, Washington, D.C.: Center for Strategic and
International Studies, November 2015. As of March 5, 2020:
https://www.csis.org/analysis/back-future
Dobbins, James, Seth G. Jones, Keith Crane, Andrew Rathmell, Brett Steele,
Richard Teltschik, and Anga R. Timilsina, The UN’s Role in Nation-Building:
From the Congo to Iraq, Santa Monica, Calif.: RAND Corporation, MG-304-RC,
2005. As of January 29, 2020:
https://www.rand.org/pubs/monographs/MG304.html
Dobbins, James, Michele A. Poole, Austin Long, and Benjamin Runkle, After the
War: Nation-Building from FDR to George W. Bush, Santa Monica, Calif.: RAND
Corporation, MG-716-CC, 2008. As of January 29, 2020:
https://www.rand.org/pubs/monographs/MG716.html
Dobbs, Michael, “One Minute to Midnight: Kennedy, Khrushchev and Castro
on the Brink of Nuclear War,” National Security Archive, George Washington
University, June 4, 2008. As of November 12, 2018:
http://nsarchive.gwu.edu/nsa/cuba_mis_cri/dobbs/gitmo.htm
“Dodging Armageddon: The Third World War That Almost Was, 1950,”
Cryptologic Quarterly, February 24, 1998. As of May 1, 2020:
https://www.nsa.gov/Portals/70/documents/news-features/declassified-documents/
cryptologic-quarterly/Dodging_Armageddon.pdf
Downes, Alexander B., and Jonathan Monten, “Forced to Be Free? Why Foreign-
Imposed Regime Change Rarely Leads to Democratization,” International Security,
Vol. 37, No. 4, 2013, pp. 90–131.
Doyle, Michael W., and Nicholas Sambanis, “International Peacebuilding: A
Theoretical and Quantitative Analysis,” American Political Science Review, Vol. 94,
No. 4, December 2000, pp. 779–801.
Easterly, William, Shanker Satyanath, and Daniel Berger, Superpower Interventions
and Their Consequences for Democracy: An Empirical Inquiry, Cambridge, Mass.:
National Bureau of Economic Research, Working Paper No. 13992, May 2008.
Edelstein, David M., “Occupational Hazards: Why Military Occupations Succeed
or Fail,” International Security, Vol. 29, No. 1, Summer 2004, pp. 49–91.
Edelstein, David M., “Foreign Militaries, Sustainable Institutions, and Postwar
Statebuilding,” in Roland Paris and Timothy D. Sisk, eds., The Dilemmas of
Statebuilding: Confronting the Contradictions of Postwar Peace Operations, New
York: Routledge, 2009, pp. 81–103.
Elbadawi, Ibrahim A., and Nicholas Sambanis, External Interventions and the
Duration of Civil Wars, Washington, D.C.: World Bank, September 2000.
174 Assessing Trade-Offs in U.S. Military Intervention Decisions
Elster, Jon, Logic and Society: Contradictions and Possible Worlds, New York: John
Wiley, 1978.
Exum, Andrew, “What’s Really at Stake for America in Yemen’s Conflict,” The
Atlantic, April 14, 2017. As of March 5, 2020:
https://www.theatlantic.com/international/archive/2017/04/
yemen-trump-aqap/522957/
Fearon, James D., “Counterfactuals and Hypothesis Testing in Political Science,”
World Politics, Vol. 43, No. 2, January 1991, pp. 169–195.
Fettweis, Christopher J., “Credibility and the War on Terror,” Political Science
Quarterly, Vol. 122, No. 4, 2007/2008, pp. 607–633.
Florig, Dennis, “Hegemonic Overreach vs. Imperial Overstretch,” Review of
International Studies, Vol. 36, No. 4, October 2010, pp. 1103–1119.
Fogel, Robert William, Railroads and American Economic Growth: Essays in
Econometric History, Baltimore, Md.: Johns Hopkins Press, 1964.
Fortna, Virginia Page, “Does Peacekeeping Keep Peace? International Intervention
and the Duration of Peace After Civil War,” International Studies Quarterly,
Vol. 48, No. 2, June 2004, pp. 269–292.
Frederick, Bryan, Stephen Watts, Matthew Lane, Abby Doll, Ashley L. Rhoades,
and Meagan L. Smith, Understanding the Deterrent Impact of U.S. Overseas Forces,
Santa Monica, Calif.: RAND Corporation, RR-2533-A, 2020. As of February 10,
2020:
https://www.rand.org/pubs/research_reports/RR2533.html
Freedman, Lawrence, and Efraim Karsh, The Gulf Conflict, 1990–1991: Diplomacy
and War in the New World Order, Princeton, N.J.: Princeton University Press,
1993.
Gartenstein-Ross, Daveed, “Successes and Failures of the U.S. and NATO
Intervention in Libya,” testimony presented before the Committee on Oversight
and Government Reform, Washington, D.C.: Foundation for Defense of
Democracies, May 1, 2014.
Gartner, Scott Sigmund, and Gary M. Segura, “War, Casualties, and Public
Opinion,” Journal of Conflict Resolution, Vol. 42, No. 3, June 1998, pp. 278–300.
George, Alexander L., and Andrew Bennett, Case Studies and Theory Development
in the Social Sciences, Cambridge, Mass.: MIT Press, 2005.
Gibbons, Phillip, “U.S. No-Fly Zones in Iraq: To What End?” Washington, D.C:
Washington Institute, July 1, 2002. As of November 12, 2018:
https://www.washingtoninstitute.org/policy-analysis/
view/u.s.-no-fly-zones-in-iraq-to-what-end
References 175
Kelly, John H., “Lebanon: 1982–1984,” in Jeremy R. Azrael and Emil A. Payin,
eds., U.S. and Russian Policymaking with Respect to the Use of Force, Santa Monica,
Calif.: RAND Corporation, CF-129-CRES, 1996. As of May 1, 2020:
https://www.rand.org/pubs/conf_proceedings/CF129.html
Kilcullen, David, The Accidental Guerrilla: Fighting Small Wars in the Midst of a
Big One, New York: Oxford University Press, 2009.
Kissinger, Henry A., “U.S. Options in Biafra Relief,” memorandum to President
Richard Nixon, Washington, D.C., January 28, 1969.
Kocher, Matthew Adam, Adria Lawrence, and Nuno P. Monteiro, The Rabbit in
the Hat: Nationalism and Resistance to Foreign Occupation, Chicago, Ill.: American
Political Science Association 2013 Annual Meeting, August 20, 2013.
Krepon, Michael, and Mishi Faruqee, eds., Conflict Prevention and Confidence-
Building Measures in South Asia: The 1990 Crisis, Washington, D.C.: Henry L.
Stimson Center, Occasional Paper No. 17, April 1994. As of November 12, 2018:
https://www.stimson.org/sites/default/files/file-attachments/
occasionalpaper17-web_1.pdf
Kreps, Sarah E., “Multilateral Military Interventions: Theory and
Practice,” Political Science Quarterly, Vol. 123, No. 4, Winter 2008/2009,
pp. 573–603.
Larson, Eric V., Casualties and Consensus: The Historical Role of Casualties in
Domestic Support for U.S. Military Operations, Santa Monica, Calif.: RAND
Corporation, MR-726-RC, 1996. As of January 29, 2020:
https://www.rand.org/pubs/monograph_reports/MR726.html
Laub, Zachary, and Kali Robinson, “What Is the Status of the Iran Nuclear
Agreement?” Council on Foreign Relations, updated January 7, 2020. As of
June 3, 2020:
https://www.cfr.org/backgrounder/impact-iran-nuclear-agreement
“Leader of Grenada Is Reported Killed by Troops,” New York Times, October 20,
1983.
Lebovic, James H., and Ashfaq Ishaq, “Military Burden, Security Needs, and
Economic Growth in the Middle East,” Journal of Conflict Resolution, Vol. 31,
No. 1, March 1987, pp. 106–138.
Lebow, Richard Ned, “What’s So Different About a Counterfactual?” World
Politics, Vol. 52, No. 4, July 2000, pp. 550–585.
LeoGrande, William M., Our Own Backyard: The United States in Central
America, 1977–1992, Chapel Hill, N.C.: University of North Carolina Press,
1998.
178 Assessing Trade-Offs in U.S. Military Intervention Decisions
Reagan, Ronald, “Remarks of the President and Prime Minister Eugenia Charles
of Dominica Announcing the Deployment of United States Forces in Grenada,”
October 25, 1983. As of May 1, 2020:
https://www.reaganlibrary.gov/research/speeches/102583a
Regan, Patrick M., “Conditions of Successful Third-Party Intervention in
Intrastate Conflicts,” Journal of Conflict Resolution, Vol. 40, No. 2, June 1996,
pp. 336–359.
Regan, Patrick M., and Aysegul Aydin, “Diplomacy and Other Forms of
Intervention in Civil Wars,” Journal of Conflict Resolution, Vol. 50, No. 5, October
2006, pp. 736–756.
Regan, Patrick M., and Allan C. Stam, “In the Nick of Time: Conflict
Management, Mediation Timing, and the Duration of Interstate
Disputes,” International Studies Quarterly, Vol. 44, No. 2, June 2000,
pp. 239–260.
Rice, Susan E., National Security Council, “Updated Talking Points on Rwanda,”
memorandum to W. Anthony K. Lake et al., Washington, D.C., May 7, 1994.
Rice, Susan E., and Nick Rasmussen, “Deputies Committee Meeting on African
Peacekeeping Issues,” memorandum to Samuel R. Berger, Washington, D.C.,
September 20, 1993.
Robinson, Piers, “The CNN Effect: Can the News Media Drive Foreign
Policy?” Review of International Studies, Vol. 25, No. 2, 1999, pp. 301–309.
Rose, Gideon, “What Obama Gets Right,” Foreign Affairs, September/October
2015. As of March 5, 2020:
https://www.foreignaffairs.com/articles/2017-07-05/what-obama-gets-right
Rudd, Gordon W., Humanitarian Intervention: Assisting the Iraqi Kurds in
Operation PROVIDE COMFORT, 1991, Washington, D.C.: Department of the
Army, 2004.
Russell, Edward T., “Cuban Missile Crisis,” in A. Timothy Warnock, ed., Short of
War: Major USAF Contingency Operations, 1947–1997, Maxwell Air Force Base,
Ala.: Air University Press, 2000, pp. 33–42.
Russett, Bruce, “Economic Decline, Electoral Pressure, and the Initiation of
Interstate Conflict,” in Charles S. Gochman and Alan Ned Sabrosky, eds.,
Prisoners of War? Nation-States in the Modern Era, Lexington, Mass.: Lexington
Books, 1990, pp. 123–140.
Salmoni, Barak, Bryce Loidolt, and Madeleine Wells, Regime and Periphery
in Northern Yemen: The Huthi Phenomenon, Santa Monica, Calif.: RAND
Corporation, MG-962-DIA, 2010. As of April 30, 2020:
https://www.rand.org/pubs/monographs/MG962.html
182 Assessing Trade-Offs in U.S. Military Intervention Decisions
Sanchez, Fernando F., Democracy After War: Causes and Consequences of the 1948
Civil War in Costa Rica, paper presented at the European Consortium for Political
Research Joint Sessions of Workshops, Uppsala, Sweden: Uppsala University,
April 13–18, 2004.
Sanger, David E., “Why Not a Strike on Iran?” New York Times, January 22,
2006.
Schelling, Thomas C., The Strategy of Conflict, Cambridge, Mass.: Harvard
University Press, 1960.
———, Arms and Influence, New Haven, Conn.: Yale University Press, 1966.
Schlesinger, Stephen, “Ghosts of Guatemala’s Past,” New York Times, June 3, 2011.
Schmitt, Eric, “Obama Is Pressed to Open Military Front Against ISIS in Libya,”
New York Times, February 4, 2016.
Seybolt, Taylor B., Humanitarian Military Intervention: The Conditions for Success
and Failure, Oxford, United Kingdom: Stockholm International Peace Research
Institute and Oxford University Press, 2007.
Siegel, Adam B., The Use of Naval Forces in the Post-War Era: U.S. Navy and U.S.
Marine Corps Crisis Response Activity, 1946–1990, Alexandria, Va.: CNA, 1991.
Simpson, Bradley R., Economists with Guns: Authoritarian Development and U.S.-
Indonesian Relations, 1960–1968, Stanford, Calif.: Stanford University Press,
2008.
Stedman, Stephen John, Peacemaking in Civil War: International Mediation in
Zimbabwe, 1974–1980, Boulder, Colo.: Lynne Reinner, 1991.
Stewart, Richard, Operation Urgent Fury: The Invasion of Grenada, October 1983,
Washington, D.C.: U.S. Army Center of Military History, 2008.
Streeter, Stephen M., “Interpreting the 1954 U.S. Intervention in Guatemala:
Realist, Revisionist, and Postrevisionist Perspectives,” The History Teacher, Vol. 34,
No. 1, November 2000, pp. 61–74.
Sullivan, Kevin, and Mary Jordan, “In Central America, Reagan Remains a
Polarizing Figure,” Washington Post, June 10, 2004.
Sullivan, Patricia L., and Johannes Karreth, “The Conditional Impact of Military
Intervention on Internal Armed Conflict Outcomes,” Conflict Management and
Peace Science, Vol. 32, No. 3, 2015, pp. 269–288.
Tetlock, Philip E., and Aaron Belkin, eds., Counterfactual Thought Experiments
in World Politics: Logical, Methodological, and Psychological Perspectives, Princeton,
N.J.: Princeton University Press, 1996.
Tetlock, Philip E., and Richard Ned Lebow, “Poking Counterfactual Holes in
Covering Laws: Cognitive Styles and Historical Reasoning,” American Political
Science Review, Vol. 95, No. 4, December 2001, pp. 829–843.
References 183
Tierney, Dominic, “Syria and the Cycle of American Intervention,” The Atlantic,
October 24, 2016.
Treaster, Joseph B., “20 Years After Dominican War, Wounds Linger,” New York
Times, May 1, 1985. As of February 8, 2018:
http://www.nytimes.com/1985/05/01/world/
20-years-after-dominican-war-wounds-linger.html
Turner, Desmond A., Civil War and (Re)Building the State: Recovery in Costa Rica,
Nicaragua, and El Salvador, thesis, Coral Gables, Fla.: University of Miami, May
2017.
United Nations, “Rwanda: A Brief History of the Country,” webpage, undated a.
As of May 1, 2020:
https://www.un.org/en/preventgenocide/rwanda/historical-background.shtml
———, “UN Comtrade Database,” online database, undated b. As of April 15,
2018:
https://comtrade.un.org/data/
U.S. Agency for International Development, “The Greenbook,” webpage, updated
April 29, 2013. As of October 12, 2018:
https://www.usaid.gov/developer/greenbookapi
———, U.S. Overseas Loans and Grants: Obligations and Loan Authorizations,
July 1, 1945–September 30, 2014, Washington, D.C., 2015.
“U.S. Aid to the Contras,” Washington Post, February 3, 1988. As of May 1, 2020:
https://www.washingtonpost.com/archive/politics/1988/02/03/
us-aid-to-the-contras/842e041b-e9b3-4741-a787-6eae1f3a34da/
U.S. Department of Defense, “Combined Joint Task Force Operation Inherent
Resolve,” webpage, undated. As of May 1, 2020:
https://www.defense.gov/OIR/
———, Summary of the 2018 National Defense Strategy of the United States of
America: Sharpening the American Military’s Competitive Edge, Washington, D.C.,
2018. As of January 30, 2020:
https://dod.defense.gov/Portals/1/Documents/pubs/
2018-National-Defense-Strategy-Summary.pdf
U.S. Department of Defense, Security Cooperation Agency, “Fiscal Year Series:
Foreign Military Sales, Foreign Military Construction Sales and Other Security
Cooperation Historical Facts,” September 30, 2016. As of November 11, 2018:
http://www.dsca.mil/sites/default/files/
fiscal_year_series_-_30_september_2016.pdf
U.S. Department of State, “Milestones in the History of U.S. Foreign Relations:
Central America, 1977–1980,” webpage, undated. As of May 1, 2020:
https://history.state.gov/milestones/1977-1980/central-america-carter
184 Assessing Trade-Offs in U.S. Military Intervention Decisions
Weber, Max, On the Methodology of the Social Sciences, ed. and trans. Edward A.
Shils and Henry A. Finch, Glencoe, Ill.: Free Press, 1949.
Weinberger, Caspar W., Fighting for Peace: Seven Critical Years in the Pentagon,
New York: Warner Books, 1990.
Western, Jon, “Sources of Humanitarian Intervention: Beliefs, Information, and
Advocacy in U.S. Decisions on Somalia and Bosnia,” in James M. McCormick,
ed., The Domestic Sources of American Foreign Policy: Insights and Evidence, 6th ed.,
New York: Rowman and Littlefield Publishers, 2012, pp. 399–416.
Western, Jon, and Joshua S. Goldstein, “Humanitarian Intervention Comes of
Age: Lessons from Somalia to Libya,” Foreign Affairs, Vol. 90, No. 6, November/
December 2011, pp. 48–59.
Wittman, Donald, “How a War Ends: A Rational Model Approach,” Journal of
Conflict Resolution, Vol. 23, No. 4, December 1979, pp. 743–763.
Woodward, Bob, Plan of Attack, New York: Simon and Schuster, 2004.
World Bank, “Mortality Rate, Infant (Per 1,000 Live Births), 1960–2017,” online
database, undated. As of January 30, 2020:
https://data.worldbank.org/indicator/SP.DYN.IMRT.IN
World Peace Foundation, “Colombia: La Violencia,” Fletcher School, Tufts
University, December 14, 2016. As of May 2018:
https://sites.tufts.edu/atrocityendings/2016/12/14/colombia-la-violencia-2/
Yacoubian, Mona, Critical Junctures in United States Policy Toward Syria: An
Assessment of the Counterfactuals, Washington, D.C.: Simon-Skjodt Center for the
Prevention of Genocide, August 2017.
Yao, Julio, “Legacies of the U.S. Invasion of Panama,” North American Congress
on Latin America, March 22, 2012.
Yates, Lawrence A., Power Pack: U.S.. Intervention in the Dominican Republic,
1965–1966, Fort Leavenworth, Kan.: Combat Studies Institute, U.S. Army
Command and General Staff College, 1988.
Yildirim, Jülide, and Nadir Öcal, “Military Expenditures, Economic Growth and
Spatial Spillovers,” Defence and Peace Economics, Vol. 27, No. 1, 2016, pp. 87–104.
Young, John M., When the Russians Blinked: The U.S. Maritime Response to
the Cuban Missile Crisis, Washington, D.C.: History and Museums Division,
Headquarters, U.S. Marine Corps, 1990.
Zartman, I. William, and Saadia Touval, “International Mediation: Conflict
Resolution and Power Politics,” Journal of Social Issues, Vol. 41, No. 2, 1985,
pp. 27–45.
AR R OY O C ENTER
W
hen faced with the outbreak of a war or the onset of an
international crisis that affects important U.S. interests,
U.S. policymakers must consider how best to respond.
One option they may consider is to directly intervene
militarily in the war or crisis. In this report, the authors
create a framework that can be used to rigorously consider the trade-offs
between intervening militarily early in a war or crisis, intervening later, and not
intervening at all, as well as the trade-offs involved in decisions regarding the
size of the potential intervention force to be employed. This framework can
provide a better understanding of the relationships between intervention timing,
intervention size, and intervention outcomes to inform future debates about
whether and when to undertake a military intervention.
The authors approached these issues in four ways. They conducted (1) an
extensive review of academic and policy literature on relevant topics, (2) a
quantitative analysis using a database of 286 crises and wars since 1945 in
which important U.S. interests were at stake, (3) a set of 45 short, focused
case studies, including both interventions and noninterventions, and (4) a set
of four in-depth counterfactual illustrations, in which U.S. intervention decisions
were altered from historical events to explore the implications. The findings
from these four research approaches lead to recommendations regarding the
contextual factors that are most essential for policymakers to consider when
making intervention decisions.
$37.00
ISBN-10 1-9774-0506-1
ISBN-13 978-1-9774-0506-7
53700
RR-4293-A