University of Southern California Dissertations and Theses

Pathways to nuclear weapon reversal: exploring mechanisms and understanding non-proliferation policy

(USC Thesis Other)

Pathways to nuclear weapon reversal: exploring mechanisms and understanding non-proliferation policy

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2023 Edward Gonzalez

Pathways to Nuclear Weapon Reversal:

Exploring Mechanisms and Understanding Non-proliferation Policy

by

Edward Gonzalez

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLITICAL SCIENCE AND INTERNATIONAL RELATIONS)

May 2023

ii

Acknowledgments
Completion of this dissertation would not have been possible without the emotional and
academic support from various individuals both within and outside the University of Southern
California.
First, I want to thank my advisor and dissertation chair Patrick James, whose support throughout
all my years at USC was crucial for my academic success. Patrick James oversaw my
dissertation at various stages of completion, read countless drafts, and provided excellent
guidance for the final product. I also want to thank my committee members, Wayne Sandholtz
and Najmedin Meshkati, who graciously accepted to be members of my dissertation committee.
Wayne Sandholtz’ comments on earlier versions of my theory and measurement challenged me
in constructive ways, for which I am most grateful.
I also give my gratitude to the past and present faculty and staff of the POIR department for
making my time at USC delightful. I would like to thank professor Gerardo Munck, who always
believed in me and offered important support during my dissertation writing. James Lo, for his
assistance with my R Studio related troubles. I also thank the USC Center for International
Studies (CIS), for their generous economic funding and feedback opportunities at the workshops.
I would not have been able to hire Jonah Buessing, my amazing RA for the Summer 2021 term,
without CIS support.
In addition to the POIR department’s generous support, I must also give thanks to the various
PhD students who I befriended at USC. I thank Kyle Reed for always being willing to discuss
ideas and read my drafts. Jackie, who frequently checked on me and pushed me to clarify my
puzzle. Paul Orner, for always being willing to help me with my R code whenever I was stuck,
and his virtual hangouts during the covid era. Alya Khemakhem, for believing in me in those
instances I doubted myself. The rest of the 2017 USC POIR Cohort, for being great colleagues
and friends throughout the process and making my time at USC a fun experience.
Professors from my previous academic institutions also deserve public recognition. My journey
began at CSUDH, where I met Wayne Martin, the first professor of international relations I had.
He instilled in me a love for the subject through his enthusiastic teaching. I also thank Richard
Palmer, Jay Kaplan, Annie Whetmore, and Hamoud Salhi, excellent professors that made my
undergraduate education memorable. I also give thanks to CSULB, where I earned my M.A. in
Political Science, for their support throughout the PhD application process back in 2015. In
particular, I owe my gratitude to Professor Larry George for his unwavering support.
Finally, I owe my success to my personal support network, family and friends who backed me
throughout the entire process. To my friends Jonathan Albornoz, Christopher Moya, Alejandra
Perez, Grace Malinao, Manuel Lomeli, and countless others-thank you for being a source of
entertainment to balance my academic studies. To my CSULB crew, Steven Gonzales, Amanda
Mendoza, Alex Reyes, Adam Collins, Joshua Casper, Spencer Lindsey, thank you for listening to
my rants. Finally, to my family, whose support was the most important of all. My mother,
Mercedes Gonzales, my father, Hector Gonzales, and my sisters, Valentine and Beverly, thank
you for your unwavering support. This dissertation is for you.
iii

TABLE OF CONTENTS
Acknowledgements ........................................................................................................................ ii
List of Tables .................................................................................................................................. v
List of Figures ................................................................................................................................ vi
Abstract ......................................................................................................................................... vii
Chapter 1: Introduction ....................................................................................................................1
1.1. Outline of dissertation .................................................................................................4
Chapter 2: Theory of Nuclear Reversals ..........................................................................................7
2.1 Short summary of theory .............................................................................................7
2.2. Existing research on nuclear proliferation and reversals ............................................8
2.3. Defining nuclear reversals and scope conditions ......................................................11
2.4. Voluntary pathway of nuclear reversals ...................................................................14
2.5. Involuntary pathway of nuclear reversals .................................................................21
2.6. How the security environment affects nuclear reversals ..........................................26
Chapter 3: Research Design ...........................................................................................................28
Chapter 4: New Measure of Nuclear Non-Proliferation Norm Acceptance and Quantitative
Models............................................................................................................................................39
4.1. Measuring acceptance of anti-nuclear weapon norms ..............................................39
4.2. The UNGA anti-nuclear voting record dataset .........................................................41
4.3. Methodology and measurements ..............................................................................51
4.4. Empirical results .......................................................................................................56
4.5. Conclusion ................................................................................................................65
Chapter 5: Iraq as a Case of Involuntary Reversal ........................................................................68
5.1. Iraq’s nuclear program: 1956-1998 ..........................................................................69
5.2. Iraq’s regime type and perception of threat ..............................................................77
5.3. Testing observable implications ...............................................................................81
5.4. Conclusion ..............................................................................................................100
Chapter 6: Brazil as a Case of Voluntary Reversal......................................................................106
6.1. Brazil’s nuclear program:1951-1994 ......................................................................107
6.2. Brazil’s regime type and perception of threat .........................................................119
6.3. Testing observable implications .............................................................................124
6.4. Conclusion ..............................................................................................................148
Chapter 7: Israel as a Case of Non-Reversal ...............................................................................154
7.1. Israel’s nuclear program: 1948-1967 ......................................................................155
7.2. Israel’s regime type and perception of threat ..........................................................163
7.3. Testing observable implications .............................................................................167
iv

7.4. Conclusion ..............................................................................................................181
Chapter 8: Conclusion: Lessons of the Two-Pathway Theory of Nuclear Reversals ..................185
8.1. Summary of findings...............................................................................................187
8.2. Contributions and lessons for future reversals ........................................................194
8.3. Next steps for future research .................................................................................199
Bibliography ................................................................................................................................201
Appendices ...................................................................................................................................210
Appendix A: Nuclear reversals and nuclear weapon states, 1945-2020 .......................210

v

List of Tables
Table 1: Nuclear Reversal Expectations ........................................................................................27
Table 2: Distribution of UNGA Voting Categories .......................................................................45
Table 3: Descriptive Statistics of State Non-Proliferation Acceptance Data ................................50
Table 4: Logistic Regression Models ............................................................................................59
Table 5: Robustness Checks ..........................................................................................................60
Table 6: NPT Membership Models................................................................................................62
Table 7: Cox Hazard Models .........................................................................................................64
Table 8: Summary of Results: Iraq ..............................................................................................105
Table 9: Summary of Results: Brazil ...........................................................................................153
Table 10: Summary of Results: Israel ..........................................................................................184

vi

List of Figures
Figure 1: Snapshot of the Data #1..................................................................................................43
Figure 2: Snapshot of the Data #2..................................................................................................44
Figure 3: Resolutions per Year ......................................................................................................46
Figure 4: Annual Proportion of “Yes” Votes .................................................................................48

vii

Abstract
Why do states reverse their nuclear weapons programs? Much has been written on why states might not
pursue nuclear weapons, but relatively little about state’s decisions to reverse course and terminate their
nuclear weapons programs, a puzzling decision considering the strategic advantages of obtaining nuclear
weapons in the international system. Theoretically, it is puzzling why states that have invested time and
money in pursuit of nuclear weapons would give up their efforts, despite the sunken costs involved and
the strategic advantages of nuclear weapon ownership. This project presents a theory of nuclear weapons
reversals to explain why countries give up their efforts to pursue the nuclear bomb, arguing that there are
two main causal mechanisms that explain nuclear reversals: an involuntary and a voluntary pathway.
Countries that follow the involuntary pathway give up pursuit of nuclear weapons due to threats or
coercive force. Countries that follow the voluntary pathway give up pursuit of nuclear weapons due to
acceptance of nuclear non-proliferation norms, norms that are accepted if two conditions are met: I
hypothesize that when states perceive a non-threatening security environment, and when states have
higher levels of democracy, states will be more likely to be influenced by international norms against
nuclear weapons acquisition, and thus more likely to terminate their nuclear weapons development
program. This project seeks to 1) develop a theory of nuclear reversals based on an analytical eclectic
approach, drawing from realist, liberal, and constructivist insights, 2) develop a proxy measure of anti-
nuclear proliferation norm acceptance, 3) use quantitative methods to show there is a relationship between
norm acceptance and reversals, and 4) show through the use of case studies how the voluntary and
involuntary mechanisms work in real world cases of nuclear reversals and non-reversals.

1

Chapter 1: Introduction

Brazil ratified the Treaty of Tlatelolco and has reiterated in international forums
its respect for its purposes and those of related international agreements…That
pioneering Latin American initiative in the field of non-proliferation of nuclear weapons
preceded all other efforts, whether regional or universal in scope, devoted to the same
objective. I believe, however, that the time has come to go one step further. Brazil today
discards the idea of any experiments that might involve nuclear explosions, even if only
for peaceful purposes. We trust other nations will consider the possibility of following the
same path.
1

With the above statement, Brazil’s President Fernando Collor de Mello publicly and
unequivocally renounced peaceful nuclear explosions, in front of the United Nations General
Assembly during the opening speech of the forty-fifth UNGA session on September 27, 1990.
The statement signaled to the international community one more step in Brazil’s integration into
the nuclear non-proliferation regime, and further solidified Brazil’s public stance against nuclear
weapon proliferation. This development was significant, for not that long ago American scholars
and the intelligence community feared that Brazil planned to pursue the nuclear bomb during its
military regime.
Brazil evolved from a country that developed a covert nuclear program in the late 70s and
was believed to be in pursuit of the bomb to a country that explicitly embraced international
norms against nuclear weapon proliferation and ownership in the 90s. Cases such as Brazil,
which are either pursuing or have the capacity to pursue the nuclear bomb, yet dismantled their
nuclear programs or abandoned their nuclear ambitions, can be categorized as cases of nuclear
reversal: defined as instances in which countries that were pursuing nuclear weapons terminate

1
Excerpt from Brazil’s President Fernando Collor de Mello UNGA address. U.N. G.A.O.R., 45
th
Sess., U.N. Doc
A/45/PV.4 (Sept. 27, 1990). Available at: https://documents-dds-
ny.un.org/doc/UNDOC/PRO/N90/640/94/PDF/N9064094.pdf?OpenElement
2

pursuit of the bomb, by their own decision or due to coercion. This phenomenon of nuclear
reversals is puzzling, due to the benefits of being a nuclear weapon state.
The phenomenon of nuclear reversals leads us to ask an important question. If nuclear
weapon ownership has strategic value in an anarchic system, why do states already in pursuit of
nuclear weapons decide to reverse course, abandoning their search for the nuclear bomb? Most
research on nuclear weapons focuses on the causes and motives of nuclear proliferation.
2

However, relatively little research exists that focuses exclusively on why countries pursuing
nuclear weapons decide to reverse course. Nuclear weapon program abandonment is puzzling for
two main reasons: first, nuclear weapon programs are costly, so abandoning them means
resources and time has been wasted.
3
Second, nuclear weapons are believed to have deterrence
value, and can be an important symbol of state power. Yet, about half to more than half—
depending on your count—of countries that pursued nuclear weapons gave up pursuit.
4
Given the
benefits of nuclear weapon ownership, studying the causes for why countries might be willing to
give up pursuit of nuclear weapons is essential.
Thus, this dissertation tackles the above underexplored area of nuclear politics, filling a
gap in current understandings of nuclear weapon reversal dynamics. In contrast to security-based
arguments for explaining abandonment of nuclear pursuit, and liberal theories emphasizing
international treaties and economic factors, this project proposes an eclectic theory that merges
insights from realist, liberal, and constructivist paradigms to explain nuclear reversals, focusing

2
Epstein (1977); Thayer (1995); Sagan (1996/97); Hymans (2006); Singh and Way (2004); Jo and Gartzke (2007);
Debs and Monteiro (2017). These are just a few of the various studies on nuclear proliferation.
3
Admittedly, progress in nuclear weapons program can have positive consequences in technological development
even after abandonment of a program, although as a by-product rather than main goal of the program.
4
Hymans (2012, 5) counts 17 cases of nuclear weapons pursuit, of which ten were successful; of these ten, South
Africa eventually dismantled its nuclear program in 1989. Mehta (2020, 21-25) counts 32 cases of nuclear weapons
pursuers.
3

on the impact of international norms on state behavior.
5
Specifically, I develop an ideational
theory of nuclear reversals that centers the effects of international norms on state decisions to
give up pursuit of the nuclear bomb, without ignoring material or institutional variables that play
a key part in the nuclear reversal process.
The theory introduced in this dissertation, which I call a Two-Pathway Theory of Nuclear
Reversals, argues that there are two main causal mechanisms leading to the same outcome of
nuclear weapon reversal: a voluntary and an involuntary pathway. Countries that are forced into
the involuntary pathway terminate pursuit of nuclear weapons due to threats from other states or
coercive force used against them. However, states that follow the voluntary pathway of nuclear
reversals decide to give up their nuclear weapon pursuit due to internalization, or at least
grudging acceptance, of international norms against nuclear weapon ownership and proliferation.
Having a strong democracy and positive perceptions of the security environment are
prerequisites for successful acceptance of anti-nuclear weapon norms.
In addition to the theoretical contribution to explain the puzzle of nuclear reversals, this
dissertation also contributes brand new data to the field of nuclear studies. In this project, I
develop a novel measurement of nuclear non-proliferation norm acceptance, using UNGA voting
records on nuclear related resolutions to build the measure. Contemporary studies attempt to
measure norm acceptance of states by looking at their membership in the Treaty on the Non-
Proliferation of Nuclear Weapons (NPT). However, membership in the NPT is an imperfect
measure of norm acceptance, due to states joining the NPT for several reasons unrelated to
accepting such norm: states might join the NPT to hide their true nuclear ambitions from the rest

5
The theory developed here draws from the analytic eclecticism approach from Sil and Katzenstein (2010),
acknowledging the need to incorporate variables from multiple paradigms into the theory for fullness of explanation.
4

of the international community or join due to coercion from powerful states.
6
Thus, a new
measure of non-proliferation norm acceptance that accounts for shifts in support for these norms
is required for better statistical analysis of how non-proliferation norms might affect the
likelihood of nuclear reversals.
1) Outline of Dissertation
In short, this dissertation tackles the puzzle of nuclear reversals, exploring why countries
that were seemingly on the path towards obtaining the nuclear bomb terminated their nuclear
weapons program or ambition. To do so, I contribute both a new measurement of nuclear non-
proliferation norm acceptance and a novel two-pathway theory of nuclear reversals, focusing on
the effect that international norms might have on state decisions to give up pursuit of the nuclear
bomb. I find that the two-pathway theory of nuclear reversals developed in this project can help
us better understand the dynamics of nuclear reversals and reveals that international norms do
have a real impact on state decision making even in the nuclear realm.
The dissertation is organized as follows: Chapter 2 presents the Two-Pathway Theory of
Nuclear Reversals. In this chapter, I justify why it is necessary to develop a new theory to
explain nuclear reversals, define what I mean by nuclear reversals and acceptance of
international non-proliferation norms, and delineate the logic of the two causal mechanisms for
nuclear reversal. Chapter 3 discusses the research design of this project and how I test the theory
of nuclear reversals. The theory is tested via a mixed-methods approach leveraging the
advantages of both quantitative and qualitative research designs. Via quantitative tests, I discover
if there is a positive relationship between non-proliferation norm acceptance and nuclear

6
These reasons are highlighted by Montgomery and Sagan (2009, 313).
5

reversals. With qualitative case studies, I can examine how the two pathways of nuclear reversal
proposed by the theory might explain real world cases. Finally, the chapter justifies the selection
of the three cases for qualitative analysis, Brazil, Iraq, and Israel.
Chapter 4, the quantitative chapter, introduces the new measure of nuclear non-
proliferation norm acceptance, providing description of the data collection process and the newly
constructed variable. After introducing the new variable, I use logistic regression and Cox hazard
models suited for observational data to test whether acceptance of nuclear non-proliferation
norms, as measured by my new variable, increases the likelihood of nuclear reversals. Aware
that quantitative tests do not show how the causal mechanisms of the two reversal pathways
work in practice, the dissertation includes three case studies to test the validity of the causal
mechanisms advanced by the theory.
Chapter 5 is a case study of Iraq. In this chapter, I analyze how the involuntary reversal
pathway can explain the process by which Iraq was forced to stop pursuit of the nuclear bomb.
Chapter 6 explores the nuclear program of Brazil, showing how the voluntary reversal pathway
might explain the puzzling case of a country that seemed to be on the path to obtaining the bomb
before deciding to dismantle their covert nuclear program.
7
Chapter 7 looks at a case of nuclear
non-reversal, Israel. If the theoretical framework is correct, then it should also be able to explain
cases of reversal failure. These three chapters rely on process tracing techniques and test
explicitly stated hypotheses—labeled observable implications throughout the dissertation—to
maintain rigor in the qualitative analysis and minimize personal biases.

7
Interestingly, and contrary to plenty of research, it appears that Brazil’s top leadership never attempted to pursue
the nuclear bomb. Why foreign analysts believed that Brazil was pursuing the bomb, and why we should still
consider Brazil a case of nuclear reversals, is explored in-depth in the chapter.
6

Finally, Chapter 8 concludes the project, providing a summary of the quantitative and
qualitative findings and how the Two-Pathway Theory of Nuclear Reversals fares after empirical
testing. Furthermore, the chapter draws policy implications from the theory and how they might
apply to the current proliferation case of Iran. The chapter ends with a discussion of potential
paths for future research and how future iterations of this project might move forward.

7

Chapter 2: Theory of Nuclear Reversals
1) Short summary of theory
I argue there are two main pathways to nuclear weapons reversals: an involuntary and a
voluntary pathway, with international norms against nuclear weapons proliferation playing a role
in both pathways, although affecting states in diverse ways. Countries along the involuntary
pathway give up pursuit of nuclear weapons due to coercion from other states, either in the form
of sanctions, threats of force, or actual use of force. Although the states being forced to reverse
their nuclear weapons program reject norms against nuclear weapons proliferation, international
norms are not entirely absent from the picture: the states coercing the proliferator do so to uphold
norms against nuclear weapons proliferation, using coercion to enforce the norm against
proliferators. In this involuntary pathway, norm internalization or acceptance may be absent from
the proliferating states, but those doing the enforcing have accepted the norm.
Countries along the voluntary pathway, on the other hand, give up pursuit of nuclear
weapons by their own volition, influenced by international norms against nuclear weapon
ownership and proliferation. These international norms will be more likely to influence a state if
two key conditions are met: the state faces a relatively benign security environment, and so does
not feel threatened by other states, and if the state considers itself a liberal state, or wants to be
included, within the liberal international order. Of course, norms against nuclear weapons do not
magically work by themselves-they have to be supplanted with benefits offered by other states so
the proliferator’s decision to abandon nuclear weapon pursuit does not lead to a total loss for the
state. International pressure to adopt non-proliferation norms might also be present as well.
However, unlike the involuntary pathway, international pressure remains limited to public
condemnation and shaming mechanisms, without including threats of sanctions or use of force.
8

In sum, the two pathways differ in degree regarding how much agency a state has in the
decision to give up nuclear weapons pursuit; the involuntary pathway attributes reversals to the
pressure exerted by other actors in the international system, while the voluntary pathway allows
for more agency in state decision to revert, given a few permissive conditions. This two-pathway
framework recognizes that there is no single explanation that could cover every single instance
of nuclear weapons reversal (or non-reversal)—case studies will reveal that the historical and
political context of each individual case will make every instance of nuclear weapons
abandonment unique. Yet, the two-pathway framework assumes that, at a broader level of
analysis, we may be able to group nuclear reversals into two general categories based on how
much agency states ultimately have in their decision to revert, a decision that will be influenced
to varying degrees by whether the state internalized or accepted nuclear non-proliferation and
non-ownership norms.
2) Existing research on nuclear proliferation and reversals
Early research on issues of nuclear weapons has focused on the motivations for nuclear
weapons pursuit. Mostly coming from the realist tradition, this line of work argued that basic
national security considerations explained why countries pursued nuclear weapons.
8
Most
famously delineated by Sagan, the security “model” of nuclear proliferation expects states that
cannot meet their security requirements by conventional means to pursue deterrence via nuclear
weapons. Either states develop their own nuclear weapon, if possible, or form alliances with
stronger, nuclear armed states that could provide a security umbrella.
9

8
Epstein (1977); Thayer (1995).
9
Sagan (1996/1997).
9

Realist theories of nuclear proliferation, however, overestimated the degree to which
nuclear weapons would spread. Contrary to predictions, the nuclear age has witnessed slow
proliferation of nuclear weapons; today, only nine countries are considered nuclear weapons
states.
10
The lack of nuclear weapons proliferation led to the rise of alternative, non-security
related explanations for why countries might want to obtain or refrain from obtaining these
weapons, focusing on psychological, economic, or domestic politics variables to explain
proliferation: State leaders might want to obtain nuclear weapons to increase their ranking in the
international system,
11
or leaders might avoid pursuing altogether if they want to integrate their
state in the international economic system.
12
Alternatively, regime type could play a role, with
authoritarian states more likely to pursue and obtain nuclear weapons.
13

However, compared to research that focuses on nuclear proliferation, little research exists
that focuses on nuclear reversals, or the decision by countries to roll back existing nuclear
weapons programs. Solingen’s work falls into this camp, for it explains not just nuclear
proliferation and non-proliferation, but also why countries might abandon their nuclear weapons
programs. Under this neoliberal perspective, a leader’s motive for economic integration and
growth, and increased interdependence with other states, mitigates the security dilemma,
lowering the incentives for obtaining nuclear weapons. A few realists have attempted to tackle
the question of nuclear reversals as well. For example, Reiss argues that when security threats
decrease, state’s nuclear programs wane, which leads to reversal of nuclear programs.
14
Most
famously within the nuclear proliferation literature, T.V. Paul’s Power and Prudence develops a

10
USA, UK, Russia, China, France, Pakistan, India, Israel, and North Korea. South Africa is the only country that
obtained and subsequently dismantled its nuclear weapons arsenal.
11
Hymans (2006).
12
Solingen (2007).
13
Way and Weeks (2014).
14
Reiss (1995).
10

realist theory that explains why countries might choose nuclear abandonment. For T.V. Paul,
leaders of states engage in cost-benefit calculations, informed by the security environment they
inhabit, before deciding if they should forgo nuclear weapons. If a state lies in a region of low
and moderate conflict, it is likely to forgo nuclear weapons to avoid starting an arms race in what
is a relatively safe regional context; the external environment matters for understanding nuclear
reversals.
15

A few works have approached the puzzle of nuclear reversals from a constructivist lens,
arguing that to understand nuclear reversals we need to look at international norms. For example,
Rublee argues that the creation of the Treaty on the Non-Proliferation of Nuclear Weapons
(NPT) in 1968, and other anti-nuclear weapons treaties, created an anti-nuclear norm that
increasingly came to define acceptable behavior of states.
16
Offering quantitative support for
Rublee’s thesis, Muller and Schmidt find that “since 1970, less than 15 percent of the
economically capable states have started nuclear weapons programs. Even more striking is that
since the inception of the nonproliferation norm, almost 70 percent of all nuclear weapons
activity states gave up those activities.”
17
Muller and Schmidt attribute the lack of nuclear
proliferation, as well as nuclear reversals, to the persistence of a non-nuclear weapon norm
growing in strength in the international system.
Finally, some theorists focus on the influence the United States has over country’s
decisions to terminate their nuclear weapons programs. For example, Mehta uses a mixed-
methods approach to show that countries are more likely to reverse their nuclear weapons
program if they are offered positive and negative inducements simultaneously, focusing on how

15
T.V. Paul (2000, 33).
16
Rublee (2009, 46.)
17
Muller and Schmidt (2010, 148).
11

American non-proliferation efforts might successfully pressure states to give up pursuit of
nuclear weapons.
18
Similarly, Schneider argues that U.S. social pressure over its allies is crucial
for successful nuclear reversals, and that these social pressures will be most effective when state
leaders perceive their nation as inferior in status compared to the United States.
19

Yet, most research on nuclear proliferation still focuses on nuclear proliferation and non-
proliferation, rather than nuclear reversals. In fact, it has not been until recently that research
began to focus on nuclear reversal itself as a separate and distinct process from nuclear
proliferation.
20
The theory presented here builds on the nascent body of research on nuclear
reversals, providing a new framework that incorporates material and normative factors in the
explanation for the puzzling phenomenon of nuclear weapons program abandonment without
ignoring or underemphasizing either material power or international norms. At the cost of some
theoretical parsimony, the theory presented here bridges the gap between different philosophical
schools of thought such as realism and constructivism, showing that rather than opposites, these
different perspectives might be complementary.
21
Rather than claim nuclear reversals occur just
because of coercion and the use of power by some states, or exclusively due to norm-
internalization, the theory claims both pathways may be correct, depending on the circumstances.
3) Defining nuclear reversals and scope conditions
An important distinction that must be made is that between countries that never pursued
nuclear weapons and those that initiated pursuit but subsequently disbanded their nuclear
weapons programs, what can be referred to as nuclear reversals. An assumption driving this

18
Mehta (2020).
19
Schneider (2020).
20
For recent work on nuclear weapons reversals, see Miller (2018), and Mattiaci and Jones (2016).
21
The theory presented here might be called “Eclectic” in its approach, formulating causal mechanisms from more
than one paradigm. See Sil and Katzenstein (2010).
12

project is that there is a qualitative distinction between countries that never pursued nuclear
weapons and countries that pursued the bomb but eventually dismantled their program. This
assumption is warranted because the causes and mechanisms behind nuclear reversals and
nuclear non-pursuit may be distinct. For instance, countries might decide not to pursue nuclear
weapons because they do not have the material capacity nor the political will to do so.
22
A state
decision to completely abstain from nuclear weapons pursuit might be based on feeling secure,
23

lack of domestic support for proliferation, or adherence to nonproliferation norms. On the other
hand, countries that initiate nuclear weapons programs indicate that they have the material
capability, political will, or both material capability and political will to pursue the bomb.
Once a country crosses the threshold between non-pursuit to nuclear weapons pursuit, our
explanations for nuclear nonproliferation become insufficient to explain nuclear reversals, for at
least two reasons. First, research on nuclear proliferation is useful for explaining the initial
decision to pursue nuclear weapons, as well as nuclear abstention. However, once a country
pursues nuclear weapons, the logic behind why a country might eventually give up an ongoing
nuclear program is not systematically covered by theories that explain the initial decision to
pursue nuclear weapons.
24
Second, we need to take into consideration that dismantling an
ongoing nuclear weapons program is costly, since countries must obtain the technological and
scientific know-how, economic funding, and invest in security of the nuclear program if they
wish to obtain the bomb. Why countries might decide that the benefits of giving up pursuit
exceed the costs of continuing the program is not something that the nuclear proliferation

22
For example, Jo and Gartzke (2007) find that technological capacity is an important determinant for a state to
form a nuclear weapons program.
23
Bleek and Lorber (2014) find that nuclear security guarantees decrease likelihood of nuclear pursuit.
24
Mehta (2020, 6) arrives at a similar conclusion: “states may choose to proliferate for specific reasons and reverse
for varied and unrelated reasons.”
13

literature has adequately explained systematically. In short, this project is not about those cases
that never considered pursuing nuclear weapons, but rather the cases that took the path towards
nuclear weapons but eventually changed course towards abandonment of their programs.
In addition, this project is not about those states that explored the possibility of obtaining
nuclear weapons but never developed a nuclear weapons program. As is well known, Singh and
Way (2004) organized states into three categories: states that explored, pursued, and obtained
nuclear weapons. Singh and Way define states that explored nuclear weapons as those in which
the political leadership seriously considered building nuclear weapons or created a task force
designed to explore the nuclear weapons option, but never created a nuclear weapons program or
never gave authorization for initiating a program.
25
Although these are important cases in their
own right worthy of research, they lie beyond the scope conditions of this project since, in
practice, they are still within the set of countries that never pursued nuclear weapons. As an
example, Japan seriously considered the possibility of obtaining nuclear weapons, but never
embarked on the journey of nuclear weapons pursuit. Logically, the universe of cases of states
that thought about pursuit is much larger than the universe of cases that pursued nuclear
weapons. However, since they ultimately decided not to pursue the bomb, their decision not to
pursue might be explained by existing research on nuclear weapons (non)proliferation.
What this project is about are nuclear reversals, which are defined as instances in
which countries that were pursuing nuclear weapons terminate pursuit of the bomb, by their own
decision or due to coercion. This definition approximates that of Levite, who provided one of the
most succinct definitions: “nuclear reversal refers to the phenomenon in which states embark on

25
Singh and Way (2004, 867).
14

a path leading to nuclear weapons acquisition but subsequently reverse course, though not
necessarily abandoning altogether their nuclear ambitions.”
26
Similar to Levite, I do not reject
the possibility that countries might want to pursue nuclear weapons again in the future. It is
theoretically possible for a country that terminated its nuclear program to restart said program in
the future. In observable terms, states that give up pursuit of nuclear weapons suspend or freeze
their nuclear weapons program, destroy their existing nuclear arsenal, relinquish their nuclear
weapons or nuclear materials to other states or non-governmental bodies, or completely stop
nuclear weapons development.
27

4) Voluntary pathway of nuclear reversals
The first of two potential pathways for nuclear reversals is the voluntary pathway.
Countries along the voluntary pathway terminate their nuclear weapons programs due to the
influence of the international normative environment, particularly international norms against
nuclear proliferation and ownership. However, whether a state is receptive to the influence of
international norms will depend on both how secure a state feels given its security environment
and its relation to the international liberal order. If a state perceives that its security environment
is relatively benign-in other words, does not perceive major security threats from rivals or
enemies-and if the state is a liberal democracy that wants to be part of the international
community of liberal states, then the state will be likely to internalize or accept as legitimate
international norms against nuclear weapons proliferation and ownership. Once the norm against
nuclear weapons has been internalized or accepted as legitimate by the state, the state will

26
Levite (2002, 61).
27
These observable implications of what nuclear reversal looks like in practice are like that of Mehta (2020, 11).
15

voluntarily give up its nuclear weapons program to adhere to the international community’s
standards of what “acceptable behavior” entails-not being a nuclear weapons state.
Importantly, a state will have to receive observable signals from the international
community that there is such a thing as an international norm against nuclear proliferation and
ownership before it internalizes or accepts such a norm. Such signaling occurs in international
forums, such as the United Nations General Assembly and World Summits, where state leaders
may argue against the proliferation or ownership of nuclear weapons. States might resort to
shaming mechanisms, such as condemning nuclear weapons programs of states, or even
diplomatic “carrots” such as promising lifting of economic sanctions or stronger security ties in
exchange for a country to give up pursuit of nuclear weapons. A state pursuing nuclear weapons,
once it has received these signals from the international community, has the option to terminate
its program so it can adhere to the international norm being signaled by fellow members of the
international community. However, the likelihood of a state internalizing or accepting such a
norm depends on its perception of the security environment as benign and if it wants to be
considered a member of the liberal order of states.
These international norms, such as the norm against nuclear proliferation and norm
against nuclear weapons ownership, are defined here as shared understandings of what is
appropriate behavior for states in the international arena. International norms are ultimately a
subcategory of norms, which I define in the standard manner as previous scholars have done: as
“standard[s] of appropriate behavior for actors with a given identity,”
28
“collective
understandings of the proper behavior of actors,”
29
and “shared expectations about appropriate

28
Finnemore and Sikkink (1998, 891).
29
Legro (1997, 33).
16

behavior held by a community of actors.”
30
All these definitions share common ground; a norm
is a shared belief of what is considered appropriate behavior for a community of actors, rather
than simply private beliefs that might not be shared by other members of a given community.
States might either internalize international norms or at minimum accept them as
legitimate. When a state internalizes an international norm, it accepts the norm as an essential
part of its identity, as a “rule” for behavior that should be obeyed to be considered a respected
member of the international community. A state that internalized the norm against nuclear
weapon ownership or proliferation will, if it hopes to maintain its identity as a member of the
international liberal order, take actions to reverse its nuclear weapons development program.
However, norm internalization might not occur in all cases of a voluntary reversal; some states
might still revert voluntarily without internalizing the norm if they accept the norm as a
legitimate rule of conduct. These states might not see an international norm as an essential part
of their identity, yet still adhere to the norm because they believe that it is advantageous to do so
or to maintain positive relations with other liberal states. Empirically, then, there is no
observable distinction between states that internalize a norm and those that accept it as legitimate
for strategic purposes. In both cases, states will voluntarily terminate their nuclear weapons
programs.
I assume that internalization/acceptance of the norms against nuclear proliferation and
ownership will be most likely to occur for liberal states, specifically liberal democracies, rather
than non-liberal or autocratic states. This assumption is held for two related reasons. First, most
democracies today are liberal democracies, which adhere to beliefs such as individual freedom,

30
Finnemore (1996, 22).
17

capitalism, the importance of human rights, and global cooperation. Second, the existing
international order is a liberal international order, a historical result of the end of the Cold War
which saw the crumbling of the Soviet Union and global communism along with it. The
international order imposed by the United States and its Western allies tended to favor the spread
of liberal democracies in the post-cold war period, leading to an international order in which
most democracies adhere to liberal ideology and norms. The norms against nuclear weapons
proliferation and ownership fall within the broader liberal international order, for these norms are
based on collective security and prevention of global destruction via nuclear weapons. If states
want to be in good standing within this liberal order, and thus be perceived as good members of
the international community, they must support the prohibition of “inhumane” weapons that
cause severe human and physical destruction.
In contrast, authoritarian states, whose values do not align to liberal international norms
emphasizing individual liberty and human rights, might also not be so receptive to international
norms against nuclear weapon proliferation and ownership. The reason for their rejection of
these international norms is that authoritarian states might perceive such norms as impositions of
the liberal West, particularly the United States, as a strategy to keep them weak in comparison to
the United States and its allies.
31
For authoritarian states, voluntarily accepting these
international norms and giving up their nuclear weapons programs would put them at a
disadvantage vis a vis their liberal rivals.
32

An important prerequisite for norm internalization or acceptance of anti-nuclear weapons
norms is the existence of a relatively benign security environment. Even democracies will be

31
Ikenberry (2011) correctly points out that the liberal order is a Western order led by the United States.
32
These fears may not be unfounded: critical scholarship perceives the Western-led liberal order as Eurocentric and
even Imperialist in practice. See: Parmar (2018).
18

hard-pressed to give up pursuit of nuclear weapons if they are surrounded by enemies or
constantly face international crises and interstate military disputes. As Debs and Monteiro
observe, “a state will attach a security benefit to nuclear weapons only when it faces a high level
of threat to its security.”
33
Logically, if the security environment perceived by a state improves,
then it loses the main incentive for obtaining a nuclear weapon and should be more likely to
revise its policy of pursuing nuclear weapons. When the security environment is perceived as
threatening, norm internalization and acceptance will be less likely, since the primary goal of a
state, survival, would be in direct competition with the international norms against nuclear
weapons ownership and proliferation.
An important question is what we should expect to observe if a liberal democracy is
pursuing nuclear weapons but is in a threatening security environment. Given a threatening
security environment, such as frequent military interstate disputes, foreign policy crises, or wars,
even liberal democracies will have difficulty adhering to international norms against nuclear
weapons proliferation or ownership. In these cases, then, the expectation is that democracies will
see themselves forced to continue pursuit of nuclear weapons for security purposes. In the
absence of a benign security environment, even liberal democracies will continue pursuit of the
nuclear bomb despite their desire to be “good” members of the liberal international order.
Yet voluntary reversals should still occur under a threatening security environment if a
key condition is met: other liberal states must offer security guarantees to the threatened liberal
state as an inducement towards nuclear reversal. Since liberal democracies tend to be friendly
towards each other, the expectation is that other liberal democracies will rely on diplomatic

33
Debs and Monteiro (2017, 7).
19

“carrots” rather than “sticks” to convince fellow liberal democracies to reverse course on their
nuclear weapons programs. Instead of threatening sanctions or use of force, positive diplomatic
inducements are expected, such as offering military economic aid or even formal military
alliance in exchange for the liberal democracy giving up its nuclear weapons program. In other
words, we should expect a voluntary reversal by a liberal democracy even under high threat as
long as other liberal democracies offer believable security guarantees to dissuade the state from
continuing with its nuclear program. Once this material incentive is offered, liberal democracies
can once again safely accept international norms against nuclear weapons ownership and
proliferation and reverse course.
In sum, the voluntary pathway predicts that when the security environment is perceived
as benign and when a state is a liberal democracy, a state will be more likely to internalize or
accept as legitimate the norm against nuclear weapons proliferation, leading to nuclear reversal.
Voluntary reversal is also expected if a liberal democracy is under high threat, but only if
security guarantees are offered by other states. Absent a threatening security environment, states
will have little incentive to pursue nuclear weapons, which is a costly and time-consuming
enterprise. Given that the norm against nuclear weapons proliferation is part of the liberal
international order, and democracies are more likely to believe in liberal norms,
34
democracies
will tend to be attracted towards these norms, incorporating them into their own state policy.
This internalization/norm acceptance should follow pressures from the liberal international
community, which should push norms against nuclear weapons ownership and proliferation via

34
Owen (1994).
20

shaming tactics in international forums or relying on positive inducements to convince states to
adhere to these norms.
Observable implications of the voluntary pathway
1) We should observe that liberal democracies are the main actors pushing norms
against nuclear weapons ownership and proliferation in international forums such as
the United Nations General Assembly and World Summits. They will advocate for
these norms by shaming nuclear pursuers or highlighting the threat nuclear weapons
pose for international security.
2) We should observe that if liberal states want other liberal states to reverse their
nuclear weapons program, they will rely on diplomatic carrots such as sanction relief
or stronger security/economic ties rather than threats or coercive mechanisms like
sanctions and use of force.
3) We should observe that democracies pursuing nuclear weapons will be more likely to
reverse their programs following a decrease in their perception of the threat
environment. Nuclear reversals should be unlikely during war or if a state feels
threatened by neighbors.
4) We should observe that democracies pursuing nuclear weapons will terminate their
programs when they perceive a high threat environment only after credible security
guarantees by other states. In the absence of credible security guarantees, the liberal
democracy will continue nuclear weapons pursuit.
5) We should observe that most states voluntarily reversing their nuclear weapons
program are liberal states rather than autocracies. In addition, they should provide
reasons related to international norms against nuclear weapons or proliferation when
21

justifying their decisions to reverse course as an indication they have internalized or
accepted these international norms as legitimate.
5) Involuntary pathway of nuclear reversals
Nuclear reversals still occur even in the absence of norm internalization/acceptance by
the proliferating state, which can be attributed to non-voluntary factors. In these cases, nuclear
reversal results from coercion by other states, who use threats of force, sanctions, or actual force
to induce countries to stop pursuit of nuclear weapons. The involuntary pathway is one in which
state’s agency is highly constrained by the international system, mainly more powerful actors
who use their power capabilities and their arsenal of diplomatic and military tools to coerce
rivals and enemy states into terminating their nuclear weapons programs.
35
However, norms are
not entirely absent from the story: states engaged in coercion might be trying to uphold norms
against nuclear proliferation and ownership. Although norms against nuclear weapons ownership
and proliferation are not internalized by the proliferating state, those states trying to coerce them
into reversal might have internalized or, at minimum, accepted the norms as legitimate. The logic
of the involuntary pathway should be familiar to realists; it centers power and coercion as the
factors driving these cases of nuclear weapons reversals.
If the security environment is perceived as threatening, states should have no incentive to
give up pursuit of nuclear weapons. On the contrary, states expecting war, experiencing major
foreign policy and international crises, or constant militarized disputes with other states, will
have an incentive to successfully complete their nuclear weapons programs.
36
In addition,
autocracies should be less likely than democracies to internalize or accept norms against nuclear

35
Mehta (2020) provides a brilliant explanation of nuclear weapons reversals resulting from external factors,
focusing on the role of the U.S. to coerce proliferators to terminate their nuclear weapons programs.
36
Epstein (1977) and Sagan (1996/1997) support the traditional view that security concerns drive nuclear weapons
proliferation.
22

weapons proliferation and ownership, since they might see these international norms as
impositions of the liberal Western world. Under these circumstances, international norms against
nuclear weapon proliferation will be ineffective at curbing state’s nuclear ambitions, particularly
for autocratic regimes. Nuclear reversal, under conditions of threat, is a suboptimal decision.
States in these threatening circumstances may reject or ignore international norms against
nuclear proliferation and ownership because they perceive it is to their advantage to own nuclear
weapons, and this should especially hold for autocratic regimes.
Yet even these states rejecting nuclear nonproliferation norms might still terminate their
programs. Nuclear reversals in these circumstances occur due to coercive actions by other states,
either by rivals or major powers, that want to deter their adversary from obtaining a nuclear
arsenal. The coercing states have a variety of tools at their disposal to force recalcitrant states to
reverse their nuclear programs; threats to use force or sanctions, sanctions, and use of force.
37

The first tool that states will likely use to pressure a potential proliferator is the threat of
sanctions or threats to use military force, since threats are less costly to make than imposing
sanctions or using military force. If the potential proliferating state refuses to back down and shut
down their nuclear program, the coercing states must choose more direct actions from their
policy toolbox, either imposing sanctions or using their military power to force the proliferator to
reverse. Since authoritarian states will not accept/internalize international norms against nuclear
weapons, and since abandoning nuclear weapons pursuit under a highly threatening security
environment would be suboptimal for security, these authoritarian states are unlikely to respond
to sanctions or threats or force. In other words, other actors in the system will likely have to

37
For a classic overview of coercion in international relations, see George, Hall, and Simons (1972).
23

resort to the use of military power to force an authoritarian state to terminate their nuclear
program in a highly threatening security environment.
What should we expect if an autocracy is facing, or perceives, a nonthreatening security
environment? These autocratic states might still pursue nuclear weapons for reasons unrelated to
security, such as higher regional status. Perhaps the security environment was initially perceived
as threatening when the nuclear weapons program was initiated, but path dependency leads the
autocratic state to continue pursuit despite a decrease in the threat environment. Regardless of
the reason for pursuit, authoritarian countries pursuing the bomb are unlikely to accept
international norms against nuclear weapons proliferation, since they will dismiss them as
impositions from liberal states. As before, nuclear reversal should only be expected if other states
coerce the authoritarian state to terminate its program, via use of diplomatic tools such as
sanctions and threats to use force. However, in contrast to cases where the security environment
is threatening, sanctions and threats of force should be sufficient to coerce an authoritarian state
to reverse its program. In the absence of a highly threatening security environment, authoritarian
states will have less of an incentive to hold steadfast to their nuclear program in the face of
sanctions and threats. Authoritarian states in a relatively secure environment still need to be
coerced to terminate their nuclear programs, but the use of force as a diplomatic tool is unlikely
to be necessary.
International norms against nuclear weapons proliferation still play a role, albeit an
indirect one. The states coercing others to terminate their programs might have internalized the
norms against nuclear weapons ownership and proliferation. Even if internalization did not
occur, some states might still accept non-proliferation norms as beneficial for their own national
security. Whether for self-interested or ideational reasons, states that accept the norm against
24

nuclear weapon ownership and proliferation will attempt to uphold the norm, using their own
power and influence to pressure other states into reversing their nuclear weapons programs.
Thus, international norms against nuclear weapons proliferation and ownership matter even in
the involuntary reversal pathway, for these norms are being upheld by states despite their
rejection by recalcitrant state actors trying to obtain nuclear weapons. Therefore, we should
observe nuclear reversals in the absence of norm-internalization or acceptance of non-
proliferation/nuclear ownership norms because of coercion by other states fearing for their own
security or trying to uphold the international non-proliferation regime.
In short, the involuntary pathway of nuclear reversal is straightforward, and perhaps less
controversial than the voluntary pathway mechanism explained above. Autocratic regimes will
be less likely to internalize the norm against nuclear weapons proliferation and ownership since
they will perceive it as an imposition of the liberal international order. Furthermore, states facing
a threatening security environment will be unlikely to terminate their programs, since obtaining a
nuclear weapon serves as a useful deterrence against potential aggressors and a symbol of state
power. In these scenarios, termination of nuclear weapons programs occurs due to coercion by
other actors in the system, specifically the use of force. Authoritarian states in a nonthreatening
security environment also need to be coerced into terminating their programs, but threats of force
or sanctions should be sufficient in these cases. These proliferating states terminate their
programs, but not due to voluntary acceptance of the international norm against nuclear
weapons, but because of external pressures coming from outside their borders.
Observable implications of the involuntary pathway
1) Authoritarian states pursuing nuclear weapons will be likely to reject norms against
nuclear weapons proliferation or ownership. We should observe state leaders
25

rejecting these norms in international forums like the United Nations General
Assembly. Importantly, these rejections should mention, in one way or another, the
imposition of a liberal order.
2) Authoritarian states will be unlikely to reverse their nuclear weapons programs
voluntarily if they are in a highly threatening environment, such as frequent territorial
disputes or war. Countries that decide to give up pursuit of nuclear weapons while
they are in a threatening environment will do so because of the use of force from
other states.
3) States will be unlikely to reverse their nuclear weapons programs voluntarily if they
are not in a highly threatening security environment. Countries that decide to give up
their nuclear weapons program will do so due to threats of force or sanctions.
4) States trying to force others to reverse their nuclear weapons programs will explicitly
mention anti-nuclear weapons proliferation norms. They should point to international
treaties during General Assembly sessions or international summits. Alternatively,
they will cite international treaties when trying to convince the proliferator to revert
their program.
5) States trying to coerce others into reversing their nuclear weapons programs should
rank-order their available tools of coercion. We should observe, threats of sanctions
or threats to use force as the first policy option. If these threats fail, we should see
economic sanctions as the second policy option. If sanctions fail, then countries
should be likely to use military force against the proliferator.

26

6) How the security environment affects nuclear reversals
As can be observed, the two-pathway theory expects that liberal democracies and
authoritarian states will be impacted by international norms against nuclear weapons
proliferation and ownership in distinct ways. Liberal democracies will be more likely to
internalize or at minimum accept as legitimate these international norms, and thus will
voluntarily give up pursuit of the nuclear bomb. Authoritarian states, on the other hand, will
reject said norms and continue pursuit of the nuclear bomb unless they are coerced to reverse by
other members of the international system.
Yet, the security environment plays an important role affecting the likelihood of nuclear
program termination as well as the strategies other states might have to use to coerce or induce
nuclear reversals. The highest likelihood of nuclear reversals is when a liberal democracy faces a
relatively benign security environment because in these circumstances norm acceptance is easiest
and nuclear program termination less costly. If a liberal democracy faces a threatening security
environment, norm acceptance is more difficult, due to having to balance adherence to the norm
against state security. In these circumstances, other actors need to offer diplomatic carrots such
as security guarantees to make voluntary nuclear reversal a safe option. Next in line are
authoritarian state facing a relatively benign security environment. These authoritarian states will
reject norms against nuclear weapons proliferation and ownership, so they must be coerced by
other states to reverse their nuclear programs. For these states, threats of force and sanctions
should be sufficient to coerce them into nuclear reversal. The most difficult cases of reversal are
those authoritarian states facing a highly threatening security environment. They have two
incentives rather than one to avoid reversal: rejection of liberal norms and a desire to obtain the
27

nuclear bomb for their security. Nuclear reversal for these states results from the use of military
force by other actors, due to threats and sanctions being insufficient.
Ultimately, the security environment affects which diplomatic options other states need to
rely on if they want to increase the likelihood of a nuclear weapons program termination. The
voluntary pathway, taken by liberal democracies, require the least amount of effort: shaming
tactics in international forums and security guarantees are all that is needed for inducing nuclear
reversal. The involuntary pathway, taken by authoritarian states, requires a harsher set of
diplomatic tools: threats of force, sanctions, and actual use of force by other states are required to
force an authoritarian state to reverse their nuclear program. Whether use of force is needed will
depend strongly on how threatening the security environment an authoritarian state faces or
perceives. Table 1 below shows how different combinations in threat environment and regime
type affect which strategies are required for a nuclear reversal to occur.
Table 1: Nuclear Reversal Expectations
Low threatening security
environment
High threatening security
environment
Liberal Democracy Voluntary reversal.
Strategies for reversal:
shaming by other states,
voluntary desire by pursuing
state to be a good member of
the liberal order.
Voluntary reversal.
Strategies for reversal:
Reversal occurs only if other
states provide security
guarantees to liberal
democracy.
Authoritarian Involuntary reversal.
Strategies for reversal: Other
states must use threats of
force or use sanctions to
coerce the authoritarian state
to reverse their program.
Involuntary reversal.
Strategies for reversal: Other
states must use actual force to
forcefully terminate an
authoritarian state’s nuclear
program.

28

Chapter 3: Research Design
As can be seen in the theory section, multiple observable implications can be derived
from the two nuclear reversal pathways. The observable implications are amenable to empirical
testing through a combination of quantitative and qualitative methods, particularly logistic
regression and process tracing of typical cases that show the theory applies to historical instances
of nuclear reversals. To test the validity of the theory, I will rely on a mixed-methods strategy
leveraging the strengths of both quantitative and qualitative analysis.
Chapter 4, the quantitative chapter, tests whether acceptance of international norms—
measured in terms of votes supportive of anti-nuclear UNGA resolutions—affects the likelihood
of nuclear reversals. These tests are carried out using various logistic regression models in
addition to Cox hazard models. I expect that as countries signal support for international norms
against nuclear weapons ownership and proliferation, the likelihood of continued nuclear pursuit
decreases, ending in termination of nuclear weapon programs.
Following standard procedure, some of the models measure norm acceptance by checking
if a state is a member of the 1968 Non-Proliferation Treaty. This measurement is not novel;
given that measuring an unobservable like international norm acceptance per country-year is not
an easy task, researchers typically rely on NPT ratification as a proxy for the nuclear
nonproliferation norm, assuming higher ratification indicates a stronger, or at least more
widespread, nuclear nonproliferation norm.
38
However, NPT ratification is a misleading
measurement of whether a country accepts nuclear nonproliferation norms. As Montgomery and
Sagan point out, states might join the NPT for a variety of reasons, such as part of a bargain with

38
For example, Jo and Gartzke (2007); Kroenig (2009); and Mattiaci and Jones (2016) measure the nonproliferation
norm by NPT membership over time.
29

rival powers, to “cheat” the international community and still pursue the bomb covertly, or
simply join to be part of the world order without ever having any intention or capability to build
nuclear weapons.
39
Since states might join the NPT for reasons unrelated to acceptance of
nuclear nonproliferation norms, using NPT ratification as a measure of the concept is highly
problematic. As a result, this project introduces a new measurement for acceptance of
international nuclear non-proliferation norms.
This project proposes the following measurement strategy: we might be able measure
how strongly states support international norms against nuclear proliferation and ownership by
looking at their voting patterns in the United Nations General Assembly (UNGA). As is well
known, UNGA resolutions are decisions or declarations voted by all member-states of the United
Nations. Except for procedural matters such as budgets and other payments, UNGA resolutions
and decisions are non-binding.
40
However, these resolutions are still meaningful because they are
public declarations by member states, which sets expectations about their future behavior in the
eyes of peers, akin to audience costs at the international level. Although states might vote for
UNGA resolutions due to pressure from allies, vote-buying influences, or as part of a political
bloc, state actions that frequently contradict how they vote in the UNGA may carry reputational
costs. Therefore, states might become bound to their public declarations via voting, restricting
the universe of actions states might take. Even if a state voted in favor of anti-nuclear weapon
resolutions due to external pressures, adherence to the international norm being advanced might
follow since states do not want to harm their reputation. This translates to norm-adherence and
acceptance in practice, which is what we can observe and measure.

39
Montgomery and Sagan (2009, 313).
40
UN General Assembly, https://ask.un.org/faq/14484
30

Most United Nations General Assembly resolutions are adopted without a vote, and in
many cases how specific states voted on a resolution is kept secret.
41
Therefore, the measurement
created here is a conservative measure of how states voted on nuclear non-proliferation and anti-
ownership resolutions. To build the measure, I searched for all UNGA votes on resolutions with
the key words “Nuclear” or “Atomic” in the title from 1950-2020, then counted how many times
a state voted against, in favor, or abstained from voting in nuclear resolutions per year. The
expectation is that the higher the proportion of “yes” votes on anti-nuclear weapons resolutions,
the stronger the state’s acceptance of nuclear weapons non-proliferation and non-ownership
norms.
Additionally, the quantitative tests also include measures of democracy and perception of
threat environment, which are theorized to be among the most important variables influencing
the likelihood of nuclear reversal alongside nuclear non-proliferation norm acceptance.
Therefore, the quantitative chapter tests the following hypotheses derived from the two-pathway
theory of nuclear reversals:
Norm acceptance hypothesis (H1): an increase in the level of nuclear non-proliferation
norm acceptance will increase the likelihood of nuclear reversals.
Democracy hypothesis (H2): greater levels of democracy will increase the likelihood of
nuclear reversals.
Threat environment hypothesis (H3): an increase in the perception of threat will decrease
the likelihood of nuclear reversals.

41
UN Documentation: General Assembly https://research.un.org/en/docs/ga/voting
31

The logistic regression models reveal how perception of threat, democracy, and most
importantly nuclear non-proliferation norm acceptance impact the likelihood of nuclear
reversals. A negative coefficient for the norm acceptance variable would imply that higher levels
of nuclear non-proliferation norm acceptance would decrease, rather than increase, the likelihood
of nuclear weapon reversals, which would run counter to theoretical expectations. Importantly, a
positive coefficient in the statistical tests, showing that nuclear non-proliferation norm
acceptance increases the likelihood of nuclear reversal, would suggest tentative evidence in favor
of the voluntary pathway, which argues international norms have an effect even in the nuclear
realm.
In addition to the quantitative chapter introducing a measurement of international norms,
the dissertation includes three qualitative chapters to carefully trace the causal mechanisms
advanced by the theory. By relying on case studies, I can show with greater detail whether and
how my theoretical pathways describe actual processes in the real world. The quantitative
chapter will reveal indirect evidence of which of the two pathways is most likely to occur,
42
but
the observable implications of my theory are better tested via case studies that trace how specific
states pursuing nuclear weapons react through time to international norms, external threats, shifts
in regime type, and other external or internal factors, and whether the theory traces the actual
mechanisms for eventual nuclear reversal.
Three cases are chosen to test the causal pathways: Brazil, Iraq, and Israel. The cases of
Brazil and Iraq serve as illustrative case studies of the voluntary and involuntary pathways,

42
The evidence presented in the quantitative chapter is indirect evidence of my pathways because it is only
measuring the correlation between higher acceptance of international norms and likelihood of nuclear reversals. It
cannot show causality due to being a non-experimental design, nor does it demonstrate that the causal mechanisms
work as they do. In addition, the quantitative chapter only reveals which pathway is more common, not how each
pathway works.
32

respectively.
43
Following the typology of Gerring, the case studies of Brazil and Iraq fall within
the diagnostic category of case studies, since the aim is to confirm/disconfirm particular
hypotheses, and most importantly show that the mechanism connecting the independent variable
(international norm acceptance\coercion, respectively) with the dependent variable (nuclear
reversals) can be found in real world cases.
44
The case of Israel, on the other hand, is chosen as a
hard case for my theory. Israel, despite being a democracy, obtained the nuclear bomb after the
international institutionalization of the NPT in 1968. If my theory is to have any merit, it should
be able to explain cases of non-reversal as well.
For discovering that the causal mechanisms are at play, process tracing will be employed,
defined here as “the analysis of evidence on processes, sequences, and conjunctures of events
within a case for the purposes of…testing hypotheses about causal mechanisms that might
causally explain the case.”
45
To properly conduct process tracing, the chosen cases will be tested
via hoop tests. Hoop tests are useful in process tracing methodology because they can be used to
evaluate hypotheses that claim a specific causal mechanism occurred and if there is a causal
connection between two or more events.
46
A hoop test proposes that specific evidence, such as a
specific causal process observation, must be found for a hypothesis to be valid. If evidence is not
found, then the hypothesis fails, but passing the hoop test does not automatically confirm the
hypothesis. An alternative type of test, not used in this project, is the smoking gun test. Smoking
gun tests pass if specific evidence of a causal process is found, confirming the hypothesis.
Failure to find such evidence, however, does not disconfirm the hypothesis.
47
In other words,

43
Gerring (2017, 105).
44
Ibid., 98-99.
45
Bennett and Checkel (2015, 7).
46
Mahoney (2012, 572).
47
Ibid., 572.
33

hoop tests can be useful for eliminating competing hypotheses, while smoking gun tests can
provide decisive support for a hypothesis. Due to the complexity of the cases, I rely exclusively
on hoop tests rather than smoking gun tests, allowing for the possibility that my interpretation of
the evidence can be contested. For each of the case studies, the hoop tests below will be applied
to test whether the causal mechanisms expected by the voluntary and involuntary nuclear
reversal pathways can be identified within the chosen cases.
In chapter 5, I test whether the case of Iraq adds support to the involuntary reversals
pathway by testing the observable implications listed below. For consistency, I will refer to these
hypotheses as “observable implications” throughout the chapter; the hoop tests are whether
evidence is found to support the expected observable implication of the involuntary reversal
pathway:
1) Iraq should reject norms against nuclear weapon proliferation and ownership. We
should observe evidence of such rejection in public statements made by Iraq’s
leadership in official public state reports, public speeches, or in international forums
like the United Nations General Assembly. Importantly, such rejection should
mention that the nuclear norms are an imposition by the liberal West.
2) It should be observed that other states in the system are trying to push Iraq towards
termination of their nuclear program. Anti-nuclear weapon proliferation norms should
be mentioned as part of the arguments advanced by states, perhaps in international
forums like the UNGA. These statements are hypothesized to be more frequent for
liberal democracies than autocracies.
3) Other states trying to push Iraq towards termination of their program should rank-
order their diplomatic tools. We should observe that threats of sanctions and
34

imposition of sanctions take priority over threats of force. In addition, threats of force
should take priority over actual use of force. We should observe that diplomatic
“carrots” are used less frequently than diplomatic “sticks” against Iraq.
4) We should observe that Iraq does not terminate its program voluntarily when
perceiving high levels of threat.
5) Iraq’s termination of their nuclear weapons program must be shown to be the result of
external coercion rather than a voluntary decision. Iraq’s termination of their program
should follow from acts of coercion by other actors such as threats of force/sanctions
or actual use of force/sanctions, rather than from diplomatic carrots such as security
guarantees and lifting of sanctions.
Successfully passing the hoop tests will show strong support for the validity of the
pathway in cases such as Iraq, demonstrating that international norms against nuclear weapons
proliferation and ownership are pushed by the international community but rejected by the target
state, and that reversal occurs due to external coercion rather than a voluntary decision. It is
important to note that support for the pathway is only evidence that the hypothesized pathway
exists in the real world, but how widespread it is can only be shown by carrying out process
tracing analysis of other similar cases. Such task is left for future research.
Chapter 6 will test whether the case of Brazil adds support to the hypothesized voluntary
reversal pathway by testing the following observable implications:
1) We should observe that liberal democracies are the main actors pushing norms
against nuclear weapons ownership and proliferation in international forums such as
the United Nations General Assembly. We should observe that these acts of
35

international pressure are directed towards Brazil after the international community
discovers that Brazil is developing a covert nuclear program.
2) We should observe that liberal states, in their attempt to push for nuclear reversal,
will rely on distinct diplomatic tools when trying to convince Brazil to reverse their
program. When Brazil is authoritarian, we should observe threats to use force or at
minimum threats of economic sanctions.
3) Following democratization in Brazil, the main diplomatic tool used by states should
be shaming mechanisms or diplomatic “Carrots” such as sanction relief or stronger
security/economic ties in exchange for nuclear reversal.
4) We should observe that Brazil’s decision to reverse their nuclear weapons program
comes about after a decrease in their perception of threat or rivalry with neighboring
countries such as Argentina.
5) If Brazil terminates their nuclear program before Brazil changes its perception of
threat, it should be because other actors promised credible security guarantees.
6) We should observe that Brazil’s voluntary decision to terminate their nuclear
weapons program occurs following democratization.
7) Importantly, Brazil should provide reasons related to nuclear nonproliferation norms
when informing the international community about their decision to reverse course.
The more hoop tests the case study of Brazil passes, the stronger support for the
voluntary nuclear reversal pathway. If at any point during the case study it is revealed that Brazil,
following democratization, terminated their nuclear program involuntarily-either because Brazil
faced damaging economic sanctions or fear of physical coercion-then the voluntary pathway is
falsified for the case of Brazil. In addition, if international nonproliferation norms do not play a
36

central role either in external actor’s efforts to push for Brazil’s nuclear reversal or as a public
reason given by Brazil’s leadership to terminate their program, then the voluntary pathway also
loses support. Once again, an important caveat must be made explicit. Successfully passing all
the hoop tests is evidence that the hypothesized voluntary pathway can apply in real world cases
but ascertaining how widespread the mechanism is requires process tracing of all other similar
cases. Such a task is left for future reach and is beyond the goals of this project.
Chapter 7, the third case study, looks at Israel’s development of the nuclear bomb, from
the inception of the program around 1950 to its completion around the 1970s. Israel, unlike
Brazil and Iraq, successfully completed its nuclear weapons program, becoming the 6
th
nuclear
armed country in the world. The inclusion of Israel in this project is important to observe if it
runs counter to theoretical expectations: As a democratic state, Israel should have followed the
voluntary reversals pathway. So why did it not terminate its program? Process tracing might
reveal why Israel diverges from Iraq and Brazil, and perhaps show that its failure to reverse is
explained perfectly by the theoretical framework poised in this project. For the case of Israel, the
following observable implications will be tested:
1) Israel’s program should begin if it perceived a major security threat from its
neighbors, and such threat should be constant throughout the nuclear weapons
program development period.
2) If the voluntary reversals pathway is correct, Israel should have been pressured to
terminate its program as soon as the international community learned that Israel was
building the bomb.
37

3) International pressure to conform to the anti-nuclear nonproliferation norm should
include shaming mechanisms and/or diplomatic “carrots” such as security guarantees,
instead of economic sanctions or threats of force.
4) Israel’s failure to conform to the norm must be justified in terms of security concerns:
Israel must justify its creation of a nuclear bomb based on security concerns rather
than rejection of nuclear norms themselves.
48

For all these hoop tests, temporal sequence of events is of utmost importance: The theory
delineated here assumes that a voluntary nuclear reversal will occur if two major conditions are
met, democratization and reduction in threat environment. If it is found that a country is taking
measures to revert their nuclear weapons programs before democratization or when security
threat levels are still high, then the case fails to support the temporal sequence proposed by the
voluntary pathway. Furthermore, evidence of reverting due to coercion rather than norm
acceptance would count as evidence against the voluntary pathway. In addition, the theory
assumes that the involuntary pathway will occur if a country is autocratic and is forced to
terminate their program by external actors. It also assumes that states will refuse to terminate
their nuclear weapons program if they perceive a threatening security environment. If it is found
that an autocratic state abandons their program willingly rather than by coercion, and that
abandonment of nuclear weapons pursuit occurs during times of high threat perception without
coercion, then the involuntary pathway loses support.
Process tracing of the chosen case studies relies on evidence drawn from a mixture of
secondary and primary sources. For tracing the history of the nuclear weapons programs of each

48
It is important to note that Israel never publicly admitted that they built the bomb. Evidence of Israel’s defense for
the bomb must come from private correspondence or declassified material.
38

case, I mainly use secondary sources such as historical monographs and scholarly research on the
countries’ nuclear programs. Primary evidence, in the form of private correspondence, UNGA
statements, speech transcripts, and newspaper reports are also identified and used as further
evidence throughout the chapters. Following good practice, each direct quotation and source for
a statement or fact is directly cited via footnotes, with a link to the source when possible. Some
justification for the sources used in the project should be highlighted. Due to the real-world
constraints of the covid-19 pandemic in addition to time constraints, the sources used for this
project are exclusively virtual or accessed via the physical library at the University of Southern
California, rather than new archival material or in-person interviews with elites or other foreign
policy makers. However, this limitation should not be an issue, given the vast amount of
evidence available online for the chosen cases, either from archival material cited by earlier
scholars or primary sources published online.
Ultimately, the process tracing methodology used within these cases will reveal whether
the hypothesized causal mechanisms for voluntary and involuntary reversals accord with the
historical record and any primary sources found. Regardless of what the findings might be, the
result should be a better understanding of the conditions under which countries might give up
pursuit of nuclear weapons, furthering our knowledge in the realm of nuclear weapons program
abandonment.

39

Chapter 4: New Measure of Nuclear Non-Proliferation Norm Acceptance and
Quantitative Models
1) Measuring acceptance of anti- nuclear weapon norms
One way the theory of nuclear reversals could be tested is via regression models, where
we could observe if there is a positive correlation between acceptance of nuclear norms and
nuclear reversals. If the correlation is statistically significant and positive, we would have
tentative evidence that greater acceptance of nuclear norms increases the likelihood of nuclear
reversals. However, measuring acceptance of nuclear norms is a difficult task. Jo and Gartzke
measure the effect of international norms on nuclear weapons proliferation by using NPT Treaty
ratification status.
49
Although seemingly an intuitive proxy for measuring anti-nuclear weapon
proliferation norms, Montgomery and Sagan point out that states might join the NPT to bargain
with rivals or as “cheating states” that are still pursuing nuclear weapons.
50
As a result, NPT
membership fails to capture whether a state accepts norms against nuclear weapons proliferation
and/or ownership because states might join the NPT for a variety of reasons unrelated to norm
acceptance.
As an alternative to using NPT membership as a measure of acceptance of anti-nuclear weapon
norms, I propose that we can measure the degree of state support for anti-nuclear weapons norms
over time by looking at their voting records on nuclear weapons-related United Nations General
Assembly (UNGA) resolutions. As is well known, UNGA resolutions are decisions or
declarations voted by all member-states of the United Nations. Except for procedural matters
such as budgets and other payments, UNGA resolutions and decisions are non-binding.
51

49
Jo and Gartzke (2007).
50
Montgomery and Sagan (2009, 313).
51
UN General Assembly, https://ask.un.org/faq/14484
40

However, these resolutions are still meaningful because they are public declarations by member
states, which sets expectations about their future behavior in the eyes of peers, akin to audience
costs at the international level. Although states might vote for UNGA resolutions due to pressure
from allies, vote-buying influences, or as part of a political bloc, state actions that frequently
contradict how they vote in the UNGA may carry reputational costs. Therefore, states might
become bound to their public declarations via voting, restricting the universe of actions a state
might take. Even if a state voted in favor of anti-nuclear weapon resolutions due to external
pressures, adherence to the international norm being advanced might follow since states do not
want to harm their reputation. This translates to norm-adherence and acceptance in practice,
which is what we can observe and measure.
Most United Nations General Assembly resolutions are adopted without a vote, and in many
cases how specific states voted on a resolution is kept secret.
52
Therefore, any measurement we
build using UNGA voting records will be conservative by nature. Yet, using UNGA voting
records as a measurement of international norm acceptance has its advantages over using NPT
treaty ratification. First, as explained above, state votes on the UNGA create reputation costs
which could affect actual state behavior. Constantly voting in favor of anti-nuclear weapon
resolutions builds a reputation for opposing proliferation or nuclear weapon ownership, and if a
state wants to maintain its image as a “good member” of the international order it will attempt to
reconcile its public image with its private policy making. In other words, the more a state votes
in favor of anti-nuclear resolutions, the more it builds a reputation as a state that opposes nuclear
weapons, creating pressures to adhere to its public image and thus terminate any nuclear
weapons pursuit. Therefore, this proxy of anti-nuclear weapons norm acceptance should have

52
UN Documentation: General Assembly https://research.un.org/en/docs/ga/voting
41

more face validity than using the standard NPT ratification measure. Second, using UNGA
voting records on nuclear resolutions allows us to measure norm acceptance over time. There
may be periods of time when a state’s adherence to and acceptance of anti-nuclear weapon
norms weakens, perhaps due to change of regime or an increasingly threatening security
environment. If so, we should expect some degree of temporal variance in state support for
UNGA anti-nuclear weapon resolutions, a variance that is missed if we use the binary NPT
ratification measure.
The rest of the chapter is organized as follows. First, I describe the UNGA antinuclear voting
record dataset, how the data was collected, and provide descriptive statistics of the data. In the
following section, I merge the data with a country-year monadic dataset that measures whether
states are pursuing/obtained nuclear weapons or reversed, provide univariate statistics of the
data, and explain how the UNGA voting record dataset could be used for statistical models
predicting the likelihood of nuclear reversal. The third section presents various logistic
regression and Cox proportional hazard models that test the relationship between anti-nuclear
norm acceptance and the likelihood of nuclear reversals. Finally, the conclusion summarizes and
analyses the findings.
2) The UNGA anti-nuclear voting record dataset
Due to the problems of using NPT membership as a measurement of anti-nuclear norm
acceptance, I developed a novel measure based on how countries voted on nuclear related
resolutions in the United Nations General Assembly. Since this project is interested solely on
those countries that obtained nuclear weapons, pursued nuclear weapons, or terminated nuclear
weapon pursuit, I gathered UNGA voting record data only for the 32 countries identified as
42

nuclear pursuers by Mehta’s research, rather than all countries in the world.
53
UNGA voting
records were gathered for 32 countries from the years 1950-2020.
To gather the data, I searched for all UNGA resolutions with the word “atomic” or
“nuclear” in their title from the United Nations Voting Dataset in the UN Digital Library.
54
I
focus on UNGA resolutions and exclude Security Council resolutions because SC resolutions are
qualitatively different from UNGA resolutions. Security Council decisions are considered legally
binding as stated by Article 25 of the UN Charter, while UNGA resolutions are not and thus
provide states more flexibility and fewer pressures to vote alongside their Great Power allies and
rivals.
55
Furthermore, the UNSC consists of only 15 member states at a given time, five of them
permanent members, while the General Assembly consists of virtually all nation-states. As a
result, UN resolutions voted on the UNGA provide a representative sample of world opinion
compared to the UN Security Council. The temporal time frame search was set from 1950-2020
to capture variance in UNGA voting records over time. The search output consisted of 1052
UNGA resolutions that contained the word “Atomic” or “Nuclear” in their title from 1950-2020.
Using the help of an undergraduate research assistant, specific metadata was hand-coded
for each resolution, creating a dataset of 1052 UNGA resolutions from 1950-2020. The
following information was collected for each resolution: Title of the resolution, the UNGA
resolution code, the link to the specific resolution, year of resolution, record of how many states
voted yes, voted no, voted abstained, or did not vote at all, how many nation-states were listed as
voting members in the UNGA on that year, and the specific date the resolution was voted on. If a

53
Mehta (2020). The list of countries can be found in the appendix.
54
The UN Voting Dataset can be accessed through this link:
https://digitallibrary.un.org/search?ln=en&cc=Voting+Data
55
https://legal.un.org/repertory/art25.shtml
43

resolution was adopted without a vote, “NA” was coded for the “voted yes/no/abstain/no
vote/total voting members” columns. The result of this data collection effort is a dataset in which
each row represents a particular UNGA nuclear resolution. Figure 1 below is a snapshot of the
first 26 rows and 10 columns as an example of what the data looks like.
Figure 1: Snapshot of the data #1

Since the interest of this project is to measure the degree of anti-nuclear norm acceptance
among states that pursued nuclear weapons and how it might affect the likelihood of nuclear
reversals, voting information on each resolution was collected for the 32 states identified as
nuclear weapon pursuers by Mehta (2020).
56
Ten of these countries successfully obtained the
nuclear weapon, although South Africa is the only one of these ten that dismantled its nuclear
arsenal, in 1993. The remaining 22 states reversed course and terminated their nuclear weapon
ambitions, sometimes willingly and other times by force. The full list of countries, including

56
The exact number of states that are coded as nuclear pursuers and their dates of pursuit differ across scholars.
However, Mehta (2020) provides one of the most updated lists of nuclear pursuers in existing research to date, so
her list of nuclear pursuers is used here.
44

their nuclear status as of 2020 and years of nuclear pursuit according to Mehta’s coding, can be
found in Appendix A.
For each of the 32 nuclear pursuer states, the following codes were used to record their
votes on each UNGA resolution: if a state voted “yes” on a resolution, it was coded 1. If the state
voted “no”, it was coded 2. An abstain vote was coded 3. If the state did not vote at all, it was
coded 4. If a state vote was anonymous, it was coded 5. If a resolution was adopted without a
vote, it was coded 9. Finally, NA was recorded for any state-year in which the state was not a
member of the UNGA and so could not be present to vote on the resolution. In sum, individual
voting records for 32 states have been recorded alongside each specific UNGA resolution.
As an illustration of how the state votes for each resolution are coded, Figure 2 below
shows a snapshot of the first 26 rows of the dataset, with each column after “Vote Date”
representing a specific state and how they voted on each resolution. For instance, Egypt is coded
as 1 (yes vote) on the resolution in row 2. Egypt’s vote for the resolution in row 14, however, is
coded as 3, an abstain vote. Germany is coded NA in the resolutions shown below because it was
not a member of the United Nations at the time; it did not join until September 18, 1973.
Figure 2: Snapshot of the data #2

45

A natural question that follows from observing the gathered data is how many resolutions were
actually voted on by states; as explained above, some resolutions were adopted without a vote,
while others were voted on anonymously. Thus, we can classify the UNGA nuclear resolutions
into three categories: UNGA resolutions adopted without a vote, resolutions voted on
anonymously, and resolutions with a public record vote. As shown on Table 2, out of 1052
resolutions, 315 were adopted without a vote, 37 received anonymous votes, and 700 were
adopted after a recorded public vote. In practical terms, this means that out of 1052 collected
resolutions, we have detailed voting information for 700 resolutions, or 66.5% of the data.
Table 2: Distribution of UNGA voting categories
Resolutions with public vote Resolutions with anonymous
vote
Resolutions adopted without
a vote
700 (66.5%) 37 (3.5%) 315 (29.94%)

Another important characteristic of the data is the distribution of resolutions per year. Figure 3
below visualizes the annual distribution of UNGA nuclear resolutions. The solid line represents
the total number of UNGA nuclear resolutions per year. The dashed line in the middle of the
graph represents the total number of resolutions adopted after a public vote, while the dotted line
at the bottom of the graph represents the total number of resolutions adopted without a vote.
57

The visualized data can be roughly divided into three time periods: 1) an early period of UNGA
nuclear resolutions between 1950-1979, in which a few resolutions are observed per year; 2) a
middle period consisting of a decade of high activity in the United Nations General Assembly
between 1980-1990; 3) and a modern period from 1991-2020 in which the average number of
nuclear resolutions decreases and remains stable. The number of nuclear resolutions is relatively

57
Resolutions adopted with anonymous votes are not graphed separately, since there were only 37 such observations
in the dataset.
46

small between the 1950-1979 period, with an average of 8.7 resolutions per year. However, the
decade between 1980-1990 saw increased activity in the UNGA regarding passage of nuclear
resolutions, evidenced by the spike observed in the middle of the graph. During that decade we
observe an average of 24.9 resolutions per year. Within this time period, we observe the highest
number of resolutions in a single year, in 1985, with a total of 31 nuclear resolutions; 24 of these
resolutions were adopted after a public vote while seven were adopted without a vote. However,
after 1990, the average number of resolutions per year decreases to 18.4.
Figure 3: Resolutions per year

Using this data, it is possible to calculate the proportion of “yes” votes for each unique resolution
by dividing the total number of “yes” votes on a given resolution by the total number of voting
UN members present. After calculating the proportion of “yes” votes for each resolution, we can
47

obtain the annual average proportion of “yes” votes as follows: All proportions of “yes” votes
obtained for each voted resolution can be summed up by year, then divided by the total number
of resolutions voted on that year. The result is the annual average of “yes” votes on UNGA
nuclear resolutions, which may serve as a proxy for international acceptance of anti-nuclear
international norms over time.
The results in figure 4 below are surprising—it appears that the annual average
proportion of “yes” votes on UNGA nuclear resolutions becomes relatively stable over time. The
average proportion of yes votes over time is 0.74. The average does not change much if we look
at the three time periods of UNGA voting separately: From 1950-1979, despite some major
fluctuations observed in the graph, the average proportion of yes votes on nuclear resolutions for
the time period is 0.75. During the peak period of voting between 1980-1990, the average is
0.77. The average decreases for the 1991-2020 period, to 0.73. In simple terms, it seems the
highest amount of support for UNGA nuclear resolutions is observed during the 1980-1990
period in which the highest number of UNGA nuclear resolutions are observed, but this average
support slightly decreases in the post-Cold War period. During the modern period, voting
patterns on UNGA nuclear resolutions appears to become more stable, as shown by the less
erratic change in the “yes” vote proportion between 1991-2020.

48

Figure 4: annual proportion of “yes” votes

Crucially, figure 4 suggests that there is no increased support by the international community for
UNGA nuclear resolutions over time, contrary to my initial expectations. Undoubtedly there has
been greater support for international non-proliferation norms as evidenced by the existence of
international treaties related to prevention of nuclear weapon proliferation and strict limits on
sensitive nuclear technology transfers. However, this support does not appear to be reflected in
the UNGA voting patterns on nuclear related resolutions. On the other hand, the graph does
show that these voting patterns become stable over time and have remained at a consistent high
of 74-77% approval for nuclear resolutions. Thus, although the data of how the international
community votes on nuclear resolutions does not point to an increase over time, it does show that
in general a majority of the international community votes in favor of these resolutions.
49

The global proportion of yes votes on UNGA nuclear resolutions over time is
informative, but what we need is a measure of norm acceptance by individual states that could be
used in monadic data analysis. In other words, how did the voting patterns change over time for
the 32 states identified as nuclear pursuers? Using the data obtained so far, we can build a
country-year measure of international nuclear nonproliferation norm acceptance for each of these
individual states. For each of the 32 states whose votes were coded, I added up how many times
a state voted yes, no, abstain, or “no vote” for each year. Then, for each country-year, I divided
the total number of yes votes over the total number of voted resolutions from that year to obtain a
proportion representing level of acceptance of international norms against nuclear weapons on an
annual basis.
For instance, out of 17 resolutions with “Nuclear” or “Atomic” in their title in 1979, 11 of
them were adopted with a vote. Of those 11, Brazil voted “yes” on six of them and voted
“abstained” in the other five. To obtain the measure of norm acceptance by Brazil for that year, I
divided the total number of yes votes, six, by the total number of voted resolutions, 11. The
output was a proportion of yes votes over total voted resolutions for Brazil in 1979, 0.54 or 54%.
The resulting output is used as a proxy of how much a state accepted international norms against
nuclear weapons for that given year. In Brazil’s case, the acceptance rate is 54% for 1979.
This measurement is not perfect and should be taken as a rough approximation of norm
acceptance over time. Some years might have more voted resolutions than others, and some
specific votes by states might be contingent on context-dependent circumstances such as
pressures by allies or even individual characteristics of state representatives voting in the UNGA.
Regardless, the measure of norm acceptance based on proportion of yes votes over total voted
resolutions should provide a rough approximation of how states are willing to vote over time in
50

favor of UNGA ratifications that condemn, restrict, or call for abolition of nuclear weapons,
proliferation, and nuclear weapons use.
The result of the data conversion described above is a variable that measures the
proportion of “yes” votes on UNGA nuclear resolutions on a country-year basis, resulting in
1903 unique country-year observations. This “proportion of yes votes” variable serves as a proxy
measuring acceptance of nuclear non-proliferation norms, for 31 countries except Taiwan, which
is not included because it is not an independent UN member state. The created variable can be
combined with other country-year datasets as an alternative measure of international nuclear
nonproliferation norm acceptance, instead of relying on the binary NPT ratification proxy
measurement. In sum, the variable measures nonproliferation norm acceptance in a 0-1 ratio
scale, the higher the “yes” vote proportion in a specific country-year, the higher the support for
non-proliferation norms in that specific year. Table 3 below shows the descriptive statistics of
the resulting variable, including years and number of countries covered, the average number of
yes, no, abstain, and non-votes, and finally the mean value of the new nuclear nonproliferation
norm acceptance proxy variable.
Table 3: Descriptive statistics of state non-proliferation acceptance data
Years States Yes votes
(mean)
No votes
(mean)
Abstain
votes
(mean)
No vote
(mean)
Norm
acceptance
(mean)
1950-
2020
31 6.43 1.74 2.21 0.41 0.54

51

3) Methodology and measurements
Having obtained the new measure of international nonproliferation norm acceptance, the
next step is to use the variable in quantitative analysis to test whether acceptance of international
norms against nuclear weapons increases the likelihood of nuclear reversals. To carry out such
an analysis, I merged the norm acceptance variable described above with a dataset covering 31
countries that attempted to pursue nuclear weapons at some point in their history according to
Mehta’s coding,
58
covering the years 1944-2020.
59
Furthermore, I merged various control
variables thought to affect the likelihood of nuclear reversals, which will be described in detail
below. Countries enter the dataset from the first year they are coded as pursuing nuclear
weapons, so the data exclusively covers countries that attempted to pursue nuclear weapons
rather than all countries in the world. The unit of analysis is country-year, for a total of 1951
country-year observations. Appendix A lists all the countries included in the dataset, based on
Mehta’s coding. Countries in the “Reversal” category terminated their nuclear weapons
programs, while countries in the “Non-Reversal” category either have nuclear weapons or were
still pursuing as of 2020. Using this constructed dataset, I employ logistic regression and Cox
hazard models to test the relationship between nuclear norm-proliferation norm acceptance and
nuclear reversals.
Dependent variable
The dependent variable is nuclear reversal, which refers to state abandonment of its nuclear
weapons program. The dependent variable is a binary variable, coded as 0 during country-year

58
Mehta (2020, 21-25).
59
Although the dependent variable, nuclear reversals, and the independent variable, international nonproliferation
norm acceptance, have data up to 2020, the temporal coverage of the control variables vary. As a result, the number
of observations will vary for each model depending on which control variables are included.
52

observations in which a state is pursuing the nuclear bomb or owns nuclear weapons, and coded
as 1 for country-year observations after termination of the program, based on Mehta’s coding.
1114 observations are coded as 0, while 837 observations are coded as 1, indicating that there are
more pursuit/nuclear ownership observations in the dataset.
Independent and control variables
The main independent variable is norm acceptance, a proxy variable that measures the degree to
which a state accepts international norms against nuclear weapons.
60
The variable, as described
earlier, measures the proportion of “yes” votes by a state on UNGA resolutions that had the word
“nuclear” or “atomic” in their titles divided by total voted resolutions per year. The assumption is
that the higher the proportion of yes votes on a given year, the more a state accepted international
norms against nuclear weapons, broadly speaking. The variable is a ratio variable varying
between 0 (zero yes votes in a year) to 1 (all votes on resolutions for that year were yes votes).
The temporal coverage of the variable is 1950-2020, including data for 1707 country-year
observations. The following hypothesis is tested using this new variable:
Norm acceptance hypothesis (H1): an increase in the level of nuclear non-proliferation
norm acceptance will increase the likelihood of nuclear reversals.
To measure democracy, one of the key control variables, this dissertation draws from the
Varieties of Democracy (V-Dem) Project, which built indexes measuring different dimensions of
democracy: electoral, participatory, egalitarian, liberal, and deliberative democracy. Electoral
democracy is defined in the standard Schumpeterian sense: electoral competition with extensive
suffrage, free operation for civil and political organizations, elections without irregularities or

60
The measure is broad by design: resolutions could be against nuclear proliferation, ownership, nuclear
development, etc.
53

fraud, and the elections affect the composition of the chief executive. To measure electoral
democracy, V-Dem created an index of electoral democracy. The index is “formed by taking the
average of, on the one hand, the weighted average of the indices measuring freedom of
association thick, clean elections, freedom of expression, elected officials, and suffrage and, on
the other, the five-way multiplicative interaction between those indices.”
61
The variable is an
interval variable ranging from 0-1. The closer the score is to 1, the more democratic the state.
The measure of democracy covers the full timespan under analysis, 1944-2020 for all countries
included in the study. For robustness checks, I include the measure of democracy from the Polity
5 project.
62
Measured from -10 to 10, the closer the score is to 10 the more democratic a country
is for a given year. Unlike the V-Dem variable, temporal coverage extends up to 2018. As
predicted by the theory, higher levels of democracy should increase the likelihood of nuclear
reversals, leading to the following hypothesis:
Democracy hypothesis (H2): greater levels of democracy will increase the likelihood of
nuclear reversals.
The second key control variable, perception of threat environment, is notoriously difficult to
measure, because perceptions are not physically observable. As a proxy variable, and following
standard practice, threat environment is measured by the total number of military interstate
disputes (MIDs) that a country was involved in for each year. I use the MID dataset version 5,
which includes militarized interstate disputes up to 2014.
63
The variable is ordinal, in which the
values indicate the number of MIDs a state was involved, regardless of initiator status, in a year.

61
Coppedge et al. (Codebook 2022, 43).
62
Marshall and Gurr (2020).
63
Palmer et al. (2020).
54

The assumption is that the more MIDs a country is involved within a year, the higher threat
perception should be for state leaders:
Threat environment hypothesis (H3): an increase in the perception of threat will decrease
the likelihood of nuclear reversals.
Other control variables are also included in the statistical analysis due to their potential effect on
nuclear reversals. Economic development might influence the likelihood of nuclear reversals
because poorer countries and countries in economic decline might face additional hurdles in
maintaining a nuclear weapons program afloat. Therefore, I include gdp per capita as a measure
of economic growth. Gdp per capita data is drawn from the World Bank, covering the years
1960-2020.
64
GDP per capita is measured as a ratio variable, higher values indicating a stronger
national economy. For robustness checks, I also include the point estimate latent variable model
from Fariss et. al. as an alternative measure of economic development, covering the years 1945-
2019.
65
The variable is a ratio variable, with higher values indicating higher GDP per capita.
The number of neighbors a state has might also influence the likelihood of nuclear reversals. One
of the underlying causes of war are territorial disputes, which are mostly fought between
neighbors.
66
Under this logic, states with several neighbors might be more prone to inter-state
conflict than states with fewer neighbors, which could negatively affect the calculus for nuclear
reversal in those states surrounded by rivals; being surrounded by several potential hostile
neighbors reduces the likelihood that a state will terminate pursuit of the nuclear bomb.
Therefore, I control for number of land border and total borders countries have per year. I draw

64
World Bank (2021).
65
Fariss et al. (2022).
66
Vasquez (2009, 136).
55

the data from the Correlates of War Project’s Direct Contiguity Data, version 3.2, which counts
how many borders a country shares with other states, either direct land borders (land border
variable) or total borders including sea borders (total borders), up to 2016.
67

The share of world power of a state might also influence the likelihood of nuclear reversals, for
stronger states might be more likely to afford a complex and expensive nuclear program. To
measure share of world power, I use the Correlates of War Project’s National Material
Capabilities (NMC) data, version 6.
68
The NMC measure of share of world power is an index
built by combining six different indicators of state power: military expenditures, military
personnel, energy consumption, total population, urban population, and iron and steel
production. The index measures for each state, on an annual basis, how much share of world
power it has up to the year 2016. The measure ranges from 0-1, the closer to 1 the greater the
share of world power a state has on a given year.
Way and Weeks’ 2014 article, “making it personal: regime type and nuclear proliferation”
argues that personalist dictatorships are more likely to pursue nuclear weapons than both non-
personalist dictatorships and democracies.
69
If so, including their variable measuring personalist
leader might be important because this factor could also affect the likelihood of nuclear
reversals. If personalist dictatorships are the most likely to pursue nuclear weapons, perhaps they
are also less likely than non-personalist dictatorships and democracies to give up pursuit of the
bomb. Way and Weeks variable, personalist leader, included in some of my models as a control,

67
Correlates of War Project (2017).
68
Greig and Enterline (2021).
69
Way and Weeks (2014).
56

is a binary variable that categorizes personalist regimes as 1 otherwise 0, covering up to the year
2000.
An additional variable that might influence the likelihood of nuclear reversals is whether a
country is receiving sensitive nuclear assistance, such as sensitive nuclear technology exports.
Countries that do not receive sensitive nuclear assistance might be more likely to terminate their
nuclear ambitions than countries that do receive such assistance, due to the increased costs
associated with starting or maintaining a nuclear weapons program without foreign support. To
measure sensitive assistance, I borrow the extended measure of sensitive assistance from Way
and Weeks data, which expands coverage of Kroenig’s binary measure of assistance up to
2007.
70
Country-years in which a state is receiving foreign assistance are coded as 1, otherwise 0.
Finally, I include two control variables, NPT membership and cold war. NPT membership is a
binary variable that codes whether a country is a signatory of the Nuclear Non-proliferation
Treaty, from the first year of the treaty, 1968, up to 2020. This variable will be used as a
robustness check in place of the new measure of norm acceptance in some of the models
presented below. Cold war is a binary measure, coding as 1 the Cold War years, 1947-1991, and
0 post-Cold War years up to 2020.
4) Empirical results
Table 4 shows the results of logistic regression models testing the relationship between
international nuclear nonproliferation norm acceptance and the likelihood of nuclear reversals.
Although not shown in the table, all models include time fixed effects. Model 1 tests the binary
relationship between norm acceptance, as measured by the UNGA voting records on nuclear

70
Kroenig’s (2009) original measure of sensitive nuclear assistance covers up to the year 2000.
57

resolutions, and nuclear reversals. The results of this basic logit model are positive and
statistically significant at the p<0.01 level, indicating that the greater the acceptance of nuclear
nonproliferation norms, the higher the likelihood of a nuclear reversal to occur. Model 2 includes
the two intervening variables theorized to matter the most for nuclear reversals, democracy
(measured using V-dem data) and threat environment. As hypothesized, all three variables are
statistically significant and point in the expected directions: as norm acceptance increases, the
likelihood of nuclear reversals also increases. Similarly, greater democracy scores increase the
likelihood of a nuclear reversal. On the other hand, when threat perception increases, the
likelihood of nuclear reversals decreases, indicated by the negative coefficient in the model.
Model 3 increases complexity by adding other control variables that might affect the
relationship between norm acceptance and nuclear reversals. The results for the main
independent variable remain unchanged, although other control variables also gain statistical
significance in expected directions. Both greater military power and sharing more land borders
decrease the likelihood of nuclear reversals. In addition, nuclear reversals were less likely to
occur during the Cold War period compared to the post-cold war period. Finally, model 4 adds
sensitive nuclear assistance into the equation, without affecting the main results of the models.
Note that the number of observations decreases as more controls are included in the models; the
number of observations available for statistical analysis drop due to the temporal limitations of
the data. For instance, in model 3, the variable GDP per capita only has data starting from 1960,
while the variables land borders and military power cover up to the year 2016, which constricts
the temporal period of analysis for model 3 from 1960-2016, leading to 1284 observations in the
model. Similarly, more observations are dropped once sensitive assistance is added to model 4,
due to the variable measuring assistance up to 2007, leading to 1063 observations in the model.
58

A concern of logistic regression, and observational data analysis broadly speaking, is that
the results will change significantly if the control variables are measured differently. Table 5
addresses this concern: each model is a robustness check in which alternative measures of
control variables are included. The first model in table 5 measures democracy using the Polity 5
scores of democracy rather than v-dem measurements. Results remain unchanged; an increase in
norm acceptance increases the likelihood of nuclear reversals. Model 2 uses the latent variable
point estimate of GDP developed by Fariss et al. instead of GDP per capita logged, which once
again reveals that results do not change for the independent variable. Model 3 includes a measure
of total borders rather than just land borders, showing that the statistical significance and
direction of effect of norm acceptance on nuclear reversal likelihood does not change.

59

Table 4: Logistic regression models
Logit Models 1-4

Dependent variable:

Reversal

M1 M2 M3 M4

(1) (2) (3) (4)
norm acceptance 0.977
***
4.190
***
3.910
***
2.212
***

(0.204) (0.447) (0.506) (0.523)
Democracy

5.296
***
5.085
***
5.165
***

(0.372) (0.514) (0.546)
threat environment

-0.813
***
-0.678
***
-0.674
***

(0.080) (0.098) (0.110)
military power

-34.153
***
-40.853
***

(5.741) (6.781)
GDPpercapita

0.005 -0.192

(0.088) (0.098)
land borders

-0.257
***
-0.323
***

(0.035) (0.040)
cold war

-5.405
***
-5.004
***

(1.287) (1.279)
Sens. Assistance

0.555
*

(0.225)
Constant -17.432 -24.000 -1.982 0.271

(835.881) (1,098.441) (1.106) (1.188)
Observations 1,707 1,519 1,284 1,063
Log Likelihood -988.937 -586.311 -475.068 -399.693
Akaike Inf. Crit. 2,111.874 1,298.622 1,070.135 907.387
Note: + p<0.1; * p<0.05; ** p<0.01; *** p<0.001

60

Table 5: Robustness Checks
Robustness checks

Dependent variable:

Reversal

polity5 FarissGDP total borders personalist alternative model

(1) (2) (3) (4) (5)
norm acceptance 3.466
***
3.888
***
4.080
***
0.115 3.600
***

(0.454) (0.473) (0.495) (0.494) (0.416)
polity5 0.143
***

0.160
***

(0.018)

(0.016)
Democracy

5.394
***
5.277
***

(0.501) (0.502)

Personalist

-0.109

(0.280)

threat environment -0.778
***
-0.708
***
-0.569
***
-0.812
***
-0.701
***

(0.099) (0.092) (0.090) (0.123) (0.086)
military power -24.252
***
-31.754
***
-32.712
***
-16.347
***
-24.282
***

(4.915) (5.175) (5.525) (4.604) (4.192)
GDPpercapita 0.301
***

0.076 0.422
***

(0.081)

(0.085) (0.092)

FarissGDP

-0.005

0.045
***

(0.011)

(0.009)
land borders -0.271
***
-0.244
***

-0.292
***

(0.034) (0.034)

(0.039)

total borders

-0.038

-0.030

(0.024)

(0.022)
cold war -4.623
***
-20.600 -4.614
***
-3.716
**
-19.364

(1.289) (1,000.494) (1.275) (1.218) (1,091.015)
Constant -2.084 -1.988
**
-3.879
***
-0.412 -1.589
*

(1.084) (0.759) (1.072) (1.146) (0.708)
Observations 1,277 1,519 1,284 815 1,508
Log Likelihood -493.959 -523.275 -503.259 -363.387 -570.977
Akaike Inf. Crit. 1,107.917 1,178.550 1,126.518 818.774 1,273.955
Note: + p<0.1; * p<0.05; ** p<0.01; *** p<0.001

61

Model 4 in table 5 switches the v-dem measure of democracy for the personalist variable
developed by Way and Weeks.
71
Interestingly, the main independent variable, norm acceptance,
is not statistically significant, although the direction of effect on nuclear reversals does not
change. However, this model involves the fewest number of observations, due to the personalist
variable covering measures of personalist leader only up to the year 2000, explaining the relative
weakness of the model compared to other models covering a longer time span. Finally, model 5
uses the Polity 5 measure of democracy, Fariss’ point estimate model of GDP, and total borders
instead of the original variables, democracy measured by v-dem, log GDP per capita, and land
borders. Once again, the substantive results seem to suggest a positive effect of nuclear norm
acceptance on increased likelihood of nuclear reversals.
As a further robustness check, I switched out the main independent variable, norm acceptance,
and include NPT membership instead to observe if there is a difference in the results. Table 6
below replicates all the models in Table 4, but with NPT membership status as the measurement
of nuclear non-proliferation norm acceptance, following standard practice. The results do not
differ substantially from Table 4; being a member of the NPT increases the likelihood of nuclear
reversals across all four models. If we compare the Akaike Information Criterion (AIC) values in
Table 6 with the AIC values in Table 4 across models, it appears that the models in Table 6
better fit the data, for the AIC values are smaller in head-to-head comparisons with their
equivalent models in Table 4. Although this comparison suggests that NPT membership might be
the better measure of nuclear non-proliferation norms than UNGA voting records, it must be
reiterated that the seemingly better measure has major theoretical problems. After all, states are
likely to join the NPT while pursuing the nuclear bomb in secret, or feel coerced to join by a

71
Way and Weeks (2014).
62

greater power. In addition, NPT membership is a dichotomous variable, which fails to capture
nuances in the level of acceptance of nuclear non-proliferation norms by states. Therefore, for
theoretical reasons this might be a case in which the seemingly weaker measure, norm
acceptance measured by UNGA voting records, conceptually captures international nuclear non-
proliferation norm acceptance better than NPT membership.
72

Table 6: NPT membership models
NPT membership models

Dependent variable:

Reversal

M1 M2 M3 M4

(1) (2) (3) (4)
NPT member 2.850
***
3.131
***
3.524
***
3.943
***

(0.206) (0.239) (0.300) (0.363)
Democracy

3.339
***
5.507
***
7.468
***

(0.252) (0.521) (0.703)
threat environment

-0.998
***
-0.762
***
-0.830
***

(0.080) (0.102) (0.121)
military power

-36.017
***
-50.231
***

(5.684) (8.925)
GDPpercapita

-0.699
***
-0.851
***

(0.110) (0.133)
land borders

-0.171
***
-0.213
***

(0.036) (0.044)
cold war

-4.861
***
-3.785
**

(1.096) (1.153)
Sens. Assistance

1.755
***

(0.294)
Constant -3.314
***
-4.824
***
3.620
***
2.220

(0.637) (0.765) (1.026) (1.151)
Observations 1,586 1,398 1,225 1,004
Log Likelihood -864.002 -521.995 -404.291 -314.639
Akaike Inf. Crit. 1,836.004 1,143.991 914.583 723.278
Note: + p<0.1; * p<0.05; ** p<0.01; *** p<0.001

72
If this explanation is correct, then it highlights one of the weaknesses of quantitative research: results are only as
good as measurements and model specifications, yet there is no guarantee that the “better model” is the best
approximation of reality. Qualitative case studies are thus needed to complement quantitative analysis and capture
nuances that might be missed by the data.
63

In addition to the logistic regression models, an additional regression modeling technique
often employed when dealing with rare events data are Cox proportional-hazard models.
73
Cox
hazard models, a type of survival analysis, focuses on the expected duration of time until
occurrence of an event—in medical research, usually time until death, while in this work time
until nuclear reversal. In sum, the Cox hazard model can be used to investigate the association
between the rare event of nuclear reversal and one or more predictor variables. Table 7 below
shows the results of the Cox hazard models. Crucially, none of the variables in all the models
approach statistical significance-with the exception of threat environment in model 2- indicating
that we cannot conclude which factors increase/decrease the risk of a nuclear reversal. These
results contrast the logistic regression models, which reveal a positive relationship between norm
acceptance and the likelihood of nuclear reversals.
This difference in results is not an issue, due to Cox and logistic regression modeling techniques
serving different purposes. The intent of the Cox-hazard models is to reveal the time to reversal
given a set of predictor variables. The results show that, given the data, we cannot make a
conclusive statement regarding which variables affect the timing of a nuclear reversal to occur.
On the other hand, logistic regression models calculate the likelihood that a reversal will occur
given a set of predictor variables, rather than the time-to-reversal as a Cox hazard model does.
Note also the smaller number of observations in the Cox hazard models compared to the logistic
regression models in the previous tables. This is the result of states dropping from analysis in
Cox hazard models once the nuclear reversal threshold is reached, which is not the case with

73
Cox (1972).
64

logistic regression models that do not drop country-year observations past the first nuclear
reversal event.
Table 7: Cox Hazard Models
Cox Hazard Models

Dependent variable:

Reversal

M1 M2 M3 M4

(1) (2) (3) (4)
norm acceptance 0.176 0.126 -0.439 -0.580

(0.693) (0.844) (0.988) (1.035)
Democracy

0.760 2.476 2.453

(0.849) (1.685) (1.822)
threat environment

-0.594
*
-0.136 -0.226

(0.284) (0.319) (0.333)
military power

-62.000 -53.039

(35.424) (33.200)
GDPpercapita

-0.176 -0.049

(0.291) (0.331)
land borders

-0.173 -0.212

(0.161) (0.173)
cold war

-1.799 -1.836

(0.946) (0.979)
Sens. Assistance

1.051

(0.792)
Observations 949 891 671 607
R
2
0.0001 0.009 0.026 0.028
Max. Possible R
2
0.106 0.107 0.113 0.117
Log Likelihood -53.027 -46.601 -31.638 -29.166
Wald Test 0.060 (df = 1) 5.840 (df = 3) 8.960 (df = 7) 8.370 (df = 8)
LR Test 0.065 (df = 1) 8.122
*
(df = 3) 17.411
*
(df = 7) 17.247
*
(df = 8)
Score (Logrank) Test 0.065 (df = 1) 6.122 (df = 3) 11.357 (df = 7) 12.707 (df = 8)
Note: + p<0.1; * p<0.05; ** p<0.01; *** p<0.001

65

5) Conclusion
This chapter introduced a novel proxy measure of nuclear non-proliferation norm acceptance
based on United Nations General Assembly voting records on nuclear resolutions. Then, the new
variable was used in statistical analysis to determine if there is a statistically significant and
positive effect on the likelihood of nuclear reversals. The results of the analysis across several
model specifications imply that, holding various control variables constant such as level of
democracy and perception of threat, greater support for nuclear non-proliferation norms—
measured as higher proportion of “yes” votes on UNGA voting resolutions—increases the
likelihood that nuclear reversals will occur. The results increase confidence that international
norms do have an independent effect on reversals apart from material and economic factors such
as material power, regime type, and economic development of the state. Among the main
intervening variables thought to affect the likelihood of reversal, level of democracy increased
while heightened perception of threats decreased the likelihood of nuclear reversals, supporting
the predictions of all three main hypotheses established at the beginning of the chapter.
However, weaknesses of the new nuclear non-proliferation norm acceptance variable must be
highlighted for transparency. As noted in the beginning of the chapter, the norm acceptance
variable was built by gathering data on UNGA resolutions that included the words “nuclear” or
“atomic” in their titles. As a result, the data gathered on UNGA voting records is conservative in
nature; there are undoubtedly several resolutions that deal with nuclear weapon proliferation,
testing, or ownership that were not included in the dataset due to lacking the key words in their
title. Furthermore, and perhaps more importantly, the measure might be accused of conceptual
overstretching: it combines in a single category several related norms such as anti-proliferation
norms, anti-nuclear weapon ownership norms, and anti-nuclear testing norms to form a broad
66

category of “anti-nuclear norms.” Further iterations of the project must categorize the different
resolutions based on the category that fits best, such as a group of resolutions specifically about
nuclear ownership, a category for resolutions accusing other actors of violating nuclear norms, a
category for non-proliferation resolutions, and so on.
74
Finally, it must be re-stated that this is an
imperfect proxy. Just like states might join the NPT due to pressures from powerful states,
country votes on UNGA resolutions can sometimes be influenced by pressures from more
powerful state actors. However, unlike the NPT, which is joined at a single iteration, countries
vote on UNGA resolutions at various points through time. Thus, states have more flexibility to
vote according to their preferences, even if some votes are cast due to external pressures. If so,
the imperfect proxy developed here is still a superior option to the simplistic binary measure of
norm acceptance based on whether a country joined the NPT or not.
The results from the statistical analysis are also tentative and suggestive rather than conclusive.
First, given the weaknesses of the new measure of anti-nuclear norms, the results might differ
with a more fine-grained measurement that includes more UNGA resolutions or breaks down
resolutions into distinct categories. Second, logistic regression models relying on observational
data cannot prove causality, only show a correlation between variables of interest. In the absence
of an experimental or instrumental variable research design, doubts may still exist about the
causal effect that acceptance of international norms against nuclear weapons might have on the
likelihood of nuclear reversals. Given these issues with the quantitative models, tests of the
theory of nuclear reversals must be complemented with case studies. Case studies, which
carefully analyze specific cases through time, may reveal the causal mechanisms theorized by the
theory that cannot be captured in a large N observational statistical model. Therefore, the

74
A stand-alone article version of this chapter will tackle these issues in the next iteration of the draft.
67

following three chapters analyze three cases of countries that were nuclear pursuers: Brazil,
which fits the voluntary reversal pathway, Iraq, which follows the involuntary reversal pathway,
and Israel as a case of reversal failure.

68

Chapter 5: Iraq as a Case of Involuntary Reversal
In this chapter, I test the observable implications derived from the involuntary nuclear
reversal pathway, using the case of Iraq as a typical case of involuntary reversals. If the theory
successfully explains the nuclear reversal of Iraq, then the theory should also apply to similar
cases of states that terminated their nuclear weapons programs or ambitions due to coercive
threats or force. To discover if the nuclear reversal theory applies to the case, I rely on process
tracing techniques: I analyze the historical record of Iraq’s nuclear program and test whether the
theorized causal mechanism of involuntary reversal can be identified in the case. The chosen
case is tested via hoop tests, which evaluate hypotheses that claim a specific causal mechanism
occurred and if there is a causal connection between two or more events.
75
A hoop test proposes
that specific evidence, such as a specific causal process observation, must be found for a
hypothesis to be valid. Each of the five tested hoop tests test a specific sequence and component
of the involuntary reversal pathway, looking for supporting or disconfirming evidence of the
hypothesized pathway. Throughout this chapter and the following two chapters, “hypothesis”
and “observable implication” will be used interchangeably: Each hypothesis is an “observable
implication” of what we should expect to observe if the theory is correct.
The chapter is organized as follows. First, I present the historical background of Iraq’s
nuclear program, highlighting the key events of Iraq’s nuclear history up to 1998. Second, I
analyze Iraq’s regime type and perception of threat, two crucial intervening variables that are
thought to affect a state’s likelihood of accepting international non-proliferation norms. The third
section introduces the five hoop tests that must be supported if the involuntary pathway is to hold

75
Bennett and Checkel (2015, 7).
69

as an accurate account of nuclear reversals. The fourth section summarizes the findings. Finally,
the conclusion elaborates the significance of the findings of the broader theoretical framework of
nuclear reversals. Ultimately, I find that Iraq, as hypothesized, is an ideal case of how the
involuntary pathway of nuclear reversals functions. As expected, Iraq was unable and unwilling
to embrace liberal international norms against nuclear weapons, given its threatening security
environment and lack of democratic governance. Thus, the only way the international
community was able to prevent nuclear proliferation in Iraq was via the use of force; first with
Israel’s surprise attack on the Osirak reactor on June 7, 1981, and later on by the United States
and the Security Council’s successful efforts to dismantle the remnants of the covert nuclear
program in the early 90s.
1) Iraq’s nuclear program: 1956-1998
76

Iraq’s nuclear history begins with the founding of the Iraqi Atomic Energy Commission
(IAEC) in 1956, using the U.S. Atoms for Peace program to fund exploration of nuclear
technology. First announced in a speech to the United Nations General Assembly on December
8, 1953 by U.S. President Eisenhower, the “Atoms for Peace” program was a policy by the
United States to share non-military nuclear technology, training, and fissionable materials with
other states around the world in order to gain more allies, undermining the influence of the
Soviet Union during the Cold War.
77
At the time, Iraq was an ally of the United States, forming
part of the anti-communist Baghdad Pact alliance. The alliance with Iraq was short lived; Iraq’s
pro-Western Monarchy was overthrown in the July 14 Revolution of 1958, in which a secret
military group led by Brigadier Abd al-Karim Qasim killed most members of the incumbent

76
This historical account of the Iraqi nuclear program is largely drawn from the research by Lobell (2020); Debs and
Monteiro (2016); and Ekeus (2016).
77
Drogan (2019).
70

Hashemite monarchy, establishing the Republic of Iraq.
78
Two months after the coup, Iraq
withdrew from the Baghdad Pact.
Iraq’s new regime continued the efforts to improve Iraq’s nascent nuclear program. By
1962, the Soviet Union approached Iraq with offers for technical assistance and training and
supplied Iraq with a research reactor for the Tuwaitha Atomic Center near Baghdad.
79
These
efforts to develop Iraq’s nuclear program were accelerated following the Baathist Coup of July
1968. By 1970, some scientists in the nuclear program even began advocating for a nuclear
weapons option, such as Abdul Razzaq al-Hashimi, who oversaw the Baath Party organization at
the IAEC; an organization for which he would eventually become vice chairman.
80
Of course,
discussions regarding plans for the creation of a nuclear weapon were unknown to Western states
at the time. For example, in 1972, Iraq reached a deal with French Prime Minister Jacques Chirac
for the sale of a nuclear reactor and uranium fuel,
81
which might have been unlikely if France
suspected Iraq of wanting to pursue the nuclear weapon option. In 1973, Saddam Hussein, who
at the time was vice-president of Iraq, became president of the IAEC, which led to even faster
advancement of the nuclear program and increased international deals for nuclear materials and
technology.
Development of Iraq’s nuclear program ramped up in December 1974, when France’s
Prime Minister Jacques Chirac negotiated the construction of an Osiris-style nuclear reactor and
a smaller Isis-style nuclear reactor, Tammuz 1 and Tammuz 2 respectively.
82
The contract also
included the sale of 70-80 kilograms of highly enriched uranium fuel enriched to a weapons

78
http://news.bbc.co.uk/onthisday/hi/dates/stories/july/14/newsid_3736000/3736391.stm
79
Debs and Monteiro (2016, 132).
80
Braut-Hegghammer (2011, 105).
81
Debs and Monteiro (2016, 136).
82
Lobell (2020, 4).
71

grade of 93%. In January 1976, French workers began construction of the infrastructure for the
nuclear reactors, and by 1979 initiated construction of the reactor core.
83
Iraq also managed to
obtain assistance from Italy. In 1976, Italy sold a reprocessing laboratory with three hot cells to
Iraq, which was capable of extracting plutonium from the spent reactor fuel-an essential
component for construction of fissile materials.
84

These fast developments began to raise some suspicions regarding Iraq’s true motives.
For example, in 1976 the new French president Valery Giscard D’Estaing tried to change the
terms of the agreement by selling Caramel fuel for the reactors, which is “slightly enriched and
not suitable for uranium-based nuclear weapons.”
85
Iraq rejected this revised offer, and after
much debate France decided to deliver the highly enriched uranium (HEU) to Iraq as agreed.
However, France was not the only country with reservations and suspicions. In 1979, following
U.S. pressures, France ultimately amended the agreement and decided to ship the HEU in “six
deliveries of 12 to 12.5 kilograms rather than a single instalment.”
86
This arrangement served to
slow down potential diversion of HEU for nuclear weapon construction—any diversions would
have been detected by French engineers on site or by IAEA safeguards.
87

Most suspicious and concerned of all was Israel. According to Israel’s analysts, it was
evident that Iraq was in pursuit of a nuclear bomb. Israel’s intelligence agencies, Mossad, and the
Military Intelligence (AMAN), estimated that Iraq’s reactor would be operational by July or
September 1981, and that by 1985 Iraq might even have an atomic weapon.
88
As further

83
Ibid., 4
84
Ibid., 4
85
Ibid., 4
86
Ibid., 4
87
Braut-Hegghammer (2011, 112).
88
Lobell (2020, 5).
72

evidence for why Israel was fearful of Iraq, Feldman’s 1983 book The Bombing of Osiraq-
Revisited highlights five indicators of Iraq’s true intentions, as perceived by Israel: 1) Iraq’s
decision to purchase the non-economical gas-graphite from France, which is capable of
producing plutonium; 2) Iraq purchased a 70 megawatt Material Testing Reactor, when Iraq
wasn’t involved in production of power reactors; 3) Iraq’s leadership insisted that the Osiraq
reactor had to be fueled by 92 percent enriched weapons grade uranium rather than less enriched
caramel fuel; 4) Iraq’s purchase of 250 tons of natural uranium only makes sense if plutonium
production was intended; and 5) the acquisition of hot cell simulators from Italy could only be
explained by a desire to build nuclear weapons.
89

Due to reasonable suspicion, Israel engaged in a long campaign to slow down and
perhaps stop Iraq’s nuclear program. Israel’s intelligence agency Mossad first identified Iraq’s
nuclear ambitions in 1974, when it was detected that there was an increase in communications
between Iraq and France.
90
By 1975, Israel made slowing down Iraq’s program a centerpiece of
their foreign policy. Israel’s efforts to push towards Iraq’s nuclear slowdown ramped up in 1977,
when Prime Minister Menachem Begin succeeded Prime Minister Rabin. Under Begin,
diplomats “were directed to raise the issue [of Iraq’s program] in every meeting of any level and
extensive dissuasion efforts were undertaken in the French parliament.”
91
In 1979, Israel created
an interagency committee to coordinate the Iraqi dossier, named “New Era,” to “consolidate all
information on Iraq’s nuclear program and to outline” how to slow it down.
92
The committee

89
Feldman (1982, 115).
90
Sadot (2016, 656).
91
Ibid., 656
92
Ibid., 657
73

concluded that Iraq’s nuclear program was an existential threat to Israel, and recommended the
destruction of the Osirak reactor as a main goal.
As highlighted by Sadot, Israel employed a years-long campaign to slow down and
terminate Iraq’s nuclear program from the moment Israel detected Iraq’s nuclear weapon
ambitions. Apart from the dissuasion efforts towards France, Israel engaged in covert action to
target key individuals and sites. A notable sabotage attempt, which may safely be attributed to
Israel’s intelligence agencies, is the April 6, 1979 sabotage operation in the French port town
Seyne-sur-Mer. Infiltrators “entered the port compound” at night and “attached plastic explosives
onto outbound containers holding two reactor core structures about to be shipped to Iraq.”
93
On a
May 14, 2008 interview, the deputy head of French internal security, the Direction de la
Surveillance du Territoire (DST) said that “the Mossad’s ‘fingerprints,’ methods, and types of
explosives were all evident in this operation.”
94
A few assassinations attempts, some successful,
were also linked to Israel via strong circumstantial evidence: On June 14 1980, Yahya el-
Mashad, an Egyptian scientists working on HEU negotiations with France, was murdered in the
Paris Meridian Hotel.
95
In addition, various European scientists scheduled to fly to Iraq received
threatening calls and threats. These threats were effective; several scientists who were supposed
to fly to Iraq for development of the nuclear program resigned.
Ultimately, Israel determined that diplomatic and coercive tactics were insufficient to
deter Iraq from developing a nuclear program. On June 7, 1981, Israel launched a preventive
strike against the Osirak reactor, successfully destroying it. The attack was perceived as a serious
setback to Iraq’s nuclear ambitions. Saddam Hussein, who became president of Iraq in 1979,

93
Ibid., 657.
94
Ibid., 657.
95
Ibid., 658.
74

himself acknowledged it, “technically, they are right in all of their attempts to harm Iraq…They
might hit Iraq with an atomic bomb someday if we reach a certain stage. And we are prepared,
and if God allows it, we will be ready to face it.”
96
However, despite the major setback to Iraq’s
nuclear weapons ambitions, Iraq’s nuclear weapons pursuit continued covertly.
Jafar Dhiya Jafar, the scientific leader of Iraq’s nuclear program, told Saddam Hussein
that it was necessary to jumpstart a clandestine nuclear weapons program, “based on the
development of indigenous skills and technologies, rather than seeking to buy key components
from foreign suppliers.”
97
On September 3, 1981, Jafar returned to the Nuclear Research Center
at Tuwaitha and established Directorate 3000, which was tasked with planning the nuclear
weapons program.
98
Iraq’s clandestine nuclear program moved at a rapid pace; according to
Jafar’s own records, “from 1983 to 1991, the program’s staff increased by 60 percent
annually….The Iraqi nuclear establishment spent 792,899,913 Iraqi dinars on the weapons
program from 1982 to 1988, and an additional 669,446,170 dinars during 1990-1991.”
99
The
covert program, to avoid a repeat of the 1981 Osirak strike, engaged in research and
development of nuclear weapons pathways that would be harder to detect by the IAEA and
Israel.
To avoid detection, the IAEC pursued the production of fissile materials using the
method of electromagnetic isotope enrichment of uranium (EMIS), which was unlikely to be
detected by foreign intelligence agencies. In 1987, following a major restructuring of the nuclear
program, Iraq began pursuit of gas centrifuge uranium technology, a major development as that

96
Quoted in Brands and Palkki (2011, 156).
97
Braut-Hegghammer (2011, 116).
98
Ibid., 117.
99
Ibid., 117.
75

technology allows enrichment of uranium, essential for the creation of a nuclear weapon.
100
By
April 1987, the vice chairman of the IAEC proposed to Saddam that “the nuclear weapons
program was ready to transition from research and development to weaponization.”
101
In their
efforts to obtain the bomb, Saddam established the Al-Hussein Project, which researched how
Iraq would produce nuclear weapons. By April 1988, the “Group Four” unit was established,
which was dedicated to weaponization. And by January 1989, the PC-3 unit was created,
focusing on weapons design.
102
Following all these developments and organized efforts, Iraq was
on the brink of achieving nuclear weapons capability by the early 1990s. For instance, Jafar
believed that production and testing of a nuclear weapon could begin by 1993.
103

However, external events deviated Iraq’s nuclearization plans. On August 2, 1990,
Saddam Hussein ordered an invasion of Kuwait. Observing international pressures to withdraw
and fearing for its security, Iraq initiated a crash program (Project 601/603) on August 17, 1990
to speed up the development of a crude nuclear weapon within six months.
104
The crash program
ended up a failure; the United States launched Operation Desert Storm on January 16, 1991 to
liberate Kuwait. As concluded by the CIA’s 2004 Duelfer Report, which reported on Iraq’s
nuclear weapons program over time, nearly all key nuclear facilities were bombed during
Operation Desert Storm. According to the report, “Many of the facilities located at Tuwaitha
were devastated, and the EMIS enrichment plants at Tarmiya and Ash Sharqat were largely

100
Debs and Monteiro (2016, 136).
101
Braut-Hegghammer (2011, 124).
102
Ibid., 124.
103
Ibid., 126.
104
Ibid., 126.
76

destroyed. Iraq’s yellowcake recovery plant at Al-Qa’im and feed material production plant at
Mosul (Al Jazira) also were bombed during the war.”
105

In addition to U.S. strikes, Iraq’s nuclear ambitions were dashed by concerted
international efforts to punish Iraq for its aggression against Kuwait. The UNSC “condemned in
Resolution 678 (1990) the Iraqi attack, imposed sanctions against Iraq (including an embargo on
Iraq’s oil exports), and endorsed the liberation of Kuwait.”
106
Following the war, UNSC
Resolution 687 “required Iraq to dismantle its nuclear-weapons program and tasked the United
Nations Special Commission (UNSCOM) to carry out inspections on Iraq’s biological, chemical,
and missile capabilities.”
107
Iraq did not easily cooperate with the UNSC; Iraq frequently
obstructed inspectors by prohibiting entrance to facilities, among other delaying tactics. Among
the most famous obstruction incidents is the “parking lot incident” of September 1991, when a
joint UNSCOM/IAEA inspection team in central Baghdad was blocked from leaving an
inspection site with documents gathered at the site. Ultimately, the inspection team was allowed
to leave without some of the documents.
108

Two of the most visible attempts to hold Iraq accountable for trying to circumvent UN
sanctions occurred in 1993 and 1998. In January of 1993, the UNSC concluded that Iraq was “in
material breach” of its obligations, which led to bombing raids of radar sites and suspected
nuclear facilities by the United States, the United Kingdom, and France.
109
In August of 1998,
Saddam Hussein once more impeded access to key facilities to inspectors, once again violating
UNSC resolutions. As a response, on December 16, 1998, the United States launched Operation

105
Duelfer Report, Vol. 2 (2004, 4).
106
Ekeus (2016, 133).
107
Debs and Monteiro (2016, 138).
108
Ekeus (2016, 138).
109
Debs and Monteiro (2016, 139).
77

Desert Fox, a four-day bombing campaign against Iraqi military targets believed to be involved
in storing, producing, or delivering weapons of mass destruction. The effects of all these
disarmament actions ultimately led to the termination of Iraq’s nuclear program. Scott Ritter, a
UN weapons inspector in Iraq between 1991 and 1998, concluded that the severity of the
sanctions led Saddam to terminate his nuclear program.
110

An exact date for when the nuclear weapons program in Iraq terminated might be
impossible to ascertain given available evidence and incentives for Saddam Hussein to hide this
fact from international observers. At the earliest, 1995 might be pointed out as the date of
termination, when UNSC Resolution 687 directed the IAEA to find and dismantle Iraq’s nuclear
weapons program, requiring Iraq to comply with the NPT by submitting to monitoring and
verification of sites.
111
Yet, continuing Iraqi recalcitrance to cooperate and doubts about
capabilities led to the U.S. strikes on suspected nuclear sites in 1998 during Operation Desert
Fox, as detailed above. Thus, at the latest 1998 might be the true date of termination, if we
conclude that there might have been existing attempts to covertly re-start the program, explaining
why Iraq refused to cooperate with the IAEA inspectors so late in the 90s. Regardless of the true
termination date, the UNSCOM and IAEA mark 1995 as the year when they reported
disarmament, which makes 1995 an appropriate year to code the end of the Iraqi nuclear
program.
2) Iraq’s regime type and perception of threat
Within the period of analysis, 1956-1998, Iraq fulfilled the two key characteristics that
make a country least likely to be receptive towards international anti-nuclear proliferation norms:

110
Ritter (1999, 34).
111
Mehta (2022, 22).
78

First, it remained an autocratic regime throughout the 1956-1998 period. Second, Iraq
consistently perceived a high level of threat from its neighbors, Iran, and Israel. Given these
facts, Iraq was unwilling to accept the international non-proliferation regime. Pursuit of nuclear
weapons was essential to establish itself as a regional power in the Middle East and ensure
national security, although some evidence suggests that ownership of nuclear weapons might
have been driven by offensive purposes towards Israel as well.
112

The historical record shows a violent history of political upheaval and forced government
transitions throughout the period of analysis. Iraq achieved independence from the British in
1932, marking the beginning of the Kingdom of Iraq. The Kingdom of Iraq was a monarchy
from 1932 until 1958, ruled by the Hashemite dynasty. The monarchy was relatively short-lived;
it came to an end on July 14, 1958, when military officers led by Colonel Abdul Salam Arif and
Abdul Karim Qasim led a coup d’état against King Faisal II and the royal family. After
assassinating the royal family in the courtyard, Qasim and Arif established the Republic of Iraq.
However, Qasim soon sidelined Arif, who was the architect of the coup, and concentrated all
power in his hands.
113
Qasim’s rule was short-lived as well: the Baath Party of Iraq organized the
so-called Ramadan Revolution in February 1963, seizing power with the help of the armed
forces, ultimately executing Qasim on February 9. On November 10, 1963, “a group of army
officers overthrew the new regime. Abdul Salam Arif took over as president. He monopolized
power and relied on a few trusted individuals. General Ahmed Hassan al-Bakr, a Baathist,
became prime minister.”
114
In 1966, Ariff died on a plane crash, and was succeeded by his
brother, Abdul Rahman Arif.

112
Brands and Palkki (2011).
113
Braut-Hegghammer (2016, 27).
114
Ibid., 34.
79

In 1968, another coup took place, when Hassan al-Bakr and his deputy, Saddam Hussein,
together with the Baath party and the armed forces, forcibly took over the government, an event
known as the Baath Coup of 1968. Saddam Hussein used his time as vice-president of Iraq to
accumulate power and build domestic alliances, increasing his influence over Iraqi politics.
Finally, Saddam Hussein forced a sick al-Bakr to resign on July 16, 1979, becoming president of
Iraq until his forced departure from power in 2003 following the U.S. invasion of Iraq.
115
As the
historical record shows, Iraq was ruled by a monarchy and then military regimes throughout the
entire period of analysis, placing Iraq solidly in the category of undemocratic regimes throughout
the 1956-1998 period.
There is also strong evidence to suggest that Iraq’s leadership, particularly Saddam
Hussein, perceived a threatening security environment, particularly in the late 70s. Iraq faced a
threat from Iran following the Iranian Revolution of 1979. Saddam Hussein feared that his
secular rule would be threatened if the Shia majority decided to mobilize against him, as his
power base consisted mainly of Sunni tribal groups, who were the minority in Iraq.
116
These
fears were not unfounded; the Islamic Republic of Iran “supported resistance by both the Shiite
majority and the Kurdish population in the north against the secular Baathist state in
Baghdad.”
117
Saddam Hussein launched a full scale invasion of Iran on September 22, 1980 to
prevent the Iranian leadership from exporting its ideology into Iraq and fomenting internal
dissent and revolution. The conflict lasted eight years, ending in August 1988 following a
ceasefire arranged by the U.N.
118

115
Ibid., 62.
116
Ekeus (2016, 132).
117
Debs and Monteiro (2016, 134).
118
Ibid.,134.
80

Perhaps more important than Iran is the threat Iraq perceived from its neighbor, Israel.
Saddam Hussein perceived Israel as a threat to Arab nations and saw Israel’s victories against
Arab nations in the 1948 Israel war of independence and the 1967 Six Day War as unacceptable.
Obtaining nuclear weapons, in Saddam’s view, would allow Iraq to correct these “historical
wrongs” and “enable a ‘patient war’-a war of attrition-that would reclaim Arab lands lost in the
Six-Day War of 1967.”
119
Saddam Hussein explicitly declared his intentions for the bomb in
1981: “We have to have this protection for the Iraqi citizen so that he will not be disappointed
and held hostage by the scientific advancement taking place in Iran or in the Zionist
entity…Without such a deterrence, the Arab nation will continue to be threatened by the Zionist
entity and Iraq will remain threatened by the Zionist entity.”
120
Clearly, obtaining nuclear
weapons was essential as a deterrence tool.
Evidence suggests, however, that deterrence was not the only reason for obtaining a
nuclear bomb. Obtaining a nuclear weapon might have been of strategic importance to allow Iraq
to fight a conventional war against Israel without fearing Israeli nuclear strikes. In 1978, at a
speech given at Bakr University, Saddam Hussein made the following comment:
When the Arabs start the deployment, Israel is going to say, ‘we will hit you with
the atomic bomb.’ So should the Arabs stop or not? If they do not have the atom, they
will stop. For that reason they should have the atom. If we were to have the atom, we
would make the conventional armies fight without using the atom. If the international
conditions were not prepared and they said, ‘We will hit you with the atom,’ we would
say, ‘we will hit you with the atom too. The Arab atom will finish you off, but the Israeli
atom will not end the Arabs.
121

119
Brands and Palkki (2011, 133).
120
Quoted in Brands and Palkki (2011, 148).
121
Quoted in Brands and Palkki (2011, 150).
81

Obtaining a nuclear strike capability, under this view, would allow Iraq to force a
conventional war with Israel without fear of nuclear retaliation, following the logic that owning a
nuclear weapon would prevent the enemy from using nuclear weapons during war.
Given Iraq’s autocratic regime type during the 1956-1998 period, Iraq was unlikely to
accept international norms advanced by the emerging liberal international order, particularly
when that order is led by the United States, a major Israel ally. Furthermore, Iraq’s security
environment was one of high threat, particularly from rivals such as Iran and Israel. Under these
conditions, internalizing or even accepting international anti-nuclear proliferation norms was
highly unlikely for Iraq. If so, a voluntary reversal due to acceptance of international norms was
a remote possibility for Iraq, which would have benefitted greatly from owning a nuclear weapon
capability.
3) Testing observable implications
As should be evident from the historical record and examination of Iraq’s regime type
and security environment context, Iraq falls within the category of cases least likely to internalize
or accept international norms against nuclear weapon proliferation and ownership. Therefore, the
theoretical expectation is that Iraq will follow the involuntary pathway of nuclear reversals:
termination of its nuclear weapons pursuit is the result of coercion and force from external
actors.
Importantly, international norms against nuclear weapon proliferation and ownership
should play a role in the reversal story. There should be evidence that Iraq explicitly rejects such
norms, and there should be evidence that external actors tried to pressure Iraq to adhere to these
international norms before resorting to the use of force. In sum, if the involuntary pathway of
82

nuclear reversals is correct, then the case of Iraq should pass the following hoop tests, which test
observable implications of the theory:
1) Iraq should reject norms against nuclear weapon proliferation and ownership. We should
observe evidence of such rejection in public statements made by Iraq’s leadership in
official public state reports, public speeches, or in international forums like the UNGA.
Importantly, such rejection should mention that the nuclear norms are an imposition by
the liberal West.
2) It should be observed that other states in the system are trying to push Iraq towards
termination of their nuclear program. Anti-nuclear weapon proliferation norms should be
mentioned as part of the arguments advanced by states, perhaps in international forums
like the UNGA. These statements are hypothesized to be more frequent for liberal
democracies than autocracies.
3) Other states trying to push Iraq towards termination of their program should rank-order
their diplomatic tools. We should observe that threats of sanctions and imposition of
sanctions take priority over threats of force. In addition, threats of force should take
priority over actual use of force. We should observe that diplomatic “carrots” are used
less frequently than diplomatic “sticks” against Iraq.
4) We should observe that Iraq does not terminate their program voluntarily when
perceiving high levels of threat.
5) Iraq’s termination of their nuclear weapons program must be shown to be the result of
external coercion rather than a voluntary decision. Iraq’s termination of their program
should follow from acts of coercion by other actors, such as threats of force, sanctions, or
83

use of force rather than diplomatic carrots such as security guarantees and lifting of
sanctions.
It is important to note that the five hoop tests testing observable implications above do
not preclude the existence of additional observable implications that may be drawn from the
theoretical framework advanced in this project. The five observable implications tested in this
section should be understood as the most obvious implications from the involuntary reversal
pathway. The observable implications will be tested via hoop tests, which propose that specific
evidence must be found to support a theorized observable implication. If evidence is not found to
support an observable implication, then the hoop test fails and might indicate that the theory
requires revision of some theoretical assumptions. However, passing a hoop test does not
automatically confirm the hypothesis, it only supports it.
122
Failure to support most of these
observable implications would greatly weaken the theoretical framework of the involuntary
nuclear reversal pathway, as the case of Iraq was identified as a most-likely case in which the
theory should apply. However, successfully passing most or all the hoop tests does not imply that
the theory is definitely confirmed; the same tests in similar cases should be done to observe if the
theoretical framework applies to cases other than Iraq. In the pages that follow, each hoop test
will be tested in sequential order.
Observable implication 1: Iraq should reject norms against nuclear weapon proliferation
and ownership. We should observe evidence of such rejection in public statements made by
Iraq’s leadership in official public state reports, public speeches, or in international forums like

122
See Chapter 3: Research Design in this project for a more elaborate discussion of hoop tests, also see Mahoney
(2012).
84

the UNGA. Importantly, such rejection should mention that the nuclear norms are an imposition
by the liberal West.
Surprisingly, the available evidence does not offer support for the observable implication
that Iraq would explicitly reject norms against nuclear weapon proliferation and ownership and
claim that these norms are part of a Western liberal order. In fact, Iraq’s rhetoric in international
forums like the United Nations General Assembly demonstrate that Iraq pretended to embrace
these anti-proliferation norms, using its status as an NPT member as leverage against Israel. For
example, following the Israeli strike against the Osirak reactor on June 7, 1981, Iraq co-
sponsored UNGA Resolution 36/27, which was adopted on November 13, 1981. UNGA
Resolution 36/27 stated that the attack on Iraq’s reactor “constituted an attack against the Agency
[IAEA] and its safeguards regime,” and “strongly condemns Israel for its premeditated and
unprecedented act of aggression in violation of the Charter of the United Nations and the norms
of international conduct.”
123

UNGA Resolution 36/27 was discussed as Agenda Item 130 in the UNGA thirty-sixth
session, 56
th
plenary meeting on November 13, 1981. In the meeting, Iraq’s representative Al-
Qaysi explained the different paragraphs of the resolution, formulating a contrast between Iraq
and Israel regarding adherence to international non-proliferation norms:
The sixth and seventh preambular paragraphs register another set of incontestable
facts…Iraq’s adherence to the Non-Proliferation Treaty, its having subscribed to the
safeguards regime of IAEA and the testimony of that Agency regarding the satisfactory

123
General Assembly Resolution 36/27, Armed Israeli aggression against the Iraqi nuclear installations and its
grave consequences for the established international system concerning the peaceful uses of nuclear energy, the
non-proliferation of nuclear weapons and international peace and security, A/RES/36/27 (13 November 1981),
https://digitallibrary.un.org/record/27621?ln=en
85

application by Iraq of those safeguards; and, in contrast, Israel’s refusal to adhere to the
Treaty and to submit its nuclear facilities to Agency safeguards.
124

Rather than publicly reject international non-proliferation norms, it appears that Iraq
leveraged its membership as NPT signatory and acceding to IAEA inspections to portray itself as
a “good member” of the international community in relation to Israel.
However, even though Iraq did not explicitly reject international non-proliferation norms
in the United Nations General Assembly, its behavior and rhetoric outside international forums
suggests that Iraq had little concern for these norms in practice. A strong rejection of non-
proliferation norms was observed on June 23, 1981, when it was reported in the New York
Times that Saddam Hussein “called on ‘all peace-loving nations of the world’ today to help
Arabs acquire nuclear weapons to balance what he described as Israel’s nuclear capacity.”
125
In
Hussein’s view, nuclear weapons in Arab hands were essential “for world peace and security.”
Nuclear weapons were important so Iraq could protect Iraqi citizens “so that [they] will not be
disappointed and held hostage by the scientific advancement taking place in Iran or in the Zionist
entity.”
126
Iraq’s justification for pursuing the nuclear bomb was framed as a necessity against
regional rivals and improve security, without mention of international norms against
proliferation. Evidently, there is a contradiction between Iraq’s speech at international forums
such as the United Nations vs Saddam Hussein’s statements regarding the necessity for the Arab
bomb in speeches directed at domestic or regional audiences. In the United Nations General
Assembly, Iraq highlights its adherence to the NPT and IAEA inspections regime, while calling
for the pursuit of the nuclear bomb in other forums.

124
U.N. GAOR, 36
th
Sess., 56
th
plen. mtg., U.N. Doc. A/36/PV.56 (13 November 1981),
https://digitallibrary.un.org/record/761434?ln=en
125
The Associated Press. “Iraq Asserts Arabs Must Acquire Atom Arms As A Balance To Israel. The New York
times. June 24, 1981.
126
Brands and Palkki (2011, 148).
86

There is also evident contradiction between Iraq’s rhetoric about adhering to the NPT and
its observable actions during the period of analysis. Despite being a member of the NPT, Iraq’s
nuclear establishment researched technology that could play a dual role in both nuclear energy
and nuclear weapons development during the 1970s, which Israel viewed with suspicion,
ultimately leading to the attack on Osirak.
127
Following the strike, Iraq initiated a covert nuclear
weapons program, which it tried to hide from IAEA inspectors multiple times throughout the 80s
and early 90s. As highlighted in the historical section, in the 90s Iraq challenged the UNSC by
interfering and obstruction UNSCOM inspectors, preventing them from pursuing vehicles
carrying sensitive nuclear material in June 1991, blocking inspectors from leaving an inspected
facility with sensitive information in September 1991, and refusing UNSCOM inspectors from
entering sites in 1997 and 1998.
128

In sum, the first hoop test fails to find evidence for the theorized implication that Iraq
would explicitly reject international norms against nuclear non-proliferation in forums such as
the United Nations General Assembly. At the same time, evidence shows that Iraq did not hide
its nuclear ambitions, which were framed in terms of national security against Israel and Iran, its
major regional rivals. Although Iraq did not specify rejection of international norms against
nuclear proliferation, even at times using its adherence to the IAEA and NPT membership to its
advantage, Iraq’s covert actions indicated a disregard and lack of internalization and acceptance
of these international norms. It appears that even though Iraq’s public speech did not reject these
international norms as an imposition of the liberal West, in practice it disregarded said norms by

127
Feldman (1982, 115).
128
Ekeus (2016, 144).
87

pursuing nuclear technology that could be redirected towards the construction of a nuclear bomb
in the 70s, and the covert existence of a nuclear weapons program in the 80s and 90s.
Observable implication 2: It should be observed that other states in the system are trying
to push Iraq towards termination of their nuclear program. Anti-nuclear weapon proliferation
norms should be mentioned as part of the arguments advanced by states, perhaps in
international forums like the UNGA. These statements are hypothesized to be more frequent for
liberal democracies than autocracies.
In the 1970s, despite the U.S. keeping a close eye on Iraq’s nuclear program, only Israel
had strong suspicions that the program was intended for nuclear weapons, as they detected Iraq’s
nuclear weapons ambitions in 1974 and made slowing down the program a national security goal
in 1975.
129
Therefore, Iraq’s nuclear program was not a major topic of discussion in the UNGA
prior to the 1981 Israeli strike on Osirak. However, normative arguments about Iraq’s violations
of nuclear non-proliferation norms and proliferation were advanced by Israel in the UNGA
thirty-sixth session, 56
th
plenary meeting of November 13, 1981, when Israel defended its
decision to strike the Iraqi reactor.
During the meeting’s discussion of Resolution 36/27, which condemned the Israeli strike,
Israel chastised the international community for not supporting its “right to self-defense,”
claiming that Iraq should be condemned instead. According to Israel, the draft resolution “totally
ignores Iraq’s attempts to bypass purposefully and methodically its obligations under the Non-
Proliferation Treaty and the safeguards agreement of IAEA and that, at the end of the day, Israel
was left with no alternative but to destroy Osirak in the legitimate exercise of its right of self-

129
Sadot (2016, 653).
88

defense.”
130
Interestingly, although Israel was not a member of the NPT, it advanced the
argument that Iraq was not adhering to its non-proliferation obligations under the NPT nor the
safeguards of the IAEA-institutions created to promote safe use of nuclear technology and,
importantly, prevent nuclear weapons proliferation. Although Israel did not utilize anti-nuclear
weapons norms to deter Iraq prior to the strike, it did rely on these norms to justify its actions
and condemn the Iraqi program.
Following the strike on Osirak, Saddam Hussein moved the Iraqi nuclear weapons
program underground, where it quickly advanced undetected by the international community.
131

Therefore, no United Nations General Assembly meetings discussed Iraq’s nuclear program in
terms of violations of the NPT or the IAEA, although the 1981 strike on Iraq was the subject of
much UNGA discussion in the years following the attack.
132
In fact, before the 1991 Gulf War, it
appears that the international community underestimated Iraq’s nuclear weapons ambitions and
was in the dark about the state of development the program underwent throughout the 80s.
According to researcher Braut-Hegghammer’s analysis of Iraq’s nuclear program, “before the
Gulf War, British intelligence assessed that Iraq was at least a decade away from the nuclear
weapons threshold. In 1989, the CIA assessed that Iraqi the [sic] program could succeed in the
mid-1990s, possibly even before.”
133
However, once Iraq invaded Kuwait, the international
community largely supported Operation Desert Storm, which revealed Iraq’s nuclear capabilities
and the extent of the nuclear program.

130
U.N. GAOR, 36
th
Sess., 56
th
plen. mtg., U.N. Doc. A/36/PV.56 (13 November 1981),
https://digitallibrary.un.org/record/761434?ln=en
131
Sadot (2016, 655).
132
UNGA Resolutions A/RES/37/18 (1982), A/RES/38/9 (1983), A/RES/39/14 (1984), A/RES/40/6 (1985), and
A/RES/41/12 (1986) were worded the exact same way as A/RES/36/27 (1981), repeatedly condemning the Israeli
strike on Iraq’s nuclear reactor.
133
Braut-Hegghammer (2016, 122).
89

The 1990s marked a significant shift in the international community’s response to Iraq’s
nuclear weapons program. Once the program was detected, the United Nations Security Council
passed various Resolutions condemning the program, several highlighting that Iraq was in
violation of the NPT and IAEA safeguards and endangering the nuclear non-proliferation
regime. These Resolutions, via condemnation and sanctions, tried to push Iraq towards nuclear
reversal, relying on international nuclear non-proliferation norms as part of their arguments to
justify the inspections and monitoring regime imposed on Iraq.
For example, the April 3, 1991 UNSC Resolution 687 demanded that Iraq, in addition to
assuring it will recognize and respect the sovereignty of Kuwait, must accept destruction of all
biological and chemical weapons, and all ballistic missiles with a range greater than 150
kilometers. Crucially, the resolution highlighted that the UNSC was concerned by reports that
Iraq tried to obtain materials for a nuclear weapons program “contrary to its obligations under
the Treaty on the Non-Proliferation of Nuclear Weapons of 1 July 1968,” recalling the “objective
of the establishment of a nuclear weapon free zone in the region of the Middle East.” The
Resolution concluded that Iraq “shall unconditionally agree not to acquire or develop nuclear
weapons or nuclear-weapon-usable material or any subsystems or components or any research,
development, support or manufacturing facilities related to the above,” and requested the
Director General of the International Atomic Energy Agency “to carry out immediate on-site
inspection of Iraq’s nuclear capabilities based on Iraq’s declarations and the designation of any
additional locations by the Special Commission” and develop a plan for the “destruction,
removal or rendering harmless” of all nuclear weapon usable materials.
134

134
United Nation Security Council Resolution 678, Security Council resolution 687 (1991) [on restoration of the
sovereignty, independence and territorial integrity of Kuwait], S/RES/687 (3 April 1991),
https://digitallibrary.un.org/record/110709?ln=en
90

A few months later, the UNSC passed Resolution 707, condemning Iraq for violating
UNSCR 687. The resolution accused Iraq with non-compliance with its IAEA and NPT
obligations. The Resolution demanded full disclosure on all aspects of Iraq’s programs to
develop weapons of mass destruction, including nuclear programs. Furthermore, it demanded
that Iraq allow UN and IAEA inspectors unrestricted access to any suspected facility, demanded
that Iraq does not conceal, move, or destroy any material related to its nuclear, chemical, and
biological programs, and demanded immediate halt to any nuclear activities not protected by the
NPT.
135
In October, The Security Council once again reiterated the need for Iraq to cooperate
with UN and IAEA inspectors.
136

From 1996 to 1998, the United Nations Security Council periodically passed Resolutions
condemning Iraq’s failure to fully adhere to IAEA and UN inspections. UNSCR 1051 (1996),
UNSCR 1060 (1996), UNSCR 1115 (1997), UNSCR 1134 (1997), and UNSCR 1137 (1997)
repeatedly condemned Iraq’s refusal to allow IAEA and UN inspectors to access sensitive sites,
demanding full cooperation with the United Nations and IAEA agencies.
137
Most strongly of all,
UNSCR 1154 (1998) demanded Iraq to adhere to its obligations and allow the IAEA and the UN
Special Commission unrestricted access, but that “any violation would have severest
consequences for Iraq.”
138
Despite repeated warnings, Iraq continued obstruction of international

135
United Nations Security Council Resolution 707, Security Council resolution 707 (1991) [on Iraqi violation of
Security Council resolution 687 (1991) with regard to inspection of its biological, chemical and nuclear weapons
capabilities], S/RES/707 (15 August 1991), https://digitallibrary.un.org/record/125533?ln=en
136
United Nations Security Council Resolution 715, Security Council resolution 715 (1991) [on plans for
monitoring and verification of Iraq's compliance with relevant parts of section C of Security Council resolution 687
(1991)], S/RES/715 (15 October 1991), https://digitallibrary.un.org/record/128235?ln=en
137
National Archives and Records Administration. “Saddam Hussein's Defiance of United Nations Resolutions,”
The White House. Accessed July 8, 2022, https://georgewbush-
whitehouse.archives.gov/infocus/iraq/decade/sect2.html
138
United Nations Security Council Resolution 1154, Security Council resolution 1154 (1998) [on endorsement of
the memorandum of understanding of 23 Feb. 1998 regarding compliance by Iraq with its obligations under
resolution 687 (1991) and other relevant resolutions], S/RES/1154 (2 March 1998),
https://digitallibrary.un.org/record/250774?ln=en
91

inspectors, which served as the basis for Operation Desert Fox, when the U.S. bombed sites
suspected to be used for development or storage of weapons of mass destructions.
From the looks of it, the available evidence appears to support the observable
implications tested by the second hoop test. Although only Israel acted when it detected Iraqi
nuclear ambitions in the 70s, the argument Israel used for justifying its attack on Osirak was
based on Iraq’s failure to adhere to international non-proliferation norms and its obligations
under international law. Similarly, once the international community discovered the extent of the
Iraqi nuclear weapons program in the 90s, the United Nations Security Council frequently
engaged in condemnations of the Iraqi program and requested Iraq to adhere to international law.
Although majority of these condemnations are observed in the Security Council rather than the
General Assembly as hypothesized, the evidence does support the theoretical expectation that
nuclear non-proliferation norms would play a role in condemnations of Iraq’s program, namely
failure to adhere to the NPT, restriction of access to IAEA inspectors, and the dangers
proliferation pose for the goal of a non-nuclear Middle East.
Observable implication 3: Other states trying to push Iraq towards termination of their
program should rank-order their diplomatic tools. We should observe that threats of sanctions
and imposition of sanctions take priority over threats of force. In addition, threats of force
should take priority over actual use of force. We should observe that diplomatic “carrots” are
used less frequently than diplomatic “sticks” against Iraq.
Although not explicit policy, it seems nation-states rank-ordered their policy options
when trying to deter Iraq from developing a nuclear weapons capacity at various points in time.
Specifically, we observe that Israel relied on diplomatic pressures before engaging in sabotage of
the Iraqi nuclear program, culminating with the 1981 Osirak strike. Furthermore, following the
92

Gulf War, the United Nations Security Council employed sanctions and threats to coerce Iraq
into allowing IAEA and UN inspectors to access suspected weapons facilities. Ultimately, Iraq’s
failure to cooperate and repeated violations of UNSC resolutions led to U.S. bombings of
suspected Iraqi cites.
Sadot and Lobell’s studies trace the diplomatic strategies used by Israel to slow down and
deter Iraq from reaching a nuclear weapons capability.
139
They reveal that Israel’s initial strategy
to deter Iraq was via diplomatic pressure towards countries supplying the Iraqi program. Israel’s
prime minister Yitzhak Rabin tried to communicate his concerns with the Iraqi nuclear program
from the moment the program was detected in 1974. For example, in 1976 Rabin raised his
concerns about the nuclear program in Iraq with Henry Kissinger, French leaders, and European
states that were collaborating with Arab states.
140
These concerns were raised more frequently
under Rabin’s successor, prime minister Menachem Begin, who tried to convince France to
replace the Highly Enriched Uranium it was going to send to Iraq with Caramel Fuel, which is
less efficient for a nuclear weapons program. French leaders “tried to assure Israel’s officials that
if uranium fuel was diverted by Bahdad, Paris would then shut the door on future purchases of
fuel, equipment, and training.”
141
Begin also tried to urge U.S. president Carter to pressure Italy
and France against sales of nuclear technology and equipment.
142
These diplomatic attempts
failed. France and Italy fulfilled the nuclear deals they established with Iraq, and the United
States did not prioritize convincing European allies to stop nuclear collaboration with Iraq,
presumably because the U.S. did not see any threats from European-Iraqi nuclear cooperation.
143

139
Sadot (2016); Lobell (2020).
140
Sadot (2016, 656).
141
Lobell (2020, 12).
142
Ibid., 12.
143
Sadot (2016, 656).
93

From 1979 onwards, Israel escalated its strategy to slow down the Iraqi program via
clandestine operations, such as threats to scientists, sabotage attempts, and assassinations. In
1979, infiltrators, suspected to be connected to Mossad, entered the port town of Seyne-sur-Mer
and “attached plastic explosives onto outbound containers holding two reactor core structures
about to be shipped to Iraq.”
144
In addition, French scientists destined for Iraq to help with
construction of the nuclear reactors received threatening letters, just like their Italian
counterparts. Another incident occurred on June 14, 1980, when Egyptian scientist Yahya el-
Mashad “was found murdered in the Paris Meridian Hotel.”
145
Mashad was a member of the
Iraqi scientific delegation visiting France to negotiate the supply of HEU shipments. France’s
intelligence services ultimately concluded that there was enough circumstantial evidence to
connect the assassination to Israel, but the French foreign minister did not react to the attack
because “we could not be sure” with full certainty about the origin of the attack.
146
Another
incident was reported on June 17, 1981 by a Brazilian newspaper, which claimed that a secret
shipment of uranium “was intercepted and hijacked on its way from Sao Paolo to Baghdad by
American-made planes.”
147
Although not all of these events are proven beyond a reasonable
doubt, there is circumstantial evidence and good reason to assume Israel was behind these
incidents.
Ultimately, Israel escalated its tactics for deterring Iraq’s nuclear program by using
military force against the Iraqi reactors. This action was taken after an Israeli committee
determined that Israel had “exhausted all ‘ideas and options’ for how to clandestinely delay

144
Ibid., 657.
145
Ibid., 658.
146
Ibid., 658.
147
Ibid., 659.
94

Osirak’s pending activation, and recommended ‘alternative strategies’ be used to do so.”
148

Although the Israeli strike destroyed the reactor, the first attempt towards destruction was done
by Iran in September 1980. At the start of the Iran-Iraq war, Iran sent two F-4 bombers to strike
Osirak, which failed to destroy the Iraqi facilities.
149
In the end, it is clear that Israel seems to
have rank-ordered its diplomatic tools for delaying and stopping the Iraqi nuclear program: first
relying on diplomacy to convince European allies to stop collaboration with Iraq, escalating to
clandestine activities such as threats, assassinations and sabotage, and finally culminating with
military force to destroy the Osirak reactor. These actions forced Iraq to move its nuclear
weapons program underground for the next decade.
Rank-ordering of policy options can also be observed in the concerted effort by the
international community to dismantle the Iraqi nuclear program after the Gulf War. Following
the adoption of UNSC Resolution 687, which established the terms of ceasefire between Iraq and
the U.S.-led coalition, the United Nations established the UN Special Commission (UNSCOM).
The UNSCOM task was to “implement the non-nuclear provisions of the resolution and to assist
the International Atomic Energy Agency (IAEA) in the nuclear areas.”
150
The UNSCOM
oversaw on-site inspections of Iraq’s biological and chemical weapons related capabilities, while
the IAEA was allowed to establish an Action Team to lead a similar effort on the nuclear
arena.
151
The UNSC inspection efforts relied on a combination of diplomatic carrots and sticks to
incentivize Iraqi compliance. If Iraq agreed to cooperate with the UNSC, then economic
sanctions imposed during the Gulf War would be lifted alongside the oil embargo, which was the
first major sanction to be lifted when Iraq accepted compliance with the UNSC inspections

148
Ibid., 661.
149
Debs and Monteiro (2016, 140).
150
Ekeus (2016, 134).
151
Ibid., 134.
95

regime. However, despite the lifting of some sanctions, Iraq consistently refused to cooperate
with on-site inspectors.
As previously described, Iraq employed obstruction and concealment tactics to prevent
inspectors from entering facilities or obtaining sensitive information on the Iraqi nuclear
program. Such obstruction led to reactions by the UNSC. For instance, on August 15, 1991,
Iraq’s failure to cooperate with UNSCOM and the IAEA led to the UNSC adopting Resolution
707, which was the first of several resolutions condemning Iraqi noncooperation. The effect of
these condemning resolutions was to prolong the economic sanctions placed on Iraq at the
beginning of the Gulf War; as long as Iraq failed to fully cooperate, the UNSC and the
international community were willing to continue the sanctions regime that was harming Iraq’s
economy. The sanctions regime was so severe, that by some accounts the sanctions “destroyed
the civilian as well as the military economy, industry and infrastructure of Iraq.”
152
A 1999
UNICEF survey of child mortality even projected that “a total of 500,000 infant deaths…could
be attributed to the sanctions.”
153

December 16, 1998, marked an explicit escalation in the U.S. attempt to dismantle the Iraqi
nuclear program. Seemingly tired of Iraqi non-compliance and the belief that the UNSC
sanctions regime was a failure, the United States and the United Kingdom initiated Operation
Desert Fox, a four-day bombing campaign of Iraq without consent from the United Nations
Security Council. President Clinton justified the attack in a speech, claiming that the attacks were
necessary due to “’stark, sobering, and profoundly disturbing’ conclusions of the Butler report,
according to which, Clinton claimed, ‘no progress was able to be made in the fields of

152
Harrer (2014, 3).
153
Ibid., 3.
96

disarmament.’”
154
The Butler report, published December 14, stated that the UNSCOM was
unable to conduct “substantive disarmament work” in Iraq due to Iraqi obstruction.
155
The four-
day operation consisted of 400 cruise missiles launched against sites suspected of housing
chemical, biological, and nuclear weapons material or laboratories.
The evidence thus far presented appears to support the rank-ordering of diplomatic tools
used by states and the international community to dismantle the Iraqi nuclear program. With
Israel, it is observed that diplomatic pressures towards France and Italy took priority before
escalating to clandestine activities to deter and slow down the Iraqi nuclear program, culminating
with the Osirak strike in 1981. The international community, represented by the UNSC, also
rank-ordered their options to dismantle the Iraqi program in the 90s. Rather than just rely on
“sticks” as hypothesized, the UNSC used a combination of diplomatic incentives and penalties
via UNSC Resolutions and IAEA/UNSCOM inspections. Actual use of force was avoided by the
Security Council, which did not authorize such an attack on Iraq. Unfortunately for Saddam, the
United States and United Kingdom, seemingly discontent with the success of the inspections and
sanctions regime, acted unilaterally to destroy the remaining nuclear-suspected sites of Iraq.
Given these findings, the observable implications expected by hoop test 3 are supported.
Observable implication 4: We should observe that Iraq does not terminate their program
voluntarily when perceiving high levels of threat.
The observable implication tested by hoop test 4 is supported by the available historical
records; Iraq did not voluntarily terminate its nuclear program during periods of high perception
of threat. For Iraq, Israel and Iran were principal security threats. For example, Israel ended

154
Ibid., 240.
155
Ibid., 239.
97

victorious in its war of independence against Arab states in 1948, the Six-Day War (1967), and
the Yom Kippur War (1973), all conflicts in which Iraq was a belligerent party. Iraq also faced a
security threat from Iran, which supported Kurdish rebels in Iraqi territory and allied with Israel
in 1965 to support these rebels.
156
The most intense moment in the Iraqi Iranian rivalry began
due to Saddam Hussein’s initiation of armed conflict with Iran. From 1980 to 1988, Iraq and Iran
fought a prolonged deadly war, which began when Iraq tried to take advantage of the fragile
internal domestic situation in Iran after the 1979 Iranian revolution.
Rather than postpone research and development of nuclear and other unconventional
weapons during the war, Iraq engaged in new exploratory and development efforts of
unconventional technology and weaponry. According to official Iraqi declarations submitted to
the UN in 2002, it appears that there were no financial restrictions on the nuclear program from
1981-1985. According to the 2002 declarations, “in 1982, the estimated budget allocation was
918,658 Iraqi dinars (ID), while the final allocation…was 948,134 ID…In 1985 2,566,470 ID
were estimated while a staggering 9,784,558 ID were spent.”
157
Nuclear scientists were given a
directive from Saddam Hussein during these war years: “avoid sensitive foreign assistance that
could alert the outside world to the nuclear weapons program.”
158
Thus, Iraqi scientists were
forced to develop an indigenous route towards nuclear weapons development to hide their covert
program from foreign threats. Unfortunately for Iraq, development of the nuclear option was
slowed down due to lack of foreign assistance, and insufficient monitoring and managing of the
nuclear weapons program.
159

156
Debs and Monteiro (2016, 133).
157
Braut-Hegghammer (2016, 80).
158
Ibid., 80.
159
Ibid., 102.
98

Similarly, the 1990 Iraqi invasion of Kuwait did not lead to slow down or termination of
the nuclear program. Rather, it incentivized nuclear scientists to develop a “crash program” to
speed up the development of a nuclear war head. The program consisted of four steps, “extract
reactor fuel; conversion of HEU into uranium metal; further enrichment of the extracted uranium
in centrifuges; and weaponization.”
160
The program was short lived—on January 16, a U.S. led
coalition initiated Operation Desert Storm, bombing several suspected sites, including sites of the
nuclear research program, putting an early end to the crash program. As can be seen, Iraq
doubled down on its efforts to obtain the bomb during moments of heightened tension and
conflict, rather than postpone development of their program. Therefore, the observable
implication that Iraq would not terminate its program during periods of high levels of threat is
supported.
Observable implication 5: Iraq’s termination of their nuclear weapons program must be
shown to be the result of external coercion rather than a voluntary decision. Iraq’s termination
of their program should follow from acts of coercion by other actors, such as threats of force,
sanctions, or use of force rather than diplomatic carrots such as security guarantees and lifting
of sanctions.
The evidence presented supports the expectation that Iraq’s termination of their nuclear
weapons program was the result of coercion and force rather than a voluntary decision. The
termination of the Iraqi nuclear program follows from a prolonged campaign in the 90s in which
the international community, represented by the United Nations Security Council, engaged in a

160
Ibid., 118.
99

sanctions and monitoring regime to keep a watchful eye on Iraq’s nuclear capabilities after the
disastrous Gulf War.
Even as late as 1991, Iraq attempted to obtain a nuclear weapon, evidenced by the fact
that Saddam Hussein approved a “crash program” in August 1990 to rapidly construct a nuclear
weapon by extracting weapons grade material from safeguarded research reactor fuel. According
to a 1997 IAEA report, the crash program, if undetected, would have placed Iraq close to the
nuclear weapons threshold within two years.
161
Yet, this program ended in failure: Operation
Desert Storm led to the destruction of the crash program and was followed by several UNSC
Resolutions condemning Iraq’s invasion of Kuwait, establishing a severe sanctions regime on
Iraq.
UNSC Resolution 687 began the coerced dismantlement of Iraq’s nuclear program,
which required elimination of the nuclear weapons program and tasked the newly created
UNSCOM to carry out inspections of biological, chemical, and missile capabilities.
162

Throughout the 1991-1998 inspections period, Iraq would remain under several international
sanctions until the UNSCOM and IAEA complete their dismantlement work.
163
As is evident,
the international community used sanctions, a “stick”, as their central policy to coerce Iraq into
giving up pursuit of the nuclear bomb. However, sanctions weren’t the only tool used to put an
end to the Iraqi nuclear program. Failure of Iraq to adhere to the sanctions regime led the United
States to use actual force against Iraq. On December 16, 1998, U.S. launched Operation Desert

161
IAEA, Fourth consolidated report of the Director General of the International Atomic Energy Agency under
paragraph 16 of Security Council resolution 1051 (1996), S/1997/779, 8 October 1997.
https://digitallibrary.un.org/record/244664?ln=en
162
Debs and Monteiro (2016, 138).
163
Ibid., 138.
100

Fox to destroy any suspected nuclear weapons research facilities in Iraq, which put a definitive
end to Iraq’s immediate nuclear weapons ambitions.
164

An important observation is that at no point during the sanctions period did the United
States offer security guarantees to Iraq as an inducement towards nuclear weapons abandonment.
Thus, the “carrot” of extending the nuclear umbrella to Iraq was never provided as an option.
Furthermore, lifting of most sanctions was dependent on Iraq’s full cooperation with the United
Nations sanctions regime, rather than pre-emptively offered as a sign of trust that Iraq would
reciprocate and voluntarily dismantle its program. This is not surprising, as Iraq covertly pursued
nuclear weapons in the 80s after the Israeli Osirak strike, in violation of its NPT responsibilities
to avoid pursuit of the nuclear bomb. As a result, it was unlikely the United Nations Security
Council and the rest of the international community would trust Iraq to voluntarily terminate its
nuclear weapons program.
4) Conclusion
Five observable implications drawn from the involuntary reversal pathway were tested
via hoop tests, using the case of Iraq as a representative case study of the theorized nuclear
reversal pathway. As seen on Table 8 at the end of the chapter, evidence was found supporting
the observable implications tested in hoop tests 2-5, which leads to greater confidence that the
involuntary nuclear reversals pathway is capturing important dynamics of nuclear reversal.
However, the failure of passing one hoop test suggests that some assumptions of the theory may
need revision to better capture empirical reality. In what follows, I summarize the hoop test
results, and propose a possible revision to the theoretical framework for future work.

164
Ibid., 139.
101

Hoop-test 2 was a success. As predicted, other states in the system did try to push Iraq
towards termination of their nuclear program, primarily Israel in the 70s and the rest of the
international community in the 90s. Importantly, it is observed that anti-nuclear weapon
proliferation norms played a part in arguments advanced by states to deter Iraqi nuclear
proliferation. In the UNGA, Israel defended its strike on the Osirak reactor by accusing Iraq of
violating non-proliferation norms and its NPT responsibilities, clearly using the language of
nonproliferation norms to justify the use of force. Then, throughout the 90s the UNSC passed
several Resolutions condemning Iraq’s lack of cooperation with international inspectors. A key
aspect of these Resolutions is the desire to avoid nuclear proliferation in the Middle East and that
Iraq should adhere to the NPT-explicit support for anti-nuclear proliferation norms.
It must be pointed out that one aspect of the observable implication did not gain support
in the Iraq case: most arguments against Iraq’s nuclear weapons program were advanced in the
United Nations Security Council in the 90s rather than the General Assembly. This finding is an
indication that the observable implication drawn from the theory should be that arguments based
on norms are not exclusive to the UNGA, but may also be found in the Security Council and
other international bodies. Although this observation sounds obvious, it must be stated that the
observable implication advanced in this case study was far too restrictive by focusing solely on
UNGA debates, which almost leads to missing UNSC Resolutions.
Hoop test 3 also passed, for we see implicit rank-ordering of diplomatic tools by states
and the international community in their attempts to deter Iraq from obtaining the nuclear bomb.
Israel’s policy throughout the 70s went through several stages: First, Israel tried diplomatic
pressures, so Italy and France stop nuclear cooperation with Iraq. Then, Israel escalated to threats
and covert actions to slow down the nuclear program. Finally, Israel resorted to the use of force
102

to destroy the Osirak reactor. Similar rank-ordering can be observed at the international level
during the 90s. Following the Gulf War, the UNSC used sanctions and inspections to dismantle
Iraq’s nuclear program. When Iraq failed to cooperate, the UNSC resorted to vague threats of
“Serious consequences” if Iraq did not cooperate fully with IAEA and UNSCOM inspections.
Finally, acting unilaterally, the United States used armed force in Operation Desert Fox
following continuing Iraqi intransigence.
Hoop test 4 and 5 also gained support. We see throughout the period under analysis that
Iraq never backed down from pursuit of the bomb, moving the program underground after it was
first detected and attacked by Israel. In fact, as expected from observable implication four, Iraq’s
nuclear weapons program advanced greatly during periods of high levels of threat, such as the
Iran-Iraq war and Iraq’s crash program during the Gulf War. Finally, it is observed that it took
external coercion to push Iraq towards nuclear weapon program dismantlement. A combination
of sanctions, force, and threats culminated with the total collapse of the nuclear program by the
mid-90s.
Interestingly, the first hoop test ended in failure. Contrary to expectations, Iraqi leaders
did not explicitly reject anti-nuclear weapons proliferation norms in public statements, speeches,
or in international forums like the UNGA. On the contrary, Iraq’s rhetoric advanced the
argument that Iraq was always adhering to international nuclear nonproliferation norms, despite
actions to the contrary. Although Saddam Hussein and others did mention at various points their
desire to obtain nuclear weapons, these claims were never couched in terms of rejection of non-
proliferation norms. Therefore, although observable actions of Iraq may indicate that Iraq’s
rhetoric of adhering to non-proliferation norms does not match reality, the expectation was that
Iraq would make normative rejections explicit.
103

The results of the first hoop test suggest that a revision is in order. Contrary to what was
originally theorized, countries along the involuntary pathway of nuclear reversals might not
always explicitly reject these norms in their speech acts. Strategic reasons could explain this: it
might not be in the best interest of a state to explicitly reject nuclear nonproliferation norms,
even if a state plans to violate said norms and build a nuclear weapon. By explicitly rejecting
said norms, a state would preemptively reveal its true intentions to the international community
and invite high scrutiny from international nuclear inspectors, rivals, and non-proliferation
actors. Thus, it is in the best interest of states to conceal their true intentions and pretend to
adhere to nuclear non-proliferation norms in public for as long as possible, and reveal their true
intentions and capabilities once the nuclear weapon has been developed and secured.
This theoretical revision to the involuntary pathway at first appears to create wider
problems for the theoretical framework developed in this project and the advanced measure of
nonproliferation norm acceptance based on UNGA voting patterns. A central argument of the
theory is that identity as a norm-adhering state can be built over time by publicly supporting non-
proliferation norms, since constant support for these norms creates an expectation that a state
will adhere to said norms. Thus, if a state wants to be a respected member of the international
community, it will match its behavior to its public identity as a non-proliferator. This expectation
drives the belief that using UNGA voting patterns is an acceptable proxy measurement of how
much a state accepts anti-nuclear proliferation norms. Yet, if states like Iraq can lie about their
norm adherence, why should we accept the assumption that public speech and voting records are
a good indicator of international norm acceptance?
The solution to the problem is nuanced. As a social science, international relations cannot
completely predict human behavior due to the problem of agency. As a result, most theories
104

within international relations are probabilistic rather than deterministic, allowing for the
existence of outliers that might not fit the predicted theoretical expectations. This truth must be
applied to the theoretical expectations set above: most countries most of the time will adhere
their behavior to their public speech acts to appear as good members of the international order,
but we must accept that there will be outliers to this expectation. Some countries will conceal
their true intentions, even in UNGA meetings and voting sessions. Thus, cases like Iraq exist,
which outwardly appear to adhere to international norms while concealing their nuclear weapons
program and true intentions. This observation drives us to the conclusion that using UNGA
voting records is also an imperfect measure, for it will capture “false positives” of countries that
appear to internalize or accept anti-nuclear norms while covertly pursuing the bomb. Yet, it
might still be a more nuanced measure than the binary NPT membership measure used in past
studies.

105

Table 8: Summary of results: Iraq
Observable Implication Supported by evidence
Iraq should reject norms against nuclear
weapons proliferation and ownership. We
should observe evidence of such rejection in
public statements made by Iraq’s leadership
in official public state reports, public
speeches, or in international forums like the
UNGA. Importantly, such rejection should
mention that the nuclear norms are an
imposition by the liberal West.
No
It should be observed that other states in the
system are trying to push Iraq towards
termination of their nuclear program. Anti-
nuclear weapon proliferation norms should be
mentioned as part of the arguments advanced
by states, perhaps in international forums like
the UNGA. These statements are
hypothesized to be more frequent for liberal
democracies than autocracies.
Yes
Other states trying to push Iraq towards
termination of their program should rank-
order their diplomatic tools. We should
observe that threats of sanctions and
imposition of sanctions take priority over
threats of force. In addition, threats of force
should take priority over actual use of force.
We should observe that diplomatic “carrots”
are used less frequently than diplomatic
“sticks” against Iraq.
Yes
We should observe that Iraq does not
terminate their program voluntarily when
perceiving high levels of threat.
Yes
Iraq’s termination of their nuclear weapons
program must be shown to be the result of
external coercion rather than a voluntary
decision. Iraq’s termination of their program
should follow from acts of coercion by other
actors, such as threats of force, sanctions, or
use of force rather than diplomatic carrots
such as security guarantees and lifting of
sanctions.
Yes

106

Chapter 6: Brazil as A Case of Voluntary Reversal
In this chapter, I test the observable implications derived from the voluntary nuclear
reversal pathway, using the case of Brazil as a typical case of voluntary reversals. If the theory
successfully explains the nuclear reversal of Brazil, then the theory is likely to apply to similar
other cases of countries that terminated their nuclear weapons program or ambitions in the
absence of coercive threats or use of force. Like the case of Iraq, I use process tracing to
carefully analyze the historical record of Brazil’s nuclear program and see if the theoretical
causal mechanisms can be identified in this case. Once again, hoop tests are employed to test
whether there is evidence in support of observable implications that follow if the voluntary
reversal pathway is an accurate description of empirical reality. These observable implications
are listed as seven hypotheses: if the voluntary pathway is correct, then we should find evidence
in support of a predicted observation.
The chapter is organized as follows. First, I present the historical background of Brazil’s
nuclear program, delineating the important developments of Brazil’s civil and, later, covert
nuclear program. Next, I analyze Brazil’s regime type and perception of threat, particularly how
these two intervening variables might have influenced Brazil’s likelihood of accepting
international non-proliferation norms and related anti-nuclear weapons norms. The third section
introduces the seven hoop tests, specifying observable implications of the voluntary nuclear
reversals pathway that we should observe if the theory applies in the case of Brazil. Finally, the
conclusion summarizes the findings. In short, I find that contrary to common misperceptions, the
top leadership of Brazil did not make the final decision to obtain the nuclear bomb, although I
argue in the conclusion that Brazil still qualifies as a case of nuclear reversal. Second, the
hypothesized observable implications tested by the hoop tests are generally supported, which
107

increases confidence that the theory of voluntary reversals may explain some cases of nuclear
reversals.
1) Brazil’s nuclear program: 1951-1994
165

Admiral Álvaro Alberto (1889-1976) is generally considered the “father” of the Brazilian
nuclear program. In 1949, then president Eurico Dutra made Alberto a counter admiral, the
highest rank for a retired navy officer in Brazil, following Alberto’s advocacy in the United
Nations for a policy of specific compensations. According to the proposed policy of “specific
compensations,” Brazil would export raw materials, such as uranium, in exchange for transfer of
technology, training of scientists, or equipment that would be essential for developing the
nuclear sector in Brazil.
166
Although his advocacy at the U.N. ended in failure, Alberto would go
on to use his new found power and influence to promote nuclear energy in Brazil, which would
eventually lead to the formation of a nuclear program in 1951.
In 1947, Alberto sketched out a plan for nuclear development in Brazil and submitted it
to the Brazilian National Security Council (Conselho de Segurança Nacional, or CSN). Based on
the plan, in April 1949, a commission consisting of various scientists, led by Alberto, submitted
to the national congress a bill that would create a national research council specialized on nuclear
energy. Two years later, on January 15, 1951, Congress passed the bill, which was signed into
law by Dutra shortly before finishing his presidential term.
167
The law established the National
Research Council (CNPq), whose “primary objective was to coordinate the domestic
development of nuclear energy and to supervise the export of nuclear minerals.”
168
The new

165
The historical account described here is drawn from Debs and Monteiro (2016) and Patti (2021).
166
Patti (2021, 18).
167
Ibid., 24.
168
Ibid., 24.
108

president, Getulio Vargas (1951-1954) approved the nuclear project in mid-1951, fully in support
of the plan to develop nuclear energy independence in Brazil.
The early years of the nuclear program under Getulio Vargas involved a lengthy process
of trying to obtain nuclear technology, training, and equipment from diverse sources. Brazil tried
to obtain a processing plant from the United States but was denied. Brazil also tried to make a
deal with France on joint development of uranium mining and extraction technologies, and
“attempted to purchase heavy water reactors from Norway.”
169
However, the most crucial step in
Brazil’s early nuclear development was when on November 25, 1953, president Vargas approved
a concrete plan to acquire all the components necessary for the nuclear energy production
process.
170
Following the plan, Brazil reached an agreement with West Germany; West Germany
would send three centrifuges to Brazil and train Brazilian scientists, in exchange Brazil would
export Uranium. However, the Brazil-West Germany deal met opposition from the United States,
which perceived the deal as a security risk for the United States.
171

The United States managed to derail the Brazil-West Germany deal. On December 1953,
US President Dwight D. Eisenhower (1953-1961) announced the “Atoms for Peace” program,
which would allow the USA to share nuclear technology, training, and materials with other
countries for peaceful use. Vargas’ successor, Joao Café Filho (1954-1955), was pro-American
and found the USA alternative more enticing than West Germany. As a result, on August 3 1955,
Brazil and the United States reached an agreement to cooperate on the peaceful use of nuclear
energy. According to the deal, the U.S. would export a nuclear reactor to Brazil and would lend

169
Ibid., 90.
170
Patti (2021, 30).
171
Ibid., 34.
109

nuclear fuel, making Brazil the first country in Latin America to have a nuclear reactor.
172
The
reactor, IEA-R1, was inaugurated on January 25, 1958 by newly elected president Juscelino
Kubitschek (1956-1961).
During the 1955-1964 period prior to the military coup, the democratic leadership of
Brazil involved itself in international debates over the peaceful use of nuclear energy and nuclear
disarmament. For example, President Kubitschek created a plan in 1956 for nuclear
independence, the “Plano de Metas” (Goals Plan), which was aimed at “producing nuclear fuel,
training nuclear personnel, and producing radioisotopes for medical purposes.”
173
Brazil’s
advocacy for nuclear disarmament was made a central point of foreign policy from 1961-1964.
In November 1961, in reaction to French nuclear tests in the Algerian Sahara, Brazil voted in the
UNGA to make Africa a nuclear weapon free zone (NWFZ).
174
Then, in January 1962, at the
Punta del Este meeting of the Organization of American States, Brazil suggested prohibiting the
presence of nuclear weapons in Latin America, although the suggestion failed to gain traction at
the time.
175
The following year Brazil, alongside Bolivia, Chile, Ecuador, and Mexico issued the
Five-Nations Declaration, calling for Latin America to become a nuclear weapon free zone. The
proposal set the stage for UNGA Resolution 1911, which called for denuclearization in Latin
America.
176

Despite early acceptance of emerging anti-nuclear norms, Brazil’s foreign policy in the
nuclear realm shifted following the 1964 military coup. On February 14, 1967, three years after

172
Ibid., 36.
173
Debs and Monteiro (2016, 91).
174
General Assembly Resolution 1652(XVI), Consideration of Africa as a denuclearized zone, A/RES/1652(XVI)
(24 November 1961), https://digitallibrary.un.org/record/662076?ln=en
175
Patti (2021, 51).
176
General Assembly Resolution 1911 (XVIII), denuclearization of Latin America, A/RES/1911(XVIII) (27
November 1963). https://digitallibrary.un.org/record/661377?ln=en

110

the military coup, the Treaty of Tlatelolco was signed by several countries, including Brazil. The
Treaty of Tlatelolco formalized the creation of a nuclear weapon free zone in Latin America and
the Caribbean. However, Brazil’s acceptance of the treaty’s terms was “only a formality” and
depended on specific instructions from the new president, Marshal Arthur Costa e Silva (1967-
1969). For Brazil to fully accede to the Treaty, Brazil required all Latin American countries to
join the nuclear weapon free zone and ratify two additional protocols to the treaty targeted at the
nuclear weapon states. According to “Additional protocol I, the ‘states having territorial interests
in the Americas agree to keep their possessions free of nuclear weapons.’ Through the
acceptance of Additional Protocol II, the NWSs (the United States, the United Kingdom, the
People’s Republic of China, France, and the Soviet Union) would commit ‘not to use or threaten
to use nuclear weapons’ against the full parties to the treaty.”
177
Not surprisingly, these terms
were unacceptable to the nuclear weapon states, and thus were rejected. As a result, Brazil did
not join the NWFZ established by the Treaty of Tlatelolco.
Brazil also opposed the creation of the Nuclear Non-Proliferation Treaty (NPT) in 1967.
Together with Argentina, Brazil opposed the NPT on the grounds that “it established a
discriminatory regime, which impinged on state sovereignty and hindered the technological
development of nonnuclear states, aiming at what Brazilian and Argentinian diplomats called the
‘disarmament of the disarmed.’”
178
Crucially for Brazil, the NPT’s prohibition of peaceful
nuclear explosions (PNE) was unacceptable, for these NPEs were perceived by Brazil as
necessary for the technological advancement of Brazil’s nuclear energy sector.
179
Ultimately, the
UNSC approved the passage of the NPT, 10 votes in favor and 0 against, with Brazil, Algeria,

177
Patti (2021, 60).
178
Debs and Monteiro (2016, 92).
179
Patti (2021, 63).
111

France, India, and Pakistan abstaining from the vote. On July 1, 1968, the NPT opened for
signatures, and entered into force on March 5, 1970.
180

Following the global adoption of the NPT, Brazil continued its path towards nuclear
independence and started plans to build its first nuclear power plant. Emílio Médici (1969-1974),
successor to Costa e Silva, started building the nuclear power plant in Angra dos Reis. Although
some Brazilian nuclear scientists wanted the development of indigenous technology, Brazil’s
leadership ultimately decided to buy the reactor from a foreign country. In 1972, “U.S. based
Westinghouse won the public bid for Angra I, a pressurized water reactor, to operate under
International Atomic Energy Agency (IAEA) safeguards.”
181
The agreement was not ideal for
Brazil; the United States Atomic Energy Commission (USAEC) retained a monopoly of uranium
enrichment services, and approved supplying Brazil with nuclear fuel only under the strict
safeguards of a trilateral agreement between Brazil, the USA, and the IAEA. This deal
effectively made Brazil’s nuclear program highly dependent on the United States, contrary to the
original goal of nuclear independence advanced by the Brazilian military leadership. Worst of all
for Brazil, the 1974 Indian nuclear test “Smiling Buddha” led the United States to tighten
worldwide nuclear controls, causing the U.S. to try to limit Brazil’s access to nuclear technology
and materials.
182

In the context of tightening global nuclear technology controls and a global oil crisis
caused by the Yom Kippur War of 1973, Brazil’s new president General Ernesto Geisel (1974-
1979) kickstarted a national plan to develop its economic self-sufficiency. Geisel’s Second
National Development Plan and the Brazilian national electricity company Electrobrás’ Plano

180
Ibid., 65.
181
Debs and Monteiro (2016, 93).
182
Ibid., 94.
112

1990 (Plan 1990) called for reducing Brazil’s dependency on external entities for power and to
develop hydropower and nuclear energy.
183
To that effect, Brazil initiated talks with West
Germany, and by mid-February of 1974 both countries seemed to be close to settling a deal in
which West Germany would train Brazilian scientists and transfer sensitive nuclear technology
and two reactors to Brazil, in exchange for Brazilian uranium and purchasing two German
nuclear power reactors.
184
However, India’s Smiling Buddha nuclear test spooked the
international community, which led to West German revision of the terms. Instead of sharing
information regarding sensitive methods for uranium isotopic separation, West Germany decided
to propose an alternative, less efficient, method for uranium enrichment: jet-nozzle
technology.
185

Brazil accepted the terms of the agreement. On June 27, 1975, the new Brazilian public
nuclear corporation Nuclebrás “reached an agreement with West German Kraftwerk Union for
‘the first sale of a complete nuclear fuel cycle and one of the largest transfers of nuclear
technology to a developing nation.’”
186
The agreement included “joint uranium mining, the
aforementioned experimental jet nozzle technology, facilities for nuclear fuel production and
spent nuclear fuel reprocessing, and the construction of four pressurized water reactor.”
187

Although the agreement with West Germany included nuclear safeguards that were more strict
than the NPT, American Senators were uncertain about the true motives of the Brazilian
government. As a result, in June 1976 the U.S. Congress approved a proposal by Senator Stuart
Symington (D-MO) to amend the 1961 Foreign Assistance Act. The measure “prohibited US

183
Patti (2021, 82).
184
Ibid., 85.
185
Ibid., 86.
186
Debs and Monteiro (2016, 95).
187
Ibid., 95.
113

economic and military assistance to any country supplying or receiving nuclear enrichment or
reprocessing equipment materials or technology unless the supplier and recipient agreed to
accept IAEA safeguards on everything transferred and on all nuclear fuels and facilities in the
recipient country”
188
The proposal endangered Brazil’s economy, since it would prohibit the U.S.
from providing any economic assistance to Brazil if the latter was deemed to be in violation of
the updated Foreign Assistance Act.
By 1978, the White House itself began to suspect Brazilian nuclear intentions. A CIA
1978 report concluded that “if [Argentina and Brazil] continue to develop reprocessing
capabilities as planned, Argentina will be within reach of nuclear weapons by 1981, and Brazil
by 1983.”
189
This skepticism led to the U.S. Congress passing the Nuclear Nonproliferation Act
in May 1978, which required full-scope safeguards and stricter rules for importation and US
assistance.
190
Increasing suspicion by the United States and the ensuing tightening of nuclear
controls, together with problems arising from the West Germany deal,
191
pushed Brazil towards
alternatives for nuclear development, such as the eventual establishment of a covert nuclear
program separate from the public civilian nuclear program.
Given the increasing number of problems plaguing the West Germany deal, such as lack
of sensitive nuclear technology, stricter international nuclear controls, and domestic criticisms
over the costs of the deal, General Ernesto Geisel decided to embark on indigenous production of
sensitive nuclear technology. On late 1978, Geisel ordered the National Commission for Nuclear
Energy (CNEN) to coordinate the activities of various research centers to produce technology

188
Patti (2016, 95).
189
Ibid., 99.
190
Ibid., 99.
191
West Germany’s construction of Angra 2 and Angra 3 was delayed due to company error, and the deal was
becoming increasingly unpopular with the Brazilian population.
114

that would convert yellowcake into uranium hexafluoride, useful for Uranium reprocessing. On
February of 1979, the Nuclear and Energy Research Institute (Instituto de Pesquisas Energéticas
e Nucleares, or IPEN) succeeded.
192
Success of IPEN convinced Geisel to approve on March 14,
1979, a working group that would develop uranium hexafluoride, which was necessary for
mastering the fuel cycle. The working group submitted Project Autonomo, which would allow
IPEN, under supervision of the Brazilian National Security Council (CSN), full capacity to
produce uranium hexafluoride. Geisel, near the end of his term, approved Project Autonomo.
Project Autonomo was a secret nuclear program, created with the explicit goal of
developing indigenous nuclear technologies free from international surveillance and control. As
a secret program, it would avoid public and international criticisms and coordination problems
extant with the West German deal. The goal of the project was “independent production of UF6
on an industrial scale,” and served as the first step “toward the Autonomous Program of Nuclear
Technology (Programa Autônomo de Tecnologia Nuclear, or PATN). The ultimate purpose was
to give Brazil total control of the nuclear fuel cycle and of peaceful uses of nuclear energy, from
naval propulsion to PNEs.”
193
Within the PATN, the three branches of the Brazilian military
pursued separate methods to try to obtain fissile materials. From the three, the navy was most
successful; by 1982, navy scientists working in the laboratories of IPEN assembled the first
Brazilian centrifuge prototype and managed their first uranium isotopic separation, a major step
in mastering control over the nuclear fuel cycle.
194

Despite the establishment of a parallel, covert nuclear program designed to indigenously
improve the nuclear sector, Brazil did not give up attempts for foreign collaboration. In fact, one

192
Ibid., 111.
193
Ibid, 112.
194
Ibid., 119.
115

of the first major steps Brazil took towards collaboration at the regional level was the nuclear
cooperation agreement signed with Argentina on May 17, 1980. After successfully ending
tensions with Argentina over the Paraná River, Brazil and Argentina reached a deal in which
Brazil would manufacture heavy components for Argentina’s nuclear plants, while Argentina
would offer fuel fabrication services and lease uranium concentrate to Brazil, marking the first
Brazilian-Argentinian agreement on the peaceful use of nuclear energy.
195
The deal symbolized
the beginning of a long-term collaboration in the nuclear sphere between Argentina and Brazil,
which settled down fears from international observers that Argentina and Brazil might engage in
a nuclear arms race.
In 1985, Brazil held its first presidential elections in 24 years. Tancredo Neves won the
national elections but died shortly thereafter before his inauguration, and was replaced by his
vice president, Jose Sarney. Sarney continued Brazil’s improved relationship with Argentina. In
November 29 and 30 of 1985, president Sarney met with Argentina’s president Alfonsín in Foz
de Iguaçu, in the Brazilian state of Parana, to sign a Joint Declaration on Nuclear Policy. The
declaration established “a standing joint committee on nuclear policy chaired by their foreign
ministers,” perhaps with the goal of strengthening their respective civilian control over their
nuclear programs and diminish the influence of the military.
196
Further deepening cooperation
between both countries, Brazil and Argentina’s joint working group (JWG) began a cycle of
mutual inspections of nuclear sites, allowing nuclear inspectors to visit previously closed off
nuclear facilities in each country, a practice that cemented trust between both nations.
197

195
Ibid., 129.
196
Debs and Monteiro (2016, 98).
197
Patti (2021, 155).
116

Under Sarney, Brazil began a process of democratization in which the armed forces were
slowly phased out of key positions of power, but not completely-the military still held powerful
positions in the cabinet.
198
Perhaps as a result of the still powerful military, the Sarney
administration refused to immediately declassify the existence of the covert nuclear military
program. Yet, the Sarney administration took steps early on to reveal the existence of the
program to the public. For instance, in December 1986, CNEN President Rex Nazaré Alves
revealed the existence of a secret nuclear program, but denied that its purpose was to build
nuclear weapons. Finally, in September 4, 1987, President Sarney announced to the world that
Brazil mastered uranium enrichment techniques that were unrelated to their deal with West
Germany.
199
The announcement made it clear to the international community about the existence
of a secret parallel nuclear program: “Sarney, surrounded by political, military, and scientific
leaders, announced to the world that Brazil had developed a method for isotopic separation
autonomously and had manufactured the centrifuges in unsafeguarded Brazilian nuclear research
centers.”
200

The announcement had important ripple effects at the domestic and international levels.
The international community, particularly the United States, were uncertain about the true
purpose of the Brazilian parallel program, leading the U.S. to refuse transfer of machinery and
materials that might be useful for the gas centrifuge program.
201
Domestically, concerned
scientists and civilians wanted to enshrine repudiation of nuclear weapon technology within the
Brazilian Constitution. At this time, a global movement called for nuclear disarmament
following the 1986 Chernobyl disaster, which no doubt influenced members of the Brazilian

198
Ibid., 140.
199
Ibid., 141.
200
Ibid., 142.
201
Ibid., 143.
117

Constituent Assembly. These activists proposed an amendment to the Constitution, which would
have explicitly limited “all nuclear activity in the national territory to peaceful aims.” However,
the amendment failed by a vote of 223 to 168, ensuring that the option for PNEs remained
open.
202
Yet, despite this victory for the military establishment that still supported the covert
nuclear program, the nuclear program underwent major changes. Most important of all, the
Brazilian Congress dismantled Nuclebrás. The CNEN was then given permission to control the
remnants of Nuclebrás, replacing it with Indústrias Nucleares Brasileiras (Brazilian Nuclear
Industries), or INB.
203

Major restructuring and reassessment of the Brazilian nuclear program began after the
election of Fernando Collor de Mello on December 17, 1989. One week after taking office,
Collor ended the Autonomous Program, shutting down PATN.
204
Then, in June 1990 the
remnants of the parallel secret program and the official nuclear program were unified. The
Brazilian congress was then tasked with full supervision over all nuclear activities.
205
A major
public blow to the military faction that wanted to maintain the covert nuclear program occurred
on September 19, 1990. On that day, President Mello flew to the Serra do Cachimbo air force
base, and symbolically threw whitewash into a shaft for nuclear tests. The previous day, Collor
“announced that he was closing the shafts. Contrary to the previous government’s declarations,
he declared that they had been designed to test nuclear devices.”
206
Further solidifying Brazil’s
acceptance of global non-proliferation norms, Collor officially announced on September 24,
1990 at the UNGA that Brazil formally gives up the right to develop peaceful nuclear

202
Ibid., 146.
203
Ibid., 146.
204
Barletta (1997, 27).
205
Patti (2021, 160).
206
Ibid., 162.
118

explosions.
207
Perhaps these major changes in Brazil’s nuclear program were the result of Collor
feeling that the military covert nuclear program undermined his political authority and fueled
suspicion by the United States and the international community that Brazil was engaged in a
nuclear competition with Argentina.
208
If so, Brazil needed to send strong costly signals that it
was now willing to enter the global nonproliferation regime.
Finally, the metaphorical nail in the coffin killing any doubts about Brazil’s nuclear
intentions occurred when Brazil fully joined the Treaty of Tlatelolco. In the end of July 1990,
Brazil and Argentina discussed the possibility of joining Tlatelolco, but this required discussing
the IAEA safeguards that were considered too severe by Brazil in the past. Eventually, both
countries agreed on a system of mutual accountability and agreed to create a Common System
for Accounting and Control of Nuclear Materials (SCCC), which would apply to all nuclear
facilities in both countries.
209
In May 1991, an international organization, ABACC, was created
to supervise implementation of the SCCC. The ABACC agreement was signed on July 18, 1991,
in Guadalajara Mexico. That day, both Argentina and Brazil explicitly renounced PNEs and
supported a ban on nuclear weapons.
210
Then, and crucially, “the bilateral pact and the
constitution of ABACC were followed on 13 December 1991 by the Quadripartite Agreement,
between Brazil, Argentina, ABACC, and the IAEA…the four parties reached an agreement to
apply international full-scope safeguards to Brazilian and Argentine nuclear activities.”
211
On
February 9, 1994, Brazil completed its accession into the nuclear non-proliferation regime with

207
Ibid., 163.
208
Goldemberg and Feiveson (1994, 13).
209
Patti (2021, 165).
210
Ibid., 168.
211
Ibid., 168.
119

the signing of the Treaty of Tlatelolco, becoming a full member of the Latin American nuclear
weapon free zone.
2) Brazil’s regime type and perception of threat
As the theoretical pathway of voluntary reversals expects, Brazil’s greatest endorsement
of international anti-nuclear weapons norms occurred during the periods in which Brazil was
governed by democratic governments and was at its lowest during the 21-year period in which
Brazil was ruled by a military regime. On the other hand, despite international perceptions that
Brazil was likely to engage in an arms race with Argentina, Brazil never experienced a
threatening security environment during the 1950-1990 period, contrary to the expectation that a
dangerous security environment would be the main driver of Brazil’s nuclear programs. As will
be shown below, Brazil’s promotion of anti-nuclear proliferation norms was more frequent
during the democratic regimes. In addition, evidence is presented that shows Brazil did not
perceive Argentina as a security threat, as some international observers believed.
From 1955 to 1964, when Brazil was a democracy, development of nuclear energy was
made a top priority. Brazilian president Kubitschek (1956-1961) presented the Plano de Metas,
which was a national plan to industrialize Brazil in a short period of time.
212
Yet, this pursuit of
nuclear energy did not translate to pursuit of a nuclear bomb. In fact, Brazil by 1955 fought to
become a founding member of the IAEA, to both “preserve sovereignty over its mineral
resources in the event of internationalization of nuclear materials” and “secure resources for
supporting the peaceful use of the atom in the Global South through the creation of an
international organization devoted to atomic energy.”
213
Importantly, between the years 1961-

212
Ibid., 39.
213
Ibid., 46.
120

1964, during the presidency of João Goulart, Brazil promoted international nonproliferation and
disarmament. For instance, on November 11, 1961, Brazil voted in the UNGA to make Africa a
nuclear weapon free zone.
214
In January 1962, Brazil, alongside Mexico, suggested creating a
Latin American nuclear weapon free zone.
215
In April 1963, Brazil alongside other Latin
American countries issued a declaration in the UNGA for a regional nuclear weapons free zone,
formalizing their 1962 call for such a NWFZ in the region.
216

However, Brazil’s explicit support for nuclear non-proliferation international norms
shifted dramatically following the 1964 military coup in Brazil. On March 31, 1964, a bloodless
coup took place in Brazil, which was supported by several elites, the military establishment, and
accepted by the general population. Several justifications are offered for the coup; President
Goulart was accused of being responsible for a socialist revolution in Brazil, which was feared
by the upper class. Furthermore, frequent inflation and slow economic growth incentivized rival
factions to overthrow the Brazilian democracy without much popular opposition and protest.
217

The military regime would last for 21 years, ending in a slow transition to democracy that began
around 1985 and solidified by 1990. During the military regime, the country’s foreign policy
towards international norms against nuclear weapons changed.
Although Brazil’s policy during the military regime was not the development of nuclear
weapons, Brazil still refused to support international treaties that served as important pillars of
the nonproliferation regime. For instance, Brazil refused to adhere to the Treaty of Tlatelolco,

214
General Assembly Resolution 1652(XVI), Consideration of Africa as a denuclearized zone, A/RES/1652(XVI)
(24 November 1961), https://digitallibrary.un.org/record/662076?ln=en
215
Patti (2021, 51).
216
General Assembly Resolution 1911(XVIII), Denuclearization of Latin America, A/RES/1911(XVIII) (27
November 1963), https://digitallibrary.un.org/record/203964?ln=en
217
Patti (2021, 57); Alston et. al. (2016, 208).
121

which would set a NWFZ in Latin America, because not all countries in the region intended to
join at the time and because the United States would not commit to never using nuclear weapons
in Latin America.
218
Furthermore, Brazil opposed the 1968 Treaty on the Non-Proliferation of
Nuclear Weapons (NPT) on the grounds that it discriminated between nuclear “haves” and “have
nots.”
219
In 1967, Brazilian President General Artur da Costa e Silva defended the right to
peaceful nuclear explosions in a National Security Meeting, stating that “nothing prevents us to
from doing research…and even developing devices that can explode. We don’t have to call it a
bomb, but a device that can explode.”
220
Although such a hypothetical device was not discussed
as a nuclear weapon, but an explosive device for “peaceful nuclear explosions,” it no doubt ran
counter to evolving international non-proliferation norms that wanted to limit the spread of
nuclear weapons and similar explosive devices.
Most worrying for international observers, the Brazilian military regime began a secret,
parallel nuclear program in the 1970s, to avoid restrictions on sensitive technology transfers and
international scrutiny. The program was revealed to the public during the period of democratic
transition in 1987, when Brazilian president Sarney declared that “we mastered the technology of
centrifuges and all peripheral equipment in an experimental cascade of uranium enrichment.”
221

The parallel covert nuclear program was an affront to international norms aiming to control
research, development, and proliferation of sensitive nuclear technology. Its covert nature meant
that International Atomic Energy Agency (IAEA) observers were unable to inspect the secret

218
Patti (2016, 60).
219
Spektor (2016, 638).
220
Ibid., 638.
221
Dalaqua (2019, 235).
122

nuclear facilities in Brazil, in direct opposition to the international norm that the IAEA would
ensure safe, secure, and peaceful use of nuclear technology.
Interestingly, Brazil rejoined the international regime against nuclear weapon
proliferation and supported related anti-nuclear norms during and after the period of
democratization. Sarney, who assumed the Presidency in 1985 after the untimely death of Neves,
the first person elected president in popular elections in 21 years, reassured the international
community that Brazil’s nuclear goals were peaceful. On September 19, 1990, Brazil’s president
Fernando Collor de Mello demonstrated via symbolic action that Brazil abandoned any lingering
nuclear ambitions by throwing whitewash into a borehole in the Serra do Cachimbo air force
base to signify the closing of the nuclear test site.
222
On that same year, Brazil publicly
renounced the right to PNEs at the United Nations General Assembly. In 1994, Brazil finally
acceded completely to the Treaty of Tlatelolco, reaffirming that Latin America will be a nuclear
weapon free zone, as originally intended by the democratic regime prior to the military coup.
223

In 1998, Brazil joined the NPT, becoming deeply integrated in the global anti-nuclear weapon
proliferation regime.
The military regime’s covert nuclear program and technological progress fueled fears
that Brazil was on the verge of developing nuclear weapons in response to a perceived
threatening security environment. At least, this was the view of the United States and foreign
observers, who assumed there was a regional competition between Brazil and Argentina. For
example, in 1975 U.S. Senator John Pastore (D-RI), chairman of the special committee on
atomic energy, wanted to block the West Germany-Brazilian nuclear deal, fearing that the deal

222
Patti (2021, 157).
223
Ibid., 172.
123

“could contribute to the fabrication of an atomic bomb if [Brazil] desires.”
224
Crucially, the CIA
itself believed that there was real potential for the development of a nuclear bomb in Brazil. In
1978, the CIA reported that “if [Argentina and Brazil] continue to develop reprocessing
capabilities as planned, Argentina will be within reach of nuclear weapons by 1981, and Brazil
by 1983.”
225
On the other hand, observant scholars should perceive that Brazil and Argentina
were not engaged in a security competition that could develop into a nuclear armed race; the last
war between Argentina and Brazil took place in 1825-1828.
Contrary to fears from other countries that Brazil and Argentina would engage in a
nuclear arms race, the historical record seems to show that Brazil perceived its security
environment as benign. At no point has any scholar found any evidence that military officers
wanted a nuclear weapon as a deterrent against Argentina.
226
In fact, the most serious dispute
between both countries was over hydraulic resources in the River Plate basin, which they settled
peacefully in 1979.
227
Even if both countries engaged in economic competition for regional
hegemony, there is no observable evidence that this competition translated to a nuclear arms race
in Latin America. Instead of competition, Argentina and Brazil engaged in deep cooperation over
the following years after settling their territorial dispute; by one scholar’s count, over the next
twelve years after 1979, Argentina and Brazil signed ten agreements publicly reaffirming the
peaceful nature of their respective nuclear programs, and coordinated their nuclear policies to
integrate into the global nonproliferation regime.
228

224
Ibid., 93.
225
Ibid., 99.
226
Debs and Monteiro (2016, 112).
227
Ibid., 112.
228
Kutchesfahani (2014, 55-56).
124

The cooperation between Brazil and Argentina reached its highest point in the 1990s,
when both countries coordinated nuclear policies to alleviate fears of a potential nuclear race
between them. As shown in previous pages, Brazil and Argentina jointly signed the Iguacu
Declaration on the Common Nuclear Policy, renouncing nuclear tests. They both renounced
PNEs and established ABACC for mutual inspection of nuclear installations. This was followed
by both countries signing the Quadripartite Agreement with the IAEA, accepting international
inspections, and finally both integrating into the Treaty of Tlatelolco and the NPT.
229
In short,
rather than competition, both countries engaged in cooperation, which deepened greatly after
democratization in both countries.
In sum, Brazil’s regime type during the period of analysis underwent two major
transitions, from democracy to military dictatorship in 1964 and then a slow transition to
democracy that began in 1985 and solidified by 1990. Although even the Brazilian military
regime publicly rejected that it intended to obtain nuclear weapons, Brazil’s greatest support for
nonproliferation norms and joining the international non-proliferation and anti-nuclear weapons
regime occurred during periods of democracy. Furthermore, although international observers
believed there was a threatening security environment for Brazil, no evidence appears to support
that Brazil perceived security threats from within its region.
3) Testing observable implications
After reviewing the history of Brazil’s nuclear program, the transitions from and to
democracy, and the constant perception of a benign security environment, make it safe to classify
Brazil as a case where we observe a voluntary nuclear reversal. The theoretical expectation is

229
Debs and Monteiro (2016, 103).
125

that a case such as Brazil would renounce nuclear weapon development and research after
accepting international norms against nuclear weapon proliferation. Such norm acceptance
should occur after democratization and if the security environment is benign—two expectations
that were met by 1985 when the democratic transition began. Granted, the historical record puts
into question the common belief that Brazil ever wanted to obtain nuclear weapons, yet as will
be demonstrated below there were some factions within the Brazilian military regime that did
discuss the nuclear option as a possibility. In addition, Brazil’s military regime rejected several
international norms against sensitive nuclear research and proliferation, norms that were not
explicitly embraced until after the end of the military regime. If, as argued here, Brazil classifies
as a case of voluntary reversal, then the case of Brazil should pass the following hoop tests:
1) We should observe that liberal democracies are the main actors pushing norms
against nuclear weapons ownership and proliferation in international forums such as
the United Nations General Assembly. We should observe that these acts of
international pressure are directed towards Brazil after the international community
discovers that Brazil is developing a covert nuclear program.
2) We should observe that liberal states, in their attempt to push for nuclear reversal,
will rely on distinct diplomatic tools when trying to convince Brazil to reverse their
program. When Brazil is authoritarian, we should observe threats to use force or at
minimum threats of economic sanctions.
3) Following democratization in Brazil, the main diplomatic tool used by states should
be shaming mechanisms or diplomatic “Carrots” such as sanction relief or stronger
security/economic ties in exchange for nuclear reversal.
126

4) We should observe that Brazil’s decision to reverse their nuclear weapons program
comes about after a decrease in their perception of threat or rivalry with neighboring
countries such as Argentina.
5) If Brazil terminates their nuclear program before Brazil changes its perception of
threat, it should be because other actors promised credible security guarantees.
6) We should observe that Brazil’s voluntary decision to terminate their nuclear
weapons program occurs following democratization.
7) Importantly, Brazil should provide reasons related to nuclear nonproliferation norms
when informing the international community about their decision to reverse course.

Like the case study of Iraq’s nuclear program, failure to pass most of these hoop tests
would weaken the theoretical model of a voluntary nuclear reversal pathway. However, passing
most if not all these hoop tests should not be interpreted as conclusive evidence that the theory
delineated here is fully supported; the theory still needs to be tested against other cases of
potential voluntary reversal to observe if the theoretical expectations travel beyond the chosen
case of Brazil. In sum, successfully passing the hoop tests increases our confidence that the
theory can be applied to real world cases successfully but is not in itself conclusive proof that it
will explain all potential cases yet to be analyzed.
Observable implication 1: We should observe that liberal democracies are the main
actors pushing norms against nuclear weapons ownership and proliferation in international
forums such as the United Nations General Assembly. We should observe that these acts of
international pressure are directed towards Brazil after the international community discovers
that Brazil is developing a covert nuclear program.
127

Evidence appears to support the hypothesis that international norms against nuclear
weapon proliferation and related anti-nuclear norms were mainly pushed by liberal democracies
during the period of analysis. As shown below, several international treaties aimed at controlling
nuclear weapon proliferation and advancing other related norms were signed and ratified
throughout the 1950-1990 period. Focusing on these treaties is important, for they serve as the
strongest examples of norm codification into international law, which can then be used as tools
against states violating international legal frameworks meant to control nuclear weapon
proliferation. By highlighting some of these treaties and which states played a major part in
pushing them, we can observe which kinds of states are pushing international norms against
nuclear weapons.
The first significant treaty that restricted nuclear behavior of states Is the Antarctic Treaty
(1959), which was proposed by the United States towards 11 nations that had either made a
claim to Antarctic territory or previously engaged in exploration of the territory. The 11 nations
plus the United States met at the Washington Conference of Antarctica from October 15 to
December 1, 1959 to discuss and sign the treaty. The treaty established that Antarctica would be
used for peaceful purposes only, that all scientific observations would be shared with all
members, and that all installations would be open to inspection. Crucially, Article V of the
Treaty states that “any nuclear explosions in Antarctica and the disposal there of radioactive
waste material shall be prohibited.”
230
Thus, the treaty established the first significant nuclear
prohibition and created precedent for future treaties to come. Of note is that the eleven countries
which signed the treaty were all democratic states at the time of signing, with the exception of
the Soviet Union.

230
The Antarctic Treaty, art. 5, December 1, 1959, 12 UST 794; 402 UNTS 71; 19 ILM 860 (1980),
https://documents.ats.aq/keydocs/vol_1/vol1_2_AT_Antarctic_Treaty_e.pdf
128

The next significant treaties were the Limited Test Ban Treaty (1963) and the Outer
Space Treaty (1967), The Limited Test Ban Treaty was first proposed by the Soviet Union, a
non-democratic state. The Limited Test Ban Treaty began as a trilateral agreement between the
United States, the United Kingdom, and the Soviet Union. The treaty prohibits tests of nuclear
devices in the atmosphere, outer space, and underwater, although it allowed tests to continue
underground.
231
The treaty, contrary to theoretical expectations, was initially pushed by a non-
democratic state, the Soviet Union, rather than a liberal democracy, yet set a norm against
nuclear weapons testing that would be upheld by the dominant liberal international regime.
Similarly, the Outer Space Treaty extended the prohibition of nuclear weapons in outer space,
banning any plans to have nuclear weapons orbit around the Earth or potential stations of nuclear
weapons on the moon.
232
Signed on January 27, 1967, the treaty was another major effort of
cooperation between the world powers to limit the scope of nuclear competition. Unlike the
Limited Test Ban Treaty, the proposal for the Outer Space Treaty was originally presented by the
United States.
In 1968, the Nuclear Non-Proliferation Treaty (NPT) was signed by several states and
went into force on March 5, 1970. The NPT, considered today a landmark treaty and a pillar of
the current international nonproliferation regime, prohibited nuclear weapon proliferation, and
promoted cooperation for peaceful use of nuclear energy. Importantly, the treaty established the
formation of a safeguard system overseen by the International Atomic Energy Agency (IAEA),
which would verify compliance with the NPT.
233
The treaty was originally negotiated by the

231
“Limited Test Ban Treaty (1963).” Limited Test Ban Treaty (1963) | Nuclear Arms Control Treaties,
https://www.atomicarchive.com/resources/treaties/ltbt.html.
232
“Outer Space Treaty (1967).” Outer Space Treaty (1967) | Nuclear Arms Control Treaties,
https://www.atomicarchive.com/resources/treaties/outer-space.html.
233
Treaty on the Non-Proliferation of Nuclear Weapons, March 5, 1970,
https://www.un.org/disarmament/wmd/nuclear/npt/text
129

Eighteen Nation Disarmament Committee, which inherited the United Nations Disarmament
Commission (UNDC) task to prepare a treaty “for the regulation, limitation and balanced
reduction of all armed forces and all armaments, including the elimination of all weapons of
mass destruction.”
234
The main membership of the eighteen nation disarmament committee was
equally divided between the liberal democratic Western bloc (Canada, France, USA, UK, and
Italy) and the Communist, non-democratic Eastern bloc (Bulgaria, Czechoslovakia, Poland,
Romania, and Soviet Union), demonstrating that the formation of the treaty was a collective
effort across regime types.
After 1968, the major nuclear-related treaties were largely bilateral agreements and
treaties between the United States and the Soviet Union designed to limit missile delivery
capabilities. The Strategic Arms Limitation Treaty I (1968), Anti-Ballistic Missile Treaty (1972),
Strategic Arms Limitation Treaty II (1979), and the Intermediate-Range Nuclear Forces Treaty
(1987) were some of the many bilateral treaties that placed various limits on delivery systems
and missile technology, further solidifying the norm against nuclear weapon use and signaling to
the international community that the two superpowers were working on international
mechanisms for oversight and enforcement of nuclear weapon non-use.
235
However, these
bilateral treaties were driven by a desire to avoid a great power conflict first and foremost, rather
than being motivated by nonproliferation norms.
Most of the states that first signed or proposed these major treaties were democratic
states, with a few exceptions. As hypothesized, it is mainly liberal democracies pushing norms

234
United Nations. United Nations Disarmament Commission. Accessed October 3, 2022.
https://www.un.org/disarmament/institutions/disarmament-commission/
235
“Intermediate-Range Nuclear Forces Treaty - Inf (1987).” Intermediate-Range Nuclear Forces Treaty - INF
(1987) | Nuclear Arms Control Treaties, accessed October 3, 2022.
https://www.atomicarchive.com/resources/treaties/inf.html.

130

against nuclear weapon proliferation during the 1950-1990 period under analysis. As shown
above, the major nuclear treaties established legal frameworks to enforce international norms
such as nuclear nonproliferation, prohibition of nuclear tests at various sites, and non-sharing of
sensitive nuclear technology. Having established that the nonproliferation and anti-nuclear
regime were mainly efforts of democratic states, the next expectation is that the developing
global anti-nuclear weapons regime would pressure Brazil to terminate their nuclear ambitions
after the covert program was revealed in 1987. Surprisingly, this expectation does not gain
support: the international community pressured Brazil to join the nonproliferation regime before
1987, several years before the world learned about Brazil’s covert parallel nuclear program.
The major driver of these pressures towards Brazil was the United States, which was
highly suspicious of Brazil’s nuclear intentions, particularly after Brazil’s rejection of the NPT in
1968. For example, in 1972 Brazil signed an agreement with the American company
Westinghouse for the sale of a nuclear power plant. However, due to Brazil’s opposition to the
NPT, the United States government pressured Westinghouse not to sell natural uranium heavy-
water reactors, which had potential to be used in nuclear weapon development. In the end,
Westinghouse sold Brazil light-water reactors instead, which were inadequate for potential bomb
production.
236

In 1975, Brazil’s attempt to collaborate with West Germany in the nuclear arena raised
fears within the United States about the actual intentions of Brazil. These fears were explicitly
expressed in a New York Times editorial titled “Nuclear Madness.” According to the editorial,
the multi-million-dollar agreement between both countries “is a reckless move that could set off
a nuclear arms race in Latin America, trigger the nuclear arming of a half‐dozen nations

236
Debs and Monteiro (2016, 106).
131

elsewhere and endanger the security of the United States and the world as a whole.”
237
The
United States government seemingly agreed; the State Department worked closely with West
Germany to restrict the sale of sensitive nuclear technology. The White House finally approved
the Brazil-West Germany deal once the agreement included “safeguards on all materials and
projects included in the technology transfer.”
238

The 1976 election of President Carter in the United States exacerbated the situation for
Brazil and might have contributed to Brazil’s decision to kickstart the parallel program. In an
October 1976 debate, Carter states the U.S. should “stop the sale by Germany and France of
reprocessing plants to Pakistan and Brazil…. If we continue under Ford’s policy, by 1985 or ’90
we’ll have twenty nations that have the capability of exploding atomic weapons. This has got to
be stopped.”
239
Then, in 1978, Carter signed Congress’ Nuclear Nonproliferation Act, which
suspended enriched-uranium transfers to non-NPT countries, an action that affected Brazil’s
access to technology given their lack of NPT status.
240
The United States skepticism of Brazil’s
nuclear intentions was also shared with West Germany, which began to deny access to specific
technology deemed sensitive. As explained in the previous section, West Germany denied
transfer of gas centrifuge enrichment technology to Brazil because it could be used for
production of weapons-grade uranium, and instead offered the inferior jet nozzle technology.
241

These barriers to technological access might have pushed Brazil towards the creation of a
parallel covert nuclear program, free from international oversight mechanisms.

237
“Nuclear Madness,” The New York Times, 13 June 1975, pp. 36.
https://www.nytimes.com/1975/06/13/archives/nuclear-madness.html
238
Spektor (2016, 641).
239
Quoted in Spektor (2016, 641).
240
Ibid., 642.
241
Ibid., 642.
132

The international community began to suspect that Brazil had a covert nuclear program in
the 1980s, concerns which were driven by Brazil’s interest in nuclear delivery systems. In 1979,
Brazil started to design and build satellites, a rocket for low Earth orbit, and a launch site. These
advancements raised suspicion and led to Brazil becoming the target of an embargo by the
Missile Technology Control Regime (MTCR), a 35-nation group designed to limit proliferation
of missiles and related technology.
242
Brazilian journalists reported in the 1980s that Brazil’s
military built two shafts in Cachimbo designed to test nuclear explosions, reports that
undoubtedly were read by foreign government intelligence agencies.
243
These reports were later
confirmed in 1987 by Brazil’s President Sarney, who confirmed the existence of a parallel
nuclear program but denied that nuclear weapons were the end goal. After the confirmation of
the existence of such a program, Sarney and proceeding presidents began a process of voluntary
integration into the nonproliferation regime, making further external pressures unnecessary.
In sum, the evidence thus far presented highlights that most international efforts to build
a nuclear non-proliferation regime came from liberal democracies, although the Soviet Union
also played a major part in the construction of this regime. However, in contradiction to the
theoretical expectation, efforts to pressure Brazil to join the regime began before the revelation
of a covert nuclear program. On the contrary, international efforts, driven by the United States, to
pressure Brazil to join the regime occurred after Brazil’s rejection of the NPT, and might actually
have pushed Brazil towards the creation of a secret nuclear program to avoid increasing
international oversight over its program. Once Brazil revealed its covert program in 1987, Brazil
rapidly embraced anti-nuclear and non-proliferation norms, which made further efforts to shame
Brazil into joining the international regime unnecessary. Given these findings, Hypothesis 1

242
Ibid., 643.
243
Ibid., 643.
133

gains only partial support: liberal democracies are the main actors pushing the international
nuclear nonproliferation regime, but these actors did not wait for confirmation of a potential
nuclear weapons program to pressure Brazil into joining the regime.
Observable implication 2: We should observe that liberal states, in their attempt to push
for nuclear reversal, will rely on distinct diplomatic tools when trying to convince Brazil to
reverse their program. When Brazil is authoritarian, we should observe threats to use force or at
minimum threats of economic sanctions.
Due to Brazil rapidly accepting international norms against nuclear weapon proliferation
and other related nuclear norms after its democratic transition in 1985, there were no signs of
meaningful diplomatic pressure from other states towards Brazil between 1985-1990. However,
as expected, we do see international coercion towards Brazil during its authoritarian era in an
attempt to push Brazil into accepting the nonproliferation regime. In particular, states pressuring
Brazil to join the international regime withheld crucial sensitive nuclear technology and expertise
in exchange for Brazilian entrance into major nuclear agreements and treaties.
Restrictions on nuclear technology transfers towards Brazil began soon after Brazil made
its opposition to the NPT public. One of the major restrictions was the effort by the Zangger
Committee, formed by Claude Zangger from Switzerland in 1971, consisting of 35 states. On
September 3, 1974, the committee established a “trigger list” of special fissionable materials that
should be subject to IAEA safeguards if the materials are supplied by an NPT member towards a
non-signatory state.
244
Although the Zangger list was not exclusively targeted at Brazil, it still
impacted Brazil’s nuclear development plans due to the country refusing to enter the nuclear

244
International Atomic Energy Agency. “Communication Received from Members Regarding the Export of
Nuclear Material and of Certain Categories of Equipment and Other Material,” 3 September, 1974. International
Atomic Energy Agency, INFCIRC/209,
https://inis.iaea.org/collection/NCLCollectionStore/_Public/44/096/44096604.pdf
134

non-proliferation treaty. As a non-NPT state, Brazil would be required to accept IAEA
safeguards if it wanted access to sensitive nuclear technology, safeguards which Brazil’s military
regime was opposed to.
As mentioned previously, West Germany itself changed its nuclear policy towards Brazil,
due in part to American pressures but also due to fears of global nuclear proliferation. As a
result, the West Germany-Brazil nuclear transfer deal was affected in 1974, leading to Brazil
receiving experimental jet nozzle technology that was considered useless for uranium
enrichment.
245
Even after conclusion of the nuclear transfer deal, the United States pressured
West Germany to delay delivery of the jet nozzle technology. The limitations and barriers extant
within the nuclear transfer deal with West Germany ensured that Brazil would not be able to
obtain sensitive nuclear technology that might be useful for a nuclear bomb.
Interestingly, there is evidence that the United States did try to use positive
reinforcements at least once to incentivize Brazil to join the NPT, contrary to the expectation that
diplomatic “carrots” would not be used against non-democratic states. In 1976, Kissinger offered
Brazil an undisclosed package of economic incentives in exchange for joining the NPT.
However, the incoming Jimmy Carter administration leaked the secret negotiations between
Kissinger and Brazil, driving Brazil to withdraw from the negotiations. Brazil’s leadership
publicly denied “that they had ever considered a proposal to trade their right to autonomous
nuclear development for US aid.”
246
Although the negotiations failed, it does show that
diplomatic incentives were not off the table in strategies to push Brazil towards the nuclear
nonproliferation regime. However, this act of diplomacy was the exception rather than the norm;

245
Debs and Monteiro (2016, 107).
246
Spektor (2016, 641).
135

coercive tactics were still central in efforts to pressure states like Brazil to join the nuclear
nonproliferation regime.
At the global level, the multitude of multilateral treaties limiting sensitive nuclear
technology transfers, enforcement of oversight measures, and general worldwide opposition to
nuclear weapons research, development, and proliferation meant that Brazil had to move its
sensitive nuclear program underground in the 80s to avoid suspicion. Yet, the international
community grew more suspicious of Brazil’s intentions, leading to an embargo imposed by the
international Missile Technology Control Regime (MTCR) in the 1980s. In 1990, the US also
imposed sanctions on Brazil, suspending sales of supercomputers because of concerns about
Brazil’s nuclear program.
247
Brazil was still not a member of the NPT or fully integrated into the
Treaty of Tlatelolco at the time, raising suspicions about Brazil’s nuclear intentions.
Overall, it appears that the main effort to pressure Brazil into joining the regime came
from the United States, but the international community also cooperated via formation of
international committees and enforcement of anti-nuclear weapon and technology treaties.
Threats of force were never directed towards Brazil, most likely because Brazil was not an
existential threat to a major power or a regional security risk. However, sanctions did occur
against Brazil, in the form of embargoes and limitations on which type of nuclear technology
Brazil could receive from partners without major global oversight. These measures ensured that
Brazil could not integrate itself in the peaceful global nuclear network until it acceded to the
NPT and related treaties, events that only took place after democratization in Brazil.

247
Ibid., 643.
136

Observable implication 3: Following democratization in Brazil, the main diplomatic tool
used by states should be shaming mechanisms or diplomatic “Carrots” such as sanction relief or
stronger security/economic ties in exchange for nuclear reversal.
In the case of Brazil, the only major instance of coercion post-democratic transition
occurred in 1990 when the United States suspended sale of supercomputers to Brazil. Apart from
that brief incident, there are no documented collective efforts to incentivize Brazil to join the
international non-proliferation regime, whether by coercion (diplomatic sticks) or positive
incentives (diplomatic carrots). There was no need to; Brazil rapidly embraced international
norms against nuclear non-proliferation and other types of sensitive nuclear controls after
democratic transition. As the historical background shows, President Collor shut down the secret
parallel nuclear program in 1990, ordered closing the Cachimbo test shafts, and formally
renounced peaceful nuclear explosions in front of the General Assembly.
248
In 1994, Brazil
entered a deal with the IAEA and ABACC which would allow full scope safeguards of formerly
secret nuclear facilities, and ratified the Treaty of Tlatelolco. In 1998, Brazil ratified the NPT,
fully immersing itself in the global anti-nuclear regime.
249

Brazil’s accession to the NPT and related treaties ensured that no diplomatic efforts were
required by other states to incentivize Brazil to join the regime. Brazil acceded to the non-
proliferation regime of its own accord, likely to join the liberal international order and avoid
becoming a pariah state outside the new global order emerging from the ashes of the Cold War.
Furthermore, by entering the global anti-nuclear proliferation regime, Brazil gained access to
previously forbidden sensitive technology and gained access to technological markets that were
out of bounds towards non-NPT members and countries suspected of trying to build a nuclear

248
Spektor (2016, 645).
249
Ibid., 645.
137

bomb. Therefore, embracing these international norms against nuclear weapon proliferation and
in favor of inspections offered Brazil material benefits, such as access to technology that would
help improve the nuclear sector. Under this perspective, the “carrots” offered by the international
community to Brazil were the promise of access to sensitive nuclear technology if Brazil adhered
to the global nonproliferation regime and IAEA inspections.
A few examples of how Brazil gained from joining the global anti-nuclear proliferation
regime illustrate the benefits Brazil accrued from cooperation with the global regime. On April
23, 1996, Brazil became a full member of the Nuclear Suppliers Group (NSG). Once a member,
Brazil was allowed to trade sensitive technology that was crucial for its public nuclear program,
technology which would have been out of reach from free trade if Brazil had not joined the
NSG.
250
Other than access to technology, joining the nonproliferation regime offered economic
benefits as a member of the liberal order. For instance, joining the NPT in 1998 allowed Brazil to
qualify for IMF aid to tackle the financial crisis Brazil was facing at the time, which would have
been denied if Brazil continued to act outside the scopes of the nonproliferation regime.
251
Thus,
voluntary acceptance of international norms against nuclear weapon proliferation and related
norms of nuclear technology controls led to other countries lifting restrictions on nuclear trade
and increased Brazil’s status as a member of the liberal international order.
In sum, the evidence does not seem to show that there were active attempts to incentivize
Brazil to join the nonproliferation regime after democratization, but this is due to the lack of
need to do so. The extant nuclear regime itself already included incentives for countries to join
the regime, such as higher standing in the liberal international order and access to sensitive
nuclear technology. By voluntarily embracing these international norms, Brazil joined the liberal

250
Patti (2021, 181).
251
Ibid., 184.
138

order, cleared fears that it wished to obtain a nuclear weapon, and finally gained access to
formerly forbidden nuclear technology.
Observable implication 4: We should observe that Brazil’s decision to reverse their
nuclear program comes about after a decrease in their perception of threat or rivalry with
neighboring countries such as Argentina.
Observable implication 5: If Brazil terminates their nuclear program before Brazil
changes its perception of threat, it should be because other actors promised credible security
guarantees.
Hoop tests 4 and 5 are grouped together because they are opposites of each other: if
Brazil terminates their covert nuclear program after a decreased perception of threat, then
logically hoop test 5 is nullified. Similarly, discovering evidence in favor of hoop test 5 would
contradict the expectation of hoop test 4. Therefore, both hoop tests must be tested together
because finding supporting evidence for one of these expectations is equivalent to finding
contradictory evidence for the other.
A common perception by the United States during the 70s and 80s was that Argentina
and Brazil were rivals with potential of initiating an arms race. During the 1970s, American
foreign policy makers were concerned that Argentina’s progress in the nuclear field, and its
supposed ability to build a nuclear device, would force Brazil to accelerate a nuclear arms race
with Argentina.
252
The June 13, 1975, New York Times editorial “nuclear madness” stated fears
that a nuclear arms race was inevitable in Latin America if Brazil built a nuclear device.
253
In
1978, the CIA reported that “if [Argentina and Brazil] continue to develop reprocessing

252
Patti (2021, 104).
253
Ibid., 81.
139

capabilities as planned, Argentina will be within reach of nuclear weapons by 1981, and Brazil
by 1983.”
254
In sum, it was taken for granted during the 70s that Argentina and Brazil were
nuclear rivals on the verge of a nuclear arms race that could endanger the nuclear
nonproliferation regime.
These fears of a regional nuclear arms race or, at minimum, nuclear competition in Latin
America were also shared by independent observers. For example, scholar Joao Resende-Santos,
in his analysis of the evolving security cooperation between Argentina and Brazil in the 70s and
80s, assumes that Argentina and Brazil were engaged in a nuclear arms race:
The race to be first to acquire nuclear weapons, which Argentina led, was
abruptly altered in 1971, when Brazil purchased a Westinghouse nuclear plant…Brazil
changed the nuclear balance, presenting Argentina with the specter of a nuclear armed,
hegemonic rival. By 1976, strategists on both sides had concluded that the two rivals
were nuclear-capable.
255

Scholar Tatiana Coutto highlights that although military officers did not think a nuclear
arms race was likely between Brazil and Argentina, both countries were involved in a
“competition to attain regional supremacy (including in the nuclear realm).”
256
These fears
weren’t completely unfounded, due to Brazil and Argentina viewing each other with suspicion in
the mid-1960s and their competition over the Parana River Basin.
257
However, these fears were
exaggerated; Brazil did not perceive a threatening security environment, contrary to the belief of
international observers.
At no point did the top leadership of Brazil state that a nuclear weapon option was
necessary as a deterrent against Argentina, even during the military dictatorship period. In fact, it

254
Ibid., 99.
255
Resende-Santos (2002, 95).
256
Coutto (2014, 308).
257
Spektor (2016, 639).
140

seems that Argentina’s nuclear program was always peaceful; a nuclear weapon option was
never an option, so Brazil had no reason to fear its regional rival.
258
Brazil did not perceive
Argentina’s nuclear program as a security threat, and might have actually identified with
Argentina’s own struggles for nuclear independence. In fact, “on a score of documents, including
the NPT and the 1967 Treaty of Tlatelolco to ban nuclear weapons from Latin America, the
positions of Brazil and Argentina were practically identical...officials on both sides thought they
were sitting on the same side of the fence.”
259
Although Brazil and Argentina did not fully trust
each other during the 60s and 70s due to the Parana River Dispute, no evidence exists that this
distrust translated into a nuclear arms race. On the contrary, mid-1979 marked a remarkable shift
in Brazil-Argentina relations, further dissipating fears that they were nuclear competitors.
By 1979, Brazil and Argentina recognized their respective rights on the Parana River,
ending the central point of tension between both countries, opening up the diplomatic space for
nuclear collaboration. On May 17, 1980, Presidents Figueiredo and Videla signed an agreement
for peaceful use of nuclear technology. The aim was to establish joint projects and supply each
other with equipment and services. The signing was also the first time a Brazilian president
visited Buenos Aires since 1935.
260
The late 80s, following democratic transitions in both
Argentina and Brazil, saw rapidly increasing collaboration in the nuclear arena between both
countries. In 1987, both countries began mutual inspections of their nuclear sites, after creation
of the Joint Working Group (JWG). Then, in 1990, both states created the SCCC, a system of
mutual accountability, and negotiated a new tripartite safeguards agreement with the IAEA that
incorporated the SCCC.
261
By November 29, 1990, Brazil and Argentina met at Foz de Iguaqu

258
Hymans (2001).
259
Spektor (2016, 639).
260
Patti (2021, 131).
261
Ibid., 167.
141

in Brazil and issued a declaration of Common Nuclear Policy, where they promised to join the
Treaty of Tlatelolco.
262
On May 1991, both countries created the ABACC, which was an
international organization whose task was to supervise implementation of the SCCC. This
barrage of cooperation treaties and the formation of an interconnected system of mutual
oversight over nuclear facilities terminated fears that both countries were in danger of starting a
Latin American nuclear arms race.
As can be observed, the analysis of the Argentina-Brazil relationship does not support the
belief that Argentina and Brazil were engaged in a nuclear arms race or that Brazil perceived
Argentina as a security threat. Therefore, we can dismiss hoop test 5; Brazil never perceived
Argentina as a security threat and so never required a security guarantee from other states to
terminate its covert nuclear program. However, hoop test 4 also faces some difficulties. Hoop
test 4 reads as follows: We should observe that Brazil’s decision to reverse their nuclear
program comes about after a decrease in their perception of threat or rivalry with neighboring
countries such as Argentina. Yet, as shown above, Brazil never perceived threat from Argentina;
it seems even the Brazilian military regime considered its security environment to be benign.
Due to security perception being perceived as benign constantly throughout the period of
analysis, we cannot conclude that security considerations were a driver of Brazilian’s shift in
policy. However, change in the rivalry relationship might have played a part in Brazil’s
willingness to engage in nuclear cooperation with Argentina after resolving the Parana River
Dispute. Therefore, support for hoop test 4 is mixed: On one hand, security threat perceptions do
not seem to explain shifts in Brazilian nuclear policy, but shifts in its rivalry relationship with

262
Ibid., 167.
142

Argentina might have contributed to Brazil’s willingness for nuclear cooperation and eventual
nuclear reversal.
Observable implication 6: We should observe that Brazil’s voluntary decision to
terminate their nuclear weapons program occurs following democratization.
Contrary to my initial expectations, which were in alignment with other scholars,
263
it
seems that Brazil’s top leadership never intended to build a nuclear weapon, despite the
existence of the covert nuclear program. Still, the historical evidence supports the hypothesis;
After the transition to democracy, Brazil explicitly renounced any plans for the nuclear bomb in
public and embraced several anti-nuclear weapon norms. As shown already, Brazil’s
renunciation of the nuclear option, including peaceful nuclear explosives, occurred under
President Collor starting in 1990 when he shut down the covert program one week after being
elected president and formally announcing in the UNGA Brazil’s renouncement of PNEs.
264

Although we do observe explicit renunciation of nuclear weapons after democratization, and
seemingly no plans for the nuclear bomb option during the military regime, why is it so common
for scholars to believe that Brazil did pursue the nuclear bomb?
The answer to such an important question is two-fold. First, Brazil’s policy during the
60s and 70s to obtain sensitive nuclear technology, during a period in which global nuclear
proliferation fears were at an all-time high, led the international community to view Brazil
nuclear intentions with skepticism. This skepticism increased following rumors-which turned out
to be true-that Brazil initiated a covert nuclear program in the late 70s to avoid nuclear
inspections from the IAEA and international observers. Second, some top members of the

263
Resende-Santos (2002) seems to believe there was a nuclear arms race at one point between Argentina and
Brazil. Krasno (1994) feared that Brazil was capable of building a nuclear bomb in a few years.
264
Debs and Monteiro (2016, 101).
143

military-even one president- openly suggested the possibility of a nuclear weapons option, which
could easily be interpreted as public declarations of nuclear intentions by the military regime.
One of the earliest signs that was interpreted as a desire to build a nuclear weapon is
President Costa e Silva’s 1967 defense of the right for peaceful nuclear explosions. At a National
Security Council meeting, Costa e Silva said that “nothing prevents us from doing research…and
even developing devices that can explode. We don’t have to call it a bomb, but a device that can
explode.”
265
This statement was interpreted by foreign observers as a statement of policy for an
eventual nuclear weapon, yet that interpretation misses important economic contexts. For
Brazil’s leadership at the time, the possible development of a peaceful nuclear explosion device
was framed as essential for technological development. In fact, at the time it was expected that
small PNEs could be used for civil engineering works.
266
The prohibition on PNEs by the Treaty
of Tlatelolco and the NPT is among the top reasons why Brazil refused to join the global nuclear
regime, for it perceived the nascent global regime as a threat to its own economic and
technological development.
The only clear statement by a Brazilian president mentioning the possibility for a nuclear
weapons option occurred on June 10, 1974. On that date, three weeks after the Indian nuclear
test, General Ernesto Geisel told his top military officers that Brazil should “develop the
technology necessary for a peaceful nuclear explosion, which will even allow us, if necessary, to
possess our own weapon.”
267
However, at no point after the statement did Geisel or another
president state the necessity of obtaining a nuclear weapon, or hint that a nuclear weapon was the
ultimate goal of the covert nuclear program established in 1979. Considering Brazil did not face

265
Quoted in Spektor (2016, 638).
266
Ibid., 638.
267
Quoted in Debs and Monteiro (2016, 94).
144

any existential or security threats from Argentina, its regional rival, then it is unlikely that a
nuclear weapon plan was in the works even during the Geisel administration.
The remaining statements that could be interpreted as a plan to obtain the bomb come
from military officers rather than the president’s office itself, hence these statements should be
interpreted as personal desires rather than official policy. In a 1967 secret meeting between
President Costa e Silva and his cabinet in which PNEs were being discussed, the Minister of
trade and industry, Brigadier General Macedo Soares e Silva, mentioned that “saying that Brazil
will not produce a nuclear weapon is an illusion.”
268
However, Costa e Silva never stated that
Brazil would pursue the nuclear bomb as part of Brazil’s nuclear policy. However, the most
concerning statements by military officers took place in 1986 and 1987, when the international
community was on high alert about the possibility of nuclear proliferation.
In 1986, one year before the public revelation of the covert nuclear program, CNEN
president Rex Nazare Alves addressed rumors that there was a covert program, denying that the
purpose of the project was an atomic bomb.
269
However, he was contradicted by some military
officers. Army minister General Leônidas Pires Gonçalves publicly declared many times that
Brazil “needed to manufacture nuclear bombs in order to deter ‘potential aggressions’ from its
neighbors.”
270
General Haroldo Erichsen da Fonseca, the army science and technology secretary,
stated that building the nuclear bomb was not the goal of the covert program but that the
possibility was open if it was necessary to build the bomb.
271
Former navy minister Admiral
Maximiano Fonseca stated in September 1986 that “if it was up to me to decide, I would make
an atomic bomb and detonate it in front of international observers to demonstrate the extent of

268
Quoted in Patti (2021, 63).
269
Ibid., 141.
270
Ibid., 142.
271
Ibid., 142.
145

national technical know-how.”
272
These statements were quickly denied by the Brazilian
government, so the public, and perhaps international observers, would not believe that these
statements reflected official nuclear policy.
Ultimately, it does seem that Brazil embraced anti-nuclear proliferation norms after
democratization, even though it never pursued the nuclear weapon option as I originally
expected. Therefore, there is partial support to pass hoop test 6: there was no nuclear weapons
program to renounce, but Brazil did terminate its covert program and joined the anti-nuclear
proliferation regime following the end of the military dictatorship. Also shown above, the belief
that Brazil tried to obtain the bomb arises from misinterpretation of Brazil’s official policy
regarding the right to carry out peaceful nuclear explosions (PNEs) and mistakenly interpreting
statements by some military officers as official policy and taking a few presidential statements
out of context.
Observable implication 7: Importantly, Brazil should provide reasons related to nuclear
nonproliferation norms when informing the international community about their decision to
reverse course.
The historical record appears to support the expectation that Brazil would provide reasons
related to nuclear nonproliferation norms when informing the international community about
their decision to reverse their nuclear ambitions. There is no better example of Brazil rejecting
nuclear weapons and embracing nuclear nonproliferation norms than President Fernando Collor
de Mello’s September 24, 1990’s opening speech at the United Nations General Assembly. On
that day, Collor de Mello addressed the United Nations for the first time after being elected
president of Brazil. After acknowledging that his country is a new democracy and promising to

272
Quoted in Krasno (1994).
146

oppose authoritarianism, Collor announced Brazil’s goal of adhering to the Treaty of Tlatelolco
and renounced peaceful nuclear explosives:

Brazil ratified the Treaty of Tlatelolco and has reiterated in international forums its
respect for its purposes and those of related international agreements. Recently Brazil and
Argentina publicly announced that an intense exchange of views is taking place among
interested parties on the implementation of the Treaty of Tlatelolco…I believe, however,
that the time has come to go one step further. Brazil today discards the idea of any
experiments that might involve nuclear explosions, even if only for peaceful purposes.
We trust other nations will consider the possibility of following the same path.
273

The above speech marked an important development in Brazil’s nuclear history: Brazil officially
renounced peaceful nuclear explosives and expressed its goal of implementing the Treaty of
Tlatelolco, an important public admission of adhering to the international nuclear regime.
Significantly, Collor de Mello re-stated Brazil’s opposition to nuclear weapons in the United
Nations General Assembly’s September 23, 1991 session. Here, Brazil once again stated its
policy of adhering by international norms against nuclear weapon proliferation use:

Last year I announced in this very forum my Government’s decision to abstain from any
nuclear explosions- including for peaceful purposes, as if it were possible to explode
atomic bombs for peaceful purposes…

The Safeguards Agreement that Brazil and Argentina will sign with the International
Atomic Energy Agency will provide all pertinent information for verification of our
commitment to the exclusively peaceful uses of nuclear energy and will also preserve the
technological advances that we have arduously attained in the realm of the nuclear energy
cycle.
274

In addition, Brazil’s ensuing cooperation with Argentina in the nuclear sector,
establishing bilateral agreements of mutual inspection, further solidified Brazil’s stance

273
U.N. G.A.O.R., 45
th
Sess., U.N. Doc A/45/PV.4 (Sept. 27, 1990). Available from https://documents-dds-
ny.un.org/doc/UNDOC/PRO/N90/640/94/PDF/N9064094.pdf?OpenElement
274
U.N. G.A.O.R., 46
th
Sess., U.N. Doc A/46/PV.4 (Sept. 24, 1991). Available from https://documents-dds-
ny.un.org/doc/UNDOC/PRO/N91/612/17/PDF/N9161217.pdf?OpenElement
147

regarding nuclear nonproliferation norms. For instance, Article I of the 1990 Quadripartite
Agreement between Argentina, Brazil, the IAEA, and ABACC rejects nuclear weapon
development: “The States Parties undertake to accept safeguards, in accordance with the terms of
this Agreement, on all nuclear material in all nuclear activities within their territories, under their
jurisdiction or carried out under their control anywhere, for the exclusive purpose of verifying
that such material is not diverted to nuclear weapons or other nuclear explosive devices.”
275

Brazil’s signature of such agreements meant de-facto and de-jure acceptance on the global
international anti-nuclear weapon regime, integrating itself within it. As evidence of Brazil’s
new-found commitment to nonproliferation norms, on May 11, 1994, Brazil’s Congress ratified
the amendments to the Treaty of Tlatelolco, becoming a full member of the Latin American
nuclear weapon free zone.
In sum, Brazil’s signature of several nuclear control and inspection agreements alongside
Argentina, joining the Treaty of Tlatelolco, and finally ratifying the NPT in 1998 established
Brazil as a state compliant with the anti-proliferation regime. Importantly, Collor de Mello’s
speeches at the United Nations General Assembly demonstrated clear intentions to adhere to
these international norms, creating audience costs at the international level. If Brazil’s leadership
violated these norms, there would be recorded evidence of contradiction between the statements
made at the UNGA by the president and actions in practice. Fortunately for the nonproliferation
regime, Brazil respected its public statements, as it integrated fully into the nonproliferation
regime. Therefore, we find evidence in support of hoop test 7, clear and obvious statements by
Brazil in which Brazil accepts anti-nuclear weapon norms.

275
International Atomic Energy Agency. “Agreement of 13 December 1991 between the republic of Argentina, the
federative republic of Brazil, the Brazilian- Argentine agency for accounting and control of nuclear materials and the
international atomic energy agency for the application of safeguards.” INFCIRC/435 (March 4, 1994). Available
from: https://www.iaea.org/sites/default/files/infcirc435.pdf
148

4) Conclusion
This chapter tested seven related observable implications drawn from the voluntary
reversal pathway. These observable implications were tested via hoop tests to analyze if the
theory can be applied to the case of Brazil. Overall, the results lean towards support for the
theory, although with some important caveats. From the seven observable implications,
observable implications 3 and 5 do not gain support, 6 and 7 gain support, and 1, 2, and 4 gain
partial support. Table 9 at the end of the chapter summarizes the tested hypotheses and their
status after analysis.
From the seven observable implications tested via hoop tests, one of them does not gain
support: Implication 3. The expectation was that following democratization, other states would
use positive diplomatic incentives or shaming mechanisms to pressure Brazil to do a nuclear
reversal. The findings of the case study do not show explicit evidence of such attempts following
the democratization process. However, this is likely because democratic Brazil promptly, and
voluntarily, joined the nuclear non-proliferation regime, which pre-empted the need for other
states to pressure the new government of Collor to join the global regime. An alternative view is
that the restrictions against trade of sensitive nuclear technology for failing to join the NPT,
Treaty of Tlatelolco, and other nuclear treaties might have served as a constant source of
pressure, thus we should interpret the lifting of these restrictions after Brazil joined the nuclear
regime as a diplomatic reward. However, in the absence of explicit quid pro quo statements
towards Brazil after democratization, the hoop test as written cannot be said to gain support.
Two observable implications passed the hoop tests with solid support, Implications 6 and
7. Observable implication 6 states that we should observe that Brazil’s voluntary decision to
terminate their nuclear weapons program occurs after democratization. Although the historical
149

section and analysis of the case reveals that it is unlikely Brazil’s top leadership ever tried to
pursue a nuclear weapon, we do see explicit rejection of the nuclear bomb and even renunciation
of peaceful nuclear explosions after democratization. Observable implication 7 expects that
Brazil should provide reasons related to nuclear nonproliferation norms as justifications for their
nuclear reversal. Evidence supports this expectation; President Collor de Mello made two
speeches in the United Nations in 1990 and 1991 explicitly renouncing any nuclear weapon
ambitions, declaring that no such a thing as a peaceful nuclear explosion is likely. Furthermore,
signing several treaties like the Treaty of Tlatelolco and the NPT can be considered tacit
agreement with international non-proliferation norms, given the context of all the other actions
Brazil undertook to preserve the nuclear non-proliferation regime, such as the bilateral treaties
signed with Argentina and the 1990 Quadripartite Agreement.
The results for the remaining hoop tests, on the other hand, reveal unexpected
observations that do not quite align with the theoretical expectations, although the historical
evidence does not lead me towards concluding that the hoop tests ended in failure. For instance,
observable expectation 1 reads as follows: we should observe that liberal democracies are the
main actors pushing norms against nuclear weapons ownership and proliferation in
international forums such as the United Nations General Assembly. We should observe that these
acts of international pressure are directed towards Brazil after the international community
discovers that Brazil is developing a covert nuclear program. As expected, we do observe that
the majority of nuclear treaties and nuclear arms control agreements were introduced or mainly
supported by liberal democracies. However, we also find that pressures towards Brazil to join the
nuclear nonproliferation regime occur before the establishment of the Brazilian covert nuclear
program in 1979. The evidence, then points towards the international community pressuring
150

Brazil to join the regime following Brazil’s rejection of the NPT, before any serious suspicions
of a nuclear weapons program. As a result, the observable implication is partially supported; the
nonproliferation regime was mainly developed by liberal states (plus the Soviet Union, for global
security reasons), but the international community did not wait until they suspected a nuclear
weapons program existed in Brazil to start international pressures towards Brazil.
Implication 2 also gains partial support. As expected, we do observe that economic and
other types of sanctions are used against Brazil during the military regime. For example, West
Germany refused to sell sensitive technology to Brazil in 1974, ultimately offering jet nuzzle
technology that would be useless for a potential nuclear bomb. Furthermore, the multitude of
international treaties prohibiting countries from trading sensitive nuclear technology to likely
proliferators restricted Brazil’s access to important nuclear technology. On the other hand,
contrary to my expectations, there is at least one documented instance of the United States using
diplomatic positive incentives to try to convince Brazil to join the nuclear regime, when in 1976
Kissinger offered economic aid to Brazil in exchange for joining the NPT, an effort that ended in
failure. Therefore, given the findings, observable implication 2 must be said to have partial
support: no threats of force were ever used against Brazil, but sanctions were applied, yet
diplomatic positive inducements were not completely off the table, either.
Observable implications 4 and 5 were opposites, evidence for one would count as
evidence against the other. Due to Brazil never experiencing a change in its perception of threat,
which seems to have been benign throughout the period of analysis, implication 5 must be
discarded. However, implication 4 does not find full support, either. Implication 4 states: we
should observe that Brazil’s decision to reverse their nuclear program comes about after a
decrease in their perception of threat or rivalry with neighboring countries such as Argentina.
151

Interestingly, and contrary to common misperceptions, Brazil and Argentina do not seem to have
viewed each other as a security threat during the period of analysis. Therefore, it cannot be said
that Brazil’s decision to reverse their nuclear ambitions is a result of decreased threat
perceptions. On the other hand, the rivalry relationship between Argentina and Brazil did
undergo an important qualitative change after both countries resolved their dispute over the
Parana River Basin. Only after both countries resolved that crisis did bilateral cooperation occur
in the nuclear realm. This results leads to the conclusion of partial support; although there was no
significant change in perception of threat, which was consistently low, there was a change in the
rivalry relationship with Argentina before Brazil reversed their nuclear ambition.
Overall, it seems there is significant support for the voluntary nuclear reversals pathway,
given the available evidence. However, some skeptics might wonder if Brazil is an ideal case to
test the theory given the absence of a nuclear weapons program or a decision by the country’s
leadership to pursue the bomb. It must be admitted that after engaging in deep analysis of the
historical record, Brazil cannot be said to have been the ideal case of nuclear reversal. Yet, the
theory also applies in the case of Brazil given the fact that Brazil could have pursued the nuclear
bomb option, that it explicitly rejected several international norms against nuclear weapons
during the military regime, and that the purpose of the nuclear program, and the covert nuclear
program that came about later, was ambiguous in the eyes of the international community.
Therefore, although the decision to pursue the bomb was not made at the highest levels of
government-at least as far as we know-Brazil should still counts as a successful case of nuclear
reversal. First, it was not until democratization that Brazil explicitly embraced nuclear
international norms, leading to voluntary termination of the parallel nuclear program. Second,
although Brazil’s top leadership never made the formal decision to pursue the bomb, a few
152

statements by military leaders and even a few presidents hinted at the possibility of pursuing the
bomb, indicating that nuclear weapons ambitions were never completely off the table, even if
they never made it past the discussion stage. Finally, the covert nuclear program could have led
to the development of a nuclear weapon, had Brazil taken that step; it remains in the
counterfactual realm if Brazil would have gone nuclear if democratization never occurred, for
there were some military officers in the Brazilian military that did have nuclear weapons
ambitions.
In conclusion, although the evidence falsifies the common misconception that Brazil was
pursuing nuclear weapons-although as shown above there were some alarming comments by
military officers regarding pursuit of the bomb-Brazil appears to follow the pathway of nuclear
voluntary reversals closely. The results of the hoop tests, on the other hand, show that some
theoretical revisions might be needed for the voluntary reversals pathway. For example,
countries might not wait to discover the existence of a covert nuclear program before strongly
pressuring a state to join the nonproliferation regime. Suspicion of nuclear ambitions or even
rejection of nuclear treaties might be sufficient to trigger international efforts to pressure a state
to join the nonproliferation regime. Second, positive inducements are also possible towards
authoritarian states, which shows that the voluntary reversals pathway was too absolutist as
originally designed; the theory should not exclude the possibility of using positive reinforcement
towards autocracies to push them towards nuclear reversals. Third, change of rivalry
relationships might have a similar effect as changes in perception of threat. If so, the theory
needs to be revised to account for benign security environments where there are serious rivalry
relationships between neighboring states, yet not so serious as to explode into international crises
or armed conflicts.
153

Table 9: Summary of results: Brazil
Observable Implication Supported by evidence
We should observe that liberal democracies
are the main actors pushing norms against
nuclear weapons ownership and proliferation
in international forums such as the United
Nations General Assembly. We should
observe that these acts of international
pressure are directed towards Brazil after the
international community discovers that Brazil
is developing a covert nuclear program.
Mixed
We should observe that liberal states, in their
attempt to push for nuclear reversal, will rely
on distinct diplomatic tools when trying to
convince Brazil to reverse their program.
When Brazil is authoritarian, we should
observe threats to use force or at minimum
threats of economic sanctions.
Mixed
Following democratization in Brazil, the main
diplomatic tool used by states should be
shaming mechanisms or diplomatic “Carrots”
such as sanction relief or stronger
security/economic ties in exchange for
nuclear reversal.
No
We should observe that Brazil’s decision to
reverse their nuclear weapons program comes
about after a decrease in their perception of
threat or rivalry with neighboring countries
such as Argentina.
Mixed
If Brazil terminates their nuclear program
before Brazil changes its perception of threat,
it should be because other actors promised
credible security guarantees.
No
We should observe that Brazil’s voluntary
decision to terminate their nuclear weapons
program occurs following democratization.
Yes
Brazil should provide reasons related to
nuclear nonproliferation norms when
informing the international community about
their decision to reverse course.
Yes

154

Chapter 7: Israel as A Case of Non-Reversal
This chapter explores a particular case that at first sight seems to contradict the
theoretical expectations of the nuclear reversal theory: the case of Israel. According to the
theoretical framework, democracies should be more likely to internalize or at minimum accept
international nuclear non-proliferation norms, leading to voluntary reversal of nuclear ambitions.
Yet, Israel obtained the nuclear bomb sometime between 1966 or 1967, one year before the NPT
treaty, at a time when non-proliferation norms were in full force. Israel, then, appears to be an
outlier, a democratic country that failed to internalize international non-proliferation norms in the
presence of a fully fledged non-proliferation global movement. Given this mismatch between the
theoretical expectations and the observable outcome of Israel as a nuclear weapons country, it is
essential to analyze Israel using the theory as a guide to truly understand if Israel does contradict
the theory, or if particularities of the case might explain why nuclear reversal failed in this case.
An important point needs justification: why do a case study of Israel using the voluntary
reversal pathway framework rather than the involuntary pathway of nuclear reversal? The
justification for this choice is that Israel, as a democracy throughout its entire history, should be
expected to face international normative pressures to give up pursuit of the bomb rather than
coercive threats or physical force from the international community, as autocracies are expected
to experience. Therefore, Israel’s unique regime characteristics should have led Israel down the
pathway of a voluntary nuclear reversal. The evidence appears to justify this choice, as well as
explain the failure to reverse in line with the theory of voluntary reversals; Israel’s perception of
threat was constantly high. Frequent existential threats from Iran and Arab neighbors ensured
that Israel could not afford to give up the nuclear option, despite international pressures to the
contrary. Upon closer inspection, the case of Israel does not contradict the theoretical framework.
155

Israel’s security threats were so high as to make nuclear reversal unlikely to this very day, even
though Israel was a democracy and despite the existence of an international norm against nuclear
weapon proliferation.
Like the previous two cases, the case study relies on process tracing techniques to test
observable implications drawn from the theoretical framework of voluntary nuclear reversals,
using the historical record and leader statements as evidence to explore why nuclear reversal
failed to materialize in this specific case. The case study is organized as follows: First, the
history of the Israeli nuclear program is reviewed. Second, the chapter analyzes two key
intervening variables that affect a country’s likelihood of accepting nuclear non-proliferation
norms, the level of democracy of Israel and Israel’s perception of threat. Third, the observable
implications of the voluntary pathway are laid out as testable hypotheses, which can then be
tested against the historical evidence of Israel to learn why nuclear reversal did not occur in this
case. Finally, the conclusion summarizes the key findings of the case and implications for the
theory of nuclear reversals.
1) Israel’s nuclear program: 1948-1967
Since its independence in 1948, Israel faced an existential threat from Arab neighbors
that opposed its existence. On midnight of May 14, 1948, the British Mandate for Palestine
terminated, and Israel formally declared independence. The Arab League, a coalition of five
Arab states (Egypt, Iraq, Jordan, Lebanon, and Syria) led an invasion of Israel in opposition to
the two-state partition plan that was supposed to establish Israel as a Jewish state and Palestine as
an Arab state. Israel, despite the numerical disadvantage, prevailed in the conflict.
276
Since

276
For a rich historical account of the 1948 Israel war of independence, see (Morris, 2008).
156

Independence, Israel frequently faced challenges to its legitimacy by its neighbors, a major
incentive for Israel’s wish to obtain the nuclear bomb as a tool for national security.
Israel’s venture in the nuclear realm began early on, following discovery of uranium
deposits in the Negev region in 1950 and the establishment of the Israeli Atomic Energy
Commission (IAEC) in 1952.
277
Two years later, Israel began nuclear cooperation with the
United States, joining Eisenhower’s Atoms for Peace program. Using this program to access
nuclear technology, Israel signed a contract with the U.S. to build a “swimming pool-type”
nuclear reactor as part of the Nahal Soreq Nuclear Centre (MAMAG) in Rehovot.
278
However,
there were limits to technology access for Israel; in August 1955, “the head of the IAEC asked
the United States for sophisticated nuclear technology, which could produce ‘new elements such
as plutonium,’ only to be rebuffed by his American counterpart.”
279
To make matters worse, In
September 1955 Egypt announced a military arms deal with Czechoslovakia, which greatly
improved Egypt’s military power.
280
Following the arms trade declaration, Egypt’s president
Nasser closed the Red Sea Straits of Tiran to Israeli shipping, and supported border attacks
against Israel. Simultaneously, Egypt, backed the Algerian revolution and nationalized the Suez
Canal, harming the economic interests of France and the United Kingdom.
281

Egypt’s actions pushed Israel, the U.K., and France to share a common interest against
Egypt, leading them to sign the Protocol of Sevres. The Protocol of Sevres was a secret
agreement among the three nations to join politically and militarily to liberate the Suez Canal
from Nasser and topple Egypt’s government. On October 29, 1956, Israel invaded the Sinai, and

277
Debs and Monteiro (2016, 221).
278
Gerlini (2010, 145).
279
Debs and Monteiro (2016, 221).
280
Ibid., 221.
281
Bahgat (2006, 116).
157

a week later British and French forces landed around the Suez Canal. The alliance successfully
defeated the Egyptian forces and took control over the canal, but the alliance met opposition
from the Soviet Union, which threatened to use military force against Israel and its allies.
282
The
United States, too, pressured France, the U.K., and Israel to withdraw their forces from Egypt
and end the conflict.
The incident taught Israel that the United States might not be a reliable ally guaranteeing
Israel’s survival. Following Soviet threats, President Eisenhower pledged to retaliate against
Soviet attacks against France or the United Kingdom, but refused to pledge the same security
guarantees to Israel. On the contrary, Eisenhower chastised Israel for its role in the war against
Egypt, warning Israel that it “risked UN opprobrium, Soviet attack, and the termination of all
U.S. aid.”
283
On the other hand, France seemed to be a reliable ally willing to share nuclear
technology and information with Israel. In September 1956, France agreed to aid Israel in the
development of an Israeli nuclear reactor and plutonium reprocessing plant. France’s nuclear
sharing was crucial for the construction of a nuclear reactor at Dimona, in the Negev desert.
Construction of the reactor began in 1958, thanks to French assistance.
Israel attempted to obtain a public security guarantee from the United States, but the
American leadership refused, perhaps to avoid angering their Arab allies and harm American
regional interests. In May 1957, President Eisenhower gave Israel a private security guarantee,
“stating that American policy ‘embraced the preservation of the state of Israel.’”
284
However,
Israel wanted a public security guarantee, which would have a greater deterrence effect against
hostile neighbors threatening Israeli security. For instance, in October 1957 Israeli Prime

282
Ibid.,116.
283
Quoted in Debs and Monteiro (2016, 222).
284
Ibid., 223.
158

Minister Ben-Gurion argued that the United States’ “NATO commitment should be extended to
the Middle East.”
285
Secretary of State John Foster Dulles replied to Israel with a rejection; “we
have made quite clear to the Soviet Union our deep interest in the maintenance of independence
and integrity of all states in the Near East, including Israel.”
286
Evidently, the United States
hesitated in providing a public security guarantee to Israel, which left Israel in uncertainty
regarding America’s commitment to Israeli national security. As a result, Israel felt pushed to
develop a nuclear weapons capability for self-defense purposes, a weapon of last resort should its
existence be threatened.
Israel’s nuclear weapon intentions were made clear at least two times in the late 1950s.
First, on February 1, 1959, Director General of the Defense Ministry Shimon Peres gave a
speech that signaled Israel nuclear ambitions. At a symposium at the Weizmann Institute, Peres
referred to a secret weapon that Israel was trying to obtain. Second, in early 1959 Israel
requested samples of liquid propellant from the UK. According to sources quoted by Levey, a
political scientist, “the missiles that the Israelis were planning were useless without a nuclear
warhead. London turned down the request.”
287
Israel’s developments at Dimona were also
suggestive of true intentions. For example, in 1958 Israel created its armament development
authority, RAFAEL, which was tasked with overseeing nuclear weapon development.
288
In 1959,
Israel acquired twenty-two tons of heavy water from Norway, supposedly for peaceful purposes.
By 1960, Dimona became a complex that “hosted facilities for uranium purification, enrichment,
and conversion, as well as nuclear fuel fabrication and reprocessing.”
289

285
Ibid., 223.
286
Ibid., 223.
287
Gerlini (2010, 146).
288
Debs and Monteiro (2016, 224).
289
Ibid., 224.
159

Suspicions regarding the true purpose of Dimona began in 1960. After taking office,
French president de Gaulle reconsidered French cooperation in the nuclear realm with many
countries, including Israel. De Gaulle demanded that Israel make the Dimona project public and
submit to international inspectors, and that Israel promise the reactor was solely for civilian
purposes rather than weapon production.
290
American intelligence also began to suspect the
military dimension of the nuclear reactor. In December 1960, the CIA concluded that Israel
began a covert nuclear program using the knowledge obtained from France, and built a
laboratory for separation of plutonium at Dimona with the purpose of obtaining a nuclear
arsenal.
291
When questioned about the purpose of Dimona, Israel provided contradictory
information. In the summer of 1960, Israel officials claimed that the nuclear facilities at Dimona
were textile plants. Three months later, they claimed that it was a metallurgical research
installation.
292

In 1961, Israel publicly acknowledged that a nuclear weapon might benefit Israeli
security. In that year, Prime Minister David Ben-Gurion told Canadian Prime Minister John
Diefenbaker that “if the Arab threat continued to grow…Israel might as a matter of self-defense
be required to develop nuclear-weapons capability.”
293
Intelligence reports from the CIA, in
addition to the suggestive statements by Israel’s leaders, made it clear that Israel was in pursuit
of a nuclear weapon capability. On that same year, John F. Kennedy became president of the
United States, replacing Dwight Eisenhower. Kennedy, aware of the proliferation problem posed
by Israel, made it his goal to stop nuclear proliferation of Israel. According to Glenn Seaborg,
chairman of the Atomic Energy Commission under Kennedy, Kennedy’s “private nightmare”

290
Bahgat (2006, 117).
291
Gerlini (2010, 148).
292
Debs and Monteiro (2016, 228).
293
Quoted in Debs and Monteiro (2016, 224).
160

was “global nuclear anarchy, a world of some 20 to 30 nuclear weapon states,” which informed
Kennedy’s nuclear policy making.
294
This major policy goal led to drastic action by the Kennedy
administration in 1963, when the United States threatened Israel with an ultimatum.
Although the Eisenhower administration had attempted regular inspections of the Dimona
site since 1960, Israel Prime Minister David Ben Gurion refused a regular schedule of
inspections. As a result, in May 1963, President Kennedy sent a letter to Ben Gurion, stating that
the United States’ “deep commitment to the security of Israel” could be “seriously jeopardized if
this Government was unable to obtain reliable information on a subject so vital to peace as
Israel’s efforts in the nuclear field.”
295
After Ben Gurion’s resignation from office, Levi Eshkol
replaced him as Prime Minister of Israel, who finally agreed by July 1963 to semiannual
inspections of Dimona. At first sight, this seemed like a victory for the United States; Israel
finally opened its nuclear facilities to American inspectors. However, even the Eshkol
administration placed obstacles in the way of American inspectors.
From 1964 to 1969, Eshkol successfully postponed various scheduled inspections from
the United States. The excuses for postponement differed; “citing political problems connected
with upcoming elections, radiation hazards in the plant, and even his own recovery from an eye
operation.”
296
These delays caused the inspections to occur only once a year between 1964 and
1969. Barriers were also present when inspections actually took place. Inspectors had to
announce they were visiting the facilities in advance, and were limited to a single day of
inspections. These barriers to successful inspections allowed the Israelis to hide the most

294
Cohen (2005, 171).
295
Engelhardt (2004, 57).
296
Ibid., 57.
161

sensitive aspects of the Dimona plant and ensure the covert nature of their nuclear weapons
program.
297

Israel successfully protected their facilities from American inspectors, which allowed
great progress in their program. The first reactor at Dimona reached critical capacity in 1963,
producing the plutonium that would later be used for the Israeli nuclear bomb. That same year,
Israel purchased nearly 100 tons of unsafeguarded uranium from Argentina.
298
In 1965, Israel’s
uranium mine allowed the country to be self-sufficient and stop imports of foreign uranium for
the nuclear plants. By January 1964, a CIA memo concluded that “Israel now had the capability
‘to redirect the program in the future toward achievement of a small nuclear weapons capability,
should they so decide.’”
299
By early 1965, a State Department memorandum concluded that the
decision to obtain nuclear weapons was already reached, and that Israel would be able to produce
nuclear weapons within a few years.
300
By early 1966, Israel was on the verge of becoming
nuclear weapons capable, to the dismay of American inspectors and policy makers.
Although Israel was on the verge of nuclear weapons capability, Israel initially hesitated
to cross the threshold into nuclear weapons ownership. According to Cohen, Prime Minister Levi
Eshkol ruled out nuclear tests, for three major reasons. First, a nuclear test would be a violation
of Israel’s “non-introduction” commitment, a pledge made in 1962 by former Prime Minister
David Ben-Gurion that Israel would not be the first to introduce nuclear weapons in the Middle
East. Second, Eshkol was aware that the world superpowers were negotiating a global treaty to
prevent proliferation of nuclear weapons. If Israel tested a nuclear weapon, it could be perceived

297
Ibid., 57.
298
Debs and Monteiro (2016, 225).
299
Ibid., 225.
300
Ibid., 225.
162

as a challenge to the international community and harm U.S.-Israel relations. Finally, it might not
have been in Israel’s best interest to demonstrate possession of a nuclear weapon, since it could
kickstart nuclear proliferation in the region.
301

Ultimately, this hesitation had to be discarded a month before the Six-Day War started on
June 5, 1967. In May, Israel perceived that the security situation was worsening and that a
conflict with Egypt was inevitable. As a result, in preparation for a worst-case scenario, “Israeli
teams assembled virtually all the components, including the handful of nuclear cores it had, into
improvised but operational explosive devices.”
302
Now armed with a nuclear weapons capability,
Israel carried out a pre-emptive airstrike on Egyptian aircrafts and an invasion of the Sinai
Peninsula. Egypt responded with an alliance with Jordan and Syria, which assisted Egypt in its
war against the Israeli military. Israel was victorious in the conflict and had now crossed the
threshold into a nuclear weapons state.
Alternative accounts state that Israel crossed the nuclear weapons threshold earlier, in
1966, when Israel supposedly carried out a “test of ‘special significance,’ which could have been
the detonation of an ultralow yield, subcritical nuclear device, or the test of a nuclear prototype
without fissile material.”
303
Regardless of the exact date, by 1968 President Johnson learned
about Israel’s nuclear status, and by the time Richard Nixon took office in 1969, the American
inspections regime terminated.
304
According to Cohen, in September 1969 Israel Prime Minister
Golda Meir reached a secret understanding with Nixon. Israel would not give up its nuclear
capability nor sign the NPT, but in exchange Israel would not publicly declare itself a nuclear

301
Cohen (2007, 13).
302
Ibid., 15.
303
At least according to Munya M. Mardor’s Rafael (1981). Cited in Debs and Monteiro (2016, 226).
304
Ibid., 226.
163

weapons power nor test a nuclear bomb. Israel, then, would keep its bomb “in the basement” and
keep the weapon as a tool of last resort.
305
On July 18, 1970, The New York Times publicly
declared that the United States intelligence community considered Israel a nuclear weapons
country, despite Israel’s refusal to confirm its nuclear status.
306
Israel successfully crossed the
threshold into nuclear weapons status, although it adopted a policy of “opacity” in which Israel
formally refused to confirm or deny that it had nuclear weapons. Thus, Israel became the 6
th

country to obtain the nuclear bomb.
2) Israel’s regime type and perception of threat
Before analyzing the observable implications from the voluntary reversal pathway and
how they might apply to the case of Israel, it is essential to discuss the two key intervening
variables thought to have the most influence on a country’s likelihood of accepting international
non-proliferation norms: perception of threat and regime type. As is evident, Israel can be
classified as a democratic state throughout the period of analysis. The theoretical expectation is
that a democratic state will be more likely to accept international norms against nuclear
proliferation if the state wants to be accepted and perceived as a law-abiding member of the
liberal international order. On the other hand, Israel faced a threatening security environment
since independence, with various neighboring states posing an existential threat. The theoretical
expectation is that a country perceiving security threats will be unlikely to accept international
non-proliferation norms, since the nuclear bomb could serve as a tool of deterrence. In sum, it
seems that the two key intervening variables pull in opposite directions.

305
Cohen (2005, 172).
306
Smith, Hedrick. “U.S. Assumes the Israelis Have A-Bomb or Its Parts.” The New York Times, July 18, 1970.
Accessed November 17, 2022. https://www.nytimes.com/1970/07/18/archives/us-assumes-the-israelis-have-abomb-
or-its-parts-us-assumes-israelis.html
164

How dangerous was Israel’s security environment exactly? From 1948-1970, within a
span of 2 years, four major wars occurred between Israel and neighboring Arab states, which
indicates that Israel’s perception of security was low. The four major wars that Israel faced
between 1948 and 1970, and that greatly influenced Israel’s need to obtain nuclear weapons, are
the 1948 War of Independence, the 1956 Suez Crisis, the 1967 Six Day War, and the 1969-1970
War of Attrition. In the War of Independence (or Arab-Israeli War), Israel successfully fought
off a coalition of Arab states consisting of Egypt, Syria, Lebanon, Iraq, and Saudi forces fighting
under Egyptian command –Arab states that were opposed to the creation of a Jewish State. In the
1956 Suez Crisis, Israel, together with France and the UK, fought against Egypt to take back
control of the Suez Canal nationalized by Nasser; a clear war guided by both Israeli fear of
Egyptian power and Western European economic interests. In the 1967 Six-Day War, Israel
fought against Egypt, Syria, and Jordan, after Israel initiated hostilities by destroying Egypt’s
entire air force in what Israel considered a pre-emptive strike against a potential aggressor.
307

Finally, in the War of Attrition, Egypt began a lengthy military campaign to force Israel to
withdraw from the Sinai, to no avail. In sum, Israel faced four major conflicts that either
threatened Israel’s existence or crucial foreign policy interests, creating the right conditions for
Israel to seek nuclear weapons as a deterrence against hostile neighbors or to help Israel project
power in the region.
Further exacerbating Israel’s insecurity, the United States refused to provide a public
security guarantee to Israel. The Eisenhower administration perceived Arab states as important
regional partners, and so the United States could not show excessive favoritism towards Israel.

307
Whether Israel or the Egyptian coalition should be blamed for initiating the war is a controversial debate among
many scholars. For a perspective that blames the Egyptian coalition due to closure of the Suez Canal, see Karsh
(2017). For an opposing view that blames Israel, see Curtiss (1992).
165

This message was relayed to Israel Ambassador Abba Eban by Assistant Secretary Henry
Byroade in June 1953, when Byroade said “[i]f only Israel and Arab relations were concerned,
the problem would be relatively easier. What is at stake, however, is the possible loss of all
Western influence in the Middle East…the decline in influence has resulted…from Western
support of Israel, and we are quite concerned about it.”
308
In 1960, Prime Minister Ben-Gurion
asked the United States for increased military aid, a request that was denied by President
Eisenhower, who “questioned the desirability of the U.S. becoming the arsenal for Israel.”
309
The
United States was willing to provide a private security guarantee to Israel, such as the 1957
Eisenhower promise to Ben-Gurion that the United States was invested in the preservation of
Israel.
310
However, in the absence of a public statement, such a security guarantee was limited
and had no deterrence effect against Israel’s enemies.
Regarding regime type, Israel stands out for being the only Middle Eastern country with a
continuous history of democratic government since its independence in 1948. Israel functions as
a parliamentary democracy; the executive power gains its legitimacy from the support of the
legislative branch of government. In Israel, the legislative branch is represented by a unicameral
body known as the Knesset, whose membership is elected democratically by the people.
Consisting of 120 seats, the Knesset is elected via a proportional representation list system,
encouraging competition between various political parties. To date, no political party has ever
won a majority of votes in elections.
311
Although Israel has a President, the position is mostly
ceremonial. The executive power lies with the Prime Minister, who is chosen by a coalition of

308
Quoted in Debs and Monteiro (2016, 227).
309
Ibid., 227.
310
Ibid., 223.
311
For a detailed description of Israel’s form of government, see Lihi Ben Shitrit’s chapter 13 in Lust (2020).
166

the Knesset and ceremoniously approved by the President. The Prime Minister serves for a four-
year term, with possibility for indefinite renewal.
Despite being classified as a democracy, controversy exists regarding treatment of Arab
citizens and the Palestinian population living under the jurisdiction of the Palestinian Authority
in the West Bank, which to some scholars calls into question the democratic tenet of equality
under the law. For instance, Smooha argues that Israel should be considered an “ethnic
democracy” rather than a “liberal democracy” due to the Jewish population holding absolute
power in the state, while granting rights to the minority Arab Palestinian population.
312
In the
early days of Israel, Arab citizens were subject to military rule and had limited democratic rights
until 1966. In addition, it was common for Arabs to experience land confiscations until the
1970s.
313
In the contemporary era, Arab citizens are considered full citizens de jure, but still live
in a country “whose symbols, flag, and national anthem are Zionist and give little expression to
their Palestinian identity and heritage.”
314
In addition, Arab citizens are not called to serve in the
military, as all Jewish citizens are asked to do, and although comprising 20% of the population as
of 2020, over 50% of those living under the poverty line are Arab citizens.
315
In effect, the de
facto two-tiered society of Israel complicates its claims to being a full-fledged democracy, and
serves as another major reason for hostile relations between Arab states and Israel.
In sum, Israel can be considered a democratic state, although with a complex and
problematic relationship with its Arab population. At the same time, Israel suffers from a
threatening security environment, whose existence has been questioned and challenged by its

312
Smooha (2002).
313
Lust (2020, 447).
314
Ibid., 447.
315
Ibid., 447.
167

neighbors at various points in its short history. The two intervening variables of perception of
threat and regime type are pointing in opposite theoretical directions—democracy should have
made Israel more likely to accept international norms against nuclear proliferation. At the same
time, a high perception of threat should lower the likelihood of successful norm acceptance and
voluntary nuclear reversal. As will be shown in the next section testing the observable
implications of the voluntary pathway on Israel, the security threats faced by Israel superseded
any effects that regime type might have had on successful norm internalization and ensuing
nuclear reversal.
3) Testing observable implications
Israel appears to be a case whose intervening variables suggest opposing outcomes for
the nuclear program. Being a democracy, Israel should be a candidate for voluntary nuclear
reversal, especially if Israel attempted to be considered part of the in-group of liberal states allied
with the West. On the other hand, Israel real security threats should make Israel unlikely to
accept or internalize international norms against nuclear weapon proliferation, since the nuclear
bomb would ensure state survival. As the historical section shows, Israel did obtain the nuclear
bomb, indicating that security concerns might have been stronger than the effects of regime type
on nuclear reversal.
In the following pages, the observable implications derived from the voluntary reversal
pathway will be tested for the case of Israel, to determine if the historical facts of the case
contradict the theoretical framework. Like the case studies of Iraq and Brazil, the observable
implications will be tested via specific hoop tests against historical data such as leader statement,
actions of states, discussions at the United Nations (if any), and historical events. If a hoop test is
successfully passed, then it indicates stronger support for the theory, while failing to pass a hoop
168

test weakens our confidence on the theory. Once again, it must be reiterated that passing all hoop
test should not be interpreted as evidence that the theory is proven, only that it is supported in
this specific case. Future case studies might reveal further support for the theoretical framework
or show that the theory does not hold in particular historical contexts. Readers at this point might
ask, what is the point of the case study, if not generalizability? The point of this chapter is to test
whether the voluntary reversal pathway can also explain cases of non-reversal, such as Israel.
Support for the observable implications might indicate that the theory could explain other cases
of nuclear non-reversals as well. Below are the observable implications that will be tested:
1) Israel’s program should begin if it perceived a major security threat from its
neighbors, and such threat should be constant throughout the nuclear weapons
program development period.
2) If the voluntary reversals pathway is correct, Israel should have been pressured to
terminate its program as soon as the international community learned that Israel was
building the bomb.
3) International pressure to conform to the anti-nuclear nonproliferation norm should
include shaming mechanisms and/or diplomatic “carrots” such as security guarantees,
instead of economic sanctions or threats of force.
4) Israel’s failure to conform to the norm must be justified in terms of security concerns:
Israel must justify its creation of a nuclear bomb based on security concerns rather
than rejection of nuclear norms themselves
316
.

316
It is important to note that Israel never publicly admitted that they built the bomb. Evidence of Israel’s defense
for the bomb must come from private correspondence or declassified material.
169

The four observable implications are listed above, tested via hoop tests, for which
evidence will be examined and analyzed in numerical order. Unlike the previous case studies, the
focus is on why Israel failed to give up pursuit of the nuclear weapon, and whether the
international community attempted to dissuade Israel from reaching nuclear capability. After
analyzing the four observable implications and whether they pass the hoop tests, the concluding
section will review the results of the case study.
Observable implication 1: Israel’s program should begin if it perceived a major security
threat from its neighbors, and such threat should be constant throughout the nuclear weapons
program development period.
The historical evidence supports the argument that Israel’s main drive for obtaining
nuclear weapons was the constant existential threat posed by hostile neighbors. Since its
independence in 1948, Israel faced military threats from Arab countries that opposed its
existence. In 1948, immediately after declaring independence on May 14, 1948, five Arab
nations launched an invasion of Israel. The reason for the invasion was opposition to the United
Nations Partition Resolution, which would have divided the Palestinian Mandate of Great Britain
into two states, a new Arab State and a Jewish state. The goal of the Arab states was to stop
Partition from taking effect and prevent the establishment of a Jewish State in the region.
317

Despite a coalition formed by Lebanon, Syria, Egypt, Saudi Arabia, and Iraq, Israel successfully
pushed back the aggressors and secured independence, in addition to territory originally granted
to the Palestinians by the planned Partition Resolution. Israel’s national security was tested once

317
Department of State, Office of the Historian. “The Arab-Israeli War of 1948.” Office of the Historian.
https://history.state.gov/milestones/1945-1952/arab-israeli-war. Accessed November 21, 2022.
170

again in the months leading up to the 1956 Suez Crisis, when Egypt’s rhetoric and military arms
buildup signaled that a potential attack from Egypt towards Israel was on the horizon.
As stated in the historical section, Egypt nationalized the Suez Canal, threatening the
economic interests of France and the United Kingdom. In 1955, Egypt announced an arms deal
with Czechoslovakia, which would have changed the military balance of power in the Middle
East in favor of Egypt. Israel, threatened by the Egyptian arms build-up, allied with France and
the U.K. to coordinate an attack on Egypt on 1956 and force the Egyptians to open the Suez
Canal. Although the Israeli alliance was militarily successful, American pressures and Soviet
threats of escalation forced the alliance to withdraw from Egypt, leading to an Egyptian political
victory.
318

The Suez Crisis showed Israel that the United States might not be such a reliable ally.
President Eisenhower, responding to Soviet threats, “pledged to retaliate against Soviet attacks
on London or Paris, while withholding similar pledges in the eventuality of a Soviet attack on
Israel and singling out the Jewish state for sharp criticism, warning that it ‘risked UN
opprobrium, Soviet attack, and the termination of all U.S. aid.’”
319
Following the Suez crisis,
Israel sought a public security guarantee from the United States, but every attempt was met with
dismissal. In the absence of a public security guarantee, Israel was forced to depend on its own
military capability for survival. As part of this strategy for survival, the nuclear bomb began to
be perceived as essential. For instance, in 1961, Israel’s Prime Minister David Ben-Gurion told
the Canadian Prime Minister that “if the Arab threat continued to grow…Israel might as a matter
of self-defense be required to develop nuclear-weapons capability.”
320
On April 1965, Israel

318
Bahgat (2006, 116).
319
Debs and Monteiro (2016, 222).
320
Quoted in Debs and Monteiro (2016, 224).
171

connected their pursuit of the bomb with the lack of a public security guarantee. In April, while
denying the nuclear program, Prime Minister Eshkol told Ambassador Walworth Barbour that
Israel “could not forever foreswear nuclear weapon development in absence (of) binding security
guarantees.”
321

The United States was aware that a security guarantee might have deterred Israel from
pursuing the nuclear bomb. A 1963 Near East Affairs Bureau report, part of the State
Department, stated that “if the U.S. were willing to give such assurances, it might be able to
present the Israelis with a fairly good case for a non-nuclear Middle East…”
322
However, the
United States still hesitated to offer such a security guarantee, due to fear of breaking relations
with the Arab countries and losing influence over them to the Soviet Union. A May 28, 1963
conversation between Shaul Bar Haim of the Israeli Embassy and Rodger Davies of the Near
East Bureau made this fear explicit. Davies told Haim that a guarantee to Israel would “cause the
Arabs to turn against the United States even at their own expense” and that “we [the U.S.] would
lose the capability to use our influence to moderate the hostility of Arab policies toward
Israel.”
323

In the absence of an American public security guarantee and the presence of security
threats from neighbors, Israel’s decision to covertly pursue the nuclear bomb as a tool of last
resort is no longer surprising. This pursuit of the bomb led Israel into crossing the nuclear
weapons threshold in 1967, in the context of the May 1967 crisis and the ensuing six-day war
with Egypt, Jordan, Syria, Iraq, and Saudi Arabia. The six-day war was initiated by an Israeli
pre-emptive air strike against Egyptian combat aircraft, carried out in advance of a suspected

321
Quoted in Engelhardt (2004, 61).
322
Ibid., 61.
323
Ibid., 61.
172

Egyptian plan to attack Israel. In the few weeks between the May crisis and the six-day war, it is
believed that Israel quickly assembled the components for a nuclear device that might have come
into use if Israel’s national existence was endangered.
324
By the end of the war, Israel is believed
by the intelligence community to have developed nuclear weapons, becoming the sixth country
in the world to own nuclear weapons. The six-day war concluded with a significant victory for
Israel, doubling its territory after taking over the Sinai, Gaza, the West Bank, East Jerusalem, and
the Golan Heights territories from Arab neighbors.
325

Observable implication 2: If the voluntary reversals pathway is correct, Israel should
have been pressured to terminate its program as soon as the international community learned
that Israel was building the bomb.
Per the theoretical expectations of the voluntary reversal pathway, other states should
pressure nuclear weapons pursuers to terminate their nuclear weapons programs and hold back
their nuclear ambitions. The historical evidence suggests that pressures to terminate nuclear
weapon pursuit did occur, but these efforts were unsuccessful in deterring Israel from obtaining
the bomb. These nonproliferation efforts were strongest in the 1960s when various Western
states began to suspect that the Dimona site was a weapons facility rather than a civil nuclear
power plant.
Suspicions from the United States began in 1958, when the Dimona site was suspected of
being a future nuclear military site. On December 9, 1960, a day after a National Security
Council Meeting in which the American president Eisenhower discussed the Dimona site with

324
Cohen (2007, 15).
325
For a short summary of the six day war, see “The Six Day War” from PBS,
https://www.pbs.org/wgbh/americanexperience/features/hijacked-wars-threats-responses/ Accessed November 24,
2022.
173

his cabinet, U.S. Secretary of State Christian Herter spoke with Israel’s ambassador Avraham
Harman for clarification about the Dimona site. According to a telegram sent to the American
Embassy in Tel Aviv to Ambassador Ogden R. Reid, the discussion that Herter had with Harman
revealed a discrepancy between the Israeli description of the Dimona site and the photographic
evidence of the site gathered by American intelligence.
326

Finally, on December 21, Israel Prime Minister Ben-Gurion publicly clarified what the
purpose of Dimona was, which was already suspected of being a nuclear weapons site by the
international community following leak reports by the London Daily Express-which reported
Israel had built their first experimental bomb-and the Times of London-which reported a non-
NATO country was building nuclear weapons.
327
According to Ben-Gurion, the Dimona reactor
would be used to train staff for a future power reactor that would be built in 10 to 15 years. The
public statement reassured some international observers, but the United States wanted further
guarantees that the Dimona site was for peaceful purposes. By January 1961, Israel stated that
there would be an opportunity for American visits to the Dimona site once the Prime Minister
chose a good time.
328

Yet, despite the promise of allowing inspections, Israel refused to allow frequent access
to the Dimona site. On May of 1963, U.S. President Kennedy sent a letter to Ben-Gurion, stating
that the relationship with the United States would be damaged if Israel did not accept terms for
inspections of the Dimona site. In July 1963, newly chosen Israeli Prime Minister Eshkol agreed
to U.S. semiannual inspections of the plant by American scientists; the first inspection occurred

326
Gerlini (2010, 149).
327
Ibid., 149.
328
Ibid., 154.
174

in January 1964.
329
However, the Dimona plant was only inspected once a year between 1964
and 1969, due to successful Israeli efforts to block the frequency of inspector visits. Eshkol
“several times succeeded in postponing them, citing political problems connected with upcoming
elections, radiation hazards in the plant, and even his own recovery from an eye operation.”
330

This lack of access allowed the Israelis to reprocess reactor waste for plutonium at a secret
location in Dimona while simultaneously avoiding discovery by American scientists sent to
inspect the site.
France, too, pressured Israel to reveal more information about Dimona to the public after
the election of President de Gaulle. De Gaulle reconsidered his relationship in the nuclear realm
with various states, including Israel, and pressured Israel to make the project public and
demanded that Israel submit to international inspections. France also demanded that the Dimona
site be used exclusively for civilian purposes rather than military.
331
De Gaulle, unhappy with
how Israel ignored requests to open up the facilities at Dimona to international inspectors, ruled
out further nuclear assistance to Israel in 1965. French Diplomats stated that “De Gaulle was
‘anxious’ to restore the French position in the Arab world and ‘he would not think of helping the
Israelis to realise their atomic ambitions.’”
332

However, these pressures were insufficient to stop Israel’s covert nuclear program.
France halted its assistance to Israel’s nuclear program in 1960, but by that period Israel had a
professional nuclear establishment capable of independently managing nuclear facilities and
even producing a nuclear bomb. The U.S. inspections efforts were weak as well, perhaps because

329
Engelhardt (2004, 57).
330
Ibid., 57.
331
Bahgat (2006, 117).
332
Rabinowitz (2019, 712).
175

the United States did not want to harm its relationship with Israel. For instance, in 1969 Atomic
Energy Commission (AEC) officials, after reviewing the history of their inspection efforts, said
that “from a number of sources, the team has drawn the inference that the US government is not
prepared to support a real ‘inspection’ effort in which the team members can feel authorized to
ask directly pertinent questions and/or insist on being allowed to look at records, logs, materials,
and the like.”
333
Although a few Western powers aware of Israel’s nuclear progress pressured
Israel, their efforts were relatively weak, allowing Israel to secretly develop the capability to
build atomic weapons around the time of the six-day war.
Observable implication 3: International pressure to conform to the anti-nuclear
nonproliferation norm should include shaming mechanisms and/or diplomatic “carrots” such as
security guarantees, instead of economic sanctions or threats of force.
Israel is a curious case of failed non-proliferation efforts, partly because these efforts
were half-hearted as seen above, but also because no country suspicious of Israel’s nuclear
program considered moving beyond shaming mechanisms in their attempts to push Israel
towards reversal. As demonstrated already, public security guarantees were never offered to
Israel, incentivizing the latter to pursue the bomb for security. Threats of force were also absent,
unsurprisingly given Israel’s close relationship with the United States, France, and the Western
Liberal bloc. Economic sanctions were not discussed either, perhaps because those countries
suspicious of Israel’s nuclear ambitions were cognizant of the existential threats faced by the
Israeli government. The main strategy by France and the United States, as shown above, was to
stop support for the nuclear program and biannual inspections of Dimona, respectively.

333
Quoted in Debs and Monteiro (2016, 230).
176

Apart from the aforementioned non-proliferation strategies, alternative means of
pressuring Israel were employed by other international actors, although they were not shaming
mechanisms as commonly understood. Rather, other countries that suspected Israel’s nuclear
program refused to collaborate with Israel unless Israel embraced the nascent global non-
proliferation regime and its rules. Two important examples of such a strategy at play are Canada
and the United Kingdom, who reacted in similar ways to Israel’s nuclear program.
In February 1961, the United Kingdom gathered information on Dimona, which it shared
with Canada. Included in the gathered information was a questionnaire presented to Israel.
According to the gathered information, the director general of Israel’s Foreign Ministry, Chaim
Yahil, said that “the Israeli government has no intentions of producing atomic weapons. It could
not be expected that any government would give firmer assurances. Circumstances may change
and meanwhile Israel is surrounded by hostile neighbors.”
334
This statement left open the
possibility of developing a nuclear weapons program, concerning both British and Canadian
policymakers. Around May and June of 1961, the United States, Canada, and the United
Kingdom engaged in talks with Prime Minister Ben-Gurion, requesting assurances that the
Dimona site was for peaceful purposes. In all instances, Israel reassured Western countries that
the program was for civil purposes. However, The United Kingdom remained unconvinced and
closely followed developments at Dimona after 1961.
Canada lost some interest in following Dimona’s development, but the United Kingdom
did not. In September 1963, the United States informed the British that the U.S. was going to
carry out the third inspection of the Dimona site, before the reactor comes into operation. The

334
Quoted in Rabinowitz (2019, 710).
177

American conclusion after the inspection was worrying; the United States concluded that they
were only “90 percent certain that nothing suspicious was going on.”
335
After an assessment
exchange between British and Canadian intelligence, one Canadian analyst at the Canadian
Directorate of Scientific Intelligence predicted that Israel “could conduct ‘one or two
underground tests of low yield’ with relatively ‘little risk of detection’ by late 1966.”
336
These
concerns regarding Israel’s nuclear intentions heavily informed Canada and the United
Kingdom’s unwillingness to sell nuclear technology to Israel.
For instance, from 1956 to 1966, Canada signed contracts with India for the export of
heavy water reactors capable of producing weapons grade plutonium. Israel was interested in
obtaining the same reactors from Canada, but this request was denied. In June 1964, Collins,
under-secretary of External Affairs, informed the Canadian ambassador to Israel, Arthur
Andrew, that the Canadian Atomic Energy Advisory Panel concluded that “it was politically
undesirable for Canada to entertain negotiations for the sale of reactors in that part of the
world.”
337
In 1966, Canada almost forbids selling devices that measure exposure to radiation,
presumably because Canada suspected why Israel requested such an item. Ultimately it was
decided that exporting the devices would be allowed, since they could be used for legitimate
medical purposes.
338

The United Kingdom also questioned its own policy of sensitive nuclear technology
transfers to Israel. For example, a 1967 Standard Ministerial guideline from the UK stated that
Britain should under “no circumstances supply material or equipment to either Israel or Arab

335
Ibid., 711.
336
Quoted in Ibid., 712.
337
Ibid., 714.
338
Ibid., 714.
178

states, ‘which could contribute to the development of a nuclear capability.’”
339
The United
Kingdom’s strongest signal of opposition to Israel’s nuclear weapons program occurred in 1973,
five years after Israel obtained a nuclear weapons capability. Israel sought nuclear reactors from
the UK, but the British nuclear establishment refused to collaborate with their Israeli counterpart.
According to Barbra Maclean of the United Kingdom Atomic Energy Agency (UKAEA),
exporting a reactor, even a safeguarded reactor that would be allowed under the terms of the
NPT treaty, “would carry an element of risk” because it could contribute to “the development of
Israel’s nuclear capacity.”
340
Although the Non-proliferation Treaty allowed nuclear exports to
non-member states, the United Kingdom treated non-membership as a red line to exclude nuclear
technology exports to Israel.
In sum, both Canada and the United Kingdom refused sensitive technological exports to
Israel, and sometimes even technology that was not usually considered sensitive, due to concerns
with Israel’s nuclear ambitions. Although these actions are not shaming techniques, they still
serve as soft ways of pressuring Israel to adhere to the international non-proliferation regime.
Although the efforts were ultimately unsuccessful, they demonstrate that other actors apart from
France and the United States acted independently to pressure Israel. Perhaps unsurprisingly,
there is no public shaming of Israel during the period under analysis, perhaps because the
Western powers respected Israel’s policy of opacity to maintain public ambiguity about the
status of their nuclear programs and avoid a Middle East nuclear arms race with neighboring
Arab States and Iran.

339
Ibid., 718.
340
Quoted in Ibid., 718.
179

Observable implication 4: Israel’s failure to conform to the norm must be justified in
terms of security concerns: Israel must justify its creation of a nuclear bomb based on security
concerns rather than rejection of nuclear norms themselves.
Evidence used to test the other observable implications makes it obvious that support for
the fourth implication exists. Israel’s motives for pursuing the nuclear bomb were driven by
security concerns rather than rejection of international non-proliferation norms themselves. In
particular, existential fear of hostile neighbors, and absence of a concrete American security
guarantee, incentivized Israel to successfully develop the nuclear bomb.
341
Israel’s own
statements support this assertion of the true motives behind the Israeli nuclear bomb.
At various points in its history, Israel made it clear that they were pursuing a public
security guarantee from the United States, considering the security threat posed by neighboring
states. In October 1957, Prime Minister Ben-Gurion requested military hardware from the United
States and “argued that Washington’s ‘NATO commitment should be extended to the Middle
East.’”
342
U.S. Secretary of State John Foster Dulles declined, arguing that it was already clear
that the United States was interested in maintaining independence of Middle Eastern countries,
including Israel. In 1961, Ben-Gurion told the Canadian Prime Minister that “if the Arab threat
continued to grow…Israel might as a matter of self-defense be required to develop nuclear-
weapons capability.”
343
On April 23, 1965, Israel’s Prime Minister Eshkol told U.S. Ambassador
Walworth Barbour that Israel “could not forever foreswear nuclear weapon development in
absence of binding security guarantees.”
344
The message was always the same: Israel needed

341
This is a conclusion in line with the broader literature on Israel’s nuclear program.
342
Debs and Monteiro (2016, 223).
343
Ibid., 224.
344
Quoted in Engelhardt (2004, 61).
180

assurance that its national security would not be threatened. Yet, the United States refused to
commit to a public security guarantee, despite Israel’s frequent requests.
Statements by Israel’s leaders suggests that Israel did not oppose non-proliferation norms.
In fact, Israel was aware that the international community was adopting an international norm
against nuclear weapon proliferation, leading Israel to develop a policy of concealment of their
program to claim publicly that they weren’t breaking the norm. For instance, in 1962 Ben-Gurion
pledged that Israel would not be the first country to introduce nuclear weapons in the Middle
East, giving birth to the “nonintroduction” commitment.
345
This policy was held by future Prime
Ministers, who were aware that the international community would not accept a nuclear Israel.
At one-point, Prime Minister Levi Eshkol ruled out conducting nuclear tests, telling some of his
inner circle “do you think that the world would congratulate us for our achievement?”
346
At least
before the 1967 Six-Day war, Israel had an interest in preventing nuclear proliferation in the
Middle East, knowing that the introduction of the nuclear bomb might drive other states in the
region to pursue the nuclear weapon as well. Although Israel built nuclear devices as a last resort
in 1967, Israel still maintained its “nonintroduction” commitment in public. In 1969, Israel Prime
Minister Golda Meir pledged to U.S. president Nixon at a secret meeting that Israel would not
become a declared nuclear power. This meant that “Israel would not test, would not declare itself
a nuclear weapon state, and would not use its nuclear capability for diplomatic gains.”
347
In other
words, Israel adopted a policy of nuclear opacity, maintaining its nuclear weapons capability
“secret” and thus support non-introduction of the bomb in the Middle East.

345
Cohen (2007, 13).
346
Ibid., 13.
347
Cohen (2005, 172).
181

In sum, the historical record supports the conventional wisdom that Israel obtained a
nuclear weapon capability for security purposes, not as repudiation of international norms. In
fact, in practice Israel appears to support international non-proliferation norms to some extent,
evidenced by the fact Israel still adheres to the non-introduction principle to avoid a nuclear arms
race in the Middle East. Although Israel has obtained the bomb, the promise to Nixon by Prime
Minister Golda Meir has been kept by future Prime Ministers, for Israel still refuses to admit the
existence of their nuclear weapons to the present day.
4) Conclusion
In this chapter, four observable implications of the voluntary reversal pathway were
tested for the case of Israel via hoop tests. Evidence was found to support three out of four
observable implications, which strengthens confidence on the theory. Table 10 at the end of the
chapter summarizes the results for each of the observable implications/hypotheses. As expected,
and not surprisingly, hoop test 1 was passed easily. Observable implication 1 reads as follows:
Israel’s program should begin if it perceived a major security threat from its neighbors, and
such threat should be constant throughout the nuclear weapons program development period. As
was observed, Israel faced a threatening security environment since independence. Within a short
period of time, Israel had three major wars with Arab neighbors: The 1948 war of independence,
1956 Suez Crisis, and the 1967 Six-Day War. The existential threats faced by Israel served as a
major incentive for Israel to develop a nuclear weapons program in the absence of public
security guarantees by Western allies.
Evidence is also found to support the second observable implication, which reads as
follows: if the voluntary reversals pathway is correct, Israel should have been pressured to
terminate its program as soon as the international community learned that Israel was building
182

the bomb. The historical evidence shows that these international pressures existed as soon as
Western countries suspected the Dimona nuclear site was a weapons facility rather than a civil
nuclear power plant. During the 60s, the United States began annual inspections of the Dimona
facility. Furthermore, France under President de Gaulle stopped collaboration with Israel after
Israel’s refusal to allow international inspectors inside the facility. However, despite these
pressures, Israel still managed to build a nuclear arsenal. A common explanation for the failure
of these pressure efforts is that they were weak to begin with; the United States, France, the
United Kingdom, and other states that knew about the program did not want to push for nuclear
reversal too strongly, perhaps because Israel’s security environment was bleak and it was
understood that Israel needed to improve its own security.
The evidence is weak for the third hoop test: international pressure to conform to the
anti-nuclear nonproliferation norm should include shaming mechanisms and/or diplomatic
“carrots” such as security guarantees, instead of economic sanctions or threats of force.
Contrary to what was expected, there were no strong diplomatic incentives for Israel to give up
pursuit of the bomb. No shaming mechanisms took place at the United Nations, perhaps because
those countries aware of the nuclear program did not want to reveal Israel’s secret. Canada and
the United Kingdom’s strategy for incentivizing Israel to join the non-proliferation regime was a
soft strategy, in which both countries refused to trade sensitive nuclear technology to Israel
unless Israel guaranteed its nuclear program was for peaceful purposes. However, these trade
restrictions were informal rather than actual economic sanctions, which once again limited their
coercive capability. From the four observable implications, the third fails to find support in the
historical record. Other means of pressuring Israel other than shaming mechanisms and security
183

guarantees actually took place: inspections of Dimona, lack of access to sensitive nuclear
technology, and requests for guarantees of the peaceful nature of the nuclear program.
Finally, evidence supports the fourth observable implication: Israel’s failure to conform
to the norm should be justified in terms of security concerns: Israel must justify its creation of a
nuclear bomb based on security concerns rather than rejection of nuclear norms themselves. As
shown above, at various points Israel made it clear to the United States that in the absence of a
public security guarantee, Israel might be forced to obtain nuclear weapons. Ultimately, that is
precisely what happened; Israel had to build the bomb as a tool of last resort. However, and
importantly, Israel did not reject international norms against nuclear weapons in public. For
Israel, obtaining the bomb was not a repudiation of non-proliferation norms themselves, but a
means for securing state survival. After obtaining the nuclear bomb, Israel adopted a policy of
nuclear ambiguity so as to maintain the appearance that it did not wish to introduce nuclear
weapons into the Middle East.
In sum, three out of four observable implications passed the hoop test, and the only one
that failed to gain support could be explained by the West hesitation to coerce Israel into nuclear
reversal. In the end, it appears that the non-reversal of Israel does not contradict the theoretical
expectations of the voluntary reversal pathway. As the theory expects, democracies under a
benign security environment will be most likely to accept international norms against nuclear
proliferation and ownership, and so more likely to fall to peer pressures to reverse course on their
nuclear pursuit. For Israel, although a democracy, the security environment was constantly
perilous. Under these circumstances, it was unlikely that international pressures towards reversal
would be successful when Israel lacked a strong security guarantee from other allies and faced
constant existential threats from neighbors. In addition, international pressures were weak in
184

comparison to the cases of Brazil and Iraq. The few countries aware of Israel’s nuclear program
refused to use economic sanctions or coercive threats to force Israel to give up development of
the nuclear bomb. No shaming mechanisms were employed in public forums like the UNGA,
and no security guarantees were offered towards Israel as an incentive to give up the bomb.
Therefore, the failure of a nuclear reversal can be explained by the half-hearted effort by the
international community and the presence of a high and constant security threat, which does not
contradict the theoretical expectations for when nuclear reversals are likely.
Table 10: Summary of results: Israel
Observable Implication Supported by Evidence
Israel’s program should begin if it perceived a
major security threat from its neighbors, and
such threat should be constant throughout the
nuclear weapons program development
period.
Yes
If the voluntary reversals pathway is correct,
Israel should have been pressured to terminate
its program as soon as the international
community learned that Israel was building
the bomb.
Yes
International pressure to conform to the anti-
nuclear nonproliferation norm should include
shaming mechanisms and/or diplomatic
“carrots” such as security guarantees, instead
of economic sanctions or threats of force.
No
Israel’s failure to conform to the norm must
be justified in terms of security concerns:
Israel must justify its creation of a nuclear
bomb based on security concerns rather than
rejection of nuclear norms themselves.
Yes

185

Chapter 8: Conclusion: Lessons of the Two-Pathway Theory of Nuclear
Reversals
This dissertation filled a gap in the current literature on nuclear weapon reversals, offering a
brand-new theory that synthesizes various insights from previous efforts to understand why
countries might dismantle their nuclear programs or give up their nuclear ambitions completely.
Rather than argue for a single causal factor for nuclear weapon reversal, the theory developed
here acknowledges the multi-causal nature of the real world, advancing an explanation that
covers multiple pathways of nuclear weapon program abandonment yet still amenable to
empirical testing and falsification. Thus, the theory straddles the fine line between parsimony
and complexity without sacrificing either.
The theory of nuclear reversals advanced in this project argues that there are two pathways
leading to the same outcome, nuclear reversals, and that whichever path a country takes is
affected by the presence of a few key intervening variables. Countries along the voluntary
pathway terminate pursuit of the bomb due to internalizing, or at least accepting as legitimate,
international norms against nuclear weapon proliferation and ownership. States are more likely
to accept these international norms against the nuclear bomb and eventually change their own
policymaking to align to these norms if they are liberal democracies and if the security
environment is not perceived as threatening. States, in an effort to align to the prevailing Western
led-order, will feel normative pressures from both domestic and international audiences to give
up pursuit of the bomb and dismantle any existing nuclear weapons program. The key, however,
is that states voluntarily terminate pursuit of the bomb in the absence of coercive threats or use of
force. On the other hand, states along the involuntary pathway terminate pursuit of the bomb due
to violent coercive strategies by other states. States in the involuntary pathway tend to be
186

authoritarian, rejecting the imposition of norms from the liberal West. In addition, states along
the involuntary pathway will not give up pursuit of the bomb if the security environment is
perceived as threatening, for the bomb can play a role in improving national security. For these
states outside the liberal community, nuclear reversals occur due to coercive pressures,
sometimes even physical force.
International norms play a role in both pathways delineated here, although more directly in one
pathway over the other. States along the voluntary pathway eventually accept international
norms against nuclear weapons as part of their identity as good members of the international
liberal community. Acceptance of these norms might not be immediate; the international
community itself pushes norms against the nuclear bomb. States, upon receiving these signals,
feel normative pressure to adhere to these norms and change their behavior, eventually joining
the broader community of states. International norms play an indirect role in the involuntary
pathway, for states alongside this pathway will tend to reject norms imposed by the liberal West.
However, the international community’s efforts to force non-proliferation on states is driven in
part by the belief that this norm against nuclear weapon non-proliferation and ownership should
be upheld, even if by force.
Overall, the theory centers international norms as an important factor that affects nuclear weapon
reversals, either directly when states voluntarily accept these norms, or indirectly as is the case
when the international community coerces states to terminate their nuclear programs. Yet, the
theory recognizes that non-ideological factors matter as well, and play a crucial role influencing
how states maneuver the prevailing normative environment. Material conditions such as the
security environment and power capabilities influence a state’s perception of threat, which in
turn affects state’s likelihood of abandoning their nuclear programs. Regime type impacts how
187

states align with the international community, affecting inter-state relations and the likelihood
that a state will accept certain norms advanced by the liberal West. Therefore, although the
theory centers international norms, it also accounts for non-normative variables that play a part
in the story, moving beyond old paradigmatic-centered approaches and embracing analytic
eclecticism.
348

1) Summary of findings
How does the theory stand up to empirical testing? The first major tests of the theory
were quantitative models shown in chapter 4, with the goal of discovering if there was a
statistically significant relationship between international norms and nuclear reversals. The
results of various logistic regression models arrived at the same conclusion: as international
norm acceptance of non-proliferation norms increases, the likelihood of nuclear reversals also
increases. The key intervening variables, democracy and threat perception, also performed as
hypothesized by the theory—the higher the level of democracy, the more likely a nuclear
reversal was to occur, while higher perception of threat was related with lower likelihood of
reversal. Although the results were suggestive of the relationship between international
acceptance of non-proliferation norms and nuclear reversal, the quantitative models were
insufficient as a test for three reasons. First, the tests do not prove causality, only correlation; due
to being tests based on observational data, it was not possible to establish a causal relationship
between international norms and nuclear reversals in the absence of an experimental design.
The second problem is that the statistical models are limited in their ability to cover a
complex two-pathway theory. The results of the logistic regression models may tell us that

348
Sil and Katzenstein (2010).
188

international norms against nuclear weapons increases the likelihood of reversals, but these
models do not highlight that not all nuclear reversals occur due to non-proliferation norm
acceptance. The statistical models, by their very nature, cannot explain two seemingly
contradictory pathways that states might take for nuclear weapon abandonment. Third, and
relatedly, the causal mechanisms connecting non-proliferation norms and nuclear reversals are
not covered by quantitative models. The models do not show us how the process of reversal is
brought about by international normative pressures, all these models can do is establish that there
is a positive effect of international norms against nuclear weapons on the likelihood of reversals.
Given these three limitations of a strictly quantitative approach, the next tests of the
theory had to be qualitative. Unlike quantitative modeling techniques, qualitative case studies
can test if the causal mechanisms of the two different pathways can be observed in real world
cases. Furthermore, case studies of most-likely cases allow the researcher to explore if there are
additional crucial factors that were overlooked by the theoretical framework, pointing to the need
for revisions in future iterations of the theory. Following contemporary best practices, the
qualitative case studies are tested rigorously via hoop tests. Observable, testable implications
drawn from each pathway are explicitly delineated prior to investigation of the cases, and then a
final stage of analysis determines which of these observable implications gain support from the
available evidence.
Chapter 5 tests the viability of the involuntary pathway of nuclear reversals through a case study
of Iraq’s nuclear program. The analysis finds support for four of the five observable implications
drawn from the involuntary reversal pathway. As predicted, it is found that the international
community relied on the language of non-proliferation norms to condemn Iraq’s nuclear program
once it was identified in the 90s. I also find that there is rank-ordering of diplomatic tools used
189

by state leaders pressuring Iraq to dismantle its program. Specifically, Israel tried to convince
European states against engaging in sensitive nuclear assistance with Iraq, then moved on to
threats of force and covert actions, and finally explicit use of force when it destroyed the Osirak
Reactor in 1981. The international community, represented by the UNSC, also followed a similar
rank-ordering of tools to dissuade Iraq from continuing its program, first using sanctions and
inspections, then moving on to threats of serious consequences if Iraq did not collaborate with
inspectors, culminating with the United States unilaterally bombing Iraq in Operation Desert
Fox. Finally, it is observed that Iraq’s nuclear weapons pursuit continued under conditions of
external threat from neighbors such as Iran and Israel, and that nuclear reversal only occurred
when military force was employed against Iraq, first by Israel in 1981 and then the United States
in the 90s.
However, one of the observable implications derived from the involuntary pathway failed to gain
support, indicating that the pathway might require some revisions. Unlike initially expected, no
evidence was found that Iraq explicitly rejected norms against nuclear weapon proliferation and
ownership. In fact, it appears that Iraq’s leadership frequently concealed their true intentions in
front of international audiences, often arguing that Iraq was adhering to existing international
non-proliferation norms. A necessary revision of the theory might be required—countries along
the involuntary pathway might conceal their true intentions in public forums, while covertly
pursuing the bomb. Yet, this too might speak of the importance of international norms, even for
autocratic states: even if authoritarian states do not accept non-proliferation norms, they might
still see it as advantageous to be perceived as if they were adhering to such norms.
Chapter 6 tests what is perhaps the most controversial aspect of the theory, the voluntary
pathway of nuclear reversals, using the case of Brazil as a most-likely case for the theory.
190

Results of the analysis were mixed—out of seven observable implications, only two of them
successfully passed the hoop tests without problems, two of them failed, and three gained mixed
support. In accordance with expectation 6, Brazil terminated its nuclear program following
democratization, finally joining the community of states supporting non-proliferation norms.
Furthermore, as predicted by observable implication 7, Brazil’s leadership used international
non-proliferation norms as justification for dismantling their nuclear program.
It is important to highlight some problems with the case selection and findings that did not
support the voluntary pathway as currently written. First, it seems that contemporary research
does not support the commonly held belief that Brazil had a nuclear weapons program; Brazil
did reject international norms against nuclear weapons during its military period, and Brazil did
establish a covert nuclear program in the 1970s, but the available evidence does not support the
belief that the top leadership intended to pursue the construction of a nuclear weapon. Although
some military leaders, even some presidents, made statements hinting at the possibility of
pursuing the nuclear bomb, the final decision was never made for such a project. Yet, it might
still be correct to classify Brazil as a case of nuclear reversal, for it did appear to be in its way to
obtaining the bomb, had it wanted to do it, and the decision to join the non-proliferation regime
did not come about until the 90s.
Two observable implications did not gain support. Implication 5, which states “if Brazil
terminates their nuclear program before Brazil changes its perception of threat, it should be
because other actors promised credible security guarantees,” can be safely discarded because
termination of the nuclear program occurred after improved relations with Argentina, widely
considered Brazil’s regional economic rival. However, observable implication 3, which also did
not gain support, is more problematic for the theoretical framework. According to the
191

expectation, the main diplomatic tools used by other states to push Brazil towards nuclear
reversal after democratization should have been shaming mechanisms or diplomatic incentives.
The available evidence does not show that the international community increased their efforts to
pressure Brazil into accepting norms against nuclear weapon proliferation. Perhaps this finding
contradicting the theoretical expectations can be explained by Brazil’s willingness to join the
nonproliferation regime under the leadership of Collor de Mello and succeeding administrations,
making it unnecessary for the international community to increase pressures for Brazil to join the
non-proliferation regime.
Support for the remaining three observable implications was mixed, but not necessarily
contradictory to the logic of the voluntary reversals pathway. The first observable implication
states that the international community should pressure Brazil to terminate its nuclear program
after discovering Brazil’s covert nuclear program and that the international norms will be
promoted by liberal states more frequently than autocracies. However, although we do observe
liberal states promoting norms against proliferation more often than non-democracies, we also
observe the international community pressured Brazil to join the non-proliferation regime when
Brazil rejected the NPT, about a decade before Brazil initiated its covert program in 1979. The
second observable implication expects rank-ordering of diplomatic tools to coerce Brazil into
joining the non-proliferation regime. As expected, economic sanctions were commonly used
against Brazil during its military period, but contrary to expectations positive inducements were
also employed, such as when the United States offered economic incentives to Brazil so it would
join the NPT in 1976. Finally, observable expectation 4 predicts that Brazil’s decision to reverse
their program will occur after decreased perception of threat. Yet, analysis of the case finds that
192

Brazil did not feel any existential threats from Argentina, although they were regional economic
rivals, highlighting that the rivalry has been overexaggerated by some American scholars.
These mixed results do not accord with the expectations of the voluntary reversal pathway, but
upon closer inspection do not present a threat to the logic of the theory. Rather, the mixed results
demonstrate that the voluntary reversal pathway needs to be flexible to future revisions if the
revisions do not harm the internal logic of the theory. First, the theory is too restrictive in its
expectation for when the international community will pressure states to join the non-
proliferation regime. The international community does not have to wait until the discovery of a
nuclear program to pressure states to accept non-proliferation norms. Second, the theory needs to
be less strict in its expectation about the diplomatic tools states will use when trying to pressure
states to join the non-proliferation regime. Although sanctions and threats of force might be more
commonly used against non-democracies, positive inducements towards such states can also be
on the table. Finally, change in perception of threat might not be the only change necessary for a
state to accept international norms; improved relations with non-threatening rivals might have
similar effects as well. Importantly, these potential improvements to the voluntary reversal
pathway do not harm the central logic of the theory, which is that states that accept international
norms against nuclear weapon proliferation may be likely to terminate their programs.
Chapter 7 tests the theory of nuclear reversals against a case of non-reversal, Israel. Unlike Iraq
and Brazil, Israel successfully launched a nuclear weapons program and covertly obtained the
bomb. Four observable implications of the theory were delineated in the chapter, with three of
them gaining support: 1) Israel’s nuclear program began in response to security threats from
neighbors, 2) Israel was pressured to terminate its program by members of the international
community as soon as it was evident that Israel had a covert nuclear program, and 3) Israel’s
193

failure to conform to non-proliferation norms was justified in terms of national security by Israeli
policymakers. The only expectation that failed was observable implication 3, which states that
international pressures towards Israel to give up pursuit of the bomb would be primarily
diplomatic incentives such as security guarantees, or at minimum shaming tactics. Contrary to
expected behavior, the international community did not employ strong diplomatic incentives to
promote nuclear reversal in Israel, perhaps because of the understanding that Israel faced a
highly threatening security environment. Overall, the chapter was effective at demonstrating the
impact that a threatening security environment may have on a democracy; norms against nuclear
weapon proliferation and ownership are unlikely to be internalized by a state under conditions of
existential threat.
In sum, the findings of the empirical tests increase my confidence that the theoretical framework
developed in this project is effective at understanding dynamics of nuclear reversal. The
statistical analysis suggests a positive relationship between international norms against nuclear
weapons and increased likelihood of nuclear weapon reversals. Additionally, the case studies
largely support the causal mechanisms delineated by the involuntary and voluntary pathways of
nuclear reversal, demonstrating how the interaction between regime type, national security, and
international norms might impact state decisions in the nuclear realm. Although not all the
observable implications derived from the pathways gained support in the case studies, none of
the results were contradictory enough as to reject the theoretical framework—although some
revisions to increase the complexity of the theory may be required.

194

2) Contributions and lessons for future reversals
Ultimately, this dissertation contributed to research on nuclear weapon reversals in two major
ways. First, I developed a novel measure of nuclear non-proliferation norm acceptance, using
UNGA voting records on nuclear resolutions as a proxy for the degree to which a state accepts
non-proliferation norms at a given point in time. Second, I built a two-pathway theory that
explains why countries trying to obtain the bomb might turn back on their original decision and
dismantle their nuclear programs. Acknowledging the multi-causal nature of the social world, the
theory incorporates insights from realism and liberalism—such as national security concerns and
regime type effects, respectively—in a broad theory that centers the effects that international
norms might have on nuclear reversal state decisions. Furthermore, the theory of nuclear
reversals accepts that there may not be a single causal mechanism at work in nuclear reversals,
thus delineating two distinct paths, the voluntary and involuntary pathways, which lead to the
same outcome.
This project adds to the growing body of work in the area of nuclear reversals, which has been
underexplored until recently. Therefore, my theoretical framework may be situated alongside
existing work such as Rupal Mehta’s Delaying Doomsday: The Politics of Nuclear Reversal,
which argues that positive inducements and negative inducements together, primarily from the
United States, can be effective tools to dissuade friendly states from pursuing the bomb,
349
and
Miller’s Stopping the Bomb: the Sources and Effectiveness of US Nonproliferation Policy, which
argues that American credible threats to cut economic and military support to states pursuing
nuclear weapons might successfully deter states from pursuing the bomb if they depend on U.S.

349
Mehta (2020).
195

aid.
350
As argued in the theory chapter, most theories of nuclear reversals focus on how the
United States uses its influence to promote nuclear reversals and non-proliferation,
underemphasizing how other actors and the international community as a whole might influence
abandonment of nuclear programs or nuclear ambitions. Thus, the theory presented here should
not be understood as a challenge to existing work, but rather a complement that fills an existing
gap in current understandings of nuclear reversals.
What lessons for policymakers might be drawn from the theory of nuclear reversals advanced in
this project? It seems to me that a few valuable lessons can be drawn from the present study,
applicable not just to American policymakers but other state leaders as well. These lessons can
be applied to current proliferation risks, such as Iran, and perhaps applied to potential future
cases of nuclear proliferation like Saudi Arabia.
351
Three policy lessons that may be drawn are
the following: 1) provide public security guarantees to minimize high perceptions of threat; 2)
promote nuclear disarmament efforts alongside non-proliferation norms and engage in shaming
mechanisms; 3) credible threats of sanctions and even force should be used only as a last resort,
and only if public security guarantees and shaming mechanism do not work.
As of this writing, Iran is the most notorious nuclear proliferation risk, largely believed by
security analysts to be a state in pursuit of the nuclear bomb. Global suspicions about Iran’s
nuclear ambitions began in 2002, when the National Council of Resistance in Iran, an exiled
opposition group, revealed the location of two undeclared nuclear sites in Iran, a uranium-

350
Miller (2018).
351
In 2018, Saudi crown prince Mohammed bin Salman threatened that Saudi Arabia might pursue nuclear weapons
if Iran obtains the bomb. (CBS interactive, 2018. https://www.cbsnews.com/news/saudi-crown-prince-mohammed-
bin-salman-iran-nuclear-bomb-saudi-arabia/)
196

enrichment complex in Natanz and a heavy-water production facility at Arak.
352
Fears about
Iran’s nuclear ambitions were temporarily relieved following the July 2015 Joint Comprehensive
Plan of Action (JCPA), informally called the Iran deal. The JCPA accord, signed between the
permanent Security Council members plus Germany on one side, and Iran on the other, imposed
restrictions on Iran’s civilian nuclear enrichment program. According to the terms of the
agreement, Iran agreed not to produce highly enriched uranium or plutonium that could be used
for a nuclear weapon. In addition, Iran agreed to allow IAEA inspectors into previously
undisclosed sites. In exchange, the United Nations, the European Union, and the United States
agreed to lift nuclear weapon-related sanctions and lift a weapons embargo within five years if
the IAEA concludes that Iran is adhering to the terms of the agreement.
353

Although initially successful in curving Iranian nuclear proliferation, the JCPA has been
on the verge of collapse ever since former President Trump withdrew the United States from the
nuclear deal in 2018.
354
Since then, Iran has breached their own side of the agreement, increasing
its highly enriched uranium stockpiles to near-weapons grade uranium, in addition to sanctions
against U.N. inspectors making it impossible for the United Nations to inspect Iran’s stockpile of
enriched uranium—according to the IAEA, by October 2022 Iran “had 62.3 kilograms of
uranium enriched to up to 60 percent fissile purity.”
355
Furthermore, the IAEA also reported that
in June 2022 Iran removed all surveillance equipment, a total of 27 cameras, that had been
installed in nuclear sites as part of the nuclear agreement.
356
These developments place Iran at
the highest level of risk of becoming the next nuclear weapon state, which may encourage further

352
NBC Universal News Group, “Exiled Group: Iran’s Hiding Nuke Site.” NBCNews.com,
https://www.nbcnews.com/id/wbna3158361 , published October 13, 2003. Accessed January 25, 2023.
353
Robinson (2022).
354
Landler (2018).
355
Liechtenstein (2022).
356
Ibid.
197

nuclear proliferation in the Middle East, particularly from Saudi Arabia, which threatened to
develop their own nuclear weapon if Iran obtains the bomb.
The three lessons that could be drawn from the theory can be applied to the case of Iran,
not just by the United States but also by other members of the international community. Since
Iran is an authoritarian state that perceives a security threat from Israel, Iran is unlikely to take
the voluntary pathway of nuclear reversals, and is more likely to terminate their nuclear program
via the involuntary reversal route. However, the lessons drawn from the theory might still allow
for nuclear reversal to occur without having to resort to actual use of military force. First, it is
crucial that the international community minimize Iran’s perception of threat by assuring Iran
that the United States will not support overthrow of the regime. From Iran’s perspective, the
security environment is threatening: Iran’s principal enemy in the region is a nuclear armed
Israel, Iran witnessed how Iraq’s Saddam Hussein regime was overthrown by the United States,
and Iran has been the frequent target of hawkish U.S. policy and rhetoric since George W.
Bush’s presidency. For Iran, obtaining a nuclear bomb might serve as a deterrent against a
possible future U.S. invasion or a militarized dispute with Israel. Assuring Iran that the United
States will not pose a security risk is easier said than done, but such de-escalatory tactic is
essential over current realpolitik logic that currently dominates U.S. foreign policy towards Iran.
The lessons from the literature on war might apply to the nuclear realm as well—if we want to
avoid future conflict with Iran, and decrease Iran’s perception of threat, then the United States
and its allies need to change their hawkish approach to Iran.
357

357
Vasquez (2009) highlights how hawkish foreign policy contributes to escalatory dynamics of conflict.
198

Relatedly, the second lesson for diminishing the probability of a nuclear Iran is for the
international community to continue promoting norms against nuclear proliferation in addition to
norms against nuclear weapon ownership. Key in this endeavor is the 2017 Treaty on the
Prohibition of Nuclear Weapons. The Treaty on the Prohibition of Nuclear Weapons is an
absolute prohibition on all nuclear weapon related activities, including acquisition, possession,
and use of nuclear weapons.
358
The ultimate goal of the treaty is eventual elimination of all
nuclear weapons, admittedly an unlikely outcome. Despite the improbability of such an outcome,
promoting anti-nuclear weapon norms might still pressure states to abide by existing nuclear
non-proliferation treaties. For instance, pressuring Israel to abandon its policy of nuclear opacity
and join the NPT would ensure that Israel’s nuclear program no longer remains a mystery to the
outside world, perhaps even allow the international community to monitor Israel’s nuclear
capability and development. In turn, a transparent Israel monitored by the international
community could appease fears that Israel might use its nuclear arsenal against Iran, while
international pressures alongside a decreased threatening security environment might increase
the chances for Iran to abandon its nuclear ambitions.
Finally, if attempts to decrease Iran’s perception of threat and normative pressures fail to
promote nuclear reversal in Iran, then the last lesson would have to be applied, credible threats of
sanctions and even force to forcibly terminate Iran’s program. This lesson does not imply that
states need to be reckless and engage in warfare to terminate Iran’s nuclear program, but rather
than states should move along the diplomatic toolkit from least to most aggressive diplomatic
tools when coercing Iran. Furthermore, coercive tactics such as economic sanctions and even

358
United Nations. “Treaty on the prohibition of nuclear weapons.” United Nations Office for Disarmament Affairs.
Accessed January 25, 2023. https://www.un.org/disarmament/wmd/nuclear/tpnw/
199

concealed destruction of nuclear capabilities should be reserved as a last resort if Iran does not
respond to “dovish” strategies such as guaranteeing Iranian national security and lifting of
current sanctions.
3) Next steps for future research
As can be seen, the two-pathway theory of nuclear reversals can offer lessons for
policymakers interested in promoting nuclear reversal worldwide. However, and perhaps more
importantly for scholars, the theory offers an account of how the process of nuclear reversals
works, broadening our understanding of the dynamics of nuclear weapon program abandonment.
The theory also opens up avenues for future research, which should improve and build on the
theoretical and data contributions of this work.
First, a logical extension of the project is to develop the dissertation into a full-length
book project, which incorporates the lessons from the case studies for a more polished theoretical
framework of the two established pathways. The book project could then either improve the case
study work by incorporating more archival sources that were unobtained due to real world
constraints of the covid-19 pandemic, or include brand new case studies and test if the causal
mechanisms of the theory apply to other cases of nuclear reversal, such as Japan, South Korea,
Libya, and Argentina. Second, another logical project that can be drawn from the dissertation is a
stand-alone article that improves on the UNGA voting records dataset—in fact, such an article is
currently in the works, and will improve on the existing data work by incorporating more UNGA
resolutions left out from the initial data construction.
Other questions for future research can be derived from this dissertation. For instance,
can the theory of nuclear reversals provide insights for nuclear weapon disarmament? To date,
200

only one country, South Africa, obtained and then dismantled its nuclear weapon arsenal.
Ukraine, Belarus, and Kazakhstan inherited nuclear weapons when the Soviet Union collapsed,
yet dismantled their inherited arsenal. Future research may examine if the theory of nuclear
reversals can also apply to nuclear weapon disarmament itself, using the aforementioned
examples as case studies. Second, we may ask what the best tools are for promoting nuclear
reversal. This dissertation shows that there is a rank-ordering of policy tools at the disposal of
policymakers, but which are more effective? Such research would have important policy
implications and further develop the lessons that might be drawn from the theoretical framework.
Finally, what are the rhetorical strategies used by countries discussing anti-nuclear UNGA
resolutions? An in-depth analysis of United Nations General Assembly discussions on nuclear
weapon related resolutions would further increase our knowledge of how international norms are
employed in international forums to pressure states to terminate pursuit of the nuclear bomb or
prevent nuclear proliferation altogether.
Overall, it can be said that academic study of nuclear reversal dynamics is in its infancy
in relation to research on nuclear proliferation. The theory of nuclear reversals introduced in this
work furthers our understanding of this phenomenon, while simultaneously broadening
possibilities for future research on nuclear reversals. Further tests are required to increase
confidence of the validity of the theory advanced in this project, but at minimum the two-
pathway theory of nuclear reversals pushes us to consider the effects of the international
normative environment seriously, even in the security arena.

201

Bibliography
Alston, Lee J., Marcus André Melo, Bernardo Mueller, and Carlos Pereira. 2016. “Why
Countries Transition? The Case of Brazil, 1964–2016,” Atlantic Economic Journal, Vol.
44, No. 2, pp. 197-224.

Bahgat, Gawdat. 2006. “Israel and Nuclear Proliferation in the Middle East,” Middle East Policy,
Vol. 13, No. 2, pp. 113-133.

Barletta, Michael. 1997. “The Military Nuclear Program in Brazil,” Center for International
Security and Arms Control, Stanford University working paper.

Bennett, Andrew, and Jeffrey T. Checkel. 2015. “Process Tracing: From Philosophical Roots to
Best Practices,” in Process Tracing: From Metaphor to Analytic Tool, pp. 3-37. Andrew
Bennett and Jeffrey T. Checkel, eds. Cambridge: Cambridge University Press.

Bleek, Philipp and Eric Lorber. 2014. “Security Guarantees and Allied Nuclear Proliferation,”
Journal of Conflict Resolution, Vol. 58, No. 3, pp. 429-454.

Brands, Hal and David Palkki. 2011. “Saddam, Israel, and the Bomb. Nuclear Alarmism
Justified?” International Security, Vol. 36, No. 1, pp. 133-166.

Braut-Hegghammer, Malfrid. 2011. “Revisiting Osirak: Preventive Attacks and Nuclear
Proliferation Risks,” International Security, Vol. 36, No. 1, pp. 101-132.

CBS Interactive. March 16, 2018. “Saudi crown prince: If Iran develops nuclear bomb, so will
we,” CBS News, from https://www.cbsnews.com/news/saudi-crown-prince-mohammed-
bin-salman-iran-nuclear-bomb-saudi-arabia/, accessed January 25, 2023.

Cohen, Avner. 2007. “Crossing the Threshold: The Untold Nuclear Dimension of the 1967 Arab-
Israeli War and Its Contemporary Lessons,” Arms Control Today, Vol. 37, No. 5, pp. 12-
16.

Cohen, Avner. 2005. “The Last Taboo: Israel’s Bomb Revisited,” Current History, Vol. 104, No.
681, pp. 169-175.

Cohen, Avner. 1998. Israel and the Bomb. New York: Columbia University Press.

Coutto, Tatiana. 2014. “An international history of the Brazilian-Argentine rapprochement,” The
International History Review, Vol.36, No. 2, pp. 302-323.

202

Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Svend-Erik
Skaaning, Jan Teorell, David Altman, Michael Bernhard, Agnes Cornell, M. Steven Fish,
Haakon Gjerløw, Adam Glynn, Allen Hicken, Joshua Krusell, Anna Lu¨hrmann, Kyle L.
Marquardt, Kelly McMann, Valeriya Mechkova, Moa Olin, Pamela Paxton, Daniel
Pemstein, Brigitte Seim, Rachel Sigman, Jeffrey Staton, Aksel Sundtr¨om, Eitan Tzelgov,
Luca Uberti, Yi-ting Wang, Tore Wig, and Daniel Ziblatt. 2022. “V-Dem Codebook
v12,” Varieties of Democracy (V-Dem) Project.

Correlates of War Project. 2017. Direct Contiguity Data, 1816-2016. Version 3.2.

Cox, David Roxbee. 1972. “Regression models and life-tables,” Journal of the Royal Statistical
Society. Series B, Vol. 34, No. 2, pp. 187-220.

Curtiss, Richard H. 1992. “Twenty-Five Years Ago This Month: Who Started the Six-Day War
on June 5, 1967? The Record is Clear,” The Washington Report on Middle East Affairs,
Vol. XI, No. 1, pp. 1-11.

Dahl, Robert. 1971. Polyarchy: Participation and Opposition. New Haven: Yale University
Press.

Dalaqua, Renata H. 2019. “’We will not make the bomb because we do not want to make the
bomb’: understanding the technopolitical regime that drives the Brazilian nuclear
program,” The Nonproliferation Review, Vol. 26, No. 3-4, pp. 231-249.

Debs, Alexandre and Nuno P. Monteiro. 2017. Nuclear Politics: The Strategic Causes of
Proliferation. New York: Cambridge University Press.

Drogan, Mara. 2019. “The Atoms for Peace Program and the Third World,” Cahiers Du Monde
Russe, Vol. 60, No. 2/3, pp. 441–60.

Duelfer, Charles. September 30, 2004. “Comprehensive Report of the Special Advisor to the
DCI on Iraq’s WMD, with Addemdums. Vol. 2,” Washington, D.C.: Central Intelligence
Agency, https://www.govinfo.gov/content/pkg/GPO-DUELFERREPORT/pdf/GPO-
DUELFERREPORT-2.pdf

Ekeus, Rolf. 2016. “The Lessons of UNSCOM and Iraq,” The Nonproliferation Review, Vol. 23,
No. 1-2, pp. 131-146.

Engelhardt, Michael J. 2004. “A nonproliferation failure: America and Israel’s nuclear program,
1960-1968,” The Nonproliferation Review, Vol. 11. No. 3, pp. 56-69.

Epstein, William. 1977. “Why States Go-And Don’t Go-Nuclear,” Annals of the American
Academy of Political and Social Science, Vol. 430, No. 1, pp. 16-28.

203

Fariss, Christopher J., Therese Anders, Jonathan N. Markowitz, and Miriam Barnum. 2022.
“New estimates of over 500 years of historic GDP and population data,” Journal of
Conflict Resolution, Vol. 66, No. 3, web.

Feldman, Shai. 1982. “The Bombing of Osiraq-Revisited,” International Security, Vol. 7, No. 2,
pp. 114-142.

Finnemore, Martha. 1996. National Interests in International Society. Ithaca: Cornell University
Press.

Finnemore, Martha, and Kathryn Sikkink. 1998. “International Norm Dynamics and Political
Change,” International Organization, Vol. 52, No. 4, pp. 887-917.

General Assembly Resolution 36/27, Armed Israeli aggression against the Iraqi nuclear
installations and its grave consequences for the established international system
concerning the peaceful uses of nuclear energy, the non-proliferation of nuclear weapons
and international peace and security, A/RES/36/27 (13 November 1981),
https://digitallibrary.un.org/record/27621?ln=en

General Assembly Resolution 1652(XVI), Consideration of Africa as a denuclearized zone,
A/RES/1652(XVI) (24 November 1961),
https://digitallibrary.un.org/record/662076?ln=en

General Assembly Resolution 1911(XVIII), Denuclearization of Latin America,
A/RES/1911(XVIII) (27 November 1963),
https://digitallibrary.un.org/record/203964?ln=en

George, Alexander L., David Hall, and William Simons. 1971. The Limits of Coercive
Diplomacy: Laos, Cuba, and Vietnam. Little, Brown and Company.

Gerlini, Matteo. 2010. “Waiting for Dimona: The United States and Israel’s development of
nuclear capability,” Cold War History, Vol. 10. No. 2, pp. 143-161.

Gerring, John. 2017. Case Study Research: Principles and Practices. 2
nd
ed. Cambridge:
Cambridge University Press.

Goldemberg, Jose and Harold A. Feiveson. 1994. “Denuclearization in Argentina and Brazil,”
Arms Control Today, Vol. 24, No. 2, pp. 10-14.

Greig, J. Michael and Andrew J. Enterline. 2021. “National Material Capabilities (NMC) Data
Documentation. Version 6.0,” Correlates of War Project.

Hymans, Jacques E. C. 2001. “Of gauchos and gringos: why Argentina never wanted the bomb,
and why the United States thought it did,” Security Studies, Vol. 10, No. 3, pp. 153-185.

204

Hymans, Jacques E. C. 2006. The Psychology of Nuclear Proliferation. New York: Cambridge
University Press.

Hymans, Jacques E. C. 2012. Achieving Nuclear Ambitions: Scientists, Politicians, and
Proliferation. Cambridge: Cambridge University Press.

IAEA, Fourth consolidated report of the Director General of the International Atomic Energy
Agency under paragraph 16 of Security Council resolution 1051 (1996), S/1997/779, (8
October 1997). https://digitallibrary.un.org/record/244664?ln=en

IAEA, “Communication Received from Members Regarding the Export of Nuclear Material and
of Certain Categories of Equipment and Other Material,” International Atomic Energy
Agency, INFCIRC/209, (3 September 1974),
https://inis.iaea.org/collection/NCLCollectionStore/_Public/44/096/44096604.pdf

IAEA, “Agreement of 13 December 1991 between the republic of Argentina, the federative
republic of Brazil, the Brazilian- Argentine agency for accounting and control of nuclear
materials and the international atomic energy agency for the application of safeguards,”
International Atomic Energy Agency, INFCIRC/435, (March 4 1994),
https://www.iaea.org/sites/default/files/infcirc435.pdf

Ikenberry, John. 2011. Liberal Leviathan: The Origins, Crisis, and Transformation of the
American World Order. Princeton: Princeton University Press.

“Intermediate-Range Nuclear Forces Treaty - Inf (1987).” Intermediate-Range Nuclear Forces
Treaty - INF (1987) | Nuclear Arms Control Treaties, accessed October 3, 2022.
https://www.atomicarchive.com/resources/treaties/inf.html.

Jo, Dong-Joon and Erik Gartzke. 2007. “Determinants of Nuclear Weapons Proliferation,”
Journal of Conflict Resolution, Vol. 51, No. 1, pp. 167-194.

Jones, Daniel M., Stuart A. Bremer, and J. David Singer. 1996. “Militarized Interstate Disputes,
1816-1992: Rationale, Coding Rules, and Empirical Patterns,” Conflict Management and
Peace Science, Vol. 15, pp. 163-213.

Karsh, Efraim. 2017.“The Six Day War: an Inevitable Conflict,” Middle East Quarterly, Vol. 24,
No. 3, pp. 1-12.

Kroenig, Matthew. 2009. “Importing the bomb: sensitive nuclear assistance and nuclear
proliferation,” Journal of Conflict Resolution, Vol. 53, No. 2, pp. 161-180.

Kutchesfahani, Sara Z. 2014. Politics and the Bomb: The Role of Experts in the Creation of
Cooperative Nuclear Non-Proliferation Agreements. New York: Routledge.

205

Landler, Mark. May 8, 2018. “Trump abandons Iran nuclear deal he long scorned,” The New
York Times. Accessed January 25, 2023,
https://www.nytimes.com/2018/05/08/world/middleeast/trump-iran-nuclear-deal.html

Legro, Jeffrey W. 1997. “Which Norms Matter? Revisiting the ‘Failure’ of Internationalism,”
International Organization, Vol. 51, No. 1, pp. 31-63.

Levite, Ariel E. 2002/03. “Never Say Never Again: Nuclear Reversal Revisited,” International
Security, Vol. 27, No. 3, pp. 59-88.

Liechtenstein, Stephanie. November 10, 2022. “U.N. agency: Iran continues to block nuclear
probe, scales up its nuclear program,” Politico. Accessed January 25, 2023,
https://www.politico.com/news/2022/11/10/un-agency-iran-nuclear-probe-00066350

“Limited Test Ban Treaty (1963).” Limited Test Ban Treaty (1963) | Nuclear Arms Control
Treaties, https://www.atomicarchive.com/resources/treaties/ltbt.html.

Little, Douglas. 1993. “The Making of a Special Relationship: The United States and Israel,
1957-68,” International Journal of Middle East Studies, Vol. 25, No. 4, pp. 563-585.

Lobell, Steven E. 2020. “Why Israel Launched a Preventive Military Strike On Iraq’s Nuclear
Weapons Program (1981): The Fungibility of Power Resources,” Journal of Strategic
Studies, pp. 1-26.

Lust, Ellen. 2020. The Middle East 15
th
ed. Thousand Oaks, California: SAGE Publications.

Mahoney, James. 2012. “The Logic of Process Tracing Tests in the Social Sciences,”
Sociological Methods and Research, Vol. 41, No. 4, pp. 570-597.

Marshall, Monty G., and Ted Robert Gurr. 2020. “Political regime characteristics and transitions,
1800-2018: Dataset User’s Manual,” Polity5.

Mattiacci, Eleonora, and Benjamin T. Jones. 2016. “(Nuclear) Change of Plans: What Explains
Nuclear Reversals?” International Interactions, Vol. 42, No. 3, pp. 530-558.

Mehta, Rupal N. 2020. Delaying Doomsday: The Politics of Nuclear Reversal. New York:
Oxford University Press.

Miller, Nicholas L. 2018. Stopping the Bomb: The Sources and Effectiveness of U.S.
Nonproliferation Policy. Ithaca, New York: Cornell University Press.

Morris, Benny. 2008. 1948: A History of the First Arab-Israeli War. Yale University Press.

206

Muller, Harald, and Andreas Schmidt. 2010. “The Little-Known Story of Deproliferation: Why
States Give Up Nuclear Weapons Activities,” in Forecasting Nuclear Proliferation In the
21
st
Century: The Role of Theory Volume 1., pp. 124-158. William C. Potter and Gaukhar
Mukhatzhanova, eds. Stanford, CA: Stanford University Press.

National Archives and Records Administration. “Saddam Hussein's Defiance of United Nations
Resolutions,” The White House. Accessed July 8, 2022, https://georgewbush-
whitehouse.archives.gov/infocus/iraq/decade/sect2.html

NBCUniversal News Group. October 13, 2003. “Exiled group: Iran's hiding Nuke Site,”
NBCNews.com. Accessed January 25, 2023, https://www.nbcnews.com/id/wbna3158361.

“Nuclear Madness,” The New York Times, 13 June 1975, pp. 36.
https://www.nytimes.com/1975/06/13/archives/nuclear-madness.html

“Outer Space Treaty (1967).” Outer Space Treaty (1967) | Nuclear Arms Control Treaties,
https://www.atomicarchive.com/resources/treaties/outer-space.html.

Owen, John M. 1994. “How liberalism produces democratic peace,” International Security, Vol.
19, No. 2, pp. 87-125.

Palmer, Glenn, Roseanne W. McManus, Vito D’Orazio, Michael R. Kenwick, Mikaela Karstens,
Chase Bloch, Nick Dietrich, Kayla Kahn, Kellan Ritter, and Michael J. Soules. 2020.
“Codebook for the Militarized Interstate Dispute Data, Version 5.0,” Correlates of War
Project.

Parmer, Inderjeet. 2018. “The US-led liberal order: imperialism by another name?” International
Affairs, Vol. 94, No. 1, pp. 151-172.

Paul, T.V. 2000. Power versus Prudence: Why Nations Forgo Nuclear Weapons. Montreal:
McGill-Queen’s University Press.

Rabinowitz, Or. 2019. “’When Pigs Fly’: Britain, Canada and Nuclear Exports to Israel, 1958-
1974,” Diplomacy and Statecraft, Vol. 30, No. 4, pp. 707-728.

Reiss, Mitchell. 1995. Bridled Ambition: Why Countries Constrain Their Nuclear Capabilities.
Washington, DC: Woodrow Wilson Center Press.

Resende-Santos, João. 2002. “The origins of security cooperation in the southern cone,” Latin
American Politics and Society, Vol. 44, No. 4, pp. 89-126.
Robinson, Kali. Updated July 20, 2022. “What is the Iran Nuclear Deal?” Council on Foreign
Relations. Accessed January 25, 2023. https://www.cfr.org/backgrounder/what-iran-
nuclear-deal
Rublee, Maria Rost. 2009. Nonproliferation Norms: Why States Choose Nuclear Restraint.
Athens, GA: University of Georgia Press.
207

Sadot, Uri. 2016. “Osirak and the Counter-Proliferation Puzzle,” Security Studies, Vol. 25, No. 4,
pp. 646-676.

Sagan, Scott D. 1996/97. “Why do States Build Nuclear Weapons? Three Models in Search of a
Bomb,” International Security, Vol. 21, No. 3, pp. 54-86.

Schneider, Jonas. 2020. “Beyond Assurance and Coercion: US Alliances and the Psychology of
Nuclear Reversal,” Security Studies, Vol. 29, No. 5, pp. 927-963.

Sil, Rudra and Peter J. Katzenstein. 2010. Beyond Paradigms: Analytic Eclecticism in the Study
of World Politics. New York: Palgrave Macmillan.

Singh, Sonali and Christopher R. Way. 2004. “The Correlates of Nuclear Proliferation: a
Quantitative Test,” Journal of Conflict Resolution, Vol. 48, No. 6, pp. 859-885.

Smith, Hedrick. July 18, 1970. “U.S. Assumes the Israelis Have A-Bomb or Its Parts,” The New
York Times. Accessed November 17, 2022.
https://www.nytimes.com/1970/07/18/archives/us-assumes-the-israelis-have-abomb-or-
its-parts-us-assumes-israelis.html

Smooha, Sammy. 2002. “The model of ethnic democracy: Israel as a Jewish and democratic
state,” Nations and nationalism, Vol. 8, No. 4, pp. 475-503.

Solingen, Etel. 2007. Nuclear Logics: Contrasting Paths in East Asia and the Middle East.
Princeton, NJ: Princeton University Press.

Spektor, Matias S. 2016. “The evolution of Brazil’s nuclear intentions,” The Nonproliferation
Review, Vol. 23, No. 5-6, pp. 635-652.

Thayer, Bradley A. 1995. “The Causes of Nuclear Proliferation and the Utility of the Nuclear
Non-Proliferation Regime,” Security Studies, Vol. 4, No. 3, pp. 463-519.

The Associated Press. June 24, 1981. “Iraq Asserts Arabs Must Acquire Atom Arms As A
Balance To Israel,” The New York Times.

The Antarctic Treaty, art. 5, December 1, 1959, 12 UST 794; 402 UNTS 71; 19 ILM 860 (1980),
https://documents.ats.aq/keydocs/vol_1/vol1_2_AT_Antarctic_Treaty_e.pdf

Treaty on the Non-Proliferation of Nuclear Weapons, March 5, 1970,
https://www.un.org/disarmament/wmd/nuclear/npt/text

United Nations. Charter of the United Nations. 1 UNTS XVI., Accessed February 18, 2022.
https://legal.un.org/repertory/art25.shtml

208

United Nations. Treaty on the prohibition of nuclear weapons. United Nations Office for
Disarmament Affairs. Accessed January 25, 2023.
https://www.un.org/disarmament/wmd/nuclear/tpnw/

United Nations. United Nations Disarmament Commission. Accessed October 3, 2022.
https://www.un.org/disarmament/institutions/disarmament-commission/

United Nations Digital Library. “Voting Data.” Accessed February 28, 2022.
https://digitallibrary.un.org/search?ln=en&cc=Voting+Data

UN Documentation: General Assembly. “Voting in the UN General Assembly.” Updated
February 16, 2022. Accessed February 20, 2022.
https://research.un.org/en/docs/ga/voting

U.N. G.A.O.R., 45
th
Sess., U.N. Doc A/45/PV.4 (Sept. 27, 1990). Available from
https://documents-dds-
ny.un.org/doc/UNDOC/PRO/N90/640/94/PDF/N9064094.pdf?OpenElement

U.N. G.A.O.R., 46
th
Sess., U.N. Doc A/46/PV.4 (Sept. 24, 1991). Available from
https://documents-dds-
ny.un.org/doc/UNDOC/PRO/N91/612/17/PDF/N9161217.pdf?OpenElement

UN General Assembly. “what is the difference between a resolution and a decision?” published
December 22, 2021. Accessed January 28, 2022. https://ask.un.org/faq/14484

United Nations Security Council Resolution 1154, Security Council resolution 1154 (1998) [on
endorsement of the memorandum of understanding of 23 Feb. 1998 regarding
compliance by Iraq with its obligations under resolution 687 (1991) and other relevant
resolutions], S/RES/1154 (2 March 1998),
https://digitallibrary.un.org/record/250774?ln=en

United Nations Security Council Resolution 715, Security Council resolution 715 (1991) [on
plans for monitoring and verification of Iraq's compliance with relevant parts of section
C of Security Council resolution 687 (1991)], S/RES/715 (15 October 1991),
https://digitallibrary.un.org/record/128235?ln=en

United Nations Security Council Resolution 707, Security Council resolution 707 (1991) [on
Iraqi violation of Security Council resolution 687 (1991) with regard to inspection of its
biological, chemical and nuclear weapons capabilities], S/RES/707 (15 August 1991),
https://digitallibrary.un.org/record/125533?ln=en

United Nation Security Council Resolution 678, Security Council resolution 687 (1991) [on
restoration of the sovereignty, independence and territorial integrity of Kuwait],
S/RES/687 (3 April 1991), https://digitallibrary.un.org/record/110709?ln=en

209

U.N. GAOR, 36
th
Sess., 56
th
plen. mtg., U.N. Doc. A/36/PV.56 (13 November 1981),
https://digitallibrary.un.org/record/761434?ln=en

Vasquez, John A. 2009. The War Puzzle Revisited. Cambridge: Cambridge University Press.

Way, Christopher Weeks, and Jessica L.P. Weeks. 2014. “Making it Personal: Regime Type and
Nuclear Proliferation,” American Journal of Political Science, Vol. 58, No. 3, pp. 705-
719.

210

Appendices
Appendix A: Nuclear Reversals and Nuclear Weapon States, 1945-2020
Reversals Non-Reversals (up to 2020)
Algeria (1983-1991) China (1955-Present)
Argentina (1968-1990) France (1946-Present)
Australia (1956-1973) India (1945-Present)
Brazil (1955-1990) Iran (1974-2015)*
Canada (1944-1969) Israel (1949-Present)
Egypt (1955-1980) North Korea (1959-Present)
Indonesia (1965-1967) Pakistan (1957-Present)
Iraq (1976-1995) Soviet Union/Russia (1945-
Present)
Italy (1955-1958) United Kingdom (1945-
Present)
Libya (1970-2003) United States (1945-Present)
Japan (1945-1970)
Norway (1949-1962)
Romania (1985-1993)
South Africa (1969-1993)**
South Korea (1959-1978)
211

Spain (1959-1975)
Sweden (1954-1969)
Switzerland (1946-1969)
Syria (1976-2007)
Taiwan (1967-1988)
West Germany (1957-1958)
Yugoslavia (1954-1998)
Notes: *Mehta codes Iran as case of reversal in 2015. However, recent events might require rethinking Iran’s
classification. **South Africa is the only country that obtained nuclear weapons and subsequently dismantled its
arsenal.

Abstract (if available)

Abstract Why do states reverse their nuclear weapons programs? Much has been written on why states might not pursue nuclear weapons, but relatively little about state’s decisions to reverse course and terminate their nuclear weapons programs, a puzzling decision considering the strategic advantages of obtaining nuclear weapons in the international system. Theoretically, it is puzzling why states that have invested time and money in pursuit of nuclear weapons would give up their efforts, despite the sunken costs involved and the strategic advantages of nuclear weapon ownership. This project presents a theory of nuclear weapons reversals to explain why countries give up their efforts to pursue the nuclear bomb, arguing that there are two main causal mechanisms that explain nuclear reversals: an involuntary and a voluntary pathway. Countries that follow the involuntary pathway give up pursuit of nuclear weapons due to threats or coercive force. Countries that follow the voluntary pathway give up pursuit of nuclear weapons due to acceptance of nuclear non-proliferation norms, norms that are accepted if two conditions are met: I hypothesize that when states perceive a non-threatening security environment, and when states have higher levels of democracy, states will be more likely to be influenced by international norms against nuclear weapons acquisition, and thus more likely to terminate their nuclear weapons development program. This project seeks to 1) develop a theory of nuclear reversals based on an analytical eclectic approach, drawing from realist, liberal, and constructivist insights, 2) develop a proxy measure of anti-nuclear proliferation norm acceptance, 3) use quantitative methods to show there is a relationship between norm acceptance and reversals, and 4) show through the use of case studies how the voluntary and involuntary mechanisms work in real world cases of nuclear reversals and non-reversals.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Lesser evils? WMD pursuit beyond nuclear weapons

PDF

The paradox of personalism: internal political constraints on nuclear posture in authoritarian regimes

PDF

When bullets are not enough: determinants of chemical weapons usage and abstention

PDF

Refugee protection and empowerment joint venture: understanding how and why public and private actors cooperate to tackle the global refugee crisis

PDF

Selling virtue: how human rights NGOs and their donors work together to create a better world ... for themselves

PDF

The IO contestation project: a multi-method approach towards understanding the ways states contest international organizations.

PDF

Things fall apart: the unraveling of international institutions through withdrawal

PDF

Taking orders from Brussels? external actors and regional organizations in the developing world

PDF

Power, profits, and politics: energy security and cooperation in Eurasia

PDF

Habits and policy: the socio-cognitive foundations of foreign policy stability

PDF

Legitimizing self-determination: advancing the sovereignty of separatist movements

PDF

From polluters to protectors: when energy firms support climate regulations

PDF

When passions run high: emotions and the communication of intentions in face-to-face diplomacy

PDF

Status, security, and socialization: explaining change in China's compliance in international institutions

PDF

The partial Good Samaritan states: China and Japan in the international relations of autocracy and democracy

PDF

Politics behind appointments: policy, patronage, and party systems in the Andes

PDF

Budget institutions and the positive theory of fiscal policy

PDF

The politics of transformation: mass media and the Northern Ireland peace process

PDF

International politics and domestic institutional change: the rise of executive war-making autonomy in the United States

PDF

Goldilocks’ signal for security cooperation in East Asia: China’s rise, hedging, and joint military exercises

Asset Metadata

Creator Gonzalez, Edward Alexander (author)

Core Title Pathways to nuclear weapon reversal: exploring mechanisms and understanding non-proliferation policy

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Degree Conferral Date 2023-05

Publication Date 03/13/2023

Defense Date 03/06/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag coercion,nonproliferation,norm acceptance,nuclear politics,nuclear reversal,OAI-PMH Harvest

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor James, Patrick (committee chair), Meshkati, Najmedin (committee member), Sandholtz, Wayne (committee member)

Creator Email edwardag@usc.edu,edwardbusiness07@yahoo.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC112764289

Unique identifier UC112764289

Identifier etd-GonzalezEd-11497.pdf (filename)

Legacy Identifier etd-GonzalezEd-11497

Document Type Dissertation

Format theses (aat)

Rights Gonzalez, Edward Alexander

Internet Media Type application/pdf

Type texts

Source 20230313-usctheses-batch-1009 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu