Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The paradox of personalism: internal political constraints on nuclear posture in authoritarian regimes
(USC Thesis Other)
The paradox of personalism: internal political constraints on nuclear posture in authoritarian regimes
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
i
THE PARADOX OF PERSONALISM: INTERNAL POLITICAL
CONSTRAINTS ON NUCLEAR POSTURE IN AUTHORITARIAN REGIMES
by
MATTHEW S. GRATIAS
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
POLITICAL SCIENCE AND INTERNATIONAL RELATIONS
May 2017
ii
Table of Contents
Figures iii
Abstract iv
Chapter One 1
The Puzzle of Nuclear Posture Formation
Chapter Two 24
Explaining Nuclear Posture Decisions: A Critical Review of the Literature
Chapter Three 75
A Theory of Internal Political Constraints on Nuclear Posture
Chapter Four 128
Concepts, Measures, and Case Selection
Chapter Five 197
The Soviet Union, 1949-1953
Chapter Six 245
Conclusion
Bibliography 256
iii
Figures
Table 1.1 Narang Typology of Nuclear Postures 7
Table 1.2 Nuclear Posture Options and Proliferation Consequences 12
Figure 3.1 Power-Sharing Relationships and Political Regime Type 113
Table 3.1 Political Constraints on Posture Dimensions by Regime Type 119
Table 4.1 Weeks Military Leadership and Personalism Index 164
Table 4.2 Civil-Military Patterns and Indicators 179
Table 4.3 Expanded Typology of Nuclear Postures 184
Table 6.1 Regime Type and Nuclear Posture 250
iv
Abstract
The importance of internal political organization for nuclear behavior has not
always been well understood. Indeed, decades of qualitative and quantitative
research on the correlates of nuclear proliferation have consistently found little to
no independent influence of political regime type on a state’s original decision to
pursue the bomb or nuclear behavior after crossing the nuclear threshold. These
studies have traditionally relied on dichotomous measures of democracy or
autocracy. This dissertation argues that the reliance on dichotomous measures of
regime type has obscured theoretically important variation in elite power-sharing
relationships among distinct types of authoritarian regimes. And it is these elite
power-sharing relationships that have tremendous influence on a regime’s nuclear
operations after crossing the nuclear threshold.
1
Chapter One: The Puzzle of Nuclear Posture Formation
In 2013 during North Korea’s “Victory Day” parade a truckload of
commandos were spotted wearing large backpacks festooned with brightly colored
radiation symbols. This curious sight – along with other regime propaganda art
depicting commando teams parachuting into enemy territory with similar
backpacks – spurred debate among nuclear security scholars as to whether North
Korea was threatening to one day develop and deploy miniaturized “suitcase”
nuclear weapons.
1
To be sure, miniaturizing a nuclear warhead to fit in a backpack
constitutes a significant technological and engineering feat – an accomplishment
that is most likely far beyond North Korea’s current technical abilities.
2
Yet the
potential threat of small-yield nuclear weapons widely dispersed to special
commando teams urgently raises a series of questions about how North Korea’s
supreme leader, Kim Jong-un, will operationalize his nascent nuclear capabilities.
What types of warheads and delivery vehicles will North Korea build? How tightly
will this emerging nuclear arsenal be managed? Will political or military authorities
retain custody of the weapons? And what is the intended purpose of the arsenal: to
deter a nuclear opponent via a threat of strategic retaliation or to prevent a
1
William C. Potter and Jeffrey Lewis, “Cheap and Dirty Bombs: Could These Creepy Chest Packs
be North Korea’s Way of Threatening Radiological War?” Foreign Policy. Available at:
http://foreignpolicy.com/2014/02/17/cheap-and-dirty-bombs/
2
https://www.defense.gov/Portals/1/Documents/pubs/Military_and_Security_Developments_Invol
ving_the_Democratic_Peoples_Republic_of_Korea_2015.PDF
2
conventional invasion by using tactical weapons against troop concentrations and
armor on the battlefield?
Taken together, these are questions of nuclear posture, the day-to-day
operationalization of state’s nuclear weapons capability. A state’s nuclear posture
is a combination of several discrete political decisions, including, but not limited
to, the allocation of economic resources to develop and build specific weapon
capabilities, doctrine about when to use nuclear weapons, how deeply to integrate
new nuclear forces with conventional forces, and management decisions about
which state institution should maintain custody of the weapons. In short, political
leaders in emerging nuclear states must decide how deeply to embrace the nuclear
revolution.
3
Importantly, not all states have adopted similar postures. Some states
have developed thousands of nuclear warheads of varying yield, myriad land, air,
and sea-based delivery vehicles, and deployed nuclear assets to far-flung locales
around the world, whereas others have only built a handful of crude devices that
were kept disassembled in storage vaults. Understanding why some states have
opted to more fully embrace the nuclear revolution than others is the central task of
this dissertation.
What is Nuclear Posture?
3
Robert Jervis, The Meaning of the Nuclear Revolution: Statecraft and the Prospect of
Armageddon (Ithaca and London: Cornell University Press, 1989).
3
There is a marked tendency for scholars of international relations to treat
nuclear weapon status as a simple binary variable: does the state possess a nuclear
weapon or not?
4
Yet a binary conception of nuclear weapon status obscures the
tremendous variation in existing nuclear capabilities among the ten nuclear
weapons states. To put it simply, some nuclear weapon states are much more
nuclearized than others in terms of the size and technological sophistication,
diversity of delivery vehicles, and degree of integration of their respective arsenals
with conventional military operations. Not only does a binary conception of nuclear
weapon status completely misses this variance, it also creates a vast and
unnecessary chasm between the academic study of nuclear security and the policy-
making, military, and intelligence communities whose mission it is to assess the
impact of small variations in nuclear capabilities.
5
Leaders of nuclear states face at least three core policy decisions when
crossing the nuclear threshold. First, they must choose how many and what types
of nuclear weapons and delivery vehicles to build. Second, they must consider
under what circumstances and for what purposes these fearsome weapons may be
used. Third, they must decide how to deploy and defend their nuclear arsenal – to
ensure that the ultimate weapon is ready to be used when properly ordered but not
4
For an example, see Michael Horowitz, The Diffusion of Military Power: Causes and
Consequences for International Politics (Princeton: Princeton University Press, 2010).
5
Francis J. Gavin and James B. Steinberg, “Mind the Gap: Why Policymakers and Scholars
Ignore Each Other, and What Should be Done About It,” Carnegie Reporter 6, no. 4 (Spring
2002), < http://carnegie.org/publications/carnegie-reporter/single/view/article/item/308/>.
4
subject to unauthorized use or theft. Taken as a whole these dimensions of nuclear
operations constitute a state’s nuclear posture – defined as the “capabilities,
deployment patterns, and command and control procedures a state uses to manage
and operationalize its nuclear weapons capability.”
6
Far from a simple binary
decision to “go for the bomb,” posturing nuclear weapons requires a continuing
process of critical policy decision-making after the first successful nuclear test. In
short, these are distinctly political decisions that affect multiple stakeholders within
the proliferating state and have major ramifications for civilian and military elites.
Recent studies in the field of nuclear security have introduced innovative
multi-dimensional conceptualizations of nuclear posture, thereby increasing the
ability of scholars to recognize theoretically important variance in nuclear behavior.
Vipin Narang has proposed a typology of nuclear postures based on four
dimensions: (1) Doctrine/Primary envisioned employment mode: plans for when,
how, and under what conditions the weapons will be used for a specific purpose
(break out to catalyze third-party intervention; nuclear retaliation; nuclear first-use
for denial); (2) Capabilities: the mix of weapons and delivery vehicles (a few
unassembled weapons; survivable second-strike forces; tactical nuclear weapons);
(3) Management: deployment patterns and command-and-control procedures for
6
Vipin Narang, Nuclear Strategy in the Modern Era: Regional Powers and International Conflict
(Princeton and Oxford: Princeton University Press, 2014). See also, Vipin Narang, “Posturing for
Peace?: Pakistan’s Nuclear Postures and South Asian Stability,” International Security 34, no. 3
(Winter 2009/10), p. 41.
5
the weapons in peace and war time (recessed and opaque; assertive civilian control;
delegative); and (4) Transparency level: how much information is provided to rivals
on both capabilities and deployment patterns (ambiguous capability and
deployment; unambiguous capability and ambiguous deployment; unambiguous
capability and deployment). The combination of values on these four dimensions
creates three categories of nuclear postures: catalytic, assured retaliation, and
asymmetric escalation.
7
For instance, an asymmetric escalation posture is a highly-
delegative, nuclear war-fighting posture which features tactical nuclear weapons
fully integrated into a regime’s military forces and intended for first-use on the
battlefield in a denial mission against an enemy’s conventional forces (armor,
infantry, or naval assets). Pakistan’s nuclear posture after its 1998 nuclear tests is
an example of an asymmetric escalation posture.
In contrast, an assured retaliation posture is a highly-centralized, defensive
nuclear posture which features strategic nuclear capabilities (often de-alerted and
de-mated), intended to be used only in retaliation following an opponent’s first
strike. Put differently, an asymmetric escalation posture tends heavily toward the
“always” side of “always/never” dilemma, eschewing negative controls on the
arsenal to ensure prompt usage when necessary, whereas an assured retaliation
posture is the reverse, skewing toward the “never” side through the use of extensive
7
Narang’s classification scheme applies only to regional power nuclear postures (the US, UK, and
USSR/Russia are excluded).
6
negative controls on the arsenal.
8
India’s nuclear posture after 1998 and China’s
nuclear posture since its 1964 test are both examples of an assured retaliation
posture.
A catalytic nuclear posture consists of an ambiguous, presumably un-tested,
nuclear weapons capability or secret arsenal, which the catalytic state threatens to
reveal to a patron – most often the U.S. – as a political signal to trigger intervention
in order to achieve de-escalation. The catalytic posture requires few, if any, actual
weapons, but critically requires that the catalytic state has, or at least perceives itself
to have, a powerful patron with regional interests significant enough to warrant the
belief that the patron will indeed intercede on its behalf in a crisis. Other important
features of this posture include: the catalytic signal is sent to the patron and not the
envisioned opponent, command and control of the nuclear assets remains highly
centralized, and nuclear forces are not integrated within the catalytic state’s military
doctrine. The catalytic state’s nuclear assets are viewed as “weapons of last resort,”
and it is the implicit threat to use these weapons in a severe crisis that incentivizes
the patron to intervene and de-escalate any crisis before such a fateful point is
reached. Empirical examples of regional nuclear powers that have adopted a
catalytic nuclear posture include Israel (1967-1991), Pakistan (1990-1999), South
Africa (1979-1990), and potentially North Korea (1994-2006).
8
Peter Feaver formulated the “always/never” dilemma in “Command and Control in Emerging
Nuclear Nations,” International Security vol. 17, no. 3 (1993): 160-187.This will be discussed at
great length in chapter two.
7
The typology of nuclear postures in displayed in Table 1.1 below.
Table 1.1 Narang’s Typology of Nuclear Postures
9
Catalytic Assured Retaliation
Asymmetric
Escalation
Primary envisioned
employment mode
(Use-Doctrine)
Break out
capabilities to
catalyze third-party
intervention
Nuclear retaliation
following highly
significant damage
Nuclear first use in a
conventional denial
mission
Capabilities
Ability to assemble
a handful of
nuclear weapons
Survivable second-
strike forces
First-use capabilities
(tactical nuclear
weapons) and
survivable secure-
second strike forces
Management
Recessed and
opaque
Assertive civilian
control
Delegative (assets
and authority
integrated into
military forces and
doctrine)
Transparency
Ambiguous
capability and
deployment
Unambiguous
capability; ambiguous
deployment
Unambiguous
capability and
deployment
9
Vipin Narang, Nuclear Strategy in the Modern Era: Regional Powers and International Conflict
(Princeton and Oxford: Princeton University Press, 2014).
8
The Significance of Nuclear Posture Formation: Why Study It?
In simplest terms, nuclear posture measures the amount and type of
destructive damage a nuclear state can impose on a rival state and under what
conditions it may do so. A security rival state that developed a massive, diverse,
and forward-deployed nuclear arsenal – like the United States or the Soviet Union
– would be orders of magnitude more worrisome to policymakers than a state – like
South Africa – that developed only seven untested and disassembled nuclear
devices.
10
To a large extent, then, estimates of the strategic consequences of nuclear
proliferation are dependent on implicit predictions of what type of nuclear posture
a state may eventually adopt. And estimates of the consequences of nuclear
proliferation have influenced a host of important international security outcomes,
including preventive war.
11
10
David Albright, “South Africa and the Affordable Bomb.” Bulletin of the Atomic Scientists,
July/August 1994. South Africa may have conducted a joint test with Israel in 1979. See “The
Vela Incident: Nuclear Test or Meteoroid, available at
http://nsarchive.gwu.edu/NSAEBB/NSAEBB190/.
11
Nuclear nonproliferation was cited as the original justification for the 2003 U.S. invasion of
Iraq. Fears that Saddam Husayn would one day use nuclear weapons either directly against Israel
or as a shield to insulate his regime from conventional reprisals following aggressive military
actions against other regional states prompted U.S. officials to discount decades of research on
deterrence stability and instead launch an immediate invasion. Yet granting that Saddam harbored
revisionist intentions, he would still have needed to develop very specific nuclear capabilities and
supporting command and control infrastructure to be able to achieve these goals. If Iraq had
merely limped across the nuclear threshold – and it is far from certain that Iraq’s scientists would
have been able to even achieve this goal – it is unclear if Saddam would have ever permitted
sufficient delegation of assets and authority to develop a first-use nuclear posture.
9
The relationship between implicit predictions of potential proliferator’s
nuclear posture and international stability is most apparent in U.S. nonproliferation
policy debate vis-à-vis the “rogue regimes” of the late twentieth and early twenty-
first century: Iraq, Iran, and North Korea. Although the basic argument applies to
all future proliferating states, the advocates of preventive military strikes against
Iran’s nuclear program have been most explicit in proposing seven potential
consequences of a nuclear-armed Iran. These are: 1) Nuclear Shield: Iran could
increase support for terrorist operations or possibly wage conventional proxy wars
from behind its new nuclear deterrent; 2) Proliferation Cascade: Iran’s acquisition
of nuclear weapons could spark a wave of new nuclear weapon programs in the
Middle East as Persian Gulf states attempt to internally balance Iran; 3)
Transfer/Loss: Iran could intentionally disperse nuclear weapons to its terrorist
proxies or a weapon system could potentially be lost and fall into the “wrong”
hands; 4) Damage the NPT and Nonproliferation Norm: Iran’s acquisition of
nuclear weapons from inside the NPT could shatter the nonproliferation norm and
fatally weaken the entire nonproliferation regime; 5) Regional Political Re-
alignment: the small states of the Persian Gulf (i.e. Gulf Cooperation Council)
could opt to bandwagon with a nuclear-armed Iran, limiting U.S. influence in the
region; 6) Inadvertent Escalation/Crisis Stability: any future Israel-Iran nuclear
deterrence relationship could be unstable and likely to end in conflict, perhaps even
10
nuclear war; 7) Nuclear War: Iran could do the unthinkable and be the first state
since the U.S. in 1945 to use nuclear weapons as an instrument of war.
12
These are serious consequences. If the likelihood of any or all of these
consequences was significantly high, a policymaker could indeed be seduced by
the apparent prudence of preventive military strikes to “solve” the Iranian nuclear
crisis.
13
Indeed, the former U.S. Ambassador to the United Nations, John Bolton,
has even formulated a doctrine of “pre-emptive self-defense” to justify military
U.S. or Israeli strikes on Iran’s nuclear program.
14
He has argued repeatedly that
the consequences of a nuclear Iran are sufficiently severe to warrant military
action.
15
Yet missing in the debate on the consequences of a nuclear Iran is any
reference to how a proliferating state’s nuclear posture may mitigate or exacerbate
the underlying risks associated with the seven proliferation concerns listed above.
In other words, what types of weapons it may build, how those weapons are
deployed and managed, and the doctrines for use – matters a great deal for any
12
Matthew Kroenig, “Time to Attack Iran: Why a Strike is the Least Bad Option,” Foreign Affairs
91, no. 1 (January/February 2012).
13
Although it is far from certain that an effective “military solution” exists. See, Kenneth M.
Pollack, Unthinkable: Iran, the Bomb, and American Strategy (New York: Simon & Schuster,
2013).
14
John R. Bolton, “Facing Reality on Iran,” National Review, September 7, 2015. Available at
http://www.nationalreview.com/article/422956/facing-reality-iran-john-r-bolton
15
It should be noted that preventive military strikes of the kind Bolton is advocating are likely
illegal under existing international law. See, Daniel H. Joyner, “The Implications of the
Proliferation of Weapons of Mass Destruction for the Prohibition on the Use of Force,” in Marc
Weller, ed. The Oxford Handbook of the Use of Force in International Law (Oxford: Oxford
University Press, 2015).
11
accurate assessment of the consequences of nuclear proliferation. Table 1.2
provides a summary of the theoretical relationships between nuclear posture and
proliferation consequences, with plus (+) signs indicating an increased likelihood
for that particular concern from a 50/50 base chance of occurrence, and minus (-)
signs indicating a decreased likelihood of occurrence.
16
16
Many of these seven concerns are disputed to begin with. Therefore, the plus/minus does not
indicate the specified is likely per se, just theoretically more likely in comparison to the other
postures.
12
Table 1.2 Nuclear Posture Options and Proliferation Consequences
As can be seen, the asymmetric escalation and full-spectrum escalation
postures increase the potential of negative consequences of nuclear proliferation.
The logic is relatively straightforward for the three issue areas exhibiting the most
variance: nuclear shield, inadvertent transfer/loss, and regional political re-
Consequence/
Posture
Catalytic
(Political)
Assured
Retaliation
(Nuclear
Defense)
Asymmetric
Escalation
(Regional Nuclear
War-Fighting)
Full-Spectrum
Escalation
(Global Nuclear-
War Fighting)
Nuclear Shield - - + +
Proliferation
Cascade
- +/- + +/-
Transfer/Loss:
Terrorist Proxy
- - + ++
Damage NPT/
Nonproliferation
Norm
- + + +
Regional Political
Re-alignment
- - + ++
Inadvertent
Escalation/Crisis
Instability
+ + + ++
Nuclear War - +/- + ++
13
alignment. For the potential nuclear shield effect, the capabilities and management
procedures associated with catalytic and assured retaliation postures reduce the
ability of a proliferating state to use nuclear weapons as a shield to protect it from
a rival’s conventional retaliation following a conventional provocation (or terrorist/
proxy attacks).
17
In other words, catalytic and assured retaliation postures do not
create a nuclear shield effect because the type and number of weapons built under
an assured retaliation posture (i.e. a handful of recessed, high-yield “city buster”
strategic weapons) – or perhaps not even fully assembled in a catalytic posture –
would be poorly suited to deter conventional reprisals and expressly not intended
to do so. Thus, only the asymmetric and full-spectrum escalation postures offer the
opportunity to use nuclear weapons as a shield against large-scale conventional
reprisal attacks because only those postures feature low-yield tactical weapons
dispersed and delegated for military use on the battlefield.
The same is true for the concern over the unintentional transfer or loss of
weapons to a third-party. Asymmetric escalation and full-spectrum escalation
postures likely increase the risks of both inadvertent transfer and accidental loss
simply because of the law of averages. A large, diverse, and widely dispersed
arsenal of weapons mated to delivery systems is far more challenging to secure
effectively than a handful of unassembled devices consistent with a catalytic
17
This is assuming the rival is also a nuclear state, e.g. Iran versus Israel. Or North Korea versus
the US.
14
posture, or even the relatively small number of recessed strategic weapons
associated with an assured retaliation posture. In short, because both catalytic and
assured retaliation postures privilege highly assertive and centralized control over
a state’s nuclear arsenal, the risk of accidental loss and rogue transfer is greatly
reduced (though clearly not eliminated entirely). Obviously, the same cannot be
said for the highly delegative asymmetric escalation posture or the global
deployments of the full-spectrum escalation posture, since such postures are
predicated on the availability and rapid first use of fully functional nuclear weapons
systems. With so many weapon systems “in play,” the probability of losing one or
having a rogue element seize one is noticeably increased.
In terms of regional political re-alignment, the asymmetric escalation and
full-spectrum escalation postures would theoretically generate the most pressure
for other states to bandwagon with an aspiring regional nuclear hegemon due to the
inherent aggressiveness and the unmistakable reliance on nuclear weapons entailed
in such posture. In other words, the more an aspiring regional nuclear hegemon
relies on nuclear weapons as war-fighting instruments to secure its interests, the
greater the pressure on regional states to react, either by bandwagoning or balancing
(either internally via bomb programs of their own or externally by deepening their
alliances with a nuclear patron, such as the US).
18
To be sure, states may resist
18
On the debate between balancing and bandwagoning, see Stephen M. Walt, The Origins of
Alliances (Ithaca, N.Y.: Cornell University Press, 1987); Randall L. Schweller, “Bandwagoning
15
band-wagoning with a powerful hegemon for a whole host of reasons (economic,
ideological, cultural). The point here is that among the various nuclear posture
options available, a full-fledged war-fighting nuclear posture – such as asymmetric
or full-spectrum escalation – generates increased potentialities for actual use, and
therefore greater impetus for regional political re-alignment since balancing the
regional nuclear hegemon is more costly and difficult.
In sum, variation in nuclear postures generate differential expectations of
proliferation consequences. Asymmetric and full-spectrum escalation postures are
by far the most destabilizing due to the potent combination of nuclear capabilities
fully integrated into military operations in a war-fighting role. In contrast, the
catalytic and assured retaliation postures ostensibly reduce many of the commonly
cited proliferation concerns. Therefore, the big question is what is the likelihood
that a proliferating state, such as North Korea, or potentially Iran, opts for an
asymmetric escalation or full-spectrum escalation posture? In other words, what are
the sources of nuclear posture formation?
Argument in Brief
Nuclear security scholars have appealed to various theoretical traditions to
explain (or sometimes ignore) variation in nuclear behavior. Structural realists have
for Profit: Bringing the Revisionist State Back In,” International Security, Vol. 19, No. 1 (Summer
1994), p. 72-107.
16
argued systemic variables are wholly determinative for nuclear behavior and that
variations in nuclear posture are simply epiphenomenal.
19
Organizational theorists
have emphasized the importance of bureaucratic biases and parochial institutional
interests in shaping a particular state’s nuclear policies.
20
Proponents of a strategic
culture approach have stressed the importance of a particular state’s culture and
history on nuclear decision-making.
21
More recently, nuclear security scholars in
the neoclassical realist tradition have proposed models of nuclear behavior that seek
to blend structural and unit-level variables – particularly civil-military relations –
to explain nuclear posture formation.
22
The central question is this: when is internal political organization –
broadly, political regime type – sufficient to determine a nuclear posture outcome?
The importance of internal political organization for nuclear behavior has not
always been well understood. Indeed, decades of qualitative and quantitative
research on the correlates of nuclear proliferation have consistently found little to
no independent influence of political regime type on a state’s original decision to
19
Kenneth N. Waltz, The Spread of Nuclear Weapons: More may be Better, Adelphi Papers no.
171 (London: International Institute for Strategic Studies, 1981).
20
Scott D. Sagan, The Limits of Safety: Organization, Accidents, and Nuclear Weapons
(Princeton: Princeton University Press, 1995).
21
Colin Gray, "National Styles in Strategy: The American Example," International Security, Vol.
6, No. 2 (Fall 1981); Colin Gray, Nuclear Strategy and National Style (Lanham, Md.: Hamilton
Press, 1986); Carnes Lord, "American Strategic Culture," Comparative Strategy, Vol. 5, No. 3
(1985); Richard Pipes, "Why the Soviet Union Thinks It Could Fight and Win a Nuclear War,"
Commentary, Vol. 64, No. 1 (July 1977), pp. 21-34. Alastair Iain Johnston, “Thinking about
Strategic Culture,” International Security Vol. 19, No. 4 (1995), pp. 32-64.
22
Vipin Narang, Nuclear Strategy in the Modern Era: Regional Powers and International Conflict
(Princeton and Oxford: Princeton University Press, 2014).
17
pursue the bomb or nuclear behavior after crossing the nuclear threshold.
23
These
studies have traditionally relied on dichotomous measures of democracy or
autocracy. But, as I will argue in subsequent chapters, the reliance on dichotomous
measures of regime type has obscured theoretically important variation in elite
power-sharing relationships among distinct types of authoritarian regimes. And it
is these elite power-sharing relationships that have tremendous influence on a
regime’s nuclear operations after crossing the nuclear threshold.
In this dissertation, I develop a theory of domestic political constraints on
nuclear posture development in authoritarian regimes. The theory postulates that
authoritarian regimes lacking functional intra-ruling elite power-sharing
institutions confront unique internal political constraints in developing, deploying,
and managing nuclear weapon capabilities required for first-use asymmetric or full-
spectrum escalation postures. This is because in the absence of authoritarian power-
sharing institutions a dictator can acquire an extreme concentration of political
power, which introduces significant distortions into multiple domains of the
regime, including civil-military relations, strategic assessment, and technological
23
Sonali Singh and Christopher Way, “The Correlates of Nuclear Proliferation,” Journal of
Conflict Resolution Vol. 48, No. 6 (2004): 859 -885. Dong-Joon Jo and Erik Gartzke,
“Determinants of Nuclear Weapons Proliferation,” The Journal of Conflict Resolution Vol. 51,
No. 1 (2007): 167-194). Karthika Sasikumar and Christopher Way, “Testing Theories of Nuclear
Proliferation: The Case of South Asia,” in Inside Nuclear South Asia (Palo Alto: Stanford
University Press, 2009): 68-105. Matthew Fuhrmann, “Spreading Temptation: Proliferation and
Peaceful Nuclear Cooperation Agreements,” International Security Vol. 34, No. 1(2009): 7-41.
Matthew Kroenig, “Exporting the Bomb: Sensitive Nuclear Assistance and Proliferation,”
American Political Science Review Vol. 103, No. 1 (2009): 113-33.
18
innovation. These distortions negatively affect each of the three core dimensions of
nuclear posture: capabilities, doctrine, and management.
First-use nuclear postures are costlier from an organizational, institutional,
and resource perspective. Political regimes vary in their abilities and willingness to
pay these costs. In severe instances, internal security trumps external security
because it is too politically costly for personalist dictators to reorganize regime
political interactions. Thus, civilian authoritarian regimes lacking functional
power-sharing arrangements not only typically opt for tighter central control of
their nuclear arsenals (management and deployment options), but also demonstrate
reduced diversity and complexity of weapon systems due to a decrease in
innovation potential (both doctrinally and technologically) and financial resources.
In other words, the absence of intra-ruling elite power-sharing institutions in
personalist authoritarian regimes poses a structural limit on the nuclear posture
options available to the dictator. This is what I term the paradox of personalism: by
purposefully degenerating power-sharing institutions to consolidate immense
power, a personalist dictator inadvertently reduces the regime’s ability to develop
a delegative nuclear posture. And without recognizing these limitations, scholars
and policy-makers consistently over predict the types of nuclear postures
personalist authoritarian regimes can develop.
19
Plan of the Dissertation
In chapter two I review the existing literature on nuclear behavior and
political regime type, starting with the legacy of the Cold War and followed by the
optimism-pessimism debate that raged in the early 1990s. The discussion then
proceeds to an analysis of the determinants of nuclear posture in the second nuclear
age.
24
With the exception of structural realism, existing models of nuclear posture
formation share a focus on a state’s civil-military relations as the most important
unit-level constraint on nuclear behavior. Yet in the quest for theoretical parsimony,
nuclear security scholars have relied on a thin version of civil-military relations
theory that fails to systematically embed the power-sharing relationship between
civilian elites and military elites within the larger power-sharing institutions and
structures of a particular political regime. This has led to the inappropriate transfer
of concepts explicitly derived from the study of civil-military relations in advanced
democratic regimes to civil-military relations under authoritarianism. In other
words, nuclear security scholars have failed to recognize that the way civilian
political elites share power with the masses and with each other has a tremendous
impact on how civilian political elites share power with military elites.
24
Paul Bracken, The Second Nuclear Age: Strategy, Danger, and the New Power Politics (New
York: Times Books, 2012). Toshi Yoshihara and James R. Holmes, eds., Strategy in the Second
Nuclear Age: Power, Ambition, and the Ultimate Weapon (Washington D.C.: Georgetown
University Press, 2012).
20
In chapter three, I develop a theory of internal constraints on nuclear posture
in authoritarian regimes by extending insights derived from the field of comparative
authoritarianism into the study of nuclear posture. I deduce the impact of specific
configurations of authoritarian political institutions on the types of nuclear weapons
and delivery vehicles a regime builds, how those forces are deployed, the doctrinal
concepts for use, and how tightly those weapons are controlled by central
authorities. I argue that binding internal constraints on nuclear operations are the
consequence of failed intra-ruling elite power-sharing institutions in authoritarian
regimes. The crux of the theory posits that personalist authoritarian regimes face
severe management and organizational challenges in adopting delegative nuclear
postures due to despotic civil-military relations that arise in the absence of intra-
ruling elite power-sharing institutions. The concentration of near absolute political
power in a single leader strongly affects intra-ruling elite interactions, severely
restricting the professional autonomy and influence of all other regime institutions,
including the military, economic, and scientific establishments. In sum, the more
broadly political power is shared throughout a political regime, the more delegative
and technologically developed a regime’s nuclear posture can be.
In chapter four, I focus on current conceptualizations of the personalist
regime concept, developing a minimalist definition of the personalist regime
concept that can be utilized as the independent variable in the empirical case studies
that follow. I also develop a suitable measurement index of a minimalist
21
conceptualization of personalism that can aid in political regime classification
schemes. Existing measurement indices are maximalist, meaning they mix multiple
conceptual dimensions of personalism (concentration of power, neopatrimonialism,
and dynastic sultanism) and include multiple indicators of civil-military relations
to classify a regime as personalist. Furthermore, I construct a four part typology of
civil-military relations that can be used as the intervening variable in the analysis
of the relationship between intra-ruling elite power sharing institutions and nuclear
posture constraints. The typology is constructed by combining different values on
the institutional identity of who exercises political control of the military in a
regime – the institutional identity of the political elite – along with the specific
methods and mechanisms of political control. In other words, the who and the how
of political control of the military are evaluated in tandem with disaggregated
authoritarian regime type. The final section of the chapter discusses research
method and case selection.
In chapter five, I apply my argument to the nuclear history of the Soviet
Union and trace the connection between power-sharing institutions, civil-military
relations, and nuclear posture in the Soviet Union under Stalin (1949-1953). The
Soviet Union’s nuclear arsenal was managed by six different paramount leaders:
Stalin (1949-1953); Khrushchev (1953-1964); Brezhnev (1964-1982); Andropov
22
(1982-1984); Chernenko (1984-1985); and Gorbachev (1985-1991).
25
I map the
intra-ruling elite power-sharing relationships under Stalin, with specific focus on
the functionality of power-sharing institutions within the Communist Party of the
Soviet Union (CPSU) and the chronology of power concentration and
consolidation. The purpose is to tease out the cyclical patterns of the power
relationships between the paramount leader and the CPSU, and to explore the
impact of the dual processes of CPSU institutional decay and regeneration on civil-
military arrangements.
Additionally, the civil-military relations under Stalin will be assessed and
categorized as delegative, assertive, despotic, or praetorian. I carefully evaluate
whether periods of despotic civil-military relations are preceded by the atrophy or
collapse of CPSU political institutions (Politburo, Central Committee, Party
Congresses, etc.). The nuclear posture milestones under each leader will be
chronicled. The link between conventional civil-military relations and nuclear
operations has been theorized to be direct. Delegative conventional civil-military
relations should predict delegative nuclear operations, all things being equal.
Likewise, assertive civil-military relations at the conventional level should
correspond to assertive nuclear operations. Therefore as I develop the narrative I
look for periods in Soviet history when nuclear operations were more delegative
25
Years of rule correspond with nuclear experience for all leaders except Stalin. He had ruled for
xx years prior to the USSR’s acquisition of nuclear weapons in 1949.
23
than conventional civil-military relations arrangements, especially any instances of
pre-delegation of nuclear assets and authority to use them.
In the sixth and concluding chapter, I reevaluate my argument and discuss
the implications of the dual trends of nuclear modernization programs and
increasing power concentration in Russia under Vladimir Putin and China under Xi
Jinping. Similarly, I will assess recent nuclear developments in North Korea under
Kim Jong-un in light of the paradox of personalism.
24
Chapter Two
Explaining Nuclear Posture Decisions: A Critical Review of the Literature
There is a long and distinguished literature on nuclear proliferation and
deterrence, but considerably less has been written about the critical nuclear posture
decisions faced by political leaders in newly proliferating states.
26
To be sure,
multiple theoretical models exist to explain why states choose to pursue nuclear
weapons in the first place.
27
However, these studies often tend to treat a single
nuclear device – or even sometimes just a significant quantity of fissile material –
as the most relevant empirical outcome of interest.
28
Of course no nuclear state has
ever built just a single nuclear weapon. On the contrary, nuclear states have
exhibited a wide variation in the size, diversity, management practices, and
deployment patterns of their respective arsenals.
26
For an excellent analysis of the deterrence literature, see Patrick M. Morgan, Deterrence: A
Conceptual Analysis (Beverly Hills, CA: Sage Publications, 1983) and Deterrence Now
(Cambridge: Cambridge University Press, 2003). See also, Complex Deterrence: Strategy in the
Global Age, T.V. Paul, Patrick M. Morgan, and James J. Wirtz, eds. (Chicago and London:
Chicago University Press, 2009).
27
Scott D. Sagan, “Why Do States Build Nuclear Weapons?: Three Models in Search of a Bomb,”
International Security 21, No. 3 (1996): 54-86; Etel Solingen, Nuclear Logics: Contrasting Paths
in East Asia and the Middle East (Princeton: Princeton University Press, 2007); Jacques E. C.
Hymans, The Psychology of Nuclear Proliferation: Identity, Emotions, and Foreign Policy (New
York: Cambridge University Press, 2006). Nuno P. Monteiro and Alexandre Debs, “The Strategic
Logic of Nuclear Proliferation,” International Security 39, Vol. 2 (2014) 7-51.
28
For a discussion of the nuclear threshold, see, Jacques E. Hymans, “When Does a State Become
a Nuclear Weapons State? An Exercise in Measurement Validation,” The Nonproliferation Review
vol. 17, no. 1 (2010), p. 161-180.
Jacques E. Hymans and Matthew S. Gratias, “Iran and the Nuclear Threshold: Where is the Line?”
The Nonproliferation Review vol. 20, no. 1 (2013), p. 13-38.
25
For instance, the Soviet Union developed thousands of nuclear warheads of
varying yield, including a full strategic nuclear triad of air, sea, and missile delivery
capabilities, tactical (battlefield) nuclear weapons, robust command-and-control
infrastructure, and eventually even space based early-warning systems. Moreover,
the Soviet Union deployed nuclear weapons outside its own territory, and may have
even briefly pre-delegated launch authority.
29
This stands in stark contrast to the
Chinese nuclear arsenal which for many decades numbered only a few hundred
large-yield strategic missiles, kept under tight control with the warheads separated
(de-mated) from delivery vehicles, and distinctly intended only to be used in a
retaliation mission.
30
What are the sources of these empirical variations in nuclear posture? Are
some states better prepared to adapt to the political exigencies of nuclear weapon
acquisition? Do certain types of political regimes face unique internal constraints
on one or more of the underlying dimensions of nuclear posture?
Nuclear security scholars have appealed to various theoretical traditions to
explain (or sometimes ignore) variation in nuclear behavior. Structural realists have
argued systemic variables are wholly determinative for nuclear behavior and that
29
For a comprehensive listing of Soviet/Russian nuclear weapons, see Pavel Podvig, ed., Russian
Strategic Nuclear Forces (Cambridge, Mass: the MIT Press, 2001).
30
M. Taylor Fravel and Evan S. Medeiros, “China’s Search for Assured Retaliation: The
Evolution of Chinese Nuclear Strategy and Force Structure,” International Security, vol. 35, no. 2
(Fall 2010), p. 48-87.
26
variations in nuclear posture are simply epiphenomenal.
31
Organizational theorists
have emphasized the importance of bureaucratic biases and parochial institutional
interests in shaping a particular state’s nuclar policies.
32
Proponents of a strategic
culture approach have stressed the importance of a particular state’s culture and
history on nuclear decision-making.
33
More recently, nuclear security scholars in
the neoclassical realist tradition have proposed models of nuclear behavior that seek
to blend structural and unit-level variables—particularly civil-military relations—
to explain nuclear posture formation.
34
Yet the importance of internal political organization – broadly, political
regime type – for nuclear behavior has not always been well understood. Indeed,
decades of qualitative and quantitative research on the correlates of nuclear
proliferation have consistently found little to no independent influence of political
regime type on a state’s original decision to pursue the bomb or nuclear behavior
after crossing the nuclear threshold.
35
These studies have traditionally relied on
31
Kenneth N. Waltz, The Spread of Nuclear Weapons: More may be Better, Adelphi Papers no.
171 (London: International Institute for Strategic Studies, 1981).
32
Scott D. Sagan, The Limits of Safety: Organization, Accidents, and Nuclear Weapons
(Princeton: Princeton University Press, 1995).
33
Colin Gray, "National Styles in Strategy: The American Example," International Security, Vol.
6, No. 2 (Fall 1981); Colin Gray, Nuclear Strategy and National Style (Lanham, Md.: Hamilton
Press, 1986); Carnes Lord, "American Strategic Culture," Comparative Strategy, Vol. 5, No. 3
(1985); Richard Pipes, "Why the Soviet Union Thinks It Could Fight and Win a Nuclear War,"
Commentary, Vol. 64, No. 1 (July 1977), pp. 21-34. Alastair Iain Johnston, “Thinking about
Strategic Culture,” International Security Vol. 19, No. 4 (1995), pp. 32-64.
34
Vipin Narang, Nuclear Strategy in the Modern Era: Regional Powers and International Conflict
(Princeton and Oxford: Princeton University Press, 2014).
35
Sonali Singh and Christopher Way, “The Correlates of Nuclear Proliferation,” Journal of
Conflict Resolution Vol. 48, No. 6 (2004): 859 -885. Dong-Joon Jo and Erik Gartzke,
“Determinants of Nuclear Weapons Proliferation,” The Journal of Conflict Resolution Vol. 51,
27
dichotomous measures of democracy or autocracy. But, as I will argue in
subsequent chapters, the reliance on dichotomous measures of regime type has
obscured theoretically important variation in elite power-sharing relationships
among distinct types of authoritarian regimes. And it is these elite power-sharing
relationships that have tremendous influence on a regime’s nuclear operations after
crossing the nuclear threshold.
Dictators and Nuclear Weapons
Studies of nuclear proliferation using disaggregated measures of
authoritarian regime type have demonstrated distinct differences in the probability
of certain types of authoritarian regimes pursuing nuclear weapons.
36
Personalist
regimes – authoritarian regimes featuring minimal institutional constraints on the
paramount leader’s decision-making authority – are more likely to pursue nuclear
weapons.
37
This is due to several factors: the lack of veto players able to block a
personalist dictator’s drive for the bomb and the powerful motive of regime – vice
No. 1 (2007): 167-194). Karthika Sasikumar and Christopher Way, “Testing Theories of Nuclear
Proliferation: The Case of South Asia,” in Inside Nuclear South Asia (Palo Alto: Stanford
University Press, 2009): 68-105. Matthew Fuhrmann, “Spreading Temptation: Proliferation and
Peaceful Nuclear Cooperation Agreements,” International Security Vol. 34, No. 1(2009): 7-41.
Matthew Kroenig, “Exporting the Bomb: Sensitive Nuclear Assistance and Proliferation,”
American Political Science Review Vol. 103, No. 1 (2009): 113-33.
36
Christopher Way and Jessica L. P. Weeks, “Making it Personal: Regime Type and Nuclear
Proliferation,” American Journal of Political Science Volume 58, No. 3 (2014): 705-719.
37
There are several different conceptualizations of “personalist” regimes. This will be discussed at
length in chapter four.
28
national – security. In short, authoritarian regime type, when specified correctly,
strongly influences the probability of nuclear proliferation.
This insight linking the internal politics of personalist regimes and nuclear
proliferation has yet to be extended into the study of nuclear posture formation.
Ironically, as I will argue in the next chapter, dictators who face the fewest
constraints on decision-making in pursuing the bomb actually face the greatest
constraints in posturing the bomb. This is a direct result of a personalist dictator’s
survival strategy, which requires reducing the autonomy and influence of all
potential elite rivals, particularly the military. This means that personalist dictators
will be less likely to adopt nuclear war-fighting postures requiring both the
deployment of nuclear weapons to front-line troops and the delegation of authority
to military commanders to use the weapons in an asymmetric role against invading
conventional forces.
Of course, no dictator – no matter how despotic – has ever carried out a
nuclear attack. This apparent nuclear restraint is all the more remarkable since
commentators on world politics have long assumed that dictators were predisposed
to aggressive behavior, both domestically and internationally.
38
But to date this
aggressiveness has not extended into the nuclear realm. It is now a truism that some
38
Recent research has supported this assumption, at least at the conventional level of conflict. See,
Jessica L. Weeks, “Strongmen and Straw Men: Authoritarian Regimes and the Initiation of
International Conflict,” American Political Science Review Vol. 106, No. 2 (2012): 326-347.
29
of the world’s most iconic dictators – Josef Stalin of the USSR and Mao Zedong of
the PRC – were remarkably cautious with their nascent nuclear arsenals.
In fact, as subsequent chapters will demonstrate, these notorious dictators
were often far more restrained in their nuclear behavior than their successors or
even their democratic counterparts; civilian dictators’ nuclear arsenals have, on
average, been kept under tighter central political control and in conditions of lower
operational readiness for more years and under more severe threat scenarios than
democratic arsenals. This discrepancy between a dictator’s conventional conflict
behavior and nuclear behavior has motivated a rich literature debating the strategic
consequences of nuclear proliferation and the sources of nuclear behavior.
39
This chapter reviews the existing literature on nuclear behavior and political
regime type, starting with the legacy of the Cold War and followed by the optimism-
pessimism debate that raged in the early 1990s. The discussion then proceeds to an
analysis of the determinants of nuclear posture in the second nuclear age.
40
With
the exception of structural realism, existing models of nuclear posture formation
share a focus on a state’s civil-military relations as the most important unit-level
constraint on nuclear behavior. Yet in the quest for theoretical parsimony, nuclear
39
The classic text, now in its third edition, is Scott D. Sagan and Kenneth N. Waltz, The Spread of
Nuclear Weapons: An Enduring Debate (New York and London: W.W. Norton & Company,
2013).
40
Paul Bracken, The Second Nuclear Age: Strategy, Danger, and the New Power Politics (New
York: Times Books, 2012). Toshi Yoshihara and James R. Holmes, eds., Strategy in the Second
Nuclear Age: Power, Ambition, and the Ultimate Weapon (Washington D.C.: Georgetown
University Press, 2012).
30
security scholars have relied on a thin version of civil-military relations theory that
fails to systematically embed the power-sharing relationship between civilian elites
and military elites within the larger power-sharing institutions and structures of a
particular political regime. This has led to the inappropriate transfer of concepts
explicitly derived from the study of civil-military relations in advanced democratic
regimes to civil-military relations under authoritarianism. In other words, nuclear
security scholars have failed to recognize that the way civilian political elites share
power with the masses and with each other has a tremendous impact on how civilian
political elites share power with military elites. And, as I will argue in the next
chapter, the absence of functional power-sharing institutions poses a severe
constraint on nuclear posture formation.
In particular, I am most interested in how the distinct challenges of
maintaining political power in dictatorships influences and constrains the way
nuclear dictators will manage their arsenals. By nuclear dictator, I mean the
paramount leader of the five nuclear authoritarian states at the time of their initial
nuclear weapons acquisition.
41
The primary reason for focusing on the early years
of a nuclear program is that the initial decision to cross the nuclear threshold is a
revolutionary political decision with unknown consequences. As Jacques Hymans
41
Dates for nuclear acquisition are from Sonali Singh and Christopher Way, “The Correlates of
Nuclear Proliferation,” Journal of Conflict Resolution, Vol. 48 No. 6 (2004): 859 -885. For an
alternate but similar set of nuclear program dates, see Dong-Joon Jo and Erik Gartzke,
“Determinants of Nuclear Weapons Proliferation,” The Journal of Conflict Resolution Vol. 51,
No. 1 (2007): 167-194.
31
has written, the decision to go nuclear is a veritable “leap in the dark.”
42
As such,
the early decisions that political leaders make about nuclear weapons – how exactly
to cross the nuclear threshold, if at all, and what capabilities to pursue, how to
manage those weapons, and under what circumstances to use them – can have a
lasting and outsized influence on subsequent nuclear posture options.
Cold War Legacy and the Optimism-Pessimism Debate
A central narrative of the Cold War was the global competition between a
handful of nuclear-armed democracies and dictatorships. Early Western analysts of
Soviet and Chinese nuclear operations were deeply concerned that the tyrannical
nature of these regimes would induce particularly reckless and aggressive nuclear
behavior.
43
Yet instead of Communist nuclear surprise attacks leading to
apocalyptic horror, the Cold War ended quietly. Nuclear deterrence was given
primary causal credit for keeping the “long peace.”
44
And the fact that the nuclear
peace has held strong for over seven decades – with multiple new entrants of into
42
Jacques E. C. Hymans, The Psychology of Nuclear Proliferation, p. 11.
43
Lawrence Freedman, The Evolution Nuclear Strategy, 3
rd
ed. (London: Palgrave Macmillan,
2003): 133-36. See also Francis J. Gavin, “Blasts from the Past: Proliferation Lessons from the
1960s,” International Security 29, no. 3 (Winter 2004/2005): 100-135. Francis J. Gavin, “Same As
It Ever Was: Nuclear Alarmism, Proliferation, and the Cold War,” International Security 34, no. 3
(Winter 2009/10): 7-37.
44
John Lewis Gaddis, “The Long Peace,” International Security 10, no. 4 (Spring 1986): 120-23;
“International Relations Theory and the End of the Cold War, International Security 17, no. 3
(Winter 1992/93): 5-58.
32
the nuclear club – has led to the conclusion that the political regime type of
emerging nuclear nations has little to no influence on nuclear behavior.
45
Yet there still remains considerable disagreement among scholars about the
dangers of nuclear weapons acquisition for international and regional security.
These conflicting views about the positive and negative consequences of nuclear
proliferation have commonly been subsumed within the “proliferation optimism-
pessimism” debate – beginning in earnest in the early 1990s and reappearing with
renewed vigor each time a new state crosses or even crawls up to the edge of the
nuclear threshold. The core of the debate addresses the question of the applicability
of the American and Soviet experience of nuclear stability (however precarious)
during the Cold War to emerging nuclear powers or whether the domestic
institutional heterogeneity of aspiring nuclear states could lead to increased risk of
deterrence failures, nuclear accidents, or unauthorized seizure.
46
In short, how and
why do the internal characteristics of nuclear states matter for nuclear behavior?
The contribution of stable nuclear deterrence to the “peaceful” resolution of
the Cold War has led prominent scholars to go so far as to advocate for the selective
spread of nuclear weapons into unstable regions regardless of a proliferators’
45
Kenneth N. Waltz, The Spread of Nuclear Weapons: More may be Better, Adelphi Papers no.
171 (London: International Institute for Strategic Studies, 1981).
46
Sagan and Waltz, The Spread of Nuclear Weapons: An Enduring Debate, 3
rd
ed. (New York and
London: W.W. Norton & Company, 2013). For an excellent critical review of the optimism-
pessimism debate, see Jeffrey W. Knopf, “Recasting the Proliferation Optimism-Pessimism
Debate,” Security Studies 12, no. 1 (2002): 41-96.
33
regime type or other unit-level characteristics.
47
This sanguine view is intimately
linked with structural realism – sharing the core theoretical assumption that
systemic variables are supreme in shaping a state’s security behavior.
48
It is
buttressed by the indisputable empirical fact that even though some of the world’s
most feared and brutal dictators have possessed nuclear weapons, they have never
used them in war. Nuclear weapons have only been used against populated areas
by the world’s oldest democracy, the US, and only under conditions of nuclear
monopoly. This leads political scientist Kenneth Waltz to conclude:
The history of the Cold War shows that what matters is not the character
of the countries that have nuclear weapons but the fact that they have
them. Differences among nuclear countries abound, but for keeping the
peace what difference have they made? Whatever the identity of rulers,
whatever the characteristics of states, the national behaviors they produce
are strongly conditioned by the world outside…In a nuclear world, any
state—whether ruled by a Stalin, a Mao Zedong, a Saddam Hussein, or a
Kim Jong Il—will be deterred by the knowledge that aggressive actions
may lead to its own destruction.
49
To be sure, Waltz’s optimism is not predicated on ignorance of the many
differences among nuclear countries. He readily admits to the remarkable
47
Waltz, The Spread of Nuclear Weapons: More may be Better. John J. Mearsheimer, “Back to
the Future: Instability in Europe After the Cold War,” International Security 15, no. 1 (Summer
1990): 5-56.
48
The link between rational deterrence and structural realism is often implied if not stated directly.
Deterrence was both a derivative of structural realism and perhaps its strongest empirical
evidence. Proliferation pessimists had to question structural realism in order to be able to question
nuclear optimism. Thus proliferation pessimism often contains an implicit critique of neorealism.
Robert Jervis, Richard Ned Lebow, and Janice Stein, eds., Psychology and Deterrence (Baltimore:
Johns Hopkins University Press, 1985).
49
Waltz, “Indian and Pakistan Nuclear Weapons: For Better or Worse?” p. 165. Emphasis mine.
34
institutional heterogeneity of existing nuclear regimes, but blithely dismisses any
independent influence such heterogeneity may have on nuclear behavior due to the
overpowering effect of external (structural) pressures.
50
According to David Karl,
the foundation for such “near-absolute claims” from proliferation optimists is the
“supposedly immutable quality of nuclear weapons: their presence is the key
variable in any deterrent situation, because fear of their devastating consequences
simply overwhelms the operation of all other factors.”
51
Similarity of nuclear
behavior among different regimes is assured by the all-powerful conditioning effect
of nuclear weapons possession.
In the language of causal inference, proliferation optimists posit a
universally applicable and strongly homogenous treatment effect for nuclear
weapons acquisition: nuclear deterrence essentially works the same way in every
possible scenario and with every possible actor. The mere fact that an opponent
possesses nuclear weapons is sufficient to deter both nuclear and conventional
conflict.
52
The “nuclear cure” overwhelms all other internal mitigating factors,
50
The affinity between neorealism and proliferation optimism, specifically the privileging of
external security pressures, has often been noted. For instance, see Feaver, “Optimists, Pessimists,
and Theories of Nuclear Proliferation Management,” Security Studies 4, No. 4 (Summer 1995),
pp. 760-762; Karl, “Proliferation Pessimism,” p. 91, fn#10.
51
Karl, “Proliferation Pessimism,” p. 91, emphasis mine.
52
There is still some controversy over the minimum number of nuclear weapons needed to obtain
the homogenous treatment effect of deterrence. Some scholars view even opaque nuclear
capabilities as sufficient. For discussion of existential deterrence, see McGeorge Bundy,
“Existential Deterrence and Its Consequences,” in Douglas MacLean, ed., The Security Gamble:
Deterrence Dilemmas in the Nuclear Age (Totowa, N.J.: Rowman and Allanheld, 1984) 3-13. See
also, Feaver, “Proliferation Optimism and Theories of Nuclear Operations.”
35
regardless of domestic institutional heterogeneity. Variance in nuclear behavior
among states should directly correlate to changes in their respective security
environments and nothing else. Proliferation optimists strongly dismiss political
regime type – or any other unit-level attribute – as an independent variable
influencing nuclear behavior.
In contrast, proliferation pessimists have long argued that the internal
characteristics of nuclear countries shape and constrain nuclear behavior and thus
affect the consequences of nuclear proliferation. Scholars working in the pessimist
tradition have emphasized how the potential organizational pathologies of certain
proliferating regimes create increased risks of deterrence failure, accidents, and
unauthorized seizure.
53
Using a combination of deductive theory and detailed case
studies of nuclear “near misses” during the Cold War, pessimists have advanced
the central claim that some future proliferators may be more dangerous than
53
The optimism-pessimism debate literature is quite large and still growing. Important works
include: David J. Karl, “Proliferation Optimism and Pessimism Revisited,” Journal of Strategic
Studies 34, no. 4 (2011): 619-641. Scott D. Sagan, ed., Inside Nuclear South Asia (Stanford:
Stanford University Press, 2009). Peter R. Lavoy, Scott D. Sagan, and James J. Wirtz, eds.,
Planning the Unthinkable: How New Powers Will Use Nuclear Biological, and Chemical
Weapons (Ithaca and London: Cornell University Press, 2000). Peter D. Feaver, “Command and
Control in Emerging Nuclear Nations,” International Security Vol. 17, no. 3 (1993): 160-187;
“Proliferation Optimism and Theories of Nuclear Operations,” Security Studies Vol 2, no. 3-4
(1993): 159-191; “Optimists, Pessimists, and Theories of Nuclear Proliferation Management,”
Security Studies Vol. 4, no. 4 (Summer 1995), pp. 754-772; “Neooptimists and the Enduring
Problem of Nuclear Proliferation,” Security Studies 6, no. 4 (Summer 1997): 93-125; Jordan Seng,
“Less is More: Command and Control Advantages of Minor Nuclear States,” Security Studies 6,
no. 4 (Summer 1997): 49-91. David J. Karl, “Proliferation Pessimism and Emerging Nuclear
Powers,” International Security 21, no. 3 (Winter 1996/97): 87-119. Peter R. Lavoy, “The
Strategic Consequences of Nuclear Proliferation,” Security Studies 4, no. 4 (Summer 1995): 695-
753.
36
others.
54
However, the metric by which proliferators are judged dangerous has
evolved over time. Cold War analyses were often marked with some degree of
ethnocentrism and perceptions of moral asymmetry between democracies and the
authoritarian regimes of Stalin and Mao (and their more politically constrained
successors).
55
Some observers believed the proliferation threats of the future would
be too poor and technologically backward to match the engineering feats of the
superpowers which ensured safe and secure nuclear arsenals.
56
Likewise, some
analysts contended the brutally repressive nature of Stalin’s Soviet Union and
Mao’s China demonstrated a dictator’s reduced sensitivity to costs (e.g. they valued
the lives of their citizens less than democratic leaders) and therefore an increased
ability to take risky action.
57
Post-Cold War nuclear pessimist scholarship shifted away from debates
about the risk-acceptance of Cold War dictators to focus on organizational
pathologies, civil-military relations, and the strategic culture of emerging nuclear
54
See footnotes #6 and #15. Although I am sympathetic to the pessimist argument, I am disturbed
by the seeming trend of each new proliferator being predicted as “the most dangerous of all.” For a
recent example, see Sagan’s (2013: 201) assessment of Iran as “the most dangerous proliferator of
all.”
55
For discussion of moral asymmetry, see Freedman, The Evolution Nuclear Strategy.
56
Waltz famously critiques ethnocentric views of early pessimists.
57
Richard Pipes, “Why the Soviet Union Thinks it Can Fight and Win a Nuclear War,”
Commentary 64, no. 1 (July 1977): 21-34. For a devastating critique of this view, see Raymond L.
Garthoff, “Mutual Deterrence and Strategic Arms Limitation in Soviet Policy,” International
Security, vol. 3, no. 1 (Summer 1978), p. 112-47. An exchange of these views was later compiled
and appeared as “A Garthoff-Pipes Debate on Soviet Strategic Doctrine,” Strategic Review, vol.
10, no. 4 (Fall 1982), p. 36-83.
37
nations.
58
In one sense this was a natural evolution following the “peaceful” end of
the Cold War; pessimist predictions about the various weaknesses of rational
deterrence theory (RDT) remained largely unfulfilled. Nuclear deterrence
apparently works, even if the theory has flaws. And, of course, the entire debate
was limited by the extremely small number of empirical cases – the classic small
“N” problem – on which analysts could test their favored theories.
Benefiting from declassified archival materials, the post-Cold War wave of
pessimist scholarship pushed the analytical focus deeper inside existing nuclear
states and simultaneously expanded the search for new outcomes of interest other
than nuclear use in war. A surprise Soviet first-strike was no longer the primary
concern for analysts. Instead, the issue at hand was to what extent emerging nuclear
nations would follow the example—both good and bad—of the U.S. and Soviet
Union. Would new nuclear nations replicate the Cold War arms race dynamic and
build large and complex arsenals? Or would the nuclear states of the future be
satisfied with just a few disassembled weapons? Would emerging nuclear nations
establish safe command and control arrangements? In short, the debate shifted to
questions of nuclear posture formation.
58
Peter R. Lavoy, Scott D. Sagan, and James J. Wirtz, eds., Planning the Unthinkable: How New
Powers Will Use Nuclear Biological, and Chemical Weapons (Ithaca and London: Cornell
University Press, 2000).
38
The Determinants of Nuclear Behavior
A major point of scholarly consensus that emerged from the classic Cold
War debates on nuclear stability is that stable deterrence at a minimum requires
nuclear states to possess secure, second-strike forces.
59
These second-strike forces,
and the command-and-control infrastructure necessary to launch them, must be
invulnerable to nuclear attack, thereby assuring the opponent that no enemy
surprise strike would go unpunished.
60
Furthermore, to achieve a modicum of
nuclear stability new nuclear states must build survivable arsenals that are not prone
to accidental or unauthorized use.
61
This leads to the conclusion that if the stability
of the Cold War is to be repeated, the three requirements of stability—invulnerable
forces, invulnerable command-and-control systems, and safe and secure arsenals—
must be replicated in emerging nuclear nations.
62
One of the central questions of the optimism-pessimism debate, then, is:
how easy or likely will it be for new nuclear states to satisfy the necessary
conditions for stability?
63
Optimists argue that it is relatively easy because obvious
incentives compel state leaders to appropriately respond to systemic pressures.
64
59
Albert Wohlstetter, “The Delicate Balance of Terror,” Foreign Affairs 37, no. 2 (January 1959):
211-34; Scott D. Sagan and Kenneth N. Waltz, The Spread of Nuclear Weapons, p. 20.
60
Bradley A. Thayer, “The Risk of Nuclear Inadvertence: A Review Essay,” Security Studies 3,
No. 3 (Spring 1994): 428-93.
61
Ibid.
62
This also requires the often unstated but critically important assumption of status quo actors.
63
Francis J. Gavin, “Politics, History, and the Ivory Tower-Policy Gap in the Nuclear Proliferation
Debate,” Journal of Strategic Studies 35, No. 4 (2012), p. 575.
64
This also assumes a status quo bias.
39
Pessimists cite a host of now familiar domestic political reasons why states may
fail to build safe and secure nuclear forces.
65
At bottom, these are competing
empirical claims about the internal or external sources of nuclear behavior.
Optimists have traditionally located the sources of nuclear behavior in
structural realism, arguing that systemic security variables shape and constrain the
nuclear policy decisions of all states.
66
It is the harsh reality of facing a nuclear rival
that compels states to adopt stability-inducing second-strike postures. Waltz writes,
“Of all the possible external forces, what can affect state behavior more strongly
than nuclear weapons?”
67
This may be true, but it is an entirely insufficient
theoretical explanation for the exhibited variation in nuclear policy among the
existing nuclear states. Of course, Waltz has never claimed that structural realism
was able to predict a given state’s specific foreign policy decisions.
68
But Waltz’s
lack of concern for any causal mechanism linking structural pressures to specific
nuclear behavior immediately raises questions about the theoretical foundations of
proliferation optimism.
The picture is further complicated by the empirical fact that nuclear states
have historically responded to similar security environments in markedly dissimilar
ways. For instance, the Soviet leadership resisted granting custody of nuclear
65
Sagan and Waltz, The Spread of Nuclear Weapons, p. 41-81.
66
Kenneth N. Waltz, Theory of International Politics, (Reading, Mass.: Addison-Wesley, 1979).
67
Sagan and Waltz, The Spread of Nuclear Weapons, p. 89.
68
Waltz, Theory of International Politics, p. 121-23.
40
warheads to the Red Army for many years longer than the U.S., under President
Truman, kept nuclear weapons in civilian (AEC) hands. This Soviet behavior
persisted for so long and under such intense nuclear threat from U.S. forces that
intelligence analysts were astounded by the behavior.
69
This leads pessimists to engage in a sustained critique of proliferation
optimism’s reliance on thin structural realist accounts of the systemic sources of
nuclear behavior. “Nuclear optimism,” as Peter Feaver writes, “…is predicated on
the assumption that states will adopt proper postures and appropriate nuclear
behavior; they might, but this begs a theory explaining how and why states choose
among different patterns of deployment and behavior.”
70
For the consequences of
nuclear proliferation to be as beneficial as optimists claim, new nuclear states must
translate systemic imperatives into very particular policy decisions about nuclear
operations. Optimists predict leaders will make the correct decisions because “they
have every incentive to do so.”
71
Pessimists doubt that systemic incentives
automatically translate into nuclear behavior. Political leaders, situated in and
potentially constrained by particular configurations of domestic political
institutions, make discreet decisions about how to deploy their state’s nuclear
forces.
69
Marc Trachtenberg, A Constructed Peace: The Making of the European Settlement 1945-1963
(New Jersey: Princeton University Press, 1999), p. 181-82.
70
Feaver, “Proliferation Optimism and Theories of Nuclear Operations,” p. 163.
71
Waltz, The Spread of Nuclear Weapons, p. 21.
41
The apparent insufficiency of systemic variables alone to account for the
full range of nuclear behavior has led pessimists to persistently call for a richer
theory of nuclear operations. “To understand variations in nuclear behavior,” Peter
Feaver argues, “a theory must thus incorporate a wider array of components,
including systemic and sub-systemic elements.”
72
In other words, Feaver –
anticipating the arrival of neoclassical realist theories of foreign policy – calls for
the blending of structural theories with unit-level variables to provide a more
predictive theory of nuclear behavior.
73
The empirical variance in nuclear postures
raises the question of why some nuclear states may select nuclear postures that are
considered suboptimal from a rational deterrence or strategic stability standpoint.
Although the nuclear postures of individual nuclear states have been subject to
tremendous scholarly attention, there have been relatively few attempts to develop
generalizable cross-national models of nuclear posture formation for emerging
nuclear powers.
74
These existing theoretical frameworks have introduced important
contributions, especially the explicit focus on civil-military relations as the most
72
Feaver, “Proliferation Optimism and Theories of Nuclear Operations,” p. 163-4.
73
Gideon Rose, “Neoclassical Realism and Theories of Foreign Policy,” World Politics 51
(October 1998): 144-172. Steven E. Lobell, Norrin M. Ripsman, Jeffrey W. Taliaferro, eds.,
Neoclassical Realism, The State, and Foreign Policy (Cambridge: Cambridge University Press,
2009).
74
Peter D. Feaver, “Command and Control in Emerging Nuclear Nations,” International Security
Vol. 17, No. 3 (1993a): 160-187; Peter D. Feaver, “Proliferation Optimism and Theories of
Nuclear Operations,” Security Studies Vol 2, No. 3-4 (1993b): 159-191; Avery Goldstein,
Deterrence and Security in the 21
st
Century: China, Britain, France, and the Enduring Legacy of
the Nuclear Revolution (Stanford, CA: Stanford University Press, 2000). Vipin Narang, Nuclear
Strategy in the Modern Era: Regional Powers and International Conflict (Princeton and Oxford:
Princeton University Press, 2014).
42
binding unit-level constraint on nuclear behavior, but they have serious
shortcomings.
First, the influence of political regime type is either woefully under-
theorized, willfully ignored, or purposely conflated with other concepts, such as
strategic culture. Second, with the exception of Feaver (1993b) and Narang (2014),
explanations tend to address only a single dimension of nuclear posture, such as
military doctrine or command and control. Third, and perhaps most importantly,
these models are built on a rather thin conception of civil-military relations theory
which fails to systematically embed the power-sharing relationship between
civilian elites and military elites within the larger power-sharing institutions and
structures of a particular political regime. This has led to an inappropriate transfer
of concepts explicitly derived from the study of civil-military relations in advanced
democratic regimes to civil-military relations under authoritarianism. This has
obscured the uniquely severe constraints civilian authoritarian regimes face when
seeking to formulate nuclear postures requiring delegation of assets and authority
to the military.
Civil-Military Relations and Nuclear Behavior
As noted above, the nature of a state’s civil-military relations has been the
unit-level variable of choice in pessimist and neoclassical realist studies of nuclear
43
behavior.
75
This reflects two complementary strands of pessimist research on civil-
military relations and nuclear weapons – one rooted in organizational theory and
the other in traditional civil-military relations theory. Both strands of research
emphasize the importance of civilian control of the military (or lack thereof) for
determining distinct empirical outcomes in the nuclear realm. Scholars favoring
organizational theory stress the importance of civilian control of the military in
constraining the assumed parochial and professional biases of military officers for
offensive doctrines, high-levels of operational autonomy, and conflict propensity.
76
The second strand of research, originating from traditional civil-military relations
theory rather than organizational theory, also recognizes the importance of civilian
control of the military for nuclear behavior, but offers new concepts of what
Huntington once termed the “varieties of civilian control.”
77
Civilian leaders in
some states grant tremendous autonomy to the military institution, but civilian
leaders in other regimes sharply curtail military autonomy. In other words, not all
civilian control is equal and how tightly (or viciously) civilians manage the military
will partially shape nuclear behavior.
75
Scott D. Sagan, ed., Civil-Military Relations and Nuclear Weapons (Palo Alto, CA: Center for
International Security and Cooperation, 1994).
76
Scott D. Sagan, “The Origin of Military Doctrine and Command and Control Systems,” in Peter
R. Lavoy, Scott D. Sagan, and James J. Wirtz, eds., Planning the Unthinkable: How New Powers
Will Use Nuclear Biological, and Chemical Weapons (Ithaca and London: Cornell University
Press, 2000): 16-46. Barry R. Posen, The Sources of Military Doctrine: France, Britain, and
Germany Between the World Wars (Ithaca and London: Cornell University Press, 1984).
77
Samuel P. Huntington, The Soldier and the State: The Theory and Politics of Civil-Military
Relations (New York: Vintage Books, 1957), p. 80.
44
Research in the organizational theory strand argues that the organizational
and institutional identity – civilian or military – of the ultimate decision-making
authority in the security sector has an independent influence on a state’s military
doctrine and command and control arrangements at both the conventional and
nuclear level. In essence, the central question is who decides security policy:
civilian political leaders or senior military officers? The underlying intuition
motivating nuclear behavior models informed by organizational theory is that the
greater influence and autonomy a military organization has in deciding and
implementing nuclear security policy, the greater likelihood that such policies will
reflect the assumed biases shared by elite military officers. As Scott Sagan has
argued, “Organizational theory … predicts that militaries are likely to hold
preferences in favor of offensive doctrines, preventive war, and decisive military
options. They are likely to support counterforce targeting doctrines and may not
build secure second-strike forces on their own accord.”
78
Taken together, these
predicted military preferences suggest that if military officers have ultimate
decision making authority for nuclear policy, the policy outcome will be uniformly
aggressive and potentially destabilizing nuclear behavior.
78
Sagan, “The Origin of Military Doctrine and Command and Control Systems,” p. 23.
45
Notwithstanding the debate over whether civilian elites or military elites are
more war prone,
79
the extent to which a military organization is able to turn its
predicted preferences for offensive nuclear doctrines, war-fighting capabilities, and
delegative command and control arrangements into actual policy is assumed to be
directly proportional to its political influence within a regime. When civilian
political leaders are the ultimate decision making authority in a polity they are
expected to function as a crucial constraint on military biases, re-introducing
important geopolitical dimensions to an otherwise excessively narrow military
strategic decision making process.
80
Yet even under civilian control, military
organizations still shape security policy via multiple pathways, ranging from
political brinkmanship, routine bureaucratic logrolling, or simply because the
technical complexity of the issues enable military leaders to exploit an inherent
informational advantage over civilian leaders.
Importantly, when a military assumes direct political control of a state, the
theoretical expectations generated by organizational theory reduce to an argument
about the importance of political regime type for nuclear behavior. But instead of
the traditional dichotomous political regime classification scheme of democratic or
autocratic, the theoretically important distinction for organizational theory is
79
For a review of the literature on the war proneness of civilian versus military elites, see Todd S.
Sescher, “Are Soldiers Less War-prone than Statesmen?” Journal of Conflict Resolution Vol. 48,
No. 5 (2004): 746-74.
80
Sagan, “The Origin of Military Doctrine and Command and Control Systems,” p. 18.
46
between military and civilian regimes. Thus, the nuclear behavior of military
regimes – particularly on issues of nuclear doctrine and command and control
arrangements – is predicted to substantively vary from civilian regimes due to the
assumed divergence of civilian and military preferences in combination with the
total lack of civilian constraints on policy formulation in a military regime. Since
military regimes are by definition autocratic regimes, nuclear security scholars
utilizing organizational theory are actually making an implicit claim that one
specific type of authoritarian regime – a military regime – should correlate with
distinct patterns of nuclear behavior.
However, simply placing military regimes in a separate analytical category
has the problematic side-effect of lumping all civilian regimes together in a rump
category. For instance, both the US and USSR have long been lumped together as
examples of strong civilian control, despite the fact that Soviet civilian political
leaders utilized radically different methods and tactics to control the Soviet military
than U.S. civilian political leaders did to control the American military. And
considering that the civil-military relations in civilian authoritarian political
regimes have historically exhibited variation over time, the question immediately
arises as to whether variation in modalities of civilian control correlate with certain
types of nuclear behavior.
It is precisely this puzzle of the empirical importance of the “varieties of
civilian control” for nuclear behavior that prompted Peter Feaver to introduce a new
47
typology of civilian control in his now classic study of U.S. nuclear operations
during the Cold War.
81
Feaver developed the concepts of “assertive” and
“delegative” control to use as a dichotomous dependent variable measuring the
amount of U.S. civilian interference in nuclear operations, particularly in the realm
of command and control arrangements. Arguing that the nuclear revolution had
reduced the applicability of the Clausewitzian and Huntingtonian division of labor
between civilians and the military, Feaver’s central claim is that the direct
involvement of U.S. civilian political leaders in military operations constituted a
new form of civilian control: “assertive control.”
82
In his original formulation of the concept, Feaver describes assertive control
as a particular type of civil-military relationship in which an advanced democracy
facing no credible coup threats from its military – in other words the state had
already solved the central civil-military relations problematique – opted to exert
greater direct oversight and control of the military insofar as nuclear weapons were
concerned. Assertive control has three characteristics: 1) it allows for qualitatively
separate civil and military institutions, but 2) it recognizes no clearly definable roles
for these separate institutions, and 3) it presumes conflict and strife between these
81
Peter D. Feaver, Guarding the Guardians: Civilian Control of Nuclear Weapons in the United
States (Ithaca and London: Cornell University Press, 1992).
82
Ibid., p. 7-12.
48
separate institutions.
83
In short, the concept of assertive control was expressly and
unabashedly tailored for the American nuclear experience.
It should be noted that Feaver explicitly created the concept of assertive
control to replace Huntington’s concept of “subjective” control of the military
because subjective control was such a poor descriptor of stable U.S. civil-military
relations.
84
As such, the assertive-delegative typology does not represent a full
theory of civil-military relations because a wide range of pathological civil-military
arrangements are purposely excluded from analytical focus.
85
Indeed, the
theoretical novelty of the assertive control concept derived primarily from its
application to U.S. civil-military relations – which prior to the nuclear revolution
had adhered to a rather strict Clausewitzian division of labor – and not from its
newness as such. Many civilian autocratic regimes have long violated a strict
division of labor between the military and civilian spheres.
Feaver and subsequent scholars have utilized the assertive vs. delegative
typology as both a dichotomous dependent and later an independent variable for
both democracies and autocracies even though it was originally tailored to describe
command and control arrangements in advanced democracies. For example, in an
83
Ibid., p. 9.
84
Ibid., p. 8.
85
In a subsequent work, Feaver offers a full theory of civil-military relations based on a principle-
agent model. But even this theory only explicitly deals with civil-military relations in established
democracies. Peter D. Feaver, Armed Servants: Agency, Oversight, and Civil-Military Relations
(Cambridge, MA: Harvard University Press, 2003).
49
early theoretical treatment of the challenges and tradeoffs emerging nuclear nations
will face, Feaver argues that the command and control arrangements of new nuclear
states will be determined by the interaction of two key factors: the time-urgency
quality of the arsenal and the pattern of civil-military relations.
86
Feaver’s
dependent variable is the assertive or delegative nature of the nuclear command and
control system, which he argues is a de facto resolution of the chief command and
control problem of interest – the always/never dilemma. He explains:
Leaders want a high assurance that the weapons will always work when
directed [positive control] and a similar assurance the weapons will never
be used in the absence of authorized direction [negative control]. Weapons
must be reliable: unlikely to fail at the moment when leaders want to use
them; safe: unlikely to detonate accidentally; and secure: resistant to
efforts by unauthorized people to detonate them.
87
The persistent problem is that measures taken to increase the ability to use nuclear
weapons when required (positive control) often simultaneously reduce the ability
to prevent unwanted or unauthorized use (negative control). The classic example is
how actions taken to reduce the decapitation threat – such as dispersal of weapons
and delegation of launch authority to military commanders – greatly increases the
risks of accidental and/or unauthorized use.
88
These tradeoffs confront political leaders in every nuclear country.
Command and control systems will either favor positive control at the expense of
86
Feaver, “Command and Control in Emerging Nuclear Nations,” p. 174.
87
Feaver, “Command and Control in Emerging Nuclear Nations,” p. 163.
88
Ibid., p. 165.
50
negative control or vice versa. Command and control arrangements that prioritize
positive control delegative and those that favor negative control are assertive. The
time-urgency quality of the arsenal is an indicator of the potential threat of a
decapitation strike faced by new nuclear nations. It incorporates the perceived
vulnerability of the arsenal – warheads, delivery vehicles, and command and
control system – to a rival’s preemptive strike and also more general perceptions
about the likelihood of conflict.
89
As such, the time urgency of the arsenal reflects
the severity of a state’s external threat environment.
New nuclear states – which are assumed to be lacking in financial resources
– facing a perceived time-urgent nuclear threat are predicted to react by dispersing
their arsenals and adopting a delegative command and control system because it is
a less expensive and technologically feasible option. In other words, the political
leaders of emerging nuclear states facing a decapitation threat will prioritize
positive control, ensuring their nuclear weapons can be used as soon as possible, if
needed. In such a scenario, analysts predict the weapons (warhead and delivery
vehicle) will be fully integrated into military operations, dispersed throughout a
state’s territory to increase survivability, with launch authority potentially
delegated to military commanders to decrease response time. Indeed, in the most
extreme situation a state may face a “use it or lose it” dilemma, creating incentives
89
As Feaver (1993a: 179) writes, “Vulnerability is a function of both one’s own capabilities and
the capabilities of one’s enemies.”
51
for a state to use its nuclear weapons at a very early stage in any unfolding conflict
or crisis situation.
However, external security pressures are only one side of the equation.
Feaver argues that internal factors – particularly a state’s civil-military relations –
constrain nuclear behavior because delegative command and control arrangements
are organizationally costly. Some leaders may not be willing to grant the necessary
level of autonomy to their military because of fears of a coup or other internal power
struggle.
90
Thus leaders of regimes with volatile civil-military relations may not be
willing to risk a delegative nuclear command and control system even if they are
under severe external threat; only regimes with stable civil-military relations or
military regimes are expected to accepts the risks of a delegative command and
control system. It is in this way that civil-military relations pose a strong internal
constraint on nuclear behavior.
This is a compelling and important insight, but the “key factors” framework
is problematic and theoretically unsatisfying because it does not provide a
resolution to the underlying question of nuclear behavior: when will internal threats
trump external threats in determining nuclear posture? With two opposing factors
pushing in opposite directions, the key factors framework does not permit an
assessment of the relative intensities of these opposing forces.
91
Moreover, Feaver
90
Feaver, “Command and Control in Emerging Nuclear Nations,” p. 174-180.
91
Feaver fully acknowledges this shortcoming.
52
does not test the framework against any empirical data on the command and control
systems of emerging nuclear powers, nor does he address other aspects of nuclear
behavior beyond command and control systems. And although Feaver hints at the
unique constraints that “totalitarian” regimes may confront in designing nuclear
command and control systems, the influence of domestic political institutions on
nuclear behavior remains considerably under-theorized.
92
Feaver subsequently offers a second model and expanded theory of nuclear
operations – defined as the “day to day management of nuclear weapons and the
doctrinal concepts by which the military use nuclear weapons in peacetime and
combat.”
93
Feaver offers this second model as an explicit counter to proliferation
optimism’s total reliance on structural/systemic variables to predict sanguine
nuclear behavior in emerging nuclear nations. A slate of nine variables are grouped
into three strategic “domains” – strategic systems, strategic environment, and
strategic culture – which each independently influence a state’s nuclear behavior.
The theoretical expectation is that variation in these three domains will shape how
a particular state responds to the nuclear revolution.
The strategic systems domain includes four variables that pertain to the
specific characteristics of a state’s nuclear arsenal: the numerical size, the type of
weapons (explosive yield amount), the diversity of delivery vehicles, and the nature
92
Feaver, “Command and Control in Emerging Nuclear Nations,” p. 176.
93
Feaver, “Proliferation Optimism and Theories of Nuclear Operations,” p. 164.
53
of the command and control system.
94
The core insight is that a state’s nuclear
behavior is independently shaped and constrained by these highly specific arsenal
and management characteristics. For instance, a state with a large and diverse
arsenal of variously sized weapons paired with both long-range and short-range
delivery vehicles has more strategic options and greater doctrinal flexibility than a
state with a handful of rudimentary weapons and only a couple of short-range
delivery systems.
Moreover, specific arsenal characteristics also interact with and influence
the nature of command and control systems, adding an additional layer of
complexity within the strategic systems domain. This is because the numerical size
of an arsenal, along with the range, promptness, accuracy, pre-launch survivability,
and post-launch survivability of the delivery systems can combine in unique ways
to generate pressure to adopt a more delegative or more assertive command and
control system.
95
For example, a state with an arsenal that features very low pre-
launch survivability would have greater incentive to adopt a more delegative
command and control system, whereas a state with high pre-launch survivability
could afford the delays associated with an assertive command and control
arrangement.
94
Ibid., p. 165.
95
Jordan Seng, “Optimism in the Balance: A Response to
Peter Feaver,” Security Studies, vol. 6, no. 4 (1997), p. 126-136.
54
Thus, as Feaver argues, “The strategic systems domain…greatly influences
the political-military behavior of a proliferating country…An analysis of nuclear
proliferation that glossed over these differences would likely misstate the range of
nuclear behavior one could expect from a proliferating country.”
96
This is
undoubtedly correct, but it obscures the fact that all four of the variables within the
strategic systems domain are outcomes of political decisions taken by a
proliferating country’s leaders. Yet it is unclear if a proliferating country first
designs and builds its arsenal to suit its preferred command and control system or
vice versa. In other words, the complex interaction of these four variables does little
to resolve the central question of the causal priority of specific variables; it only
raises awareness that these factors are in the causal mix.
Similarly, the strategic environment domain – consisting of geopolitical
stability and crisis stability – influences nuclear behavior by creating incentives to
adopt specific types of doctrines and command and control arrangements. A state
confronted with a precarious geopolitical situation would have clear incentives to
prepare for the worst by keeping open the option of using nuclear weapons early in
a crisis. However, early-use (or time-urgent) doctrines undermine crisis stability
and exacerbate geopolitical tensions since such doctrines generate strong incentives
for an opponent to likewise adopt an early-use posture to reduce its vulnerability to
a first-strike. And, as Feaver has long argued, time-urgent doctrines require
96
Feaver, “Proliferation Optimism and Theories of Nuclear Operations,” p. 167.
55
delegative command and control systems to be credible, which further increases the
risks of accidental or unauthorized use in a crisis.
Taken together, the strategic systems and strategic environment domains
illustrate the way in which external security pressures and the technological
particulars of a proliferating state’s nuclear arsenal can interact to create pressure
for early-use doctrines and delegative command and control systems. It is both a
challenge to proliferation optimism’s claim that the incentives for proper nuclear
behavior are always unidirectional, meaning the adoption of stability-inducing,
second-strike postures can be taken for granted, and a relatively sophisticated
account of the how external security threats can influence nuclear policy. Yet just
because the potential incentives exist to adopt early-use doctrines and delegative
command and control systems exist does not mean that particular proliferating
states will choose to do so. Indeed, what is actually under debate is not that a state’s
external security situation influences nuclear behavior – which is a shared
consensus among nearly all nuclear security scholars – but it is that internal factors
can likewise shape nuclear operations in important ways. In other words, a state’s
external security situation is not a sufficient explanation of nuclear behavior;
external security threat alone is often indeterminate.
For this reason, the third domain in Feaver’s model of nuclear operations is
explicitly focused on three domestic factors combined under the nebulous heading
of “strategic culture.” The strategic culture domain refers to: “1) the pattern of civil-
56
military relations; 2) the character of and relations among other governmental
institutions, and 3) the long shadow of military and political history which
influences attitudes about the usefulness of military force (including nuclear
weapons) as an instrument of state policy.”
97
To a large extent the strategic culture
domain is a repackaging of the central insights on civil-military relations and
nuclear behavior from Feaver’s earlier model. As he argues, “The way a new
proliferator wields its nuclear power will depend, in part, on the way civilian and
military leaders interact…[D]ifferent patterns of civil-military relations will
suggest different day-to-day management styles and, as a consequence, different
nuclear behaviors.”
98
The importance of this insight cannot be overstated.
99
Nuclear
behavior is directly constrained by the institutional interaction of civilian and
military elites.
However, in this second model Feaver redefines his original concepts of
assertive and delegative civilian control from a narrow dependent variable
describing nuclear command and control systems, and instead uses these terms as
an independent variable to describe overall patterns of civil-military relations. Now
assertive civil-military relations are simply “those in which the civilians exercise
close and direct control over military affairs.”
100
And delegative civil-military
97
Feaver, “Proliferation Optimism and Theories of Nuclear Operations,” p. 169.
98
Ibid.
99
It is the foundation for the theoretical propositions of the next chapter.
100
Feaver, “Proliferation Optimism and Theories of Nuclear Operations,” p. 170
57
relations “are characterized by a high degree of civilian trust and military
autonomy.”
101
This is a far thinner conceptualization of civilian control than Feaver
originally presented and it hardly amounts to a full theory of civil-military relations.
As applied to nuclear operations, the assertive-delegative pattern of civil-military
relations maps directly onto the assertive-delegative continuum of command and
control. States with assertive patterns of civil-military relations are expected to
adopt assertive arsenal management procedures and vice versa with delegative
patterns. The risk of conceptual confusion and theoretical tautology are quite
obvious when using such similar concepts as both independent and dependent
variables within the same model.
Furthermore, linking the thinner concept of assertive patterns of civil-
military relations with assertive command and control procedures actually obscures
Feaver’s original insight concerning the history of U.S. nuclear command and
control. This is because at different times the U.S. adopted assertive management
procedures in the nuclear realm even though overall civil-military relations
remained highly delegative. It was this divergence that prompted Feaver to develop
the assertive-delegative continuum in the first place. And it is this divergence
between patterns of civil-military relations and nuclear operations that is of most
concern when attempting to predict the nuclear behavior of emerging nuclear states.
Furthermore, it raises the theoretically important question of when and under what
101
Ibid.
58
conditions a state’s overall pattern of civil-military relations is a poor predictor of
nuclear behavior.
In other words, the central tension between internal and external factors not
only remains unresolved in Feaver’s second model but is actually considerably
exacerbated by the problematic reconceptualization and application of the overly
simplified assertive-delegative continuum. The same goes for the issue of relative
preference intensities: the definition of assertive control as simply “close and
direct” civilian control is too broad a concept to measure the intensity of preferences
and, most importantly, when external threat provides sufficient stimuli to alter those
preferences.
This has led to such gross conceptual stretching of the assertive concept that
it is now used to describe overall civil-military relations in stable democracies and
also authoritarian regimes that routinely execute senior military leaders for political
reasons. The concept of assertive civil-military relations has been used to describe
civil-military arrangements in diverse cases such as the US under President
Kennedy, China under Mao and his successors, the USSR under Stalin and his
successors, in Iraq under Saddam, and in contemporary democratic India. Whatever
the original intent of the concept, it is clear that assertive is currently used as a rump
category to describe civil-military relations in all but the most delegative regimes.
And since current models of nuclear behavior explicitly rely on civil-military
relations as a binding constraint on nuclear behavior, the continued use of the
59
muddled assertive concept leads analysts to consistently over-predict the
probability of new nuclear states, particularly civilian authoritarian regimes,
adopting delegative nuclear postures.
This problem resurfaces when, in the most comprehensive and sophisticated
model of nuclear posture formation offered to date, Vipin Narang adopts Feaver’s
re-conceptualization of assertive-delegative civil-military relations as the key
internal constraint restricting the adoption early-use postures.
102
Narang’s “rational
optimization theory of nuclear posture” focuses on what he terms regional powers,
the six nuclear states below the plane of the Cold War superpowers: France, China,
India, South Africa, Israel and Pakistan.
103
Narang defines posture as “the
capabilities, deployment patterns, and command and control procedures a state uses
to manage and operationalize its nuclear weapons capability.”
104
The classification
scheme includes four dimensions: (1) Primary envisioned employment mode (break
out to catalyze third-party intervention; nuclear retaliation; nuclear first-use for
denial); (2) Capabilities (a few unassembled weapons; survivable second-strike
forces; tactical nuclear weapons); (3) Management (recessed and opaque; assertive
civilian control; delegative); and (4) Transparency level (ambiguous capability and
102
Narang, Nuclear Strategy in the Modern Era, p. 37-39.
103
The United Kingdom and North Korea are excluded, as are the superpowers (US and USSR).
104
Narang, “Posturing for Peace?”, p. 41. See also, Narang, Nuclear Strategy in the Modern Era,
p. 14-23.
60
deployment; unambiguous capability and ambiguous deployment; unambiguous
capability and deployment).
The combination of values on these four dimensions creates three
identifiable postures: catalytic, assured retaliation, and asymmetric escalation. For
instance, an asymmetric escalation posture is a highly-delegative, nuclear war-
fighting posture which features tactical nuclear weapons fully integrated into a
regime’s military forces and intended for first-use on the battlefield in a denial
mission against an enemy’s conventional forces (armor, infantry, or naval assets).
In contrast, an assured retaliation posture is a highly-centralized, defensive nuclear
posture which features strategic nuclear capabilities (often de-alerted and de-
mated), intended to be used only in retaliation following an opponent’s first strike.
Put differently, an asymmetric escalation posture tends heavily toward the
“always” side of “always/never” dilemma, eschewing negative controls on the
arsenal to ensure prompt usage when necessary, whereas an assured retaliation
posture is the reverse, skewing toward the “never” side through the use of extensive
negative controls on the arsenal.
The management dimension is of particular significance for the current
project because it is this dimension that is most directly and clearly affected by the
nature of a regime’s civil-military relations. To be sure, nuclear posture is best
conceived as the integration of employment mode, capabilities, and management,
because all three dimensions are required to make a particular posture credible. This
61
is especially true for asymmetric escalation posture, which requires tactical (or
battlefield) nuclear capabilities to be forward-deployed to front line units along with
delegated authority to launch under particular conditions.
The causal model consists of four independent variables that regulate a
regional state’s nuclear posture: security environment, availability of a third-party
patron, civil-military arrangements, and relative resource endowment. The
variables are arranged into a hierarchical decision-tree according to causal priority,
with some variables affecting all nodes of the framework and others intervening
only at specific nodes. Consistent with a neoclassical realist approach, the security
environment variable is given pride of place. When the external security variable
takes its maximal value, it is wholly determinative for nuclear posture formation.
The primacy of the external security environment variable is best understood to
mean that this independent variable influences all four dimensions of nuclear
posture. In other words, when an emerging nuclear nation is faced with a nuclear-
armed conventionally superior proximate offensive threat the state’s leaders have
no choice but to adopt an unambiguous early-use posture with tactical weapons
capabilities widely dispersed and launch authority delegated to the military—or an
asymmetric escalation posture in his terminology.
105
In essence, this means that the external threat is so severe that it overwhelms
all potential internal political constraints. A state’s leaders will agree that the goal
105
Narang, Nuclear Strategy in the Modern Era, p. 35.
62
of their emerging nuclear arsenal should be to deter both conventional and nuclear
attacks, allocate adequate financial resources to build tactical nuclear weapons for
use on the battlefield, adopt appropriately delegative civil-military arrangements
(requiring endogenous reorganization of regime institutions, if necessary), and then
publicly reveal that they have taken these steps to establish credible deterrence.
This is quite an extraordinary claim. Yet it is tempered by the strict
specification of when the external security variable is wholly determinative. The
strict and falsifiable operationalization of the external security environment
variable is commendable because it is an important rejoinder to earlier
realist/optimist accounts that leave external security so ill-defined that it
approximates a mystical force. However, Narang’s specification of the external
security variable is so strict that only one historic case – Pakistan after 1998 – even
qualifies.
106
Thus, the nuclear postures of the remaining five regional nuclear
powers included in Narang’s universe of cases are determined by factors other than
external security threat. External security environment is given pride of place, but
as an explanatory variable it accounts for only a tiny portion of observed variance
in nuclear posture, a single case. It is a de facto resolution of the question of relative
intensity (external versus internal threat), but only applicable in the most extreme
106
Indeed, it is not clear how well it would apply to any emerging nuclear states on the near
horizon. South Korea faces a proximate, nuclear-armed rival, but its conventional forces are
superior in qualitative, if not quantitative, measures than North Korea. Japan is separated from its
rivals by bodies of water (stopping power of water). Israel and Iran are separated by two countries
and almost a thousand miles at their most proximate points.
63
case – Pakistan – whose behavior few analysts would dispute is motivated by
security concerns. For all other values of the security environment variable, it
remains unclear under what conditions external threat trumps internal threat.
The second independent variable in the model, the availability of a third-
party patron, is conceptualized as both a system-level and unit-level variable since
it may be reflective of formal alliance patterns or just a false perception held by the
leaders of an emerging nuclear state. The third-party patron variable allows states
to adopt a very particular nuclear posture – catalytic. Rather than developing
nuclear weapons for deterrence or denial, a catalytic state may not even build
nuclear weapons, but instead uses the ambiguity surrounding its nuclear energy
program to catalyze third-party, historically only U.S., political intervention for the
purposes of de-escalating a crisis. A state’s opaque nuclear capabilities are not
meant to be used militarily, but exclusively as a political signal to a third party and
not to their rival.
107
The weapons are publicly untested, not deployed or integrated
with conventional military forces, and perhaps not even assembled.
Narang’s model suggests the adoption of a catalytic posture was a rational
response for certain proliferating states – Israel, South Africa, and Pakistan pre-
1998 have all been considered opaque proliferators at one point in time – to the
107
In one sense, all nuclear weapons are inherently political. Yet a catalytic or opaque posture is
unique because the nuclear capability will be untested and not be weaponized. In other words, a
state might have real difficulty in executing a nuclear attack. This stands in stark contrast to other
nuclear postures that feature weapon systems designed for use, even if such use would be
catastrophic and therefore unlikely.
64
systemic constraints imposed on them by the United States. Indeed, the historical
record is clear that the U.S. actively sought to curtail the nuclear progress of these
states by threatening various economic penalties if these states publicized their
nuclear weapons capability via an explosive test, since such a test would contravene
U.S. nonproliferation policy.
108
In this sense U.S. opposition to open proliferation
could conceivably be viewed as a system-level constraint, since the proliferating
states in question were forced to develop their nascent nuclear postures under
significant external political pressure. But it also suggests that a catalytic posture is
a temporary option, a political compromise that is highly sensitive to both the
unique historical circumstances of the Cold War alliance system and to emerging
security threats; no proliferating state has persisted in the catalytic posture once a
rival has unmistakably gone nuclear.
109
Nor has any security rival of the U.S.
adopted a catalytic posture. A catalytic posture requires more than a “permissive
security environment” – by which Narang means the nuclear state does not face a
nuclear-armed conventionally superior proximate offensive threat – it requires an
outright benign nuclear security environment, meaning no nuclear-capable rival.
This makes sense because a catalytic posture is the most severe tradeoff of potential
deterrence, placing faith in a third party to resolve the state’s security crises.
108
Indeed, the U.S. imposed sanctions on Pakistan and India following their 1998 tests.
109
There is a widespread assumption that Israel will end their policy of opacity if Iran eventually
nuclearizes. See, Avner Cohen and Marvin Miller, “Bringing Israel's Bomb Out of the Basement:
Has Nuclear Ambiguity Outlived Its Shelf Life?” Foreign Affairs Vol. 89, No. 5
(September/October 2010), p. 30-44.
65
Moreover, the three proliferating states that have adopted catalytic postures
share another common feature that Narang fails to properly highlight: none of them
were signatories to the Treaty on Nuclear Nonproliferation (NPT) when they
developed nuclear weapons. In fact, the catalytic nuclear posture is perhaps best
viewed as a posture that is adopted primarily to reduce the political costs of
proliferation in the NPT era and not to catalyze U.S. political intervention. It is true
that during the Cold War the U.S. was intensely concerned that any nuclear use by
an ally might inadvertently trigger World War Three, which created incentives for
the U.S. to intervene and de-escalate various crises. Yet it is difficult to dismiss the
notion that the real purpose of a catalytic or opaque posture was to utilize U.S.
patronage to shield the proliferating state from a negative reaction of the
international community. As such, it is a continuation of proliferation hedging
strategy as much as a unique nuclear posture, and one that has only been of long-
term appeal to Israel since it has not faced a nuclear rival in the Middle East.
110
It
is a posture that is adopted when state leaders are attempting to avoid making
decisions about nuclear policy for as long as possible.
Thus the mere possible availability of a powerful patron is insufficient to
compel a proliferating state to adopt a catalytic posture; other conditioning factors
must be present to increase the attractiveness of this option, namely an extremely
110
This raises the question of whether a long-term strategy of opacity requires a linked policy of
preemptive military strikes against opponents nuclear programs.
66
benign nuclear security environment (no nuclear rival), a particular orientation to
the NPT (non-signatory), and a certain type of political relationship with the U.S.
(an arms-length ally).
111
In short, a permissive security environment and the mere
availability of a third-party patron alone are not sufficient explanations for the
nuclear behavior of Israel, South Africa, and pre-1998 Pakistan.
The third variable in the rational optimization theory – the nature of a state’s
civil-military relations – represents the primary intersection point between
Narang’s model and earlier accounts of nuclear behavior. For Narang, the nature of
a state’s civil-military relations is the most important unit-level variable regulating
the choice between a stabilizing assured retaliation posture and a potentially de-
stabilizing asymmetric escalation posture. States with assertive civil-military
relations “simply do not have the domestic structures or willingness necessary to
implement an asymmetric escalation posture – an assured retaliation posture is the
only option for states with assertive civil-military relations that lack a third party
patron.”
112
Only nuclear states with delegative civil-military relations can afford to
grant the requisite autonomy to the military institution to manage an asymmetric
escalation posture. Thus, in Narang’s model civil-military relations regulates
111
Narang (2014: 206) freely admits the rational optimization theory struggles to accurately
predict Israel’s continuing catalytic posture (e.g. post-Gulf War I), confirming to some extent
Avner Cohen’s claim that Israel is truly a sui generis proliferator. Avner Cohen, “Israel: A Sui
Generis Proliferator,” in Muthia Alagappa, ed., The Long Shadow: Nuclear Weapons and Security
in 21
st
Century Asia (Stanford, CA: Stanford University Press, 2008): 241-68.
112
Narang, Nuclear Strategy in the Modern Era, p. 37.
67
nuclear posture specifically at the management node, or command and control
arrangements. Civilian elites do not trust military elites to manage the nuclear
arsenal without strict constraints on deployment and launch authority.
Yet the lack of appropriate institutional support structures for delegative
civil-military arrangements is a separate and distinct problem from the willingness
to adopt delegative arrangements when adequate political structures exist. And it is
on this point that Narang’s uncritical adoption of Feaver’s reconceptualization of
the assertive-delegative continuum poses significant problems for the rational
optimization model. While noting the possibility that civil-military relations might
be a proxy for some other underlying structural feature that influences the political
relationship between civilian elites and military elites, Narang is surprisingly
unconcerned with those deeper causes. As he writes, “The key indicator of interest
with respect to choices about nuclear posture is whether the state evinces assertive
or delegative control over the military, not necessarily the features that might
produce the arrangement (particularly regime type; not all democracies evince
delegative structures, and not all autocracies have assertive civil-military regime
types).”
113
Such a view is only accurate in the narrowest sense when dichotomous
measures of regime type are used in combination with the previously discussed thin
version of the assertive-delegative continuum.
113
Narang, Nuclear Strategy in the Modern Era, p. 38-39.
68
To be sure, even democracies have at times been hesitant to grant full
autonomy to the military in the nuclear realm, as evidenced by Feaver’s early study
of U.S. nuclear behavior.
114
Conversely, military autocratic regimes by definition
cannot be labeled as featuring assertive civilian control, since the military is
exercising direct political rule. Therefore, in this narrow sense democracies have
been assertive at certain times and specific types of autocracies have been
delegative. But Narang’s formulation obscures the fact that civilian autocratic
regimes have assiduously avoided delegative civil-military arrangements.
Advanced democracies may have the option of seamlessly oscillating between
assertive and delegative civil-military relations, but civilian autocracies certainly
do not have the same luxury. As will be explained at length below, for a civilian
autocracy to shift from assertive control to delegative control would often require
changes to the internal institutional infrastructure of the regime.
115
And in the
special case of personalist leaders, such institutional reforms are extremely costly
from a political survival perspective. So costly, in fact, that such dictators simply
choose to accept a greater degree of external vulnerability than sacrifice reduced
internal control.
114
As President Truman noted, “You have got to understand that this isn’t a military weapon. It is
used to wipe out women, children and unarmed people, and not for military use. So we have to
treat this differently from rifles and cannon and ordinary things like that.” Harry S. Truman to
David Lilienthal. Quoted in Freedman, The Evolution Nuclear Strategy, p. 49.
115
For an account of how institutional stress creates incentives for institutional change, see Eric J.
Hamilton, “International Politics and Domestic Institutional Change: The Rise of Executive War-
making Autonomy in the United States,” Ph.D. Dissertation, University of Southern California,
Los Angeles, 2015.
69
In other words, it matters a great deal why a particular state would adopt
assertive civil-military arrangements because without interrogating the underlying
structural causes of particular civil-military arrangements, there is little ability to
understand how intensely civilian authoritarian leaders might resist shifting into a
different nuclear posture under varying levels of external threat. Narang is generally
unconcerned with the root causes of assertive civil-military relations because, as he
writes, “the same civil-military arrangements can be produced by a variety of
conditions – not just fear of coups – and vary across or within regime types,
[therefore] it is this indicator which is most relevant to predicting the types of
stewardship procedures a state might employ in managing its nuclear arsenal.”
116
Narang is clearly searching for the most parsimonious regulatory variable. But this
comes at the expense of explanatory power because removing civil-military
relations from the larger institutional context of the political regime while
simultaneously conceding that the political regime writ large influences the nature
of the civil-military relationship results in a severely muddled concept of civil-
military relations that obfuscates the relationship between civil-military relations
and a regime’s political power-sharing institutions.
The bottom line is that whole classes of political regimes, those lacking
functional power-sharing institutions, are extremely unlikely to exhibit delegative
civil-military arrangements, thus precluding them from adopting an asymmetric
116
Narang, Nuclear Strategy in the Modern Era, p. 38.
70
escalation posture in the absence of major institutional reform. And since such
reforms are essentially inconceivable in certain regimes, this is a far more severe
constraint than has been understood to date.
117
The focus on internal power-sharing
institutions has several implications for the study of nuclear posture formation. The
first implication is that democracies should have a distinct advantage in developing
delegative civil-military structures that many civilian authoritarian regimes lack.
Assertive civil-military relations in democracies are likely to be far less severe and
persistent than in civilian authoritarian regimes, posing only a minimal and
temporary constraint on nuclear posture decision-making. In terms of empirical
cases, this means that although Narang codes both China’s and India’s civil-
military relations as highly assertive – and thus the reason for their assured
retaliation postures – it is far more likely that India will shift toward a more
delegative posture under a similar level of external threat than China.
A second implication is that the exclusive focus on the narrow assertive-
delegative continuum as the primary unit-level regulator of nuclear posture is
misplaced. Instead, analysts seeking to predict the nuclear posture of future
proliferators should be examining two internal factors in tandem: the level of
military influence and the nature of power-sharing institutions in a political regime.
As will be discussed at great length in subsequent chapters, civil-military relations
operate much differently in political regimes featuring a full panoply of functioning
117
This is because such reforms would likely conflict with a dictator’s survival strategy.
71
elite power-sharing institutions than in those regimes where power is effectively
concentrated in a single individual leader.
The final variable in Narang’s rational optimization theory is relative
resource endowment. This variable influences nuclear posture at the capabilities
node. Relative resource endowment regulates nuclear posture options insofar as it
may be rational for a state with significant capital or technological constraints but
a comparative advantage in manpower or territory vis-à-vis its primary rival to
adopt an assured retaliation posture instead of the more far more expensive
asymmetric escalation posture. In other words, regional states with a relative
resource endowment of territory or manpower can opt out of the type of arms racing
behavior – characteristic of the Cold War – that could potentially bankrupt the
state.
118
However, states that have capital or technological resource advantages
against their primary rivals have the option of pursuing an asymmetric escalation
posture, assuming the state’s civil-military relations are sufficiently delegative.
Only democratic France post-1990 makes it all the way down the decision tree to
this node, although the model also predicts that Israel could one day move toward
an asymmetric escalation posture. A state that makes it this far down the decision
tree is essentially unconstrained in its nuclear posture options. And, as will become
clear in the discussion below, it is no accident that the states that reach this level
118
Narang, Nuclear Strategy in the Modern Era, p. 40.
72
feature political systems with established power-sharing institutions and open
economies.
To recap, the rational optimization model represents a significant theoretical
achievement over earlier models of nuclear behavior. It specifies under what
conditions it may be rational for states to “optimize” their nuclear postures by
accepting increased vulnerability in order to reap “savings” in organizational and
financial costs. In essence, the model is a series of potential tradeoffs between
external and internal threats, except for the single binding security environment of
facing a nuclear-armed conventionally superior proximate offensive threat which
compels the adoption of an asymmetric escalation posture. States adopting catalytic
or assured retaliation postures sacrifice the enhanced deterrence of an asymmetric
escalation posture because they essentially choose to prioritize internal political
factors over external threats. The core insight of the rational optimization model is
that given specific security environments it may be rational for a nuclear state to
sacrifice deterrence potential and adopt a more restricted nuclear posture.
The key is to understand how these tradeoffs function at various values of
both external and internal threat. One de facto resolution is to specify when external
or internal threat should dominate – the outer poles of continuum so to speak. This
can be done relatively easily for external threat and, indeed, Narang does precisely
this by specifying the binding condition of a nuclear-armed conventionally superior
73
proximate offensive threat. But what is the internal pole? When, if ever, should
internal concerns be determinate even when an external security is severe?
Currently, it seems that for theorists like Narang, a little bit of assertive
civilian control goes a long way in constraining nuclear posture options at relatively
low external threat levels. Yet, as discussed above, the definition of assertive
civilian control has been expanded almost to the point of incoherence and applied
so broadly across political regime types that its theoretical usefulness has been
degraded. This is made amply clear when the universe of cases is expanded beyond
Narang’s focus on regional powers and the U.S. and Soviet Union are included as
subjects of analysis. Despite having what Narang would no doubt code as highly
assertive civil-military relations, in addition to conventional military superiority
over the U.S., the Soviet Union amassed a massive, diverse, and widely-dispersed
nuclear arsenal. Although Soviet nuclear posture never exhibited the extreme levels
of delegation that characterized U.S. nuclear posture, it certainly exceeded the level
that would be expected to correspond with its assertive civil-military arrangements.
However, to fully understand internal political constraints on nuclear
posture formation requires a theory of nuclear posture formation that takes political
regime type seriously. The benefit of a political regime approach is that it can
provide analytical leverage on questions of how internal balances of power
influence multiple dimensions of nuclear posture formation, not just the
management dimension. In fact, the institutional structures of a political regime can
74
shape doctrinal decisions, innovation potential, decisions on weapon capabilities,
and even strategic assessment of threats. In other words, the internal institutional
structure of a political regime matters tremendously for nuclear posture decision-
making. Any model of nuclear posture formation that excludes political regime type
unnecessarily sacrifices analytical leverage on the question of when internal politics
can trump external security concerns.
Thus what is needed is a more comprehensive formulation of how civil-
military relations constrains nuclear posture in various types of authoritarian
regimes. This will require a more precise concept of political control of the armed
forces that is embedded within the larger institutional structures of a political
regime and is more sensitive to permutations and fluctuations in the internal balance
of power, enabling analysts to discern when shifts from specific patterns of civil-
military relations require additional internal institutional reforms. In short, we need
to better understand the unique way that the balance of internal political power
under specific types of authoritarian regimes shapes civil-military relations. It is to
this task of formulating a theory of internal political constraints on nuclear posture
that we now turn.
75
Chapter Three: A Theory of Internal Political Constraints on Nuclear Posture
When states cross the nuclear threshold the fearsome new weapons must be
integrated into existing political structures. As the previous chapter illustrated, the
question of how different political structures affect nuclear operations has been a
source of considerable disagreement among scholars of nuclear security. Decades
of research have isolated the importance of civil-military relations as one potential
internal constraint on nuclear operations. However, existing accounts of nuclear
posture formation often divorce civil-military relations from other political
structures. No theory exists that integrates civil-military relations with other
political power-sharing institutions to offer a coherent account of how regime type
shapes and constrains the nuclear posture options available to political leaders. This
chapter seeks to provide such a theory.
In this chapter I extend insights derived from the field of comparative
authoritarianism into the study of nuclear posture to deduce the impact of
configurations of specific authoritarian political institutions on the types of nuclear
weapons and delivery vehicles a regime builds, how those forces are deployed, the
doctrinal concepts for use, and how tightly those weapons are controlled by central
authorities. I argue that binding internal constraints on nuclear operations are the
consequence of failed intra-ruling elite power-sharing institutions in authoritarian
regimes. The crux of the theory posits that personalist authoritarian regimes face
76
severe management and organizational challenges in adopting delegative nuclear
postures due to despotic civil-military relations that arise in the absence of intra-
ruling elite power-sharing institutions. The concentration of near absolute political
power in a single leader strongly affects intra-ruling elite interactions, severely
restricting the professional autonomy and influence of all other regime institutions,
including the military, economic, and scientific establishments. In sum, the more
broadly political power is shared throughout a political regime, the more delegative
a regime’s nuclear posture can be. This means that it is unlikely a personalist
dictator will develop tactical nuclear weapons, integrate those weapons with
conventional military forces, and delegate first-use launch authority to military
commanders in the field.
A Political Regime Approach: Internal Political Constraints on Nuclear
Posture
Do civilian authoritarian leaders face unique internal political constraints
preventing them from adopting certain nuclear postures? The answer offered by
international relations scholars, as last chapter reviewed, is a resounding “maybe.”
Given the stakes involved with nuclear weapons, such an answer is obviously
insufficient. The muddled answer is reflective, at least partially, of a noticeable
impulse to avoid explicitly framing the nuclear posture debate in terms of political
regime type. This impulse was likely a corrective to decades of lackluster findings
77
in quantitative nuclear proliferation research using dichotomous measures of
regime type and also the lingering dominance of structural realism in international
security studies. Rather than jettisoning regime type as an analytical lens, however,
the better course of action is to develop a more suitable conceptualization of how
specific institutional configurations shape and constrain nuclear operations.
The chapter will procced as follows. I begin by presenting a
conceptualization of political regime type based on three fundamental power-
sharing relationships: elite-mass, intra-ruling elite, and civil-military relations.
Political regimes vary in the scope and scale of power-sharing among the masses,
elites, and the military. Importantly, the interconnections between these power-
sharing relationships will be mapped. Political regimes that lack elite-mass power-
sharing – authoritarian regimes – are at a distinct structural disadvantage in crafting
functional intra-ruling elite power-sharing arrangements. And the lack of functional
intra-ruling elite power-sharing relationships, in turn, restricts the possible range of
civil-military relations in that regime. In practice, unconstrained dictators often
establish despotic control over the military as a preventive measure to protect their
rule. Thus, civil-military relations in authoritarian regimes is directly affected by
the amount of power the paramount leader can amass.
The second part of the chapter discusses how specific configurations of
political institutions in authoritarian regimes shape and constrain nuclear posture
decisions. The motivating question is how does the institutional configuration of an
78
authoritarian regime affect its ability to surmount the financial, technical, and
management challenges in developing and deploying nuclear weapons? Do all
regimes have similar problems? Or are there specific challenges that only
authoritarian regimes confront?
I argue that the lack of elite-mass and intra-ruling elite power-sharing
strongly affects nuclear posture. I generate a series of hypotheses connecting
variation in political power-sharing relationships in regimes with the three
dimensions of nuclear posture: capabilities, doctrine, and management. Political
regimes lacking intra-ruling elite power-sharing institutions are likely to confront
the most severe challenges in developing the capabilities, doctrine, and
organizational management procedures for a “first-use” nuclear posture. This is
because extreme personalism reduces the efficiency and success of nuclear
weapons development programs due to arbitrary political interference, distorts the
strategic assessment process, and transforms civil-military relations through
despotic control of the military. The end result is the fundamental paradox of
personalism: when dictators free themselves of political constraints by subverting
institutionalized power-sharing arrangements they create binding internal
constraint on nuclear posture. Without intra-ruling elite power-sharing, delegative
nuclear posture options are highly unlikely, if not unattainable.
The chapter concludes with a summary of the theoretical predictions and a
brief discussion of case selection for the case study chapters that follow.
79
Three Power-Sharing Relationships: Elite-Mass, Intra-Ruling Elite, and
Civil-Military Relations
All political regimes feature three observable power-sharing relationships
of significant importance: elite-mass, intra-ruling elite, and civil-military relations.
These power-sharing relationships form the core of existing definitions of political
regime types and sub-types. In simplest terms, political regimes that feature all
three functional power-sharing relationships are considered democratic. Political
regimes that lack elite-mass power-sharing arrangements are considered
authoritarian. Authoritarian regimes can be further classified by the existence or
absence of intra-ruling elite power-sharing arrangements. Authoritarian regimes
that lack intra-ruling elite power-sharing arrangements, whether the ruling elite is
civilian or military, are at risk of degenerating into personalist regimes—defined
by the extreme concentration of power in a single leader.
Power-sharing relationships can be sustained by a panoply of formal
political institutions (or more accurately, specific equilibria of power-sharing
relationships can be sustained by institutions) or through informal balances of
power among various political actors. The role of formal institutions in facilitating
power-sharing will be discussed at length below, but it is important to note that
power-sharing can exist without formal institutions—although in the absence of
80
institutions power-sharing is likely to be far less stable.
119
Indeed, in some cases
formal institutions may simply reflect existing power-sharing relationships rather
than create them ex nihilo. Power-sharing in the absence of formal institutions
resembles a type of feudalism, where the paramount leader lacks the power to
subjugate rivals and therefore must live in a tenuous balance with them.
Conceptualizing regimes according to these three power-sharing
relationships affords the analyst an ability to interrogate the impact of the absence
of one power-sharing relationship upon the others and also the impact of specific
configurations of power-sharing institutions on outcome variables of interest, such
as nuclear operations. In other words, the presence or absence of functional power-
sharing at the elite-mass level shapes and constrains intra-ruling elite power
relations. And intra-ruling elite power-sharing sets the broad parameters for civil-
military relations in a particular political regime. Power-sharing relationships not
only influence who decides nuclear policy and with what type of input and
oversight from other political actors in a regime, but also determine the type and
severity of organizational and management challenges faced by the regime as it
attempts to build a usable nuclear arsenal.
119
This point is made by Milan W. Svolik, The Politics of Authoritarian Rule (New York:
Cambridge University Press, 2012).
81
Elite-Mass Power-Sharing
The first and most fundamental power-sharing relationship in a political
regime is between the political elite and the masses. Historically, this has been the
relationship of primary concern for political regime theorists as it demarcates
representative regimes from those that restrict political power to only a select few.
The lack of elite-mass power-sharing in authoritarian regimes creates the
foundational and iconic political environment of such regimes: a single dictator or
small clique ruling over disenfranchised masses. Although the exclusion of the
masses from the political process is an important component of authoritarian
politics, it has little direct impact on nuclear operations since even in the most
representative democracies the masses have extremely minimal input into nuclear
decision-making.
120
The true importance of the lack of elite-mass power-sharing on
nuclear posture evolution is the way in which it makes power-sharing among
civilian ruling elites and between civilian and military elites much more difficult.
The lack of elite-mass power-sharing shapes the political dynamic of
authoritarian regimes in four distinct ways. First, it normalizes the principle of
forcible exclusion within the political system.
121
The masses are forcibly excluded
120
Robert Dahl, Controlling Nuclear Weapons: Democracy Versus Guardianship (Syracuse, N.Y.:
Syracuse University Press, 1985). The point here is that the nuclear domain is defined by
specialized expertise and extreme secrecy even in otherwise open regimes.
121
All political regimes have some form of exclusionary policy, e.g. youth or noncitizens.
However, the presumption in democratic regimes is that only a small numerical minority are
excluded from the political process.
82
from meaningful political competition; they do not have the ability to participate in
the selection of leaders or policies nor become leaders themselves.
122
Importantly,
since forcible exclusion from political competition is an institutionalized feature of
authoritarian regimes, political elites are also at risk of being banished into the “out
group” if they are seen as challenging the leader’s position or favored policies.
123
In other words, the principle of forcible exclusion is flexible in so far as it can
always expand to exclude ever larger numbers of people, both masses and dissident
elites.
Second, the lack of elite-mass power-sharing creates the need for a robust
internal repressive apparatus to enforce the principle of mass exclusion. The
repressive apparatus—normally an internal security service (paramilitary) or secret
police organization—monitors and punishes anti-regime (or anti-leader) sentiment
from both the masses and dissident elites. An internal security service, in
combination with a politicized criminal justice system, provides dictators with a
unique capability to forcibly deal with challenges from the masses and rival elites
(including military elites). Indeed, the ability to purge civilian and military elite
rivals for “political” crimes is a key instrument in a dictator’s survival toolkit. The
122
Totalitarian regimes featured mandatory mass mobilization, but no control over political
outcomes. Mass involvement in regime politics, mobilization, is not the same as power-sharing in
decision-making. Regimes can be highly mobilized and still very authoritarian. There can be high
levels of mandatory mass participation in the rituals of the regime, but no mass influence on how
leaders are selected or policies decided.
123
Of course, it is quite common for elites to be rehabilitated back into power as fortunes change
or the situation warrants
83
use of violence (or the threat thereof) to sustain mass exclusion creates a dynamic
in which force is the ultimate arbiter of disputes among the ruling elite and
masses—and by extension among the ruling elite themselves.
Third, the lack of elite-mass power-sharing creates structural possibilities
for the extreme concentration of power in a single leader. This is due to the
interaction effect of the principle of exclusion with the existence of a robust
repressive apparatus. Violence (or threat thereof) can be used to not only repress
the masses and eliminate or marginalize important rivals, but also to alter the
institutional framework of the regime. Without elite-mass power-sharing, the
“people” cannot be an effective check on a dictator’s power grabs. Only rival
elites—either civilian or military—can potentially constrain a dictator. But these
elites are themselves held at risk by the mechanisms cited above, inculcating a
prudent docility. Any move to remove or replace an aspiring dictator is fraught with
danger. Furthermore, resistance to an aspiring dictator’s institutional reforms also
carries risk for political elites. This enables some dictators to openly flaunt
constitutional restraints and even write new constitutions when it suits their
purposes. This accounts for the plasticity of political institutions in authoritarian
regimes. The rules of the game can quickly change and common politics cannot be
insulated from power politics.
Fourth, the lack of elite-mass power-sharing in a regime dramatically alters
the relationship between the military institution and both the masses and civilian
84
elites. Militaries in civilian authoritarian regimes encounter a particularly perverse
civil-military problem: the sheer size of the excluded masses often means the
military is the only security organization with sufficient force to quell a mass
uprising seeking to overthrow the regime (or a particular dictator). As such, the
support of the military is a key prerequisite for leaders of civilian authoritarian
regimes. But it means the military – the experts of violence – must also contemplate
using force against their fellow citizens, the very same group that they must
ostensibly defend from foreign foes. In authoritarian regimes, the military may be
called upon to serve as repressor of last resort. And the role of the military as
repressor of last resort in civilian authoritarian regimes creates the potential for the
military to assume political leverage over the civilian ruling elites. Or to use
Svolik’s terms, to become a “politically pivotal military.”
124
The more pivotal a
military is within a civilian authoritarian regime, the more likely it will attempt to
use that position to extract concessions (budgetary increases, for instance) or
special prerogatives (increased autonomy, immunity from prosecution) from the
civilian ruling elite.
125
This is especially the case if the civilian ruling elite is
factionalized, as the military may attempt to play factions off each other to extract
prerogatives or concessions.
124
Svolik, The Politics of Authoritarian Rule, p. 125.
125
On military prerogatives, see Alfred Stepan, Rethinking Military Politics: Brazil and the
Southern Cone (Princeton, New Jersey: Princeton University Press, 1988) p. 93-127.
85
Of course, a politically pivotal military poses a more fundamental threat to
civilian authoritarian leaders than bargaining over resource allocation or other
concessions. Given the principle of exclusion and the role of violence as the
ultimate arbiter of disputes in authoritarian regimes, the military elites can
themselves grab the reins of power and forcibly exclude civilian elites from
political decision-making. In other words, the lack of elite-mass power-sharing in
authoritarian regimes lowers the barriers of entry for the military to directly enter
politics because such a coup is simply the replacement of a civilian exclusionary
elite with a military exclusionary elite.
126
Indeed, the increased potential for
military coups in civilian authoritarian regimes is well-understood by civilian
dictators – who often take dramatic prophylactic actions to curtail the political
influence and autonomy of their respective militaries. The motivation to “coup
proof” a civilian authoritarian regime flows directly from the comparatively low
barriers to entry for military coups in regimes lacking elite-mass power-sharing.
Taken as a whole, the four effects of the absence of elite-mass power-
sharing arrangements in authoritarian regimes – the principle of forcible exclusion,
the need for a robust repressive apparatus, the plasticity of authoritarian
institutional constraints, and the authoritarian civil-military problematique – create
a particularly toxic political environment for both masses and elites. One
implication of the above argument is that elite-mass power-sharing arrangements
126
As such, it would be an intra-authoritarian regime change.
86
can actually provide elites some intrinsic level of protection from other elites, and
not just protection for the masses from predatory elites. But when elite-mass power-
sharing institutions are non-existent, the ruling elite must look to other pathways to
protect their interests, influence, and often their very lives from elite rivals. In other
words, they attempt to establish intra-ruling elite power-sharing arrangements.
Intra-Ruling Elite Power-Sharing
The second fundamental power-sharing relationship in political regimes is
among the ruling political elite who compete for leadership positions, policy
influence, and spoils. In civilian authoritarian regimes, the intra-ruling elite power-
sharing relationship is between the dictator and elites within the ruling coalition. In
this sense, intra-ruling elite power-sharing institutions can be seen as purposeful
attempts to place limits on the dictator’s arbitrary exercise of power. Four key
questions will guide the subsequent discussion: 1) What are the various forms of
intra-ruling elite power-sharing institutions in authoritarian regimes? 2) How do
these institutions function to constrain the dictator? 3) How do they fail? 4) What
are the consequences of failure?
Because intra-ruling elite power-sharing can exist in both democracies and
authoritarian regimes, it is useful to compare the connection between elite-mass
power-sharing and intra-ruling elite power-sharing in both types of regimes. In
democratic regimes, intra-ruling elite power-sharing is a direct extension and
87
enhancement of elite-mass power-sharing. The tight integration of elite-mass and
intra-ruling elite power-sharing in mature democratic regimes is a result of the
mechanism for executive selection: frequent and competitive popular elections for
fixed terms in office. Incumbents cannot remain in office through force alone.
Access to power must be renewed at regular intervals in accordance with specified
institutional rules. Thus, in democracies the combination of elite-mass and intra-
ruling elite power-sharing regulates both initial and continuing access to power for
political elites elected to office. As a result, the potential for elite rotation in office
is frequent and broad within democracies.
Beyond regulating how political elites retain power, intra-ruling elite
power-sharing in democratic regimes creates constraints on the scope of a leader’s
decision-making and implementation power when in office. As such, various
“checks and balances” among political institutions in democracies strictly limit the
ability of chief executives to amass and exercise political power. Formal
institutional and normative constraints prescribe how power is properly wielded.
Institutional mechanisms also exist to remove officials from office before their
elected term expires in cases of illegal conduct. In mature democracies, elite-mass
and intra-ruling elite power-sharing institutions are mutually reinforcing. Intra-
ruling elite power-sharing is built on the foundation of elite-mass power-sharing.
Yet even in democracies intra-ruling elite-power sharing can fail, leading to
democratic breakdown. Indeed, in the sequencing of democratic breakdown the
88
theoretical expectation is that intra-ruling elite power-sharing collapses prior to
elite-mass power-sharing. A president or prime minister may come to power via
legal means but then execute an “autogolpe” by assuming extraordinary or extra-
legal powers and then dissolving representative political institutions (national
legislatures, suspend elections, etc.). Thus, even in political regimes with elite-
mass power-sharing the institutional checks on a leader’s ability to accumulate
power may fail to constrain an aspiring autocrat.
127
In authoritarian regimes, the problem of limiting the authority of an aspiring
autocrat is far more challenging. The primary power-sharing relationship is
between the dictator and elites within the ruling coalition. Dictators need elite
support to rule over the disenfranchised masses, but given the structural
environment of authoritarianism these elites also pose a threat to the dictator.
Without the need for popular consent and given the utility of violence to settle
leadership disputes, the possibility of a rival from within the ruling coalition
overthrowing the dictator is quite high. The threat of removal by rival elites
incentivizes dictators to accumulate and consolidate sufficient power to render
sterile potential threats from within their ruling coalition.
The specific actions dictators take to address threats from within their ruling
coalitions are well-known: periodic violent purges, packing institutions with loyal
supporters, rigging institutional rules to avoid competition for leadership positions,
127
Ostensibly, democratic citizens would object and protest such a maneuver.
89
circumventing existing organizations and creating new institutions and
organizations to increase personal authority.
128
Furthermore, dictators often engage
in the frequent rotation of senior administrators and military officers, task the secret
police to surveil and monitor rival elites, and cultivate personality cults through
ostentatious displays of their political preeminence. Taken as a whole, these
mechanisms of power concentration reduce the autonomy and influence of rival
elites within a dictator’s ruling coalition. If successfully implemented, a dictator
can significantly reduce the threat of removal. In other words, authoritarian regimes
rule without the consent of the masses but a personalist authoritarian regime is ruled
by a dictator without the consent of the elite.
Elites within the ruling coalition have competing incentives to prevent the
dictator from consolidating extreme power. Elites within the ruling coalition seek
to preserve their influence within the regime and retain sufficient autonomy to
credibly threaten the dictator with removal. The implicit threat of removal must be
credible to act as a constraint on the dictator’s actions. This is the conflict at the
heart of every authoritarian regime. And it is the reason why establishing and
maintaining intra-ruling elite power-sharing relationships in regimes lacking elite-
mass power-sharing is so difficult. Authoritarian regimes are plagued by structural
128
For a discussion of packing, rigging, and circumventing, see Dan Slater, “Iron Cage in and Iron
Fist: Authoritarian Institutions and the Personalization of Power in Malaysia,” Comparative
Politics 36, No. 1 (October 2003), p. 86-91.
90
characteristics that increase the need for intra-ruling elite power-sharing while
simultaneously facilitating their failure.
The implication of the above argument is twofold. First, intra-ruling elite
power-sharing is supported by, but independent of, elite-mass power-sharing. Intra-
elite power-sharing may independently fail even in democratic regimes, heralding
a regime change from democratic to authoritarian. Second, power-sharing can exist
among the ruling elite in authoritarian regimes even when it does not exist at the
elite-mass level. However, intra-ruling elite power-sharing is far more fragile and
always subject to reversal when it is untethered from elite-mass power-sharing.
The political challenges facing elites in authoritarian regimes can be
partially ameliorated through the adoption of specific power-sharing institutions.
Such institutions come in a variety of forms—contributing to the institutional
heterogeneity of authoritarian regimes—but all designed to achieve the similar
purpose of regulating and stabilizing elite interactions. And although many intra-
ruling elite power-sharing institutions in authoritarian regimes outwardly resemble
institutions in democratic regimes, authoritarian power-sharing is not evidence of
partial democratization. Rather the sole purpose of authoritarian power-sharing is
to stabilize, preserve, and prolong authoritarian rule.
Authoritarian power-sharing institutions can be classified into several main
types: institutions that regulate access to power (leader selection and transfer of
power), and institutions that regulate decision-making processes – how binding
91
policy decisions are made (unanimity, majority, etc.), by whom, and with what level
of input from other actors or institutions (consultation). It should be noted that
although I am separating these power-sharing institutions intro two categories for
analytical purposes, decision-making institutions in authoritarian regimes can be
critically important for regulating access to power. This is because personnel
decisions, including executive selection, are made in accordance with the regime’s
decision-making procedures. In other words, the breakdown of power-sharing
institutions regulating the decision-making process are likely to reduce the
effectiveness of the political institutions governing the access to office.
Access to Power. Institutional procedures governing executive selection
and transfer of power in authoritarian regimes regulate the two dimensions of the
access to power: entry to office and exit from office. Institutional procedures
regulating how leaders enter office in authoritarian regimes may take a variety of
forms, including executive selection by a small group of unelected elites, single-
party elections or plebiscites, and even idiosyncratic mechanisms such as
primogeniture in traditional monarchies. These institutional procedures vary in
their effectiveness as power-sharing constraints on dictators.
They may also be completely irrelevant to how power is actually gained or
lost. One problem is that formal institutional processes can often be circumvented
or otherwise rendered moot, as in the obvious case of a coup or other non-
constitutional seizure of power. But even in less dramatic circumstances,
92
institutional procedures of leader selection may simply function as a rubber stamp
on a prior decision made by a small inner circle rather than the legally specified
larger body of elites. Due to the possibility of power concentration, the hierarchy
of power-sharing responsibility may be subverted.
Furthermore, in many authoritarian regimes the boundaries of a particular
office may be ill-defined. As such, it is not only not always clear where true
authority is located. New organizations and leadership positions can be created ad
hoc as institutional instruments of rule and power accumulation. Similarly, unless
otherwise restricted, a dictator may hold multiple executive offices simultaneously,
further concentrating authority. To make matters even more complex, power may
exist outside of formal office entirely. In this sense, leadership selection procedures
may not function as independent institutional constraints in any meaningful sense
unless power is first channeled into government offices (or other ruling institutions,
such as a single party).
Although entry to office and exit from office are clearly theoretically
connected – an incumbent leader must vacate a position in order for new leader to
be installed – in practice, a regime may have institutional procedures for executive
selection without any formal procedures in place for the rotation, removal, or
retirement of the paramount leader. A fixed term of office may not be specified at
the time of accession into power. Indeed, the primary reason that procedures for
leadership selection in authoritarian regimes fail to function as binding constraints
93
is that they often are untethered from processes regulating the rotation, removal, or
retirement of leaders.
If strictly enforced, institutional term limits and mandatory retirement ages
for executive offices are effective mechanisms sustaining authoritarian power-
sharing. Term limits promote stable and regular transfers of power. They also
reduce the negative incentives that inhibit an authoritarian leader from developing
and designating successor.
129
Another benefit of term limits is that they act as a
very transparent signal to the ruling coalition of a dictator’s desire to accumulate
greater power. A dictator’s attempt to stay in power at the expiration of a specified
term is an unambiguous sign of a dangerous power grab.
However, in regimes lacking executive term limits or mandatory retirement
ages there is no legal or normative restriction on how long a leader may remain in
office. In such a case a dictator may remain in office for as long as he or she can
continue to dominate their ruling coalition – be it through a combination of the
threat of force, spoils and patronage, or effective performance. Unlimited terms of
office exacerbate underlying structural opportunities for the concentration of
power. Long tenures can result in even stronger leaders, as extended time in office
provides the dictator an opportunity to slip constraints and grow more powerful.
129
In the absence of term limits in authoritarian regimes it is dangerous for a dictator to develop
and designate a successor because there are no restrictions that would prevent the successor from
usurping the position. A powerful successor waiting in the wings is a constant threat to the
dictator.
94
The uncoupling of institutional procedures for entry into and exit from office also
reduces the frequency of elite rotation in authoritarian regimes. This can result in a
sclerotic leadership, with attendant negative impact on policy innovation and
implementation. Finally, without regularized power transfers, succession often
becomes a political crisis taking years to resolve.
Decision-making. A second set of authoritarian power-sharing institutions
regulate how policy decisions are made by the dictator and the ruling coalition.
These political institutions generally take the form of ruling councils or executive
committees, and are often located within authoritarian political parties or
legislatures.
130
Intra-ruling elite power-sharing institutions regulating decision
making provide a venue for the discussion of private information, expert opinion,
and alternate views on policy issues. These power-sharing institutions form the
functional predicate for the principle of “collective leadership” in authoritarian
regimes.
The power-sharing benefit of these institutions is derived from their formal
structure and regular interactions, which increases transparency among the
authoritarian elite and provides the ruling coalition an ability to monitor the
dictator’s attempt to accumulate more power. This monitoring partially eases the
collective action problem elites would otherwise face in attempting to remove the
130
Svolik, The Politics of Authoritarian Rule.
95
dictator, thereby increasing the credibility of the threat of removal. It is in this way
that authoritarian power-sharing institutions strengthen constraints on the dictator.
When consultative and decision-making institutions are properly
functioning, authoritarian elites within the ruling coalition are better able to
mitigate or block a dictator’s arbitrary impulses. This applies to preventing policy
blunders, but also extends to the personal security of the ruling coalition members.
Authoritarian power-sharing allows elites within the ruling coalition to build their
own power networks and thereby preserve their influence and protect themselves.
The more a dictator needs elite allies to stay in power, the greater opportunity for
the ruling coalition to exert influence and carve out partially autonomous domains
of influence. Moreover, the autonomy associated with functional power-sharing
institutions extends beyond just the ruling coalition and into the lower
administrative elite tiers of the regimes as well. This decentralization of power
contributes to the “rationalization” of the bureaucratic apparatus and increases the
efficiency of policy-making and public administration throughout the regime.
Authoritarian power-sharing institutions regulating decision-making are
also important for another critical issue area: personnel appointment. To the extent
that personnel appointments are made in accordance with the principle of collective
leadership, elites within the ruling coalition can resist a dictator’s attempt to pack
the regime’s most important decision-making bodies with loyal followers.
Conversely, if the dictator is able to manipulate the personnel appointment process,
96
the factional balance of power within key decision-making councils will change
and the end result will be the subversion of power-sharing.
Failure of Authoritarian Power-Sharing. Authoritarian power-sharing is
most likely to fail when a dictator accumulates sufficient power to rule without elite
allies. This begs the question of how aspiring autocrats acquire enough power to
dispense with allies in the ruling coalition. The simple answer is that ruling elites
fail to stop the dictator while they still possess the ability to do so. If incremental
power grabs by the dictator go unpunished, the dictator can eventually become
powerful enough to begin targeting members of the ruling coalition. In other words,
authoritarian power-sharing institutions quickly decay if elites in the ruling
coalition do not capitalize on the monitoring and collective action benefits provided
by these institutions and take appropriate action against the dictator.
The collapse of consultative or collective decision-making institutions in
authoritarian regimes signals the return to a particularly pernicious and precarious
political environment. Elites in the ruling coalition cannot protect themselves or
each other from purges or extra-legal punishments. Indeed, the more powerful the
members of the ruling coalition are the greater potential threat they are to the
dictator. As such, elites within the ruling collation often bear the brunt of an
aspiring personalist dictator’s wrath. Importantly, purges of senior members of the
ruling coalition also provide the dictator an opportunity to appoint new personnel
who would ostensibly be weaker and thus personally loyal to the dictator. The cycle
97
of purging and replacement can repeat until the dictator has removed any and all
threats from within the ruling coalition.
The impact of the failure of intra-ruling elite power-sharing is not confined
to the civilian members of the ruling coalition alone. It spreads throughout the entire
regime because an unconstrained dictator has the ability to intervene anywhere in
the regime for wholly arbitrary reasons. Notably, this includes the military. The
failure of intra-ruling elite power-sharing in civilian authoritarian regimes has a
profound and lasting effect on civil-military relations.
Civil-Military Relations: The Degeneration Into Despotic Control
The third fundamental power-sharing relationship in a political regime is
between political and military elites. The power-sharing relationship between
civilian and military elites can take a variety of institutional forms in both
democratic and authoritarian regimes – reflecting rough parity or more often the
political dominance of either civilian or military elites. Importantly, civil-military
relations is structured by the other power-sharing relationships within a regime.
Specific configurations of elite-mass and intra-ruling elite power-sharing
institutions set the broad parameters of civil-military relations in a regime.
All political regimes with military forces face the threat of a potential
military coup. This is the core civil-military problematique. Civilian political
leaders – in both democratic and authoritarian regimes – and even military leaders
98
in military authoritarian regimes are confronted with the challenge of ensuring the
military institution (or some segment thereof) does not attempt the direct seizure of
power.
131
Political control of the military is the central issue of civil-military
relations. This is often reduced to the concept of “civilian control” of the military,
but even military dictators must ensure their political control of the military
institution.
132
Political elites face the challenge of how best to achieve and maintain
political supremacy without compromising military effectiveness on the battlefield.
Political leaders – in both democracies and authoritarian regimes – may adopt
institutional arrangements and organizational practices that either reduce the ability
of the military to stage a coup (thereby reducing battlefield effectiveness) or
enhance battlefield effectiveness (potentially increasing the risk of a coup).
Managing the inherent tradeoff between coup prevention and battlefield
effectiveness is a universal challenge facing political leaders in all types of regimes.
The universality of the civil-military challenge, however, does not mean
that all political regimes confront the political control problem in equal intensity or
possess equal capacities to optimize civil-military power-sharing arrangements. As
we have seen so far, political regimes vary in both elite-mass and intra-ruling elite
131
This can be a hierarchical or non-hierarchical coup. A coup is hierarchical when lead by the
highest ranking officers in the military hierarchy. A coup is non-hierarchical when led by junior
officers or even enlisted personnel.
132
Stepan introduced the distinction between “military-as-government” and “military-as-
institution” to highlight this tension. See, Alfred Stepan, Rethinking Military Politics: Brazil and
the Southern Cone (Princeton, New Jersey: Princeton University Press, 1988).
99
power-sharing arrangements. The impact of elite-mass and intra-ruling elite power-
sharing on civil-military relations will be analyzed for each configuration of regime
type.
Democracies. Democratic regimes enjoy several advantages in developing
stable civil-military relations due to robust power-sharing institutions at the elite-
mass and intra-ruling elite level. First, the barriers to entry for direct seizure of
power are extremely high due to the existence of elite-mass power-sharing. Popular
consent also removes the military’s potential role as repressor of last resort, as well
as obviating the need for a repressor of first resort, such as a secret (political) police.
With elite-mass power-sharing in place, the military’s primary relationship to the
polity and political elites is professional protector, not potential predator.
Second, functional intra-ruling elite power-sharing institutions provide a
foundation upon which civil-military power-sharing institutions can be developed.
The institutional processes for decision-making and accession to office for civilians
can be naturally extended into civil-military relations. How civilian political elites
share power with each other sets the stage for how those civilians interact with
military elites. Thus, the institutional constraints that limit arbitrary exercise of
executive power in civilian political appointments (e.g. ministers, secretaries,
judges, etc.) are likely to extend to military appointments at the senior officer level.
Promotions to senior rank (general officer) and command appointments are likely
to require some type of legislative confirmation, thus increasing the scope of
100
civilian oversight and increasing the likelihood that promotions are for professional
merit rather than political loyalty. Fixed terms of political office and limits on
tenure are likely to translate in predictable and fair command rotations for elite
officers, leading to established career paths and the opportunity to develop
considerable professional expertise.
Similarly, intra-ruling elite power-sharing institutions structure the
decision-making process for reaching binding policy decisions on the use of
military force, military budgets, weapons procurement, strategic doctrine and other
defense issues. Military officers often have the opportunity to voice preferences at
multiple points in the policy formulation stage, even though they do not participate
in the actual policy decision-making. And because political power is relatively
widely dispersed in regimes with intra-ruling elite power-sharing institutions, a
greater number of civilian political elites are able to interact with senior officers
and develop expertise on military issues and the military institution, generating a
capability to independently assess and understand military advice and substantive
concerns.
The advantages of political regimes with both elite-mass and intra-ruling
elite power-sharing institutions translates into an increased ability to grant
autonomy to the military institution without fear of increasing the risk of a coup
Functional intra-ruling elite power-sharing institutions means civilian political
elites are constrained and military elites have ample opportunity for joint
101
consultation on policy issues and multiple outlets to communicate expert military
opinion. Civil-military arrangements in democratic regimes can be highly
delegative because the relationship between political leaders and military leaders
takes place within a dense institutional framework of power-sharing.
Civilian Authoritarian Regimes with Intra-Ruling Elite Power-Sharing
(Machine). Authoritarian regimes lack elite-mass power-sharing and are at a
distinct structural disadvantage in establishing and maintaining functional intra-
ruling elite power-sharing institutions. The lack of elite-mass power-sharing
exacerbates the basic civil-military problem by lowering the barriers for entry to
direct seizure of power and creating the potential role of repressor of last resort.
Under such conditions there is an increased risk that any grant of autonomy to the
military may be turned against the regime or leader. As a result, civilian
authoritarian regimes are induced to curtail military autonomy, even at the risk of
reducing battlefield effectiveness. Civilian-authoritarian regimes simply lack the
institutional infrastructure to support highly delegative civil-military arrangements.
The concept and practice of civilian political control of the military in
authoritarian regimes is markedly different than in democracies. This is because
civilian authoritarian leaders attempt to compensate for the disadvantages arising
from the lack of elite-mass power-sharing by centralizing command authority and
adopting a variety of intrusive control measures to monitor and restrict the
military’s institutional autonomy. Moreover, authoritarian leaders also have the
102
repressive apparatus – internal security organization and/or secret police – at their
disposal to “counterbalance” the military and ensure the political loyalty of senior
officers through invasive surveillance and extra-legal punishments. Strong internal
security services also reduce the regime’s reliance on the military for repression,
thereby reducing the military’s political leverage. This results in a situation where
civilian authoritarian elites have both the increased need for political control of the
military and greater (often extra-legal) capabilities to establish strict control.
However, the severity of the political control measures used in authoritarian
regimes varies with the strength of intra-ruling elite power-sharing institutions.
When authoritarian regimes have robust and functional intra-ruling elite power-
sharing institutions civilian political elites have some capacity to constrain the
arbitrary impulses of the paramount leader. As I argued above for democracies,
intra-ruling elite power-sharing institutions form the foundation for civil-military
power-sharing arrangements. The same is true for authoritarian regimes.
Authoritarian power-sharing among civilian political elites creates the institutional
infrastructure necessary for limited military autonomy. Senior officers may still be
viewed as an internal political threat, but authoritarian power-sharing creates
opportunities for civil-military arrangements that are more balanced (in terms of
coup prevention and battlefield effectiveness) than found in regimes lacking intra-
ruling elite power-sharing. The end result is that civilian authoritarian regimes with
103
functional intra-ruling elite power-sharing can reach the upper range of assertive
civil-military relations.
Civilian Authoritarian Regimes without Intra-Ruling Elite Power-Sharing
(Personalist-Boss). Authoritarian regimes lacking intra-ruling elite power-sharing
institutions have the most restrictive civil-military arrangements. Indeed, civil-
military relations deteriorate rapidly as a political regime degenerates into
personalism. The breakdown of authoritarian power-sharing in a civilian
authoritarian regime heralds the subjugation of the ruling coalition at the hands of
the dictator. If unchecked by elites within the ruling coalition, the dictator’s active
subversion of intra-ruling elite power-sharing institutions – through the practices of
packing, purging, and rigging – results in an extremely high concentration of
decision-making power. The practical implication of such an extreme concentration
of political power is that the dictator is essentially freed from all institutional
constraints and can intervene with impunity in every sector of the regime.
The impact of an unconstrained dictator on civil-military relations is severe.
The decay of intra-ruling elite power-sharing institutions leads to despotic civil-
military relations. The dictator has the ability to intervene into all aspects and levels
of military operations. Military budget allocations, weapons procurement
decisions, promotions in rank, appointments and command assignments are all
subject to the dictator’s political meddling. The processes for strategic assessment
and formulating doctrine are subject to serious distortion. The dictator has the
104
ability to single-handedly determine the regime’s military doctrine and then force
senior officers to not only accept but praise the doctrine even if it runs against the
principles of sound military science. Moreover, an unconstrained dictator has the
sole prerogative to authorize the use of military force, either internally or externally.
And since the dictator is unconstrained by institutions or elite rivals there is no
possibility of holding the dictator accountable for any military blunders that result
from doctrinal shortcomings or poor military organizational practices.
The key feature of despotic control is the high concentration of political
power in a paramount leader. The military is so powerless under conditions of
despotic control that it cannot even protect its senior officers from extreme extra-
legal punishments, such as arbitrary arrest, imprisonment, and executions, which
are meant to enforce near-total military subordination to the paramount political
leader. Furthermore, unlike assertive civil-military relations which requires some
degree of institutional distinctiveness between the civilian and military spheres,
133
the paramount leader under despotic control recognizes no boundaries to his
authority. Under despotic control the military is not just fully politicized, but
personalized. Indeed, the paramount political leader often dons a uniform,
irrespective of prior military service, and appoints himself to the highest military
133
As Feaver writes (1992: 9): “Assertive control does not presuppose that the military will
conform to society or that the professional officer corps will be completely undermined. Rather it
allows for qualitatively separate institutions, civil and military, which remain philosophically and
sociologically distinct.”
105
rank possible—even inventing new ranks such as supreme commander, if
necessary. Despotic control represents a condition of a near-total breakdown and
subversion of “traditional” civil-military distinctions.
In this sense, despotic control shares theoretical affinities with Huntington’s
original concept of subjective civilian control, which Feaver abandoned in his
influential study of U.S. nuclear behavior because it had such limited applicability
for advanced democratic regimes.
134
This limited applicability derived from the
theoretical understanding that in advanced democratic regimes, the parameters of
civilian elite and military elite interactions were largely set by law or custom and
widely accepted by both elite groups. Thus neither group would resort to
extraordinary or extra-legal actions to advance preferred policy positions.
American officers were not in danger of being executed for political disagreements
and civilian political elites were not in danger of being overthrown in a coup d’état.
As such, the power relations among elite civilian groups were not a primary concern
for scholars of nuclear behavior.
But as Huntington noted, subjective civilian control of the military
“involves the power relations among civilian groups. It is advanced by one civilian
group as a means to enhance its power at the expense of other civilian groups.”
135
134
Feaver (1992:8): “Subjective control is not a useful concept in explaining the American post
war experience—at least not with respect to nuclear weapons—because by and large, civilians
have opted for a different method of control.”
135
Huntington, Soldier and State, p. 80.
106
Thus, in one sense, despotic control can be viewed as an extreme case of subjective
control because it only obtains when the paramount political leader is able to
enhance his power so completely that he can no longer be credibly threatened with
forcible removal from power by political rivals. As such, despotic control emerges
when political power-sharing among elite civilian groups has completely collapsed,
enabling the leader to act against both civilian and military elite rivals with virtual
impunity.
However, despotic control differs from assertive and subjective civilian
control in two critical ways: it cannot exist in a democracy and it need not be
civilian control.
136
In one sense, this is purely definitional. The underlying
conditions that give rise to despotic control – extreme concentration of power – are
simply inconsistent with existing conceptualizations of democracy. Moreover, the
extra-legal methods associated with despotic control, particularly violent purges of
senior military officers, clearly violate the rule of law principle at the heart of
democracy. Despotic control is simply not an option for democratic leaders.
137
136
For Huntington, subjective control was expressly civilian control (maximization of civilian
power by reducing the military’s power) and was not exclusive to a particular regime type.
137
Huntington (1957: 82) argues that “subjective civilian control …is not the monopoly of any
particular constitutional system,” meaning that civilian control was not the exclusive domain of
democratic forms of government. In democracies, such as the U.S., subjective control represented
the struggle between the executive and legislative branch, with a wily military institution being
able to play the two governmental institutions off each other through legitimate democratic
processes to maximize its own political power and influence. In totalitarian regimes, however,
where “terror, conspiracy, surveillance, and force are the methods of government…terror,
conspiracy, surveillance, and force are the means by which the civilians in such a state control
their armed forces.” Huntington develops this insight on totalitarian regimes to its logical
conclusion, noting that “If employed sufficiently ruthlessly, these means may virtually eliminate
military political power.” In other words, forms of subjective civilian control in various countries
107
In short, a personalist dictator has an untrammeled ability to subjugate the
military institution and render it politically sterile. To achieve such ends, however,
requires such intense control measures that military battlefield efficiency is
sacrificed. Military morale is also degraded since senior officers may be purged for
many reasons, including battlefield success, lest they begin to accrue popular
acclaim. Senior officers are subject to harassment, intrusive surveillance, public
dismissals, and, quite frequently, execution. The process of personalist power
consolidation is incredibly damaging for the military.
Civil-military arrangements typically reflect the underlying balance of
internal versus external threat within a political regime. The expectation is that
regime elites – both civilian and military – jointly determine the proper balance.
But the condition of extreme concentration of political power gives the dictator an
unparalleled ability to skew the threat assessment process. As a result, civil-military
relations under a personalist dictator reflect the near total breakdown of the threat
assessment process. Civil-military relations under personalism prioritize internal
political threats from within the military institution at the expense of military
effectiveness.
would reflect how political power is practiced in the respective country, with democratic processes
permitting greater military influence than often recognized and totalitarian methods of control
reducing military influence to extraordinary low levels. In both regimes the purpose of subjective
civilian control was to maximize civilian power at the expense of military power—the efficacy of
the control tactics simply varied, with civilian control in a totalitarian regime potentially far more
effective than in a democracy.
108
Military Authoritarian Regimes with Intra-Ruling Elite Power-Sharing
(Junta). In the aftermath of a military coup, military leaders may develop intra-
ruling elite power-sharing institutions, such as processes for the rotation of
executive office among senior officers from each military service (e.g. army, navy,
etc.). Similarly, consultative and collective decision-making procedures may also
be instituted among the senior officers in a military regime and embedded within
various councils or committees (e.g. a national security council, etc.). Although
military officers will exercise political control over the regime, civilian elites may
be brought in as political allies and even organized within a political party system.
In military regimes with functional intra-ruling elite power-sharing institutions the
upper range for civil-military relations is praetorian.
138
Praetorian civil-military
relations denotes the political supremacy of the military institution within the
regime. Simply put, all aspects of national security and defense policy-making – as
well as domestic policy – are in the hands of the military.
Praetorian civil-military relations are unique insofar as the traditional
power-sharing relationship between military and civilian elites is inverted. The
policy influence and participation of civilian political elites and civilian political
institutions within a military regime may range from near total exclusion to a
symbiotic synthesis that harnesses civilian elite expertise across all policy sectors,
138
Military regimes with functional intra-ruling elite power-sharing institutions are unique in so
far as they are the only political regimes in which the value for civil-military relations is
determinative of regime type.
109
including defense policy. And just as civil-military relations in a civilian
authoritarian regime are dependent upon the institutional core of civilian intra-
ruling elite power-sharing, military-civilian relations in a military regime are
dependent upon military intra-ruling elite power-sharing institutions. In other
words, military regimes that feature robust power-sharing institutions among the
senior officers leading the regime are far more likely to be able to facilitate civilian
elite participation in the regime.
Another important aspect of praetorian civil-military relations is that the
level of military autonomy is variable, and, in fact, may be more restricted than
otherwise would be assumed in a military regime. This is due to two factors. First,
a military coup may be hierarchical (led by the military’s most senior officers) or
non-hierarchical (led by junior officers or even non-commissioned officers). In a
hierarchical coup the most senior officers in the military move the military
institution into politics as a corporate entity and military hierarchy is preserved.
When a military regime is founded through a non-hierarchical coup, however, the
coup plotters subvert the military chain-of-command in the process of seizing direct
political power. As such, the junior officers leading the new regime have incentives
to restrict the autonomy of their colleagues, lest the senior officers – who may
object to the original coup – broker a transition with civilian elites and subsequently
undertake an extrication coup to return the military to the barracks. And, of course,
110
having come to power in a coup means the junior officers are intimately aware of
the possibility of another coup.
It is for this reason that praetorian civil-military relations encompasses the
greatest range of military autonomy of all categories. Under conditions of robust
intra-ruling elite power-sharing and a hierarchical military government the military
institutions is likely to have extremely high levels of autonomy extending far down
the chain of command. However, autonomy for commanders may be severely
restricted in the absence or collapse of power-sharing institutions, especially if the
military regime is non-hierarchical. In other words, the slide into military
personalism has tremendous impact on the military institution’s autonomy and
influence, even though the leader of the regime is an active or former officer.
Military Authoritarian Regimes without Intra-Ruling Elite Power-Sharing
(Personalist-Strongman). Military regimes are susceptible to the same
degeneration into personalism that plagues all authoritarian regimes. An officer
who originally was primus inter pares within a junta’s ruling coalition may attempt
to subvert existing power-sharing arrangements and subordinate military elite
rivals. If successful, the leader may become a military personalist dictator
(Strongman). The process and pace of personalist power consolidation may differ
in military regimes because the aspiring strongman may need to build institutional
capacity outside of the military institution in order to subjugate the military.
Whereas a civilian personalist leader (Boss) can ostensibly draw on civilian
111
political institutions to control the military, the military personalist leader may first
need to reconstitute civilian political institutions in order to counterbalance the
political influence of the military institution. In this sense, the military personalist
dictator must reduce the military institution’s political influence in order to
consolidate personal power. Strengthening civilian political institutions is a
prerequisite for an aspiring strongman.
There is a certain irony in the civilianization of a military regime at the
hands of a strongman. After utilizing the military’s institutional strength to seize
political power from ostensibly weak civilian political institutions, the aspiring
strongman then builds institutional capacity to prevent being removed by other
military elites in the junta’s ruling coalition. Indeed, the leader may even take off
the uniform to signal the civilianization of power in the regime. To the extent that
the process of personalization and civilianization of the regime work in tandem, the
theoretical expectation is that the aspiring strongman will develop the same control
capabilities of a civilian personalist, namely an internal security service or special
paramilitary organization to monitor the military. Political parties may be created
to augment the infrastructural power of the regime and facilitate the diminishment
of the military’s role in government. The military may still remain a privileged
institution in the regime, receiving benefits and perquisites that inculcate
continuing loyalty or at least docility. But command authority is centralized and
concentrated in the hands of the strongman.
112
The end result of the personalization of power in a military regime is the
same as in a civilian authoritarian regime: despotic civil-military relations. Since
political power-sharing arrangements among military elite rivals within military
dictatorships can be subverted just like similar arrangements in civilian
dictatorships, the processes of purging, packing, and rigging extends to the military
institution.. Command appointments, promotions, and dismissal are all subject to
the strongman’s calculus for political loyalty. Fellow military officers are not
immune to vicious purges. The paramount political leader may resort to despotic
control over his recent colleagues in an attempt to purge potential elite rivals from
the senior ranks of the military institution, just as civilian personalist leaders have
often done when consolidating power. To be sure, the paramount political leader in
such a situation – most likely a former general or sometimes even a colonel – may
eventually choose to take off the uniform and even attempt to create a civilian
political party, thereby blurring the distinction between a military and civilian
regime, but the larger point remains. Despotic control is a potential option for all
personalist dictators, both civilian bosses and military strongmen.
Figure 3.1 visually depicts the theoretical interaction between and among
the three power-sharing relationships and disaggregated political regime type.
113
Figure 3.1 Power-Sharing Relationships and Political Regime Type
114
The starting point of the conceptual argument is elite-mass power-sharing
(represented by the block in the middle of the diagram). Elite-mass power-sharing
is the fundamental dividing line between democratic and authoritarian regimes.
Democratic regimes have functional elite-mass power-sharing institutions
(represented by the “yes” next to the up arrow). Political regimes that lack elite-
mass power-sharing (represented by the “no” next to the left and right arrows) can
be further classified as civilian authoritarian or military regimes, reflecting that the
institutional composition of the ruling elite may be civilian or military.
Intra-ruling elite power-sharing is the next node in the diagram. In
democratic regimes the ruling political elite is civilian by definition, and therefore
intra-ruling elite power-sharing denotes a civilian-civilian institutional
relationship.
139
As discussed above, intra-ruling elite power-sharing in mature
democratic regimes is supported by a dense and robust set of integrated political
institutions (represented by the “yes” next to the second up arrow).
In authoritarian regimes, the intra-ruling elite power-sharing institutions of
greatest importance are located in the primary institutional base of the regime. If
the regime’s core institutional infrastructure is civilian (e.g., a single or hegemonic
party regime), then intra-ruling elite power-sharing denotes a civilian-civilian
139
To be sure, political officeholders in democratic regimes may be recently retired senior
officers. It is not the professional background of the officeholder that is of concern. It is the
institutional identity of the ruling elite. The military may rule as a corporate elite in authoritarian
regimes.
115
relationship. In other words, the dictator’s institutional identity would be civilian
and the ruling coalition would be predominantly if not exclusively civilian, as well.
Similarly, if the regime’s core institutional infrastructure is the military, then intra-
ruling elite power-sharing denotes a military-military relationship. The dictator’s
institutional identity would be military (e.g. a senior officer) and the ruling coalition
would be predominantly composed of other senior officers (e.g. military service
branch chiefs). Functional intra-ruling elite power-sharing institutions may exist in
both types of authoritarian regimes (represented by the “yes” next to the second set
of arrows pointing left and right).
The third node in the diagram is civil-military relations. There are four
categories of civil-military relations: delegative, assertive, praetorian, and despotic.
Democracies are the only political regimes with sufficiently robust power-sharing
infrastructure to reach delegative civil-military relations. The existence of multiple
interlocking power-sharing institutions – at the elite-mass and intra-elite levels –
increases barriers to entry for a military coup and thus reduces the potential for
coups. As such, democracies can afford extremely delegative civil-military
relations.
For authoritarian regimes the institutional identity of the dictator and the
ruling coalition are important indicators of the conceptual starting point for the
evolution of civil-military relations. Whereas civil-military arrangements in
civilian authoritarian regimes are expressly designed to prevent a military coup, in
116
military authoritarian regimes the coup has already taken place. As such, civilian
authoritarian regimes and military regimes quite expectedly have conceptual
starting points on opposite ends of the civil-military relations spectrum. Indeed, in
military regimes the “civil” dimension of the civil-military problematique
essentially reduces into intra-military conflict (military-as-government versus the
military-as-institution).
In both instances, however, intra-ruling elite power-sharing determines the
possible upper range of civil-military relations arrangements in those regimes. A
civilian authoritarian regime with a full suite of functional intra-ruling elite power-
sharing institutions (access to power, decision-making) is able to reach the upper
range of assertive civil-military relations. Civilian leaders will exercise tight and
centralized control over the military ensuring civilian political supremacy, but the
military institution will have sufficient autonomy and influence to protect senior
officers from arbitrary and politically motivated purges. Personal safety for senior
officers in the regime will be roughly equivalent to that of civilian elites within the
larger ruling coalition. Sufficient autonomy will be granted to the military to at least
approximate the adoption of professional military organizational practices. Political
oversight of the military will be far more vigorous than found in democracies or
juntas, but far less severe than in personalist regimes.
Military regimes with functional and robust intra-ruling elite power-sharing
institutions can reach the upper range of praetorian civil-military relations. Juntas
117
will exhibit the highest degree of internal military autonomy, especially if the
hierarchy of the military institution is preserved.
Both civilian and military dictators may attempt to take advantage of the
structural incentives for power arrogation which exist in authoritarian regimes.
(The failure of intra-ruling elite power-sharing institutions in both military and
civilian authoritarian regimes is represented by the “no” next to the diagonal
arrows). The degeneration of intra-ruling elite power-sharing institutions leads to
despotic civil-military relations in both civilian and military authoritarian regimes.
(The dotted lines indicate the permeability of the lower boundaries in both
machines and juntas). The personalist dictator faces few or no constraints on the
arbitrary exercise of power. Neither civilian nor military elites are able to protect
themselves from the dictator’s power consolidation.
Putting It All Together: Political Power-Sharing and Nuclear Posture
Nuclear posture is a multidimensional variable. It includes capabilities (the
mix of weapons and delivery vehicles), doctrine (plans for when, how, and under
what conditions the weapons will be used for a specific purpose), and management
(deployment patterns and command and control procedures for the weapons in
peace and war time).
140
The combination of values on these three dimensions create
140
Vipin Narang, “Posturing for Peace?: Pakistan’s Nuclear Postures and South Asian Stability,”
International Security 34, no. 3 (Winter 2009/10), p. 41. I exclude Narang’s fourth dimension of
118
the basic categories of nuclear postures: assured retaliation or asymmetric
escalation.
141
For instance, an asymmetric escalation posture is a highly-delegative,
nuclear war-fighting posture which features tactical nuclear weapons fully
integrated into a regime’s military forces and intended for first-use on the battlefield
in a denial mission against an enemy’s or conventional forces (armor, infantry, or
naval assets). In contrast, an assured retaliation posture is a highly-centralized,
defensive nuclear posture which features strategic nuclear capabilities (often de-
alerted and de-mated), intended to be used only in retaliation following an
opponent’s first strike. Put differently, an asymmetric escalation posture tends
heavily toward the “always” side of “always/never” dilemma, eschewing negative
controls on the arsenal to ensure prompt usage when necessary, whereas an assured
retaliation posture is the reverse, skewing toward the “never” side through the use
of extensive negative controls on the arsenal.
Ideally, a state’s nuclear posture is internally consistent: a doctrine calling
for early battlefield use of nuclear weapons will be paired with the appropriate
capabilities, such as tactical nuclear weapons integrated into military forces, and a
command and control system featuring high levels of delegation of both the nuclear
assets and launch authority. But an internally consistent nuclear posture may not be
transparency because I doubt the utility and validity of the “catalytic” or opaque category of
nuclear posture. It is a proliferation strategy rather than a nuclear weapon posture.
141
Ibid. Narang applied his classification scheme only to regional power nuclear postures. He
excludes the US, UK, and USSR/Russia.
119
achievable for all political regimes due to internal political constraints. Figure 3.2
displays the relationship between political regime type and internal political
constraints on each dimensions of nuclear posture.
Table 3.1 Political Constraints on Nuclear Posture Dimensions by Regime Type
Nuclear Posture
Regime Type
Capabilities Doctrine Management
Democracy None None None
Machine Minor Minor Moderate
Junta Minor None None
Boss Severe Severe Severe
Strongman Severe Severe Severe
Political regimes that lack both elite-mass and intra-ruling elite power-
sharing face the most severe internal constraints on nuclear posture. Binding
internal constraints on nuclear posture are generated by the extreme concentration
of political power in a single leader. When authoritarian power-sharing is subverted
by a personalist dictator, the ramifications for nuclear posture formation are
significant across each dimension. The evolution of nuclear posture in personalist
regimes is inhibited by the dictator’s primary goal of power consolidation. In
personalist regimes internal political concerns shape and constrain nuclear
operations to the exclusion of all other factors.
120
Capabilities. In the absence of institutional constraints, a personalist dictator
(boss; strongman) may intervene at will into the regime’s nuclear weapons
development program. If political interference is malign, persistent and severe, the
program may even fail. But even if a regime succeeds in developing a nuclear test
device, political meddling is likely to drastically reduce the efficiency of the
weapons development cycle (including nuclear materials, warheads, and delivery
vehicles).
Nuclear Program Inefficiency Hypothesis: Personalist regimes are likely,
ceteris paribus, to have inefficient nuclear weapon development programs.
The personalist dictator is also likely to dominate all procurement decisions
for weapons and delivery vehicles, prioritizing weapon systems that are easier to
control and less likely to disrupt the dictator’s political survival strategy. As such,
personalist dictators are likely to possess less sophisticated and smaller nuclear
arsenals. And since warhead miniaturization is also one of the most technologically
challenging feats, it is unlikely that personalist regimes will develop tactical nuclear
weapons suitable for use on the battlefield.
121
Tactical Nuclear Weapons Hypothesis: Personalist dictatorships are
unlikely, ceteris paribus, to develop tactical nuclear weapons for use on the
battlefield.
Doctrine. The concentration of power in a single leader also introduces bias
into the development of military doctrine. A personalist dictator can exert
tremendous influence over doctrinal discussions by restricting pertinent
information about nuclear weapons from regime elites, curtailing professional
military debate, and enforcing the acceptance of their preferred doctrine. The result
is likely to be a convoluted synthesis emphasizing the importance of possessing
nuclear weapons as a symbol of regime strength and a continued diminishment of
the strategic challenges of the nuclear revolution. Personalist dictators will be slow
to anticipate or embrace the full implications of the nuclear revolution, preferring
to dismiss both the likelihood and utility of nuclear war.
Doctrine Hypothesis: Personalist dictators are unlikely, ceteris paribus, to
develop first-use or nuclear war-fighting doctrines.
Management. Despotic civil-military relations in personalist regimes
constitute a severe constraint on the management of a personalist dictator’s nuclear
arsenal. Personalist dictators are likely to favor restrictive negative controls on the
122
regime’s nuclear arsenal so that the weapons are rendered inaccessible for early use
on the battlefield. Given the overall distrust of the military, personalist dictators
will resist integrating nuclear weapons with conventional military forces and
instead may the place the arsenal under the control of a more politically reliable
organization, such as an internal security service or special paramilitary unit.
Integration Hypothesis: Personalist dictators are unlikely, ceteris paribus,
to integrate nuclear weapons with conventional military forces.
Delegation Hypothesis: Personalist dictators are unlikely, ceteris paribus,
to delegate nuclear assets and launch authority to military commanders.
Each of the above three hypotheses isolate core dimensions required to
implement a viable asymmetric escalation nuclear posture: tactical nuclear
weapons for use on the battlefield, nuclear doctrines for first-use in a denial
mission, and delegative command and control arrangements. A political regime
facing a severe internal constraint on a single dimension of an asymmetric
escalation posture is likely sufficient to prevent the adoption of such a nuclear
posture. But the adoption of an asymmetric escalation posture by a personalist
regime facing severe constraints on all three constituent dimensions is essentially
impossible without endogenous institutional political reform.
123
Personalist Posture Hypothesis: Personalist dictators are highly unlikely,
ceteris paribus, to adopt asymmetric escalation nuclear postures.
Whereas personalist authoritarian regimes face the most severe internal
constraints on multiple dimensions of nuclear posture, other authoritarian regimes
confront only moderate or minor internal constraints. Civilian and military
authoritarian regimes with functional intra-ruling elite power-sharing encounter
minor internal constraints on the capabilities dimension due to the probability of
authoritarian mismanagement of human capital and endemic corruption in their
respective nuclear weapons programs. These minor constraints on nuclear
capabilities can be overcome, especially under conditions of nuclear rivalry or other
external threat, and would likely merely slow down the development timeline rather
than prevent the adoption of nuclear postures requiring diverse nuclear capabilities.
However, civilian authoritarian regimes (machines) also confront moderate
internal political constraints on the management dimension. This is due to the
difficulty of establishing delegative civil-military arrangements in civilian
authoritarian regimes. The reduced military autonomy associated with assertive
civil-military relations in civilian authoritarian regimes translates into strict
negative controls on the nuclear arsenal, limiting nuclear posture options.
Moreover, civilian authoritarian regimes may also experience minor constraints on
124
the nuclear doctrine dimension due to assertive civil-military relations. Although
doctrinal discussions between civilian political elites and military elites would be
expected to be far more open and professional than in personalist regimes, civilian
authoritarian leaders are still likely to evaluate doctrine through a prism of distrust
and suspicion of military influence. The end result is a civilian wariness of offense-
oriented nuclear doctrines. The combination of minor constraints on the capabilities
and doctrine dimensions, as well as moderate constraints on the arsenal
management dimension, results in a reduced likelihood of civilian authoritarian
regimes adopting an asymmetric escalation nuclear posture.
“Machine” Posture Hypothesis: Civilian authoritarian regimes
(machines) are less likely than democracies or juntas, ceteris paribus, to
adopt asymmetric escalation nuclear postures.
As mentioned above, military regimes with intra-ruling elite power-sharing
institutions (juntas) face only a single minor internal constraint on the dimension of
nuclear capabilities. This single constraint is insufficient to prohibit the adoption of
an asymmetric escalation nuclear posture, but it may temporarily impede the
weapons development cycle.
142
If a junta is able to surmount the technical,
142
To some extent this will be dependent on whether the military regime is able to develop the
scientific expertise internally in the officer corps or bring in civilian scientists from outside the
military.
125
scientific, and economic hurdles in developing tactical nuclear weapons suitable
for battlefield, military leaders are capable of adopting an asymmetric escalation
posture. This is because the military institution’s political dominance ensures it
controls both the doctrinal debate and the arsenal. As such, there are no structural
restrictions on the how nuclear weapons are integrated into military forces and
doctrine.
In contrast to authoritarian regimes, democratic regimes do not face any
structural political constraints on nuclear posture formation. The lack of any
internal political constraints on nuclear posture formation in democratic regimes
does not necessarily indicate that those regimes will always opt for highly
delegative offensive nuclear postures – it simply means they have sufficient
institutional infrastructure to do so if they choose. Nuclear posture formation is a
complex process that is driven by both internal political and external military
threats, as well as being dependent upon the technological expertise and economic
potential of the regime. The robust institutional processes for managing political
conflict in democratic regimes means that democracies can be more sensitive and
responsive to external threats since there are few major internal political threats.
Furthermore, it is unlikely that external threats will be willfully ignored by
both civilian political and military elites in a democratic regime – as they can be in
personalist regimes – because the relatively free flow of information and multiple
opportunities for military officers to voice their professional assessments. External
126
threats may actually be politicized for electoral gain in democracies, which creates
incentives for civilian elites to hype a threat. And the hyping of a threat can lead to
an increase in the economic resources allocated to the development of new nuclear
technologies and capabilities, which are then likely to be deployed given that
democratic regimes do not have internal constraints on doctrine or management. It
is in this way that democratic regimes can dominate the weapons development-
deployment cycle and therefore can truly excel at arms racing.
Conclusion
In this chapter I have introduced my basic theoretical explanation of
authoritarian constraints on nuclear posture evolution as a consequence of failed
intra-ruling elite power-sharing institutions in personalist regimes. The theory
suggests that personalist authoritarian regimes face severe management and
organizational challenges in adopting delegative nuclear postures due to despotic
civil-military relations that arise in the absence of functional intra-ruling elite
power-sharing institutions. A central argument of this chapter is that specific
configurations of elite-mass and intra-ruling elite power-sharing institutions
determine the upper boundaries of civil-military arrangements available to political
leaders. The concentration of near absolute political power in a single leader
strongly affects intra-ruling elite interactions, severely restricting the autonomy and
influence of all other regime institutions, including the military and scientific
127
establishments. In sum, the more broadly political power is shared throughout a
political regime, the more delegative a regime’s nuclear posture can be.
In next chapter, I turn to the task of developing a minimalist definition of
the personalist regime concept that can be utilized as the independent variable in
the empirical case studies that follow. I also develop a suitable measurement index
of a minimalist conceptualization of personalism that can aid in political regime
classification schemes. Existing measurement indices are maximalist, meaning they
mix multiple conceptual dimensions of personalism (concentration of power,
neopatrimonialism, and dynastic sultanism) and include multiple indicators of civil-
military relations to classify a regime as personalist. Furthermore, I construct a four
part typology of civil-military relations that can be used as the intervening variable
in the analysis of the relationship between intra-ruling elite power sharing
institutions and nuclear posture constraints. The typology is constructed by
combining different values on the institutional identity of who exercises political
control of the military in a regime – the institutional identity of the political elite –
along with the specific methods and mechanisms of political control.
128
Chapter Four: Concepts, Measures, and Case Selection
In the previous chapter, I proposed a theory of domestic political constraints
on nuclear posture development in authoritarian regimes. The theory postulated that
authoritarian regimes lacking functional intra-ruling elite power-sharing
institutions confront unique internal political constraints in developing, deploying,
and managing nuclear weapon capabilities required for a first-use asymmetric
escalation posture. This is because in the absence of authoritarian power-sharing
institutions a dictator can acquire an extreme concentration of political power,
which introduces significant distortions into multiple domains of the regime,
including civil-military relations, strategic assessment, and technological
innovation. These distortions negatively affect each dimension of nuclear posture:
capabilities, doctrine, and management. In other words, the absence of intra-ruling
elite power-sharing institutions in personalist authoritarian regimes pose a
structural limit on the nuclear posture options available to the dictator. This is what
I term the paradox of personalism: by purposefully degenerating power-sharing
institutions, a politically unconstrained dictator reduces the regime’s ability to
develop a delegative nuclear posture.
The purpose of this chapter is threefold. First, it is to develop a minimalist
definition of the personalist regime concept that can be utilized as the independent
variable in the empirical case studies that follow. Current use of the personalism
129
concept exhibits considerable conceptual ambiguity and inconsistency. A
minimalist definition of personalism focuses exclusively on the concentration of
decision-making and appointment power in an authoritarian regime. Moreover, a
minimalist conceptual definition of personalism allows maximum variance of civil-
military relations, without jeopardizing or infringing on the conceptual consistency
of the personalist regime concept. A minimalist definition of personalism is
necessary because extensive research agendas are being built upon the concept of
personalist regimes, but few scholars deeply engage with the concept itself.
The second task of this chapter is to develop a suitable measurement index
of a minimalist conceptualization of personalism that can aid in political regime
classification schemes. Existing measurement indices are maximalist, meaning they
mix multiple conceptual dimensions of personalism (concentration of power,
neopatrimonialism, and dynastic sultanism) and include multiple indicators of civil-
military relations to classify a regime as personalist. However, using a “wide net”
approach to personalist regime measurement and classification introduces
substantial ambiguity and increases the potential for measurement error. The end
result is a “class” of personalist regimes that potentially vary dramatically in
political institutionalization and civil-military relations, but without any ability to
recognize which portions of the index were responsible for the classification.
The third task of the chapter is to develop a four part typology of civil-
military relations that can be used as the intervening variable in the analysis of the
130
relationship between intra-ruling elite power sharing institutions and nuclear
posture constraints. The typology is constructed by combining different values on
the institutional identity of who exercises political control of the military in a
regime – the institutional identity of the political elite – along with the specific
methods and mechanisms of political control. In other words, the who and the how
of political control of the military are evaluated in tandem with disaggregated
authoritarian regime type.
The fourth section of the chapter discusses how to measure the dependent
variable of nuclear posture. I expand on Narang’s typology of regional nuclear
postures by adding the category for full spectrum denial that encompasses the
global nuclear warfighting postures of the superpowers. I focus on the specific
underlying indicators that form the core dimensions of a state’s nuclear posture:
doctrine, capabilities, and management.
The final section of the chapter discusses research method and case
selection. I propose to conduct a plausibility probe of my theory and illustrate its
central hypotheses with qualitative analytical methods. Specifically, I discuss my
decision to conduct an in-depth case study of the original nuclear authoritarian
regime: the Soviet Union under Stalin, (1949-1953). Qualitative research methods
are particularly appropriate for generating hypotheses in this dissertation because
my goal is to map the interaction of the nuclear dictator and the regime’s ruling
131
elites, both civil and military, within the context of specific configurations of
authoritarian political institutions.
Conceptualizing Personalism
The debate on how to properly conceptualize, classify, and measure
authoritarian political regimes, has evolved over many decades.
143
Much of this
debate has sought to aid in the analysis of democratic transitions by classifying
different types of non-democratic regimes according to prominent institutional
features.
144
The remarkable institutional heterogeneity of authoritarian regimes has
often posed a significant challenge to these efforts. Indeed, as Geddes observed,
“different kinds of authoritarianism differ from each other as much as they differ
from democracy.”
145
Despite these challenges, the use of political regime
classifications in cross-national comparative analysis has yielded a range of
important insights, such as differential regime breakdown probabilities, autocratic
survival rates, and the purpose of nominally democratic political institutions –
143
For an excellent review, see Paul Brooker, Non-Democratic Regimes 2
nd
ed. (London: Palgrave
Macmillan, 2009). Also, David Art, “What Do We Know About Authoritarianism After Ten
Years?” Comparative Politics vol. 44, no. 3 (April 2012), p. 351-373.
144
Samuel Huntington, The Third Wave: Democratization in the Late Twentieth Century (Norman,
OK: University of Oklahoma Press, 1991). Barbara Geddes, “What Do We Know About
Democratization After Twenty Years?” Annual Review of Political Science 2 (1999). Barbara
Geddes, Paradigms and Sandcastles: Theory Building and Research Design in Comparative
Politics (Ann Arbor, MI: University of Michigan Press, 2003).
145
Geddes, “What Do We Know About Democratization After Twenty Years?” Annual Review of
Political Science 2 (1999): 121.
132
political parties, parliaments, and elections – in distinctly authoritarian regimes.
146
The pronounced focus on the institutional structures of authoritarian regimes has
led to a far more nuanced understanding of authoritarian politics which has, in turn,
enabled international relations scholars to fruitfully search for empirical
relationships between configurations of domestic political institutions and foreign
policy outcomes.
147
In particular, one type of authoritarian regime – personalist –
has received considerable attention. Personalist regime type has been linked to
decisions to fight wars (war selection/initiation),
148
how those wars are fought
(military organizational practices), and the outcome on the battlefield (battlefield
effectiveness).
149
And, importantly, personalist regimes have been correlated with
higher probabilities of pursuing nuclear weapons.
150
Given the burgeoning
utilization of the personalist regime concept in security studies, extending regime
analysis to the study of nuclear posture formation is a natural next step.
146
Jennifer Gandhi, Political Institutions under Dictatorship (New York: Cambridge University
Press, 2008); Jennifer Gandhi and Adam Przeworski, “Authoritarian Institutions and the Survival
of Autocrats,” Comparative Political Studies Vol. 40, No. 11 (2007): 1279-1301; Jennifer Gandhi
and Ellen Lust-Okar. “Elections Under Authoritarianism,” Annual Review of Political Science vol.
12, no. 1 (2009). Jason Brownlee, Authoritarianism in an Age of Democratization (New York :
Cambridge University Press, 2007).
147
Brian Lai and Dan Slater, “Institutions of the Offensive: Domestic Sources of Dispute Initiation
in Authoritarian Regimes, 1950-1992,” American Journal of Political Science, vol. 50, no. 1
(January 2006), p. 112-126.
148
Jessica Weeks, “Strongmen and Straw Men: Authoritarian Regimes and the Initiation of
International Conflict,” American Political Science Review Vol. 106, No. 2 (2012): 326- 347.
149
Caitlin Talmadge, “The Puzzle of Personalist Performance: Iraqi Battlefield Effectiveness in
the Iran-Iraq War,” Security Studies, vol. 22, no. 2, (2013).
150
Christopher Way and Jessica Weeks, “Making it Personal: Regime Type and Nuclear
Proliferation,” American Journal of Political Science vol. 58, no. 3 (2014).
133
Nevertheless, multiple significant issues remain unresolved in the
conceptualization, classification, and measurement of personalist regimes. First,
what precisely is personalism? And how does personalism relate to other cognate
political regime concepts, such as neopatrimonial and sultanistic regimes.
151
Although often lumped together each of these concepts have distinct empirical
connotations that must be addressed. Second, what is the relationship between
personalism and political institutionalization in authoritarian regimes? Personalism
has often been considered antithetical to robust political institutions, especially in
some regional applications of the concept.
152
Yet the relationship between
personalism and political institutionalization is ill-defined, especially since
prominent empirical examples of personalist regimes were highly institutionalized.
Third, does the concept of personalism require pathological civil-military relations?
It is well understood that personalist leaders can emerge in both civilian and
military regimes, but should extensive indicators of civil-military relations be
embedded within the concept of personalism? Finally, how should personalist
regimes be classified? Is personalism a secondary property that is present to some
151
Max Weber, Economy and Society: An Outline of Interpretive Sociology, ed. Guenter Roth and
Claus Wittich (Berkeley: University of California Press, 1978), p. 231-232. H.E. Chehabi and
Juan J. Linz, eds. Sultanistic Regimes (Baltimore and London: The Johns Hopkins University
Press, 1998).
152
The personalist concept was originally applied to the study of politics in the Third World. See
Guenter Roth, “Personal Rulership, Patrimonialism, and Empire-Building in the New States,”
World Politics 20 (January 1968).
134
degree in all regimes or is it the primary feature, meaning it is the defining feature
of a political regime subtype?
Each of these questions will be briefly dealt with in turn. The goal is to
develop a theoretically consistent but minimalist definition of personalism that can
be utilized to explore the influence of authoritarian power-sharing institutions on
civil-military relations and nuclear posture formation in the empirical chapters to
follow.
Dimensions of Personalism. Existing conceptual definitions of personalism
commonly combine two distinct dimensions: 1) the dictator’s concentration of
decision-making and appointment power, and 2) the personalization of political
interactions in a regime. The first dimension refers to the immense discretionary
power wielded by some dictators to monopolize decision-making power in a regime
and to appoint loyal associates to high office. Although relatively rare in empirical
terms, some dictators amass sufficient political power to arbitrarily hold both
civilian and military elites at risk, thereby reducing rival elites’ ex ante and ex post
ability to remove the dictator from power. As such, neither civilian political
institutions (such as a political party) nor the military “exercises independent
decision-making power insulated from the whims of the ruler.”
153
Simply put,
regime elites – and the political institutions they inhabit – cease to be a functional
153
Geddes, “What Do We Know About Democratization After Twenty Years?” p. 121-122.
135
constraint on the dictator’s exercise of power. The concentration of political power
in a single leader is the essence of the personalist concept.
The second dimension of personalism relates to the degree of
institutionalization (or rationalization) of political power within a regime.
Personalization in this sense indicates the fusion of public and private domains:
“the right to rule is ascribed to a person rather than an office.”
154
The state and
government – to whatever extent it exists – becomes the personal property of the
personalist dictator. Political power is not concentrated in the formal institutions of
the regime. Instead, the dictator rules through the use of informal patronage
networks rather than formal political institutions.
155
Viewed in this way,
personalization of power is consistent with the absence or degeneration of a
regime’s political institutions. Not surprisingly, the end result of this second
dimension of personalization is often the reduction or retardation of state
institutional capacity and governing authority.
156
The dictator essentially subverts
the state’s existing political institutions in the process of consolidating power,
thereby reducing the state’s level of political development.
154
Michael Bratton and Nicolas van de Walle “Neopatrimonial Regimes and Political Transitions
in Africa” World Politics Vol 46, No 4 (July 1994): p. 458. See also, Michael Bratton and Nicolas
van de Walle, Democratic Experiments in Africa: Regime Transitions in Comparative Perspective
(Cambridge and New York: Cambridge University Press, 1997).
155
Rational bureaucracy versus traditional authority.
156
Braut-Hegghammer defines state capacity as “professionalism of the state bureaucracy.” It has
two dimensions: 1) ability to carry out specialized functions, 2) independence from state elites and
political leadership.” See, Malfrid Braut-Hegghammer, Unclear Physics: Why Iraq and Libya
failed to Build Nuclear Weapons (Ithaca and London: Cornell University Press, 2016), p. 7.
136
It is on this second dimension that the theoretical affinities between
personalism and neopatrimonialism and sultanism are most apparent. The
distinction between the concepts is the degree to which political power has been
institutionalized in a rational bureaucracy. The three concepts – personalism,
neopatrimonialism, and sultanism – exist on a spectrum of political
institutionalization, with sultanism best understood as an extreme form of
neopatrimonialism.
157
Therefore these concepts are not interchangeable.
158
This
point has often been obscured because the personalist concept was originally
developed in tandem with and substantially reliant on the concepts of
neopatrimonialism and sultanism.
159
However, a minimalist definition of personalism focuses solely on the first
dimension: concentration of power. It is the immense discretionary power over
appointment and removal from office (civilian and military) and policy decision-
making that is the defining conceptual feature of a personalist regime.
160
The fusion
of public and private domains and the personalization of political interactions are
157
H.E. Chehabi and Juan J. Linz, eds. Sultanistic Regimes (Baltimore and London: The Johns
Hopkins University Press, 1998), p. 6.
158
As Chehabi and Linz (1998: 9-10) write: “A regime in which some sultanistic tendencies are
present, but where the circle of clients is wider and the discretion of the leader less extensive,
should be called neopatrimonial. Personalist rulers whose regimes penetrate society by means of a
political party…do not fit the pure type of sultanistic regime.”
159
This is most evident in the “chain” of citations for Geddes’ personalist concept (1999). She
primarily cites works on neopatrimonialism. And in Paradigms and Sand Castles. More on this
below.
160
There is considerable confusion in using the term “personalism” in a minimalist sense due to
connotation of the personalization of political interactions. For this reason Svolik (2012) adopts
the terminology of “contested” autocrat versus “established” autocrat. However, Svolik’s
terminology has yet to catch on in the international relations literature.
137
secondary aspects that do not necessarily reflect the primary analytical distinction
of a dictator that has freed himself from the credible threat of removal from power
by rival elites.
161
For the purposes of this study, a personalist regime is defined as
one in which the dictator dominates policy decision making and can arbitrarily
remove rival elites from office, punishing them at will and often with considerable
violence.
Personalism and Political Institutionalization. The neopatrimonial
dimension of personalism raises the larger theoretical question of the relationship
between personalism and political institutionalization. The issue of what it means
for an authoritarian regime to be well-institutionalized has been a central debate in
the study of authoritarian institutions. From Huntington onward, a well-
institutionalized authoritarian regime has often been synonymous with a single-
party regime.
162
A strong political party was viewed as a cornerstone of regime
stability.
163
Although it was understood that parties could very in level of
institutionalization, and hence strength, the precise relationship between a
paramount leader’s concentration of power and the party was less clear.
164
Any
161
This is often framed in terms of accountability or audience costs.
162
For a recent and notable example of this approach in international relations, see Caitlin
Talmadge, The Dictator’s Army: Battlefield Effectiveness in Authoritarian Regimes (Ithaca and
London: Cornell University Press, 2015).
163
Huntington (1968: 91): “The prerequisite of stability is at least one highly institutionalized
political party. States with one such party are markedly more stable than states which lack such a
party.”
164
For recent attempt to chart the relationships between leader, party, and military, see Wright,
Honaker, Geddes, “Measuring What You Can’t See: The Latent Characteristics that Structure
138
increase in a leader’s personal power was theorized to come at the expense of the
party.
165
This led to the common understanding that “personalization is typically
seen as antithetical to institutionalization.”
166
Seen in this way, personalism should
only develop from within weakly institutionalized or undeveloped regimes.
167
Yet the operative question is whether an authoritarian regime can be both
highly institutionalized and personalized?
168
Slater answers in the affirmative by
formulating two dimensions of institutional power in authoritarian regimes:
infrastructural and despotic.
169
In the context of a single-party authoritarian regime,
the party’s power to control and order society may remain strong even as the party
Autocratic Rule,” Paper Presented at APSA 2014.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2451510
165
In a similar context, Huntington (1968: 26-27) offers a theory of Stalinism: “Stalinism can be
defined as a situation in which the personal interest of the ruler take precedence over the
institutionalized interest of the party. Beginning in the late 1930s, Stalin consistently weakened the
party. No party congress was held between 1939 and 1952. During and after World War II the
Central Committee seldom met. The party secretariat and party hierarchy were weakened by the
creation of competing organs. Conceivably this process could have resulted in the displacement of
one set of governing institutions by another…Such, however, was neither the intent nor effect of
Stalin’s action. He increased his personal power not the governmental power. When he died, his
personal power died with him.” Yet the CPSU remained in power after Stalin’s death. Stalin
weakened CPSU power-sharing institutions but the regime retained firm control over society.
166
Dan Slater, “Iron Cage in an Iron Fist: Authoritarian Institutions and the Personalization of
Power in Malaysia,” Comparative Politics Vol. 36, No. 1 (October 2003) p. 81.
167
Huntington (1968: 145-146) elaborated on the differences between concentration and
expansion of power. “Typically, the first challenge of modernization to a dispersed, weakly
articulated and organized, feudalistic traditional system is to concentrate the power necessary to
produce changes in the traditional society and economy. The second problem is then to expand the
power in the system to assimilate the newly mobilized and politically participant groups, thus
creating a modern system. This challenge is the predominant one in the modernizing world today.
At a later stage the system is confronted with the demands of the participant groups for greater
dispersion of power and for the establishment of reciprocal checks and controls among groups and
institutions. Many of the communist states of Eastern Europe are grappling with the problem of
adaptation to the pressures for the dispersion of power.” Personalism, an over-concentration of
power, is thus a problem of the third (late) stage of political development.
168
Or stated differently, can a regime be personalist but not neopatrimonial.
169
Slater, “Iron Cage in an Iron Fist,” p. 82.
139
loses the ability to constrain the paramount leader. As Slater writes, “Despotic
power (the power to decide) can become highly personalized, even as
infrastructural power (the power to implement) remains highly institutionalized.
Institutions to curtail the chief executive may falter while institutions to curtail
political opposition remain formidable.” In such a case, the overall macro-
institutional setting of the regime would still remain institutionalized, yet the leader
is unconstrained by intra-ruling elite power-sharing institutions.
This distinction between authoritarian infrastructural institutions and
authoritarian power-sharing institutions is important because it points to the two-
tiered nature of authoritarian political institutions. Personalism, as I define it,
requires the degeneration of only a specific set of authoritarian politics institutions,
what I have termed intra-ruling elite power-sharing institutions. Importantly, these
power-sharing institutions can fail independently of the overall political regime.
For instance, the power-sharing institutions embedded in the Communist Party of
the Soviet Union (CPSU) failed to constrain Stalin from concentrating immense
political power, yet the CPSU did not lose its political grip on Soviet society. The
point here is that the infrastructural strength of the regime can be maintained while
simultaneously being repurposed by the dictator to enhance his or her own power
consolidation. As Slater argues, “In authoritarian regimes, high levels of
infrastructural power facilitate the effective concentration of despotic power.
Institutionalization along one dimension [regime infrastructural power] ironically
140
abets deinstitutionalization along another [elite power-sharing].”
170
In effect, a
personalist dictator in a regime with high infrastructural strength can become a
more powerful leader than a personalist dictator in a regime with low infrastructural
strength.
171
The bottom line is that even otherwise well-institutionalized regimes—
those with strong infrastructural power, such as single-party regimes—can become
personalized.
Distinguishing between infrastructural [implementation] institutions and
power-sharing institutions in authoritarian regimes permits a minimalist definition
of personalism to be focused exclusively on the concentration of power rather than
the nature of political interactions.
172
It is the extreme concentration of power in the
hands of a single leader that is the common conceptual thread that runs through all
three concepts: personalism, neopatrimonialism, and sultanism.
173
Thus, political
power in a personalist regime is concentrated in a single leader but political
interactions need not be primarily informal or personal; a personalist dictator can
opt to rule through formal institutions, subverting and repurposing power-sharing
institutions to consolidate his or her own rule.
170
Ibid.
171
The combination of a powerful dictator and a powerful regime is also at the heart of the
totalitarian concept. In the classic Linz and Stepan (1996: 46) definition of totalitarianism the
leader is unconstrained: “Totalitarian leadership is unconstrained by laws and procedures and is
often charismatic.” The shift from totalitarianism to post-totalitarianism is primarily a change on
this dimension. As they write (1996: 42), “Early post-totalitarianism is very close to the
totalitarian ideal type but differs from it on at least one key dimension, normally some constraints
on the leader.”
172
It also aligns with Brooker’s (2009) notion of ruling institution versus instrument of rule.
173
Svolik (2012) makes this argument.
141
In other words, there is a clear differential in state capacity between
personalist dictators who destroy both infrastructural and power-sharing
institutions and those who preserve infrastructural institutions and only degenerate
elite power-sharing institutions. Thus, infrastructural weakness should not be
included in the definition of personalist regime type. It can vary even within
personalist regimes. In empirical terms, this is the primary difference between
“strong” personalist dictators such as Stalin and Mao (who were able to
successfully build the bomb) and “weak” personalists such as Muammar Qaddafi
and Saddam Husayn (who failed in their quest for the bomb).
174
It is for this reason
that assumptions of reduced state capacity due to institutional decay should not be
included in a minimalist conceptualization of a personalist regime. An authoritarian
regime can be personalist without being neo-patrimonial.
Personalism and Civil-Military Relations. Developing a minimalist
definition of personalism also confronts the challenge of how to conceptualize civil-
military relations under a personalist dictator. On one hand, there is a compelling
rationale to avoid embedding any civil-military relations indicators into the concept
of personalism in order to preserve the ability to empirically assess the relationship
between personalist dictators and their military forces.
175
Yet on the other hand, the
174
Malfrid Braut-Hegghammer, Unclear Physics: Why Iraq and Libya failed to Build Nuclear
Weapons (Ithaca and London: Cornell University Press, 2016).
175
The “transitology” literature often focused on variance in the ruler-military relationship to
explain post-transition outcomes. That such variance existed within the various categories of
personalist, neopatrimonial, and sultanistic regimes is indicative of the potential for different civil-
142
dictator’s control over all relevant institutions in the regime is the inescapable sine
qua non of personalism, and thus inevitably extends to the military institution. As
such, even a minimalist conceptualization of personalism requires an assessment of
civil-military relations; it is simply unavoidable. The conceptual challenge is to
include only those factors that are necessary and thereby retain the maximum ability
to empirically observe potential variation in civil-military arrangements in
personalist regimes. This is important since civil-military relations can be affected
by other factors than domestic institutions, namely external security threat.
176
Two main questions are operative when assessing civil-military relations in
personalist regimes. First, what is the conceptual range of civil-military relations in
a personalist regime? In essence, this question seeks to apply the traditional focus
of civil-military relations theory – who controls the military and how they do it –
to personalist regimes. This also requires addressing a more fundamental question
of whether the leader’s political power base is civilian, military, or some amalgam
of the two. Personalist leaders may emerge from either civilian or military regimes
and this can create civil-military and military-military interactions that are not
present in democracies.
military arrangements in these regimes. See, Richard Snyder, “Paths out of Sultanistic Regimes:
Combining Structural and Voluntarist Perspectivees” in H.E. Chehabi and Juan J. Linz, eds.
Sultanistic Regimes (Baltimore and London: The Johns Hopkins University Press, 1998): 49-81.
176
Michael C. Desch, Civilian Control of the Military: The Changing Security Environment
(Baltimore and London: The Johns Hopkins University Press, 1999).
143
Second, is a specific pattern or method of political control of the military
conceptually necessary for a personalist regime? Put differently, should personalist
regimes be defined, at least in part, by their civil-military arrangements? This
question connects to the larger theoretical issue of whether it is justifiable to
construct concepts of regime type utilizing indicators of civil-military relations.
The current trend of developing disaggregated concepts of political regime type
adds urgency to the question because civil-military relations indicators have played
a prominent role in the classification of personalist and non-personalist regimes.
177
The conceptual task is particularly complicated because contemporary civil-
military relations theory is essentially an off-shoot of democratic theory.
178
It is
primarily concerned with the institutional structures and organizational practices
that preserve and strengthen civilian democratic control, which has often meant de-
politicizing or reducing the political influence of the military institution.
179
Simply
put, civilian control of the military is considered a necessary conceptual condition
177
An alternative approach would be to use a maximalist definition of personalism as a proxy for
specific civil-military relations arrangements. In other words, if civil-military relations indicators
are included in the conceptual definition of personalism, all regimes classified as personalist
would obviously exhibit some or all of those specific (pathological) civil-military arrangements.
The problem with such an approach is that it becomes impossible to empirically disentangle the
effect of a concentration of political power from those of civil-military arrangements. This matters
because civil-military arrangements are increasingly being used as an independent variable in
causal models of a range of international/foreign policy outcomes.
178
As Feaver (1999: 215) writes, “Civil-military relations are just a special, extreme case of
democratic theory, involving designated political agents controlling designated military agents.”
179
The classic work on how the varieties of civilian control affect outcomes such as coup
propensity and professionalism is, of course, Samuel P. Huntington, The Solider and State: The
Theory and Politics of Civil-Military Relations (New York: Vintage Books, 1957).
144
of a democratic regime.
180
Importantly, the linkage between democratic theory and
civil-military relations theory demonstrates that it is common for political regime
conceptualizations to include underlying assumptions of particular civil-military
arrangements.
181
The institutional locus of civilian control of the military in a
democratic regime – civilian political leaders – cannot vary without violating the
conceptual consistency of the definition of democracy. However, the methods of
how civilians control the military can and do vary within democratic regimes. Yet
even the variety of civilian control methods in a democracy are limited by other
aspects of the democratic process, namely the rule of law. Extralegal actions, such
as purges, imprisonment, and executions of senior military leaders, would be
simply out of bounds as a method of civilian control in a democratic regime.
182
Civil-military relations are thus conceptually bounded in democratic
regimes, both in terms of who controls (civilian political leaders) and how they
180
Dahl (1989: 245) writes: “In order for a state to be governed democratically, evidently two
conditions are required: 1) If military and police organizations exists, as they surely will, then they
must be subject to civilian control. But civilian control, while necessary, is not sufficient, for
many nondemocratic regimes also maintain civilian control. Therefore, (2) the civilians who
control the military and police must themselves be subject to the democratic process.” Robert A.
Dahl, Democracy and its Critics (New Haven and London: Yale University Press, 1989).
181
For instance, achieving “civilian supremacy” is a key task in regimes transitioning to
democracy. Felipe Aguero, Soldiers, Civilians, and Democracy: Post-Franco Spain in
Comparative Perspective (Baltimore and London: The Johns Hopkins University Press, 1995).
See also, Issues in Democratic Consolidation: The New South American Democracies in
Comparative Perspective (Notre Dame, IN: Notre Dame Press, 1992).
182
Feaver discusses “extralegal civilian action taken against specific military personnel” as a
miscellaneous category of potential civilian punishment mechanisms used to control the military,
but leaves largely unexplored why such punishments are unlikely in a democracy. See, Peter D.
Feaver, Armed Servants: Agency, Oversight, and Civil-Military Relations (Cambridge, MA:
Harvard University Press, 2003), p. 93-94.
145
control the military (with ostensibly legal methods). In this sense, civil-military
relations in democratic regimes should operate within a relatively narrow spectrum
– bounded and grounded by democratic theory and process – with the level of
professional autonomy and influence afforded to the military organization varying
across different democratic regimes.
183
Or, in Feaver’s terms, the difference
between delegative and assertive civilian control. However, even within this
relatively narrow range, variations in civil-military arrangements generate
observable outcomes of theoretical interest. Allowing civil-military relations to
vary independently of democratic regime type, even if only in the narrow band of
delegative-assertive, has facilitated empirical analysis of specific military-related
outcomes, such as military organizational patterns, battlefield effectiveness,
doctrinal debates, and nuclear management.
184
In contrast to the study of civil-military relations in democratic regimes,
there is no comparable foundation of autocratic theory on which to ground a theory
183
In discussing the legacy and longevity of Huntington’s theory of civil-military relations, Feaver
(2003:9) correctly notes the link between democratic theory and Huntingtonian civil-military
relations. “It is democratic theory that distinguishes between the instruments of coercion and the
people, and requires the former to be the servant of the latter. Thus, democratic theory motivates
not only the distinction but also the focus on ‘control’…Huntington survives these challenges,
then because he grounded his theory in a deductive logic derived from democratic theory while his
critics have not.”
184
Peter D. Feaver, Guarding the Guardians: Civilian Control of Nuclear Weapons in the United
States (Ithaca and London: Cornell University Press, 1992). Deborah D. Avant, “The Institutional
Sources of Military Doctrine: Hegemons and Peripheral Wars,” International Studies Quarterly,
vol. 37, no. 4 (1993), p. 409-430. Elizabeth Kier, Imagining War: French and British Military
Doctrine between the Wars (Princeton: Princeton University Press, 1997). Stephen Biddle and
Robert Zirkle, “Technology, Civil-Military Relations, and Warfare in the Developing World,” The
Journal of Strategic Studies, vol. 19, no. 2 (June 1996), p. 171-212.
146
of civil-military (or military-military) relations in authoritarian regimes. This lack
of theoretical grounding has not prevented scholars from extending the concept of
civilian control of the military into authoritarian regimes, particularly single-party
or communist regimes.
185
The emphasis on civilian control, however, has obscured
the fundamental issue in authoritarian regimes is political control of the military,
not civilian control per se. To be sure, civilian and military authoritarian leaders –
similar to democratic leaders – all confront the core issue of how to prevent being
overthrown by their own militaries in a classic coup d’état. But specifying the
conceptual range for authoritarian regimes is more challenging than democracies
because political control of the military is not reducible to civilian control nor do
control mechanisms need to be consistent with any legal rules or norms, democratic
or otherwise. As such, authoritarian leaders, especially personalist dictators, have a
wide range and combination of available civil-military arrangements and political
control mechanisms.
That said, answering the conceptual question of who controls the military
in a personalist regime is relatively straightforward: the leader should personally
control the military.
186
Recall that a minimalist definition of personalism focuses
exclusively on the concentration of appointment-removal and decision-making
185
This was especially noticeable during the Cold War studies of Soviet civil-military relations, a
topic which will be discussed in considerable detail in chapter five.
186
How to measure a dictator’s control over the military is an altogether more difficult challenge.
This will be discussed at length below.
147
power within the regime. The underlying concept suggests that a personalist
dictator exercises near-total control over the military, meaning it truly is an “army
of one” insofar as policy decision-making and appointments and removal from
office are concerned. It is the dictator who has the power to decide military and
national defense policy, appoint and remove commanders at will, develop doctrine,
dominate strategic assessment processes, oversee weapons procurement, etc. The
dictator’s discretion in these decisions and policy domains should not be subject to
collective decision-making procedures or review.
The question of how a personalist dictator controls the military is
considerably more complex. This is because the nature of personal political control
of the military is intricately linked with the dictator’s survival strategy. As
Huntington pithily noted, “In the personal dictatorships, the ruler did everything he
could to ensure that the military was permeated by and controlled by his cronies
and agents, that it was divided against itself, and that it served his purpose of
keeping a tight grip on power.”
187
Personal control of the military is thus marked
by the prime objective of sustaining the dictator’s survival in power and not
necessarily building military capabilities to confront external threats. Military elites
– by virtue of possession of arms and knowledge of violence – pose a unique
obstacle to the dictator’s attempts to concentrate and consolidate political power.
187
Samuel P. Huntington, “ Reforming Civil-Military Relations." Journal of Democracy, Vol. 6
No. 4, 1995, p. 10.
148
For this reason a personalist dictator is often expected to control the military in a
quite specific manner to mitigate the risk of a coup: sharply curtailing military
autonomy, centralizing command authority, and frequently rotating military leaders
to ensure the loyalty of the officer corps.
This raises the second question of whether the particular methods of
achieving and maintaining total personal control of the military are unique to
personalist regimes because they reflect the underlying distribution of political
power in the regime. In other words, are there certain control methods that can only
be used in authoritarian regimes in which political power is highly concentrated in
the hands of a single leader, and thus are indicative of a personalist regime? The
answer is “yes” and such methods are epitomized by a personalist dictator’s
discretionary ability to subject senior military officers to purges with extreme
prejudice– or what I term despotic control. As I explained in the previous chapter,
of all possible political control mechanisms the violent purge is the most harmful
to the military institution and the most dangerous for a dictator to employ, lest it
trigger a coup in response. It is a clear, and often public, signal of the dictator’s
power over the military institution. If the military institution does not possess the
ability to protects its own senior leadership from arbitrary arrest, imprisonment, or
execution, it is unlikely to have any independent ability to constrain the dictator
through a priori or ex post punishment (e.g. threat of removal by coup). The
dictator’s ability to utilize despotic control methods is a direct consequence of the
149
lack of political constraints and the failure of authoritarian power-sharing
institutions.
188
Although violent purges represent the most severe form of political control,
few personalist dictators rely exclusively on such control methods to manage their
militaries. A dictator can adopt several management approaches, including making
the military a core pillar of the regime and “buying” military officers’ loyalty and
compliance through material transfers and special perquisites. In fact, a dictator can
utilize multiple control mechanisms simultaneously or in sequence: co-opting or
coercing senior officers as necessary. The dictator can also redefine his institutional
relationship with the military: appointing himself generalissimo and putting on a
uniform (bedecked with a chest full of medals, of course) even if he never served
in the military, or taking off a uniform – if he was formerly a military officer – and
ruling as a civilian. Indeed, military personalist leaders (strongman) confront
unique challenges in managing their former colleagues and consolidating power. A
strongman’s military professional background does not ensure that other senior
military officers will be spared from despotic control methods. In fact, senior
military colleagues are likely the greatest potential threat to a strongman’s rule. As
such, they are the most likely targets for removal, be it through forced retirement
or more violent means.
188
The institutional arrangements that Stalin used to achieve personal control of the military –
such as personal secretariats, office of the leader, or special security forces – will be discussed in
chapter five.
150
The conceptual argument being advanced here is that despotic control
should only be available to personalist dictators because only they have amassed
such extreme discretionary power to enable them to periodically execute senior
officers.
189
After consolidating power, however, personalist leaders may instead opt
for more benign management practices. Thus, despotic civil-military relations are
most likely to be evident during the dictator’s power consolidation phase and during
political crises. The precise relationship of a personalist dictator to the regime’s
military forces is often in flux. Civil-military relations are rarely static in
authoritarian regimes, especially personalist regimes. Instead, civil-military
arrangements and specific military organizational practices are dynamic variables
that fluctuate in response to distinct internal and external threat environments.
In short, personalist dictators have a wide range of civil-military approaches
that they can employ at distinct times and in response to distinct internal and
external threats. Some of these approaches may be pathological, others more
professional. Indeed, realist logic would suggest a strong role for external threats
in shaping a state’s civil-military relations.
190
How a personalist dictator responds
189
Authoritarian regimes with functional power-sharing institutions, whether civilian (machines) or
military (juntas), do not wield the same discretionary power to arbitrarily execute senior military
officers. Indeed, often they do not exhibit the ability to remove/replace particularly distinguished
military officers. This is because power-sharing institutions in authoritarian regimes create
opportunities for multiple elite coalitions to potentially form. If the leader in one of these types of
regimes moves against the military in a vicious manner, the military institution may align with
rival elites to form an alternate coalition and remove the leader. This is the very nature of power
balancing in an authoritarian regime. The threat of an alternate elite coalition induces caution.
190
Peter Gourevitch, “The Second Image Reversed: The International Sources of Domestic
Politics,” International Organization, vol. 32, no. 4 (Autumn 1978), p. 881-912.
151
to such stimuli is best left for empirical analysis rather than being included in the
concept of personalism. Indeed, it is a far better approach to empirically evaluate
how a personalist leader adapts to external threat stimuli by adopting differing
military organizational practices (or not) than to a priori assume a specific civil-
military arrangement. That said, evidence of a dictator using despotic control
methods is a sure sign that the dictator has achieved sufficient personal control over
the military to qualify as a personalist dictator.
Classifying Personalist Regimes. There are two main approaches to
classifying personalism: 1) classifying personalist as a “pure” regime type, meaning
it is the defining feature of a political regime and thus constitutes a distinct
authoritarian regime subtype,
191
and 2) classifying personalism as a secondary
feature of a political regime, a property that may be more or less present in any
regime.
192
In other words, is personalism a sufficient defining feature of a regime
to warrant being a distinct analytical subtype? Put differently, the issue is whether
to treat personalism as a dichotomous or continues variable.
For application of structural realist logic to civil-military relations, see Michael C. Desch, Civilian
Control of the Military: The Changing Security Environment (Baltimore and London: The Johns
Hopkins University Press, 1999). Staniland develops a multi-causal account of civil-military
relations. However, he is primarily concerned with explaining the maintenance of civilian control,
e.g. coup prevention, rather than the varieties of civilian control patterns. See, Paul Staniland,
“Explaining Civil-Military Relations in Complex Political Environments: India and Pakistan in
Comparative Perspective,” Security Studies, vol 17, no. 2 (2008), p. 322-362.
191
Huntington (1991); Geddes (1999; 2003). Also the Weberian inspired “Sultanism” and
“Neopatrimonialism” of Chehabi and Linz (1998); Snyder (1998); Bratton and van de Walle
(1994;1997).
192
Brooker (2009); Hadenius and Toerrell (2007); Svolik (2012).
152
The first approach is most commonly associated with Geddes’ typology of
personalist, military, single-party, or amalgams of pure types.
193
Geddes defines
personalist regimes as distinct from “both military and single-party [regimes] in
that access to office and the fruits of office depends much more on the discretion
of an individual leader. The leader may be an officer and may have created a party
to support himself, but neither the military nor the party exercises independent
decision-making power insulated from the whims of the ruler.”
194
The conceptual
core of the Geddes quadripartite typology is built on differentiating personalist
regimes from other types authoritarian regimes, namely single-party and military
regimes.
195
This approach requires an explicit specification of the dictator’s
relationship to pre-existing and newly created political institutions.
Importantly, Geddes frames the issue of a political party’s role under
personalism primarily as a supporting institution the dictator may create after rising
to power. In this framing, the party may not exist prior to the dictator calling it into
existence. This is consistent with Geddes’ emphasis on the neopatrimonial
193
Barbara Geddes, “Authoritarian Breakdown: Empirical Test of a Game-Theoretic Argument,”
Working Paper (1999); “What Do We Know About Democratization After Twenty Years?”
Annual Review of Political Science 2 (1999): 115-44; Paradigms and Sandcastles: Theory
Building and Research Design in Comparative Politics (Ann Arbor, MI: University of Michigan
Press, 2003).
194
Geddes, “What Do We Know About Democratization After Twenty Years?” p. 121-122.
195
Huntington (1991: 110) was the first to propose a tripartite classification scheme. “The regimes
that moved to and toward democracy in the third wave generally fell into three groups: one-party
systems, military regimes, and personal dictatorships.” Moreover, Huntington (1991:111) defines
personalism in the following way: “The distinguishing characteristic of a personal dictatorship is
that the individual leader is the source of authority and that power depends on access to, closeness
to, dependence on, and support from the leader.”
153
dimension of personalism by revealing the presumption that personalism primarily,
if not exclusively, afflicts poorly institutionalized regimes (e.g. those lacking a
single/dominant political party).
196
Thus, for Geddes, it matters a great deal when
the party was created because it is a signal of the party’s strength and potential
ability to constrain an aspiring personalist dictator. There is no presumption of
political independence for a party that was created by a personalist dictator. Indeed,
political parties may be serially abolished and created according to the dictator’s
whims.
197
The dictator’s relationship with a pre-existing political party is more
nuanced. Even if a political party exists, it may not be strong enough to constrain
the dictator. As Geddes writes, “Institutionally, what [personalist regimes] have in
common is that although they are often supported by parties and militaries, these
organizations have not yet become sufficiently developed or autonomous to prevent
the leader from taking personal control of policy decisions and the selections of
regime personnel.” The presumption is that a regime with a strong political party
will not degenerate into personalism. This is reflected in Geddes’ coding of both
the Soviet Union under Stalin and the People’s Republic of China under Mao as
single-party rather than personalist regimes. The existence of communist party
institutions with functioning local level organizations in both regimes was
196
Geddes (2003: 53).
197
This would no doubt affect the overall institutional (infrastructural) power of the regime.
154
sufficient to mitigate a personalist coding in the Geddes typology despite the
massive concentration of power that both Stalin and Mao amassed.
The challenge of specifying a personalist dictator’s relationship to pre-
existing regime institutions also applies to the dictator’s relationship with the
military. A personalist dictator may emerge from a military authoritarian regime,
and since few, if any, personalist dictators abolish their regime’s military forces
after reaching the pinnacle of power, the continued existence of a military
institution is a near certainty. The role of the military institution in a personalist
regime, however, varies significantly from regime to regime and leader to leader
(discussed above at length). As Geddes notes, even though military officers may
ostensibly staff a regime’s most important political offices, they may be completely
excluded from decision-making if a personalist dictator arises. The personalist
dictator’s precise power relationship to the officer corps must be specified.
The key point is that in both instances a functional assessment of the
dictator’s relationship to the regime’s prominent and pre-existing political
institutions is required for regime classification. In practice, this requires difficult
judgments on how formal political institutions actually function in the regime, a
task which is complicated by the secretive nature of most authoritarian regimes.
198
Moreover, the process of concentrating power is rarely quick or easy for an aspiring
198
Geddes (2003:52-53) discusses the difficulties of distinguishing between “real” and nominal
parties.
155
personalist dictator. Instead personalization of a regime may unfold in multiple
phases as the dictator maneuvers around and manipulates existing political
institutions, often pitting prominent institutions against each other to sequentially
weaken both and further concentrate power in his or her own hands (e.g. using the
party to purge the military and vice versa). The point here is that the dictator’s
relationship to extant political institutions is often in flux, further complicating
classification decisions.
The result is that there are relatively few clear cut empirical cases of “pure”
personalism. This becomes quite evident due to the necessity of the amalgam or
hybrid category in Geddes’ typology.
199
And classifying “pure” personalist regimes
seem to rely on the absence of robust political institutions, thereby emphasizing
(intentionally or not) the connotation of overall regime institutional decay or
weakness in the concept of personalism. Thus, Geddes relies on several hybrid
classifications for regimes that continue to feature political institutions with some
ability to exercise power over the dictator: military-personalist, party-personalist,
and military-party-personalist regimes. These hybrid categories reflect a process of
incomplete or ongoing personalization; the dictator has concentrated significant
power but either the party or military or some combination of both still retain
sufficient autonomy to function as a check on the leader’s personal power. It makes
199
Nearly a quarter of cases were hybrids. As she (2003: 51) writes: “I had to add intermediate
categories to this classification scheme after discovering how many of the cases simply resisted
being crammed into one or another of the original categories.”
156
intuitive sense that not all aspiring personalist dictators possess the requisite
political skills to concentrate power and slip the constraints imposed on them by
their elite rivals. Some dictators will fall short.
However, the use of hybrid categories reveals a central shortcoming in the
Geddes typology: a regime cannot be personalist and non-personalist at the same
time. If personalist regime is to be used as a distinct regime type, it must reflect the
underlying equilibrium change of the dictator’s position vis-à-vis the regime elite
and existing political institutions. Simply put, a regime on the margin between
personalist and non-personalist is not a personalist regime. To be of analytical value
as a category of regime type, a personalist regime it must have a sharp conceptual
cutoff, clearly distinguishing between a personalist and a non-personalist regime.
Hybrids blur the conceptual line between personalist and non-personalist regimes
and, as such, introduce significant ambiguity and confusion.
Critics of the Geddes typology adopt an alternate classification approach
and treat personalism as a continuous trait that is more or less present on all
authoritarian regimes.
200
This second approach to classifying personalism eschews
the practice of creating a distinct personalist regime category and instead focuses
on how personalism varies within and across dictatorships. As Brooker writes,
“Variations in degree of personalist rule from one personal dictatorship to another
200
Axel Hadenius and Jan Toerell, “Pathways from Authoritarianism,” Journal of Democracy
Vol. 18, No. 1 (2007): 143-57.
157
clearly constitute a range, scale or continuum.” This approach attempts to avoid the
difficult classification issues raised in the Geddes typology by treating personalism
as a secondary or supplemental feature of an authoritarian regime, but, as Brooker
readily admits, “it is not so clear how to frame these variations conceptually.”
201
This is an understatement. The problem is that a continuous scale of personalism
does not solve the conceptual issue of a “tipping point” at which juncture the
dictator’s concentration of power becomes the most dominant and prominent
political feature in the regime. This is apparent when Brooker adopts a principle-
agent analytical framework, wherein a non-personalist or weak personalist dictator
starts as the agent of a ruling institutional principle (party or military) but over time
reverses the relationship and becomes the sole principle with the supporting
institutions then becoming mere agents of his rule. This reversal of power-sharing
relationships in the regime clearly constitutes a conceptual tipping point, even if
intermediate categories could be demarcated along the continuum of personalism
leading up to it.
202
Similarly, other critics of the Geddes typology also note the theoretical and
importance of the conceptual tipping point between personalist and non-personalist
regimes. As Svolik writes, ”Even if the distribution of power between the dictator
and his allies spans a continuum…there are, in fact, only two qualitatively distinct
201
Brooker, Non-Democratic Regimes p. 126.
202
Brooker, Non-Democratic Regimes, p. 125-129. He includes an intermediate category on the
continuum where the party or military could still function as a “veto player”
158
power-sharing regimes: Contested autocrats [non-personalists] can be credibly
threatened with removal; established autocrats [personalists] have effectively
monopolized power.”
203
Once the tipping point is crossed, removing the dictator
likely requires exogenous action outside the ruling elite coalition, meaning popular
revolt or foreign intervention. Otherwise the personalist dictator is likely to only
leave office by natural causes, namely death at a ripe old age. In other words, there
is no return from personalism once the tipping point has been crossed.
204
It is this
change in the power-sharing equilibrium of the regime that is the conceptual
justification for the analytical distinction between personalist and non-personalist
regimes.
Thus, a regime in the upper range of a continuum of personalism – one that
has crossed the point of no return – has transformed so significantly that it
conceptually warrants its own regime type category. Conceptualizing personalism
as a continuum to avoid difficult classification decisions is a theoretically deficient
solution. So too, however, is the use of hybrid categories of personalism. To achieve
203
Milan W. Svolik, The Politics of Authoritarian Rule (New York: Cambridge University Press,
2012), p. 55. Despite its widespread usage in empirical studies, Svolik advances two major
criticisms of the Geddes typology: 1) it is neither mutually exclusive nor collectively exhaustive
and 2) it “requires difficult classification judgments that weigh incommensurate aspects of
authoritarian politics.” Svolik proposes an alternate conceptual framework for analyzing and
categorizing the institutional heterogeneity of authoritarian regimes based on four dimensions:
military involvement in politics, restrictions on political parties, and concentration of executive
and legislative power.
204
Svolik (2012: 63) explains: “Whereas the transition from contested to established autocracy
happens with a positive (if small) probability, there is no return from established to contested
autocracy. Once a dictator is established, he may, of course, still lose power, but such instances
should be rare and occur primarily by a process that is politically divorced from the interaction
between the dictator and his ruling coalition.”
159
conceptual consistency, the transition between a non-personalist and personalist
regime must be understood as a sharp break. It may not be a sudden break, meaning
the process of personalization is often slow and subject to reversal up until the
tipping point, but once crossed, the personalist dictator is no longer accountable to
elite domestic rivals in the regime and thus can engage in specific foreign policy
behaviors (war initiation, nuclear proliferation, etc.) without fear of punishment
(removal from office).
Minimalist Conceptual Definition of Personalism. In summary, a
minimalist definition of personalism should focus exclusively and explicitly on the
dictator’s concentration of decision-making and appointment and removal power.
It should be understood that the overall level of regime institutionalization may
remain relatively robust, even as the personalist dictator erodes or subverts
authoritarian power-sharing institutions. Moreover, the collapse of intra-ruling elite
power-sharing institutions in an authoritarian regime signals the potential for
despotic civil-military relations (or military-military relations in a strongman
regime).
205
For this reason, using indicators of pathological civil-military relations
is conceptually justified as political control methods such as violent purges are a
clear indication of a dictator’s power over the military. Finally, a minimalist
conceptual definition supports a sharp analytical distinction between personalist
205
The extreme concentration of political power in a single leader, however, will effectively
restrict the upper range of potential civil-military relations from reaching the most delegative
arrangements.
160
and non-personalist regimes. It reflects the profound change in the power-sharing
equilibrium of the personalist regime which justifies classifying it as a distinct
authoritarian regime subtype.
Measuring Personalism: Independent Variable
The preceding section discussed the multiple challenges in developing a
minimalist conceptual definition of personalism. However, developing
measurement indicators consistent with a minimalist and dichotomous concept of
personalism presents a second set of challenges. Existing measures of personalism
suffer from three problems: 1) they are exclusively focused on formal political
institutions and therefore unable to discriminate between personalist and non-
personalist regimes, or 2) they utilize leader tenure as a crude proxy of personalism,
or 3) are maximalist, mixing multiple dimensions of personalism.
For example, Svolik (2012) stresses that the rise of an unconstrained leader
is a systematic but rare phenomenon in authoritarian regimes. He devotes an entire
chapter to explicating the logic underlying the transition from collective
authoritarian rule to personal autocracy, emphasizing the qualitative distinctions
between personalist and non-personalist regimes. Yet his measure of the
concentration of power in authoritarian regimes is entirely focused on formal
161
institutions, in particular legislative and executive selection.
206
A focus on formal
institutions alone, however, is insufficient to distinguish if a leader has
subordinated or subverted the power-sharing function of these institutions, and is
therefore unconstrained by elite rivals. Thus, despite elaborating at length on how
Stalin personified the concept of an unconstrained/personalist dictator, Svolik’s
dataset on authoritarian regimes is unable to distinguish between Stalin and his
successors (Malenkov, Khrushchev, Brezhnev, etc.).
207
A second Svolik dataset on
leadership change in authoritarian regimes does distinguish among the Soviet
leaders but only by the duration of their respective tenures and how they entered
and exited office, which although important, is insufficient. Leader tenure and exit
from office data does not tell the full story about the actual distribution or
concentration of political power in the Soviet regime during leaders’ respective
periods of rule.
208
A similar criticism applies to the measurement of personalism offered by
Hadenius and Toerell (2007), who treat personalism as a continuous trait present in
206
Svolik measures the concentration of power by focusing on two dimensions: legislative and
executive selection. For the legislative dimension, he includes six categories: none; unelected or
appointed legislature; one party or candidate per seat; largest party controls more than 75% of
seats; largest party controls less than 75% of seats; nonpartisan legislature. For executive selection
dimension, he incudes five categories: unelected; one party or candidate; selected by a small,
unelected body; elected by more than 75% of the vote; elected by less than 75% of the vote.
207
All Soviet leaders from Stalin to Gorbachev have the concentration of power coding of “one
party or candidate per seat” for the legislative dimension and “one party or candidate” for the
executive dimension.
208
Stalin ruled for 30 years, whereas Khrushchev and Brezhnev were in power for 11 and 18
years, respectively. Khrushchev exited office as the result of elite removal/coup in 1964, as did
Gorbachev in 1991. All other Soviet-era leaders died in office of natural causes, except Malenkov
who resigned after just two weeks in power following Stalin’s death in 1953.
162
all regimes.
209
To measure personalism they utilize leader tenure, in (mean) years,
arguing that the logic of personalism requires infrequent leadership changes. In
other words, the longer a leader stays in power the greater the likelihood that the
leader can concentrate power. Persistence in power is indeed a hallmark of
personalism. However, even they seem to be aware that leader tenure is a too crude
to function as an accurate measure of personalism.
210
Personalism no doubt
correlates with long leader tenure, but leader tenure is a wholly insufficient measure
of personalism.
What is needed, then, is a measure of authoritarian regime type that is
sensitive enough to register internal power shifts between dictators and their allies
and rivals within the institutional structures of the particular regime over time.
Weeks (2012) seeks to provide such a measure by creating indices of personalism
and military leadership based on the “raw” data from Geddes’ original classification
typology.
211
Weeks adopts Slater’s (2003) labels to distinguish among the four
authoritarian regime types created by combining values for military leadership
dimension and personalism: non-personalist civilian regimes (machines), non-
209
Axel Hadenius and Jan Toerell, “Pathways from Authoritarianism,” Journal of Democracy
Vol. 18, No. 1 (2007): 143-57.
210
As they write (2007: 149), “Admittedly, this is a crude proxy for the concentration of power in
the hands of a single person.”
211
Geddes originally used a thirteen question index to measure personalism. Weeks reduces this to
eight questions. The questions that Weeks omitted are the following: 1) Does the leader lack the
support of a party? 2) If there is a support party, was it created after the leader’s accession to
power? 3) If there is a support party, is it limited to a few urban areas? 4) Does the leader govern
without routine elections? 5) If there are elections, are they essentially plebiscites, i.e., without
either within-party or interparty competition?
163
personalist military regimes (juntas) personalist civilian regimes (bosses), and
personalist military regimes (strongmen).
212
Weeks uses a 0.5 cutoff score for each
index to assign regimes into their corresponding categories (e.g. affirmative
answers to half of the above questions), so a regime that scored above 0.5 on the
personalism index and 0.5 on the military leadership index would be coded as a
strongman for that country-year. Table 4.1 displays the questions for both indices.
212
Weeks, “Strongmen and Straw Men.”
164
Table 4.1 Weeks Military Leadership Index and Personalism Index
Military Leadership Dimension:
(1) Whether the leader was a current or former high-ranking military officer?
(2) Whether officers hold cabinet positions not related to the armed forces?
(3) Whether the military high command is consulted primarily about security (as opposed
to political) matters?
(4) Whether most members of the cabinet or politburo-equivalent are civilians?
(5) Whether the Banks dataset considers the government to be military or military-civilian?
Personalism Dimension:
(1) Whether access to high government office depends on the personal favor of the leader?
(2) Whether country specialists viewed the politburo or equivalent as a rubber stamp for the
leader’s decisions?
(3) Whether the leader personally controlled the security forces?
(4) If there is a supporting party, does the leader choose most of the members of the
politburo equivalent?
(5) Was the successor to the first leader, or is the heir apparent, a member of the same
family, clan, tribe, or minority ethnic group as the first leader?
(6) Has normal military hierarchy been seriously disorganized or overturned, or has the
leader created new military forces loyal to him personally?
(7) Have dissenting officers or officers from different regions, tribes, religions, or ethnic
groups been murdered, imprisoned, or forced into exile?
(8) If the leader is from the military, has the officer corps been marginalized from most
decision making?
165
The Weeks personalism index is a major step in the right direction in terms
of measuring personalism as a dichotomous equilibrium shift in an authoritarian
regime. However, the Weeks personalism index is clearly maximalist, meaning it
mixes multiple conceptual dimensions of personalism (concentration of power and
neopatrimonialism and dynastic sultanism) and includes multiple indicators of
civil-military relations to classify a regime as personalist. The ambiguity of the
maximalist approach to measuring personalism is further exacerbated by the
arbitrary 0.5 cutoff: any mix of four positive answers to the eight questions is
sufficient to render a “personalist” coding.
213
This, in effect, substantially lowers the bar for a personalist coding,
resulting in a less restrictive coding scheme than originally employed by Geddes,
which is readily apparent by the increased numbers of personalist regimes in
Weeks’ dataset – including regimes that Geddes considered hybrids or even non-
personalist. Moreover, it also increases the potential for measurement error because
it blends dimensions of personalism and fails to specify the relationship between
the dimensions or even between the various questions. As a result it is unclear
213
Weeks (2012: 39) defends this decision: “Choosing the particular .5 cutoff did not affect the
substantive results, nor did weighting certain subcomponents of the index more heavily than
others or revising components of the index in reasonable ways.” However, for this to be true
certain questions of the index must be picking up the core concept of personalism, whereas others
are observable implications of the foundational concept. This means that many of the questions are
superfluous or are actually outcomes of the foundational concept. I would argue that questions 1-3
likely drive the answers to the remaining questions. In other words, the concentration of power is
the core concept and neopatrimonial and civil-military relations outcomes are the result.
166
which specific questions, if any, are required to be positive to ensure a personalist
coding.
For instance, the inclusion of three civil-military relations indicators in the
index (questions #6-8) means that only a single additional question needs to be
positive to potentially classify the regime as personalist.
214
Yet in another
combination it is possible for a regime to be coded personalist without any reference
to civil-military questions at all (e.g. a positive answer to the four questions not
dealing with civil-military relations). Furthermore, at least two of the questions
(#5,7) refer to kinship dynamics that reflect the neo-patrimonial dimension of
personalism. And only two questions (#2,4) specifically reference authoritarian
political institutions capable of imposing constraints on a dictator. Quite simply,
the Weeks personalism index is a jumble of indicators lacking conceptual clarity.
The primary implication is that the ability to analyze and draw inferences
from a “class” of personalist regimes is actually reduced by the ambiguity of the
“wide net” approach to personalist regime measurement and classification. In other
words, the Weeks index can render many different types of personalist regime
classifications. The end result is a “class” of personalist regimes that potentially
vary dramatically in political institutionalization and civil-military relations, but
without any ability to recognize which portions of the index were responsible for
214
Question number three concerning the control of security forces could also be broadly
interpreted as concerning civil-military relations. Thus a personalist coding could be achieved by
solely focusing on the leader’s control of the military and security sectors.
167
the classification. To some extent, the source of the ambiguity of the Weeks
personalism index is due to the using the same index to measure personalism in
both civilian and military regimes. Thus, some questions in the index are more
applicable, and perhaps necessary, depending on whether the ruling institution of
the regime is military or civilian.
215
But, again, weighting all the questions in the
index equally while only requiring positive answers to an unspecified combination
of half of them to render a personalist classification opens the door to substantial
conceptual confusion and measurement error.
216
A better approach is to build on the Weeks index by reformulating it to
reflect a minimalist conceptualization of personalism, focusing exclusively on
measuring the concentration of power dimension, and then require that all primary
questions/indicators in the index are positive/affirmative to justify a classification
as a personalist regime. A reformulated personalism index would thus seek to
directly measure the leader’s ability to remove rival elites from power-sharing
institutions and decision-making bodies embedded in the regime’s existing political
institutions. This would reduce the ambiguity in the current measurement index and
215
This could also be the result of Weeks’ formulation of audience costs in authoritarian regimes
being military or civilian and not both. In other words, a military leader in a junta faces audience
costs only from a military audience. A civilian dictator only from a civilian audience. In practice,
however, a civilian dictator often faces dual audiences: both civilian and military.
216
The measurement error would be an increase in false positives, meaning coding regimes as
personalist that in fact were not.
168
would produce a more consistent measurement of personalism since regimes would
not need to exhibit neopatrimonial traits in order to receive a personalist coding.
The reformulated index of personalism consists of six primary and two
supplemental questions that similarly must be affirmative (when applicable) to
render a coding as a personalist regime:
(1) Whether the leader personally controls appointments to and removal from
the regime’s highest level decision-making institutions (politburo, ruling
council, or military equivalent)?
(2) Whether country specialists viewed the regime’s highest level decision-
making institutions (politburo, ruling council, or military equivalent) as a
rubber stamp for the leader’s decisions?
(3) Whether the leader personally controls the internal security forces or secret
police?
(4) Have dissenting elite political rivals been murdered, imprisoned, or forced
into exile?
(5) Whether the leader personally controls the appointment to and removal
from command of the regime’s senior military officers?
(6) Have dissenting military officers been murdered, imprisoned, or forced into
exile?
(7) If formal term limits exist, has the leader exceeded or circumvented them?
169
(8) If the leader is from the military, has the officer corps been marginalized
from most decision making?
The reformulated personalism index is consistent with both the audience cost
and equilibrium shift conceptualizations of personalism. In other words, the index
focuses on what I will term the “removal game” – who can remove whom from
power. In the absence of constitutional or other rule-based mechanisms for the
routine removal and replacement of the paramount leader, the only option for rival
elites to punish the dictator is to band together and force the leader from power by
threat or actual use of force. In practice, such palace coups require significant
coordination among rival elites, including both civilian and military elites. The
primary opportunity for such coordination to take place in an authoritarian regime
is within intra-ruling elite power-sharing institutions due to the routine interaction
among the political elite. Without routine interaction, overcoming the elite
coordination problem inherent in such an endeavor as likely to be challenging.
Thus, even when true power is not located within formal institutions, the
opportunity for routine elite interaction in such institutions generates a secondary
check on the leader’s power.
217
217
In the DPRK under Kim Jong-il, senior regime leaders could not gather together without his
approval. For example, Jang Song-taek received a reprimand for attending wedding without
permission. See, Ken E. Gause, North Korean House of Cards: Leadership Dynamics under Kim
Jong-un (Washington D.C.: Committee for Human Rights, 2015), p. 46-47, fn#102.
170
It is for this reason that when assessing if a regime’s highest level decision-
making institutions (politburo, ruling council, or military equivalent) act merely as
a rubber stamp for the leader’s decisions, it is necessary to focus on the specific
interactions within these institutions. In other words, even if rival elites cannot
effectively challenge or veto a personalist leader’s decisions, routine interactions in
formal institutional settings can serve a secondary purpose as coordination sites for
elites.
218
Therefore, the meeting frequency, attendance, and even location of a
regime’s formal decision-making institutions are all germane to the task of
understanding the balance of power between the leader and the ruling coalition.
Furthermore, it is not always clear which formal decision-making
institutions are the true power centers in the regime. This is because aspiring
personalist dictators often create new committees, institutions, and organizations to
increase and consolidate their own power since they can better control who is
appointed to non-legacy institutions. In the process, they can shift and restrict
responsibility among various institutions and committees. Moreover, personalist
dictators often hold more than a single leadership position within a regime; indeed,
it is quite common for a personalist leader to continue to formally hold top positions
in defunct decision-making institutions even as they proceed to create new ruling
committees. It is critical, therefore, to not only isolate where power resides in a
218
Milan W. Svolik, The Politics of Authoritarian Rule (New York: Cambridge University Press,
2012).
171
regime, but also to understand how many levers of power a particular leader
controls. As such, in the empirical chapters, particular attention will be paid to
counting the number of leadership positions held by the paramount leader.
Measuring Civil-Military Relations: Intervening Variable
As discussed in previous chapters, the nature of a regime’s conventional
civil-military relations is a key variable in the two extant predictive models of
nuclear posture.
219
Feaver argues that emerging nuclear nations will select
command and control procedures based on a rough balance of internal threats (fear
of coups) and external threats (decapitation attacks).
220
The more a regime’s leaders
fear a military coup, the less autonomy will be granted to military organizations
controlling the nuclear arsenal. In a more extensive and sophisticated framework
for predicting the nuclear postures of regional powers, Narang also focuses on
military autonomy and organizational practices, arguing that assertive
(conventional) civil-military relations are a key constraint for the type of nuclear
posture a regional power may select. Assertive civil-military arrangements
generally foreclose an asymmetric escalation posture due to restrictions on arsenal
219
Peter D. Feaver, “Command and Control in Emerging Nuclear Nations,” International Security
Vol. 17, No. 3 (1993): 160-187. Vipin Narang, “Posturing for Peace?: The Sources and Deterrence
Consequences of Regional Nuclear Postures,” Ph.D. dissertation, Harvard University, 2010. Also,
Vipin Narang, “Posturing for Peace? Pakistan’s Nuclear Postures and South Asian Stability,”
International Security Vol. 34, No. 3 (2009): 38-78. Vipin Narang, Nuclear Strategy in the
Modern Era: Regional Powers and International Conflict (Princeton and Oxford: Princeton
University Press, 2014).
220
Feaver, “Command and Control in Emerging Nuclear Nations.”
172
management, specifically forward deployment with military units and delegation
of release authority to military commanders.
221
In both models, however, civil-
military relations are decoupled from any analysis of the nuclear state’s political
regime type.
In fact, Narang argues explicitly against an analytical focus on regime type.
As he writes, “Regime type does not provide the appropriate variable because not
all democracies possess delegative structures, and not all autocracies have assertive
civil-military regime types.”
222
In other words, regime type is not a reliable
predictor of civil-military relations because there is substantial variation of civil-
military relations within and across regime types.
223
However, Narang’s analysis is
limited by his use of the traditional dichotomous variable of regime type
(democracy/autocracy) and is based wholly on the empirical referents of
democratic India’s assertive civil-military relations and authoritarian Pakistan’s
delegative arrangements. By focusing on Pakistan’s military regime, he obscures
the major challenge civilian authoritarian regimes have in developing delegative
structures.
In contrast, this study views civil-military relations as an intervening
variable that is significantly shaped by a political regime’s power-sharing
institutions and subsequently affects two distinct dimensions of nuclear posture:
221
Narang, Nuclear Strategy in the Modern Era.
222
Ibid. p. 39.
223
Ibid. p. 38.
173
doctrine and management. The argument advanced here is that civil-military
relations must be analyzed in tandem with regime type in order to understand the
strength of the potential internal constraint on a regime’s nuclear posture options.
This is because there is considerable overlap between current disaggregated regime
type classifications and civil-military indicators. Distinguishing between civilian
and military authoritarian regimes, and between personalist and non-personalist
authoritarian regimes creates additional analytical leverage to understand the
interaction between disaggregated regime type and civil-military relations.
To evaluate the civil-military arrangements of the empirical cases under
study, a four part typology of civil-military relations will be utilized (ranging from
lowest to greatest military autonomy): despotic, assertive, delegative, and
praetorian. Coding decisions will be based on a set of qualitative indicator
questions crafted to determine the nature and extent of political control of the
military in the each regime. The Weeks Military Leadership Index (above) will be
used to code whether a regime’s political leadership is civilian or military, as well
as the extent of the military institution’s infrastructural role in the regime. The focus
will be on both the who and the how of political control. In terms of who controls
the military, there will be four possible options:
Civilian control (collective leadership)
Military control (collective leadership)
174
Personalist control (civilian dictator)
Personalist control (military dictator)
As to the how of political control, four categories of control mechanisms will be
evaluated:
Promotions and Assignments: Are promotions and assignments to senior
command positions based primarily on professional merit or on political
loyalty (ideological or personal)? Is military hierarchy broadly respected in
promotion policy? Are there predictable patterns of career advancement?
Command Authority: Is command authority concentrated with central
political authorities or is it dispersed to lower echelon commanders? Does
tactical command authority reside on the battlefield or elsewhere? Are there
dual command arrangements, such as commissars, or other constraints on
command authority? How involved are political leaders in military
operations? Are there restrictions on training exercises or internal troop
movements?
Monitoring Mechanisms: How intrusive are oversight mechanisms? Are
senior officers subject to surveillance by secret police or another security
organization? Do other institutional checks exist (e.g. competing
175
paramilitary forces)? Are there fire alarms, police patrols, or other forms of
monitoring?
224
Punishment: Are senior officers punished for political crimes or solely for
poor performance/professional misconduct? How severe are potential
punishments (e.g. reductions in rank, discharge, forced retirement,
imprisonment, or death)? Is punishment arbitrary or predictable (e.g. subject
to regulation, due process, or other legal norms)? How frequent and
widespread are purges (e.g. narrowly targeted at specific officers or
throughout the rank-and-file)? Does punishment occur under the legal
auspices of the military institution (e.g. military tribunals or courts martial)
or in civilian/political tribunals? Are tribunals secret or public? Do political
leaders use material disincentives (e.g. budget cuts) to enforce the political
quiescence of the military?
The typology of civil-military arrangements is constructed by combining different
values on the four categories of political control mechanisms (the how) with the
four possible options of the institutional identity of political control (the who). This
approach yields the following typological categories:
224
See Feaver (2003:86) for a list of monitoring mechanisms derived from principle-agent
theoretical approaches.
176
Despotic. The dictator (civilian or military leader) personally controls the
military. Personal loyalty is primary metric for promotion and assignment.
Command authority is extremely concentrated. There is extensive and persistent
political involvement in military operations. Training and troop movements are
highly restricted. Monitoring mechanisms are exceptionally intrusive. Punishments
are the most severe, including execution, and will often be politically motivated.
Military autonomy is essentially non-existent under despotic civil-military
relations.
Assertive. A collective civilian political elite control the military under
assertive civil-military relations. Promotion and assignments may be based on a
mix of professional merit and loyalty. Military hierarchy is likely respected, though
not determinative, in promotion and assignment procedures. There is moderate to
high centralization of command authority, with dual command arrangements, such
as commissars, or other constraints on command authority. There is extensive
civilian political involvement in military operations and restrictions on training
exercises or internal troop movements. There is the likelihood of significant
monitoring and surveillance mechanisms. Wide-scale purges and executions are
unlikely, though removal from service and dismissal for political reasons is quite
likely. The military institution can achieve periods of professionalism, though this
subject to reversal. As such, some degree of institutional separation between the
177
military and political elite should be noticeable.
225
In sum, assertive civil-military
relations will exhibit low to moderate military autonomy.
Delegative. A collective civilian political elite also controls the military
under delegative civil-military relations. However, promotions and assignments are
based on professional merit. Command authority is widely dispersed and there is
relatively low political involvement in military operations. Delegative civil-
military relations is best described as a principle-agent relationship, where the
civilian principle can reclaim delegated authority from the military agent at will.
Monitoring mechanisms will be non-intrusive (such as fire alarms and police
patrols) and should focus on professional compliance rather than political loyalty.
Punishments will be both predictable and mild (reduction in rank, discharge) and
solely for poor professional performance or misconduct, and not for political
loyalty.
Praetorian. A collective military-political elite controls the military
institution under praetorian civil-military relations.
226
Indeed, the defining feature
of praetorian arrangements is that military officers exercise political control of the
regime. In other words, command authority is alienated from civilian leaders. The
role of military hierarchy in promotions and assignments is dependent on whether
225
As Feaver writes (1992: 9): “Assertive control does not presuppose that the military will
conform to society or that the professional officer corps will be completely undermined. Rather it
allows for qualitatively separate institutions, civil and military, which remain philosophically and
sociologically distinct.”
226
As such, civil-military relations is actually a misnomer.
178
the original coup that brought the military into political power was hierarchical (led
by highest ranking general officers) or non-hierarchical (led by lower ranking
officers or enlisted personnel). If military hierarchy was subverted in the process of
taking power, then promotion and assignment patterns are likely to reflect loyalty
to the junta. Command authority is likely to be widely dispersed, per military
professional norms. Monitoring mechanisms are likely to be more intrusive than in
delegative civil-military arrangements, and may feature counter-balancing
institutions to guard against another coup. Table 4.2 reviews the characteristics of
the civil-military relations typology and indicators.
179
Table 4.2 Civil-Military Patterns and Indicators
Promotions
&
Assignments
Command
Authority
Monitoring
Mechanisms
Punishment
Despotic
(Personal
Control)
Loyalty
Extremely
Concentrated
Most
Intrusive
Most Severe;
Arbitrary;
Political
Crimes
Assertive
(Collective
Civilian
Control)
Loyalty +
Merit
Moderate-to-
Highly
Concentrated
Intrusive
Moderately
Severe; Semi-
Predictable;
Political and
Professional
Performance
Delegative
(Collective
Civilian
Control)
Merit
Widely
Dispersed
Non-intrusive
Mild;
Predictable;
Professional
Performance
Praetorian
(Collective
Military
Control)
Merit +
Loyalty
Widely
Dispersed
Potentially
Intrusive
Mild to
Moderate;
Predictable;
Political and
Professional
Performance
Measuring Nuclear Posture: Dependent Variable
The cross-national study of nuclear posture is a relatively recent endeavor,
meaning there are few “off-the-shelf” measures of nuclear posture that can be used
for comparative analysis. As discussed earlier, the “key factors” framework offered
by Feaver (1992) primarily focuses on the narrower concept of nuclear command
and control in emerging proliferators rather than a more expansive
180
conceptualization of nuclear posture. Command and control is certainly a key
element of nuclear posture, but it must be combined with capabilities and doctrine
to present a more holistic view of a regime’s nuclear posture.
Narang has proposed a typology of nuclear postures based on four
dimensions: (1) Primary envisioned employment mode (break out to catalyze third-
party intervention; nuclear retaliation; nuclear first-use for conventional denial); (2)
Capabilities (a few unassembled weapons; survivable second-strike forces; tactical
nuclear weapons); (3) Management (recessed and opaque; assertive civilian
control; delegative); and (4) Transparency level (ambiguous capability and
deployment; unambiguous capability and ambiguous deployment; unambiguous
capability and deployment). The combination of values on these four dimensions
creates four categories of nuclear postures: catalytic, assured retaliation, and
asymmetric escalation.
227
For instance, an asymmetric escalation posture is a
highly-delegative, nuclear war-fighting posture which features tactical nuclear
weapons fully integrated into a regime’s military forces and intended for first-use
on the battlefield in a denial mission against an enemy’s conventional forces
(armor, infantry, or naval assets). Pakistan’s nuclear posture (1998-present) and
France (1972-present) are examples of an asymmetric escalation posture.
227
Narang’s classification scheme applies only to regional power nuclear postures (the US, UK,
and USSR/Russia are excluded).
181
In contrast, an assured retaliation posture is a highly-centralized, defensive
nuclear posture which features strategic nuclear capabilities (often de-alerted and
de-mated), intended to be used only in retaliation following an opponent’s first
strike. Put differently, an asymmetric escalation posture tends heavily toward the
“always” side of “always/never” dilemma, eschewing negative controls on the
nuclear arsenal to ensure prompt usage against invading conventional forces when
necessary. Whereas an assured retaliation posture is the reverse, skewing toward
the “never” side through the use of extensive negative controls on the nuclear
arsenal, rendering it incapable of quick reaction.
228
India’s nuclear posture after
1998 and China’s nuclear posture since 1964 are both examples of an assured
retaliation posture.
229
A catalytic nuclear posture consists of an ambiguous, presumably un-tested,
nuclear weapons capability or secret arsenal, which the catalytic state threatens to
reveal to a patron – most often the U.S. – as a political signal to trigger intervention
in order to achieve de-escalation. The catalytic posture requires few, if any, actual
weapons, but critically requires that the catalytic state has, or at least perceives itself
to have, a powerful patron with regional interests significant enough to warrant the
belief that the patron will indeed intercede on its behalf in a crisis. Other important
features of this posture include: the catalytic signal is sent to the patron and not the
228
Feaver, “Command and Control in Emerging Nuclear Nations.”
229
Narang, Nuclear Strategy in the Modern Era.
182
envisioned opponent, command and control of the nuclear assets remains highly
centralized, and nuclear forces are not integrated within the catalytic state’s military
doctrine. The catalytic state’s nuclear assets are viewed as “weapons of last resort,”
and it is the implicit threat to use these weapons in a severe crisis that incentivizes
the patron to intervene and de-escalate any crisis before such a fateful point is
reached. Narang cites Israel (1967-1991), Pakistan (1990-1999), and South Africa
(1979-1990), as empirical examples of regional nuclear powers that have adopted
a catalytic nuclear posture.
230
These three categories of nuclear posture are a useful starting point, but they
do not exhaust all the potential posture options available to proliferating states.
Indeed, the expansive nuclear postures of the superpowers are explicitly excluded
by Narang due to his focus on regional nuclear powers.
231
Thus, a fourth nuclear
posture category needs to be added, what I will term “full-spectrum escalation.”
Full-spectrum escalation is an asymmetric escalation posture on steroids. The goal
of full-spectrum escalation is to have dominance (or at least parity) over a nuclear
rival at each rung of the escalation ladder. This includes robust targeting
capabilities, including real-time intelligence and early-warning assets, such as over-
the-horizon (OTH) radar or space-based satellites to be able to effectuate a
230
I disagree with Narang on the concept of catalytic posture. It is a nuclear proliferation strategy,
a way to ambiguously develop nuclear weapons, not a persistent nuclear posture.
231
Narang (2014: 22-23) directly addresses this point. He refers to it as a “splendid first-strike”
posture, and notes that only the superpowers have adopted such a posture.
183
symmetric denial (counterforce) or blunting mission. Moreover, full-spectrum
escalation is also a global strategy, meaning the extra-territorial deployment of
nuclear weapons on land (regional bases) or at sea (naval vessels and submarines).
The US (1970-present) and the Soviet Union (1982-1991) are empirical examples
of the full spectrum escalation posture. The typology of nuclear postures in
displayed in Table 4.4 below.
232
232
Derived from Narang, Nuclear Strategy in the Modern Era.
184
Table 4.3 Expanded Typology of Nuclear Postures
Catalytic
Assured
Retaliation
Asymmetric
Escalation
Full-Spectrum
Escalation
Primary
envisioned
employment
mode
(Use-Doctrine)
Break out
capabilities
to catalyze
third-party
intervention
Nuclear
retaliation
following
highly
significant
damage
Nuclear first use
in a conventional
denial mission
Nuclear first use in
a conventional
and/or nuclear
denial (counter-
force) mission
Capabilities
Ability to
assemble a
handful of
nuclear
weapons
Survivable
second-strike
forces
First-use
capabilities
(tactical nuclear
weapons)
First-use
capabilities
(tactical nuclear
weapons), global
survivable secure-
second strike
forces (ICBM), and
early-warning
infrastructure
(space-based or
OTH)
Management
Recessed
and opaque
Assertive
civilian
control
Delegative (assets
and authority
integrated into
military forces and
doctrine)
Delegative (assets
and authority
integrated into
military forces and
doctrine)
Transparency
Ambiguous
capability
and
deployment
Unambiguous
capability;
ambiguous
deployment
Unambiguous
capability and
deployment
Unambiguous
capability and
deployment. Extra-
territorial/global
deployment
185
The expanded Narang typology is a useful approach to organizing the
overall nuclear postures of each nuclear state. However, two issues should be kept
in mind. First, significant variance is possible within each posture category because
some states will operationalize certain dimensions (capabilities, especially) of each
posture category in different ways. Thus, a state adopting an assured retaliation
posture may strongly emphasize one leg of the nuclear triad over others, or avoid
building a triad altogether, and instead focus exclusively on one type of survivable
delivery platform (such as a road-mobile ICBM or ballistic missile submarine).
Similarly, a state adopting an asymmetric escalation posture may emphasize one
type of tactical nuclear weapon system exclusively (SRBM or air-dropped bombs),
while others will develop multiple tactical nuclear delivery systems (mines,
torpedoes, artillery, etc.) and survivable second strike nuclear assets, too. In both
cases the overall size of the nuclear arsenal (and the amount of possible damage
inflicted) may vary dramatically among various states who have adopted similar
postures (e.g. two states may have adopted a similar assured retaliation posture but
one state has 20 ICBMs and the other 200 ICBMs.)
Second, the variance within the posture categories and multiple dimensions
of Narang’s conceptualization of nuclear posture underscores the importance of
understanding the interplay among the indicators which determine the mutual
186
exclusivity of the assured retaliation and asymmetric escalation postures.
233
Here I
will focus on the issue of tactical nuclear weapons. The distinguishing
characteristics are as follows: 1) the presence of tactical nuclear weapons versus
survivable second strike weapons, 2) whether command-and-control is
characterized as assertive/centralized or delegative, and 3) whether the weapons are
designed to be used as a first strike against an opponent’s conventional military
forces in a denial mission or against counter-value targets (e.g. cities) in a
retaliatory punishment mission. The underlying assumption is that the values for
each of the three dimensions develop in unison organically. In other words, tactical
nuclear weapons will always be paired with a first-use doctrine and delegative
command-and-control arrangements. The question, however, is how to resolve
discrepancies in the typology when this is not the case – if, for instance, a state
developed tactical nuclear weapons to be used in a retaliatory counter-force
mission against an opponent’s tactical nuclear weapons.
234
Or if a state develops
tactical nuclear weapons but never adopts delegative command-and-control
arrangements.
235
233
I focus on these two postures rather than catalytic posture because I view Narang’s catalytic
posture as a proliferation strategy and not a weaponized posture.
234
In the late 1950s and throughout the 1960s the Soviet Union seemed to conceptualize its
tactical nuclear weapons for use in a retaliatory blunting mission against any remaining NATO
tactical nuclear weapons deployed in Europe following a NATO first strike.
235
The Soviet Union is the clear example of a state developing tactical nuclear weapons but
continuing to prioritize tight central control. See, Bruce Blair, Logic of Accidental Nuclear War, p.
59-114.
187
The assured retaliation posture is defined, in part, by the lack of tactical
nuclear weapons. “The clearest indicator of an assured retaliation posture,” Narang
writes, “is the development of secure second-strike forces, but the absence of
tactical nuclear weapons within a state’s arsenal.”
236
This suggests that the
capabilities dimension, namely the possession or absence of tactical nuclear
weapons, is really the determining factor between an assured retaliation and
asymmetric escalation posture. However, the conceptual problem is further
complicated because capabilities alone are insufficient to render an asymmetric
escalation posture coding in the typology. Tactical nuclear capabilities must be
paired with delegative command-and-control procedures and envisioned to be used
in a specific war-fighting role against conventional forces to qualify as an
asymmetric escalation posture.
237
Thus, tactical nuclear weapons are inconsistent
with an assured retaliation posture, and necessary but not sufficient for an
asymmetric escalation posture. There is no posture category for a nuclear state that
develops tactical nuclear weapons but keeps them under tight control and
envisioned to be used in a retaliatory mission.
The larger point here is that is possible that each dimension of nuclear
posture may be on an independent “developmental” track, meaning weapon design
and production can occur independent of doctrinal development or the construction
236
Narang, Nuclear Strategy in the Modern Era, p.18. Emphasis mine.
237
Ibid., p 19.
188
of command-and-control organizations. Ideally, all three dimensions would move
in tandem to a specific posture goal, but empirically this has not always been the
case. In fact, doctrinal development often lags behind the other dimensions,
sometimes by many years.
238
Moreover, two dimensions (capabilities and
management) are considered determinative, in part, because they are “sticky,”
meaning they cannot be quickly changed, even under crises conditions since they
require significant lead times to develop.
239
The third dimension of use-doctrine or
primary employment mode only becomes determinative when both suitable
capabilities and command-and-control structures exist. But doctrine cannot be
determinative when it does not exist.
This raises the central question of how political regime type may
independently affect each dimension of nuclear posture. Narang’s “Posture
Optimization Theory” addresses only a single political constraint on the
management dimension: civil-military relations. This implies that the management
dimension can be the determining factor in nuclear posture formation. However,
there is no discussion of sequencing or the potential differential developmental
timelines among the multiple dimensions of nuclear posture. Or, to put it more
simply, how constraints on one dimension of nuclear posture may affect other
238
Indeed, neither the Soviet Union under Stalin or China under Mao developed formal nuclear-
use doctrines. Doctrinal development under Stalin will be discussed in the empirical case study
chapter.
239
Narang, Nuclear Strategy in the Modern Era, p. 21-23.
189
dimensions. To explore these questions of developmental sequencing across the
dimensions of nuclear posture in the empirical cases that follow, I will utilize a
series of supplemental indicators for each dimension in addition to those covered
by Narang’s typology of nuclear postures:
Capabilities. Regarding capabilities, I ask the following supplemental questions:
When was the first explosive test of a tactical nuclear warhead?
How many different types of tactical warheads are in the arsenal?
How many different types of tactical delivery platforms (e.g. nuclear-
capable strike aircraft, SSMs, SRBMs, SLCMs, unguided rockets, artillery,
land mines, torpedoes, depth charges, naval mines, etc.) are in the arsenal?
Was the decision to pursue tactical nuclear weapons a top-down or bottom-
up decision (e.g. did it originate with senior political leaders or with nuclear
scientists or the military)?
Does the state have real-time intelligence targeting capabilities?
Does the state have space-based or over-the-horizon (OTH) early-warning
radar capabilities?
Doctrine/Employment Mode. For doctrinal development, I analyze the following:
190
When was a formal nuclear use doctrine articulated and promulgated? Was
there a lag between the introduction of new nuclear capabilities and doctrine
articulation? How long (months, years)?
Was nuclear use doctrine primarily developed by civilian political leaders
or military leaders? Were civilian non-governmental strategists involved in
formulating the nuclear use doctrine?
What is the process for doctrinal reform? Who initiates the process? Who is
in charge (civilian or military leaders)? How often does it occur? Does
doctrinal reform correlate temporally with internal political developments
or external security developments?
Management. Nuclear command-and-control is at the crux of existing causal
theories of nuclear operations. As such, it is the most critical nuclear posture
dimension for the evaluation of internal political constraints. I ask the following
questions:
Who has physical custody of a state’s nuclear weapons, civilian
organizations (e.g. atomic agency) or the military?
Are warheads and delivery vehicles kept separate (de-mated)? If so, are
warheads managed by a different entity (e.g. atomic agency, secret police,
paramilitary organization) than delivery vehicles?
If the military has physical custody, when were the weapons transferred?
191
How long after the first nuclear test (months, years) until weapons were
transferred to the military?
Was a new civilian institution or military service/organization created to be
in charge of nuclear weapons? How long after the first explosive test was
the new entity formed?
Are nuclear weapons integrated with conventional military forces? If a
state has multiple military service branches (army, navy, air force, etc.),
does each service have its own nuclear weapons?
Are nuclear weapons deployed only in the homeland or extra-territorially
(on land or at sea)?
Are remote locks, permissive action links (PAL), or other physical control
measures utilized? When were they adopted? Are such control measures
uniform across all nuclear platforms or are some nuclear platforms under
less stringent controls?
Are political controls in place (dual key launch procedures, bifurcated
approval, separated chains of command, etc.)?
Is there evidence of pre-delegation of launch authority? How far down the
chain of command did such authority reside? Was pre-delegation granted
or rescinded during crises?
Does a formal plan for the devolution of launch authority exist? When was
it developed?
192
Are launch procedures tightly couples with early warning capabilities?
Based on the coding criteria of Narang’s typology of nuclear postures and the
answers to the supplemental questions, it becomes possible to piece together the
narrative of how changes in domestic political institutional configurations within
an authoritarian regime translate into political constraints which inhibit the
adoption of specific nuclear posture options. Moreover, the focus on sequencing
among nuclear posture dimensions allows for an assessment of whether the
adoption of specific nuclear capabilities drive doctrinal and management decisions
or whether management and doctrinal concerns prevent the adoption of specific
capabilities.
Method and Case Selection
I illustrate my theory with qualitative analytical methods. Specifically, I
conduct an in-depth case study of the original nuclear authoritarian regime: the
Soviet Union under Stalin, 1949-1953. Qualitative research methods are
particularly appropriate for hypothesis generation in this dissertation because my
goal is to map the interaction of each nuclear dictator and the regime’s ruling elites,
both civil and military, within the context of specific configurations of authoritarian
political institutions. Analytical process tracing provides the level of depth required
193
for such a task.
240
And it allows me to evaluate whether variation in the underlying
dimensions of nuclear posture were the result of changes in civil-military relations
that were in turn the result of the degeneration and regeneration of authoritarian
power-sharing institutions.
The formal rationale for selecting this case is as follows. First, the Soviet
Union is an “influential” case, defined as “a case that appears at first glance to
invalidate a theory, or at least cast doubt upon a theory.”
241
In terms of the sheer
size and sophistication of its nuclear arsenal, the Soviets reached a level of nuclear
capability not replicated by any other authoritarian regime (or most democratic
regimes for that matter). As such, the Soviet case would seemingly argue against
the idea of internal political constraints on nuclear posture in authoritarian regimes.
Indeed, if Stalin, Khrushchev, and Brezhnev were all personalist dictators, my
theory would be flatly rejected. But, as I will argue in subsequent chapters, only
Stalin amassed the concentration of power to be considered a personalist dictator;
his successors did not, as they were constrained by the post-Stalin renewal of
power-sharing institutions within the CPSU. Thus the history of Soviet nuclear
evolution is actually a story of how authoritarian power-sharing reform post-Stalin
enabled a revolution in the USSR’s nuclear posture.
240
Alexander L. George and Andrew Bennet, Case Studies and Theory Development in the Social
Sciences Cambridge: BCSIA Studies in International Security, MIT Press, 2005): 205-232.
241
John Gerring, Case Study Research: Principles and Practices (New York: Cambridge
University Press, 2007) 108.
194
Additionally, there are four other reasons for selecting this case. First, it is
important to select cases with variance on the primary independent variable: intra-
ruling elite power-sharing arrangements. Of the five authoritarian regimes
(USSR/Russia, China, South Africa, Pakistan, and North Korea) that have
successfully developed and tested nuclear weapons, only three (USSR/Russia,
China, and North Korea) have experienced historical periods of institutional
degeneration sufficient to potentially trigger the theorized internal political
constraints on nuclear posture. Indeed, each of these three countries crossed the
nuclear weapon threshold under the rule of a personalist dictator (Stalin, Mao, and
Kim Jong-Il, respectively). And as expected for personalist dictators, each of these
leaders died in office (1953, 1976, 2011, respectively), thus ending the personalist
regime and sparking a potential intra-authoritarian regime change (e.g. from boss
to machine). It is this type of regime transition that presents an ideal opportunity to
assess the strength of political constraints on nuclear posture during and
immediately after personal dictatorship.
Second, the Soviet Union/Russia is the authoritarian regime with the earliest
and now longest nuclear experience, meaning sufficient time has passed for
pertinent details to emerge about its early nuclear posture and elite power-sharing
arrangements. Research on nuclear weapons programs and postures has always
been hampered by the intense secrecy surrounding all things nuclear. There is
always considerable ambiguity about the developmental status of nuclear weapons
195
programs since there a powerful incentive for leaders in emerging nuclear regimes
to intentionally misinform their rivals – often by inflating their capabilities. The
research limitations of secrecy and misinformation are severe in regimes that are
still in the early stages of nuclear weapons development, such as North Korea.
242
To some extent, the challenges of secrecy and misinformation dissipate with time
as security environments evolve, prompting the declassification of documents and
the opening of government archives to researchers.
The final reason for selecting the Soviet Union/Russia for the case study is
simply because it was and continues to be important case for U.S. foreign policy.
As noted above, the Soviet Union under Stalin is the iconic example of a personalist
regimes that successfully developed deliverable nuclear weapons.
243
Studying the
impact of personalism on nuclear posture development in this regime presents the
opportunity to establish a “baseline” understanding of political constraints. This is
increasingly important as Russia is currently undergoing simultaneous processes of
nuclear modernization and potential personalization. Russian president Vladimir
Putin has embarked on ambitious and costly nuclear modernization programs,
while also amassing tremendous personal political power. Although the outcome
of the arsenal modernization and personalization processes is far from certain,
242
There is simply insufficient reliable data on North Korean nuclear developments to conduct an
in-depth empirical case study of its nuclear posture at this time. Instead, it will be treated as a
“shadow case” in the concluding chapter of this dissertation.
243
North Korea may achieve this feat sooner than later, but it has yet to demonstrate a deployable
or deliverable nuclear weapon.
196
focusing on the early periods of personalism and nuclear development may yield
important insights about the possible tradeoffs and political constraints such leaders
may face as they attempt to subvert existing power-sharing arrangements.
197
“When Stalin says dance, a wise man dances.” – Khrushchev
244
Chapter Five: The Soviet Union, 1949-1953
In chapter three, I introduced my theoretical explanation of authoritarian
constraints on nuclear posture evolution as a consequence of failed intra-ruling elite
power-sharing institutions in personalist regimes. The crux of the theory suggests
that personalist authoritarian regimes face severe constraints on each dimension of
nuclear posture. The concentration of near absolute political power in a single
leader strongly affects intra-ruling elite interactions, restricting the autonomy and
influence of all other regime institutions, including the military and scientific
establishments. In sum, the more concentrated political power is in a political
regime, the greater the management and organizational challenges faced in
adopting a delegative nuclear posture. Now, to assess the empirical value of that
theory, I turn to an in-depth historical case study of a nuclear authoritarian regime:
the Soviet Union from 1949-1953.
The evolution of the Soviet Union’s nuclear posture represents an important
test for the theory introduced in this dissertation for two main reasons. First, the
Soviet Union is a unique case among nuclear authoritarian regimes. It was the first
244
Khrushchev Remembers, trans.and ed. Strobe Talbott (Boston: Little, Brown & Company,
1970), p. 301.
198
authoritarian regime to build nuclear weapons and it was the only authoritarian
regime known to have eventually stationed nuclear weapons outside its own
territory. It developed the largest and most technologically diverse nuclear arsenal
of any authoritarian regime, including a full strategic nuclear triad of air, sea, and
missile delivery capabilities, tactical (battlefield) nuclear weapons, robust
command-and-control infrastructure, and eventually even space based early-
warning systems. In short, the Soviet Union was the world’s first and only
authoritarian nuclear superpower. Given the eventual size and sophistication of its
nuclear arsenal, the Soviet case demonstrates the upper boundary of nuclear posture
achievable for a civilian authoritarian regime. And it provides an opportunity to
assess the varying strength of internal political constraints on nuclear posture due
to variations in the strength intra-ruling elite power-sharing institutions.
Second, the case is ideal for an applied assessment of the personalist regime
concept. The Soviet case presents an opportunity to evaluate the relationship
between variations in intra-ruling elite power-sharing and civil-military relations.
Although Stalin is generally understood to be a personalist dictator from 1936
onwards, scholars have debated whether Stalin’s successors were personalist
dictators.
245
The degree of personal power wielded by Khrushchev and Brezhnev is
245
Geddes did not classify Stalin as personalist in her early work. Weeks (2014) and Weeks and
Way (2014), using Geddes raw data, subsequently re-classify Stalin as personalist. Svolik argues
that Stalin was prototypical personalist, but his dataset itself cannot differentiate between Stalin,
Khrushchev, or Brezhnev. As I argued in Chapter Four, one distinctive threshold for classifying
personalist dictators is the use of physical violence against both civilian and military elites. Thus
199
a point of considerable controversy in both Soviet political studies and the field of
comparative authoritarianism. If the paramount leaders of the Soviet Union during
the years of maximum arsenal development and deployment were personalist
dictators, then the theory I elaborated in previous chapters is seriously flawed.
246
If,
however, the history of the case shows that nuclear autonomy for the Soviet military
evolved only after intra-ruling elite power-sharing institutions were reestablished
in the post-Stalin era – meaning military professionalism followed political
development – this would be tentative confirmation of the basic validity of the
theory.
Plan of the Chapter
This chapter will trace the connection between power-sharing institutions,
civil-military relations, and nuclear posture in the Soviet Union under Stalin. The
Soviet Union’s nuclear arsenal was managed by six different paramount leaders:
Stalin (1949-1953); Khrushchev (1953-1964); Brezhnev (1964-1982); Andropov
(1982-1984); Chernenko (1984-1985); and Gorbachev (1985-1991).
247
The intra-
ruling elite power-sharing relationships will be mapped under Stalin, with a specific
focus on the functionality of power-sharing institutions within the Communist Party
executions of senior civilian and military elites is a primary indicator of the concentration of
power associated with personalism, as it demonstrates that elites cannot be ensured of their
physical safety and therefore cannot credibly constrain the dictator.
246
As will be argued at length below, neither Khrushchev nor Brezhnev were personalist dictators.
247
Years of rule correspond with nuclear experience for all leaders.
200
of the Soviet Union (CPSU) and the chronology of power concentration and
consolidation. The purpose is to tease out the cyclical patterns of the power
relationship between the paramount leader and the CPSU, and to explore the impact
of the dual processes of CPSU institutional decay and regeneration on civil-military
relations.
Second, the civil-military relations under Stalin’s rule will be assessed and
categorized as delegative, assertive, despotic, or praetorian. I will carefully evaluate
whether periods of despotic civil-military relations are preceded by the atrophy or
collapse of CPSU political institutions (Politburo, Central Committee, Party
Congresses, etc.). As previously discussed in chapter three, despotic civil-military
relations may develop once intra-ruling elite power-sharing institutions have failed,
but assertive and delegative civil-military relations require functional power-
sharing arrangements.
Third, nuclear posture milestones under Stalin will be chronicled. The link
between conventional civil-military relations and nuclear operations has been
theorized to be direct. Delegative conventional civil-military relations should
predict delegative nuclear operations, all things being equal. Likewise, assertive
civil-military relations at the conventional level should correspond to assertive
nuclear operations. Therefore as I develop the narrative I will be looking for periods
in Soviet history when nuclear operations were more delegative than conventional
201
civil-military relations arrangements, especially any instances of pre-delegation of
nuclear assets and authority to use them.
To recap, the argument is that to the extent that civil-military relations is a
strong and binding constraint on nuclear posture selection, it is due to the lack of
functional intra-ruling elite power-sharing institutions within a political regime. In
other words, nuclear constraints are the result of specific authoritarian regime types.
For internal nuclear constraints to be truly operative, the dictator must have freed
himself from peer rivals and have established despotic control over the military. In
such cases, it would be highly unlikely for the dictator to grant the necessary degree
of autonomy to the military to implement an asymmetric escalation posture. But in
all other situations the military is assumed to have structural opportunities and be
institutionally capable of promoting its prerogatives, to include demands for
increased nuclear autonomy. This is the fundamental paradox of personalism: when
the leader has freed himself of political constraints by subverting power-sharing
arrangements he creates a binding internal constraint on nuclear posture.
In elaborating the above argument, I will utilize secondary sources on the
nature of Soviet political institutions during the Cold War, along with the large
volume of scholarly works that emerged after the Soviet Union’s collapse. The
post-1991 scholarship benefited from access to long restricted Soviet archives and
declassified American intelligence assessments, greatly increasing the level of
detailed knowledge of intra-ruling elite interactions during the period of Stalin’s
202
rule. I will augment this with analysis based on recent refinements in the study of
comparative authoritarianism and the substantial literature on Soviet civil-military
relations. This provides the analytical leverage to understand the impact of intra-
authoritarian regime change (from boss to machine) on civil-military relations and
on each dimension of nuclear posture (capabilities, management, and doctrine).
The Cycle of Intra-Ruling Elite Power-Sharing in the USSR
Joseph Stalin’s rise to unconstrained dictator of the Soviet Union is in
essence the story of his successful attempts to consolidate his position as Lenin’s
heir while freeing himself from the Leninist legacy of collective leadership of the
Communist Party of the Soviet Union (CPSU). The CPSU under Lenin featured a
broad array of elite power-sharing institutions, rules, and procedures. To be sure,
Lenin personally dominated the CPSU until 1922, enforcing strict Party discipline,
but “for all the rigid central discipline imposed by Lenin, the institutions of the
Party were not destroyed and continued to function with some degree of vigor and
genuine debate.”
248
Moreover, Lenin’s preference for a strong party and collective
leadership rather than fuhrerism (strong leader, weak party) was what separated
him not only from other non-communist “totalitarian” leaders such as Hitler and
Mussolini, but also from Stalin and Mao.
249
248
Schapiro, Leonard, and John W. Lewis. 1969. The roles of the monolithic party under the
totalitarian leader. The China Quarterly 40 (40): 43.
249
Ibid., p. 41-44.
203
Stalin’s inheritance was a dense panoply of overlapping CPSU (Politburo,
Central Committee, Party Congress) and government institutions (Council of
People’s Commissars, Supreme Soviet, etc.). The Soviet Union is often considered
one of the most stable authoritarian regimes in modern history, boasting a regime
ruling-coalition spell of 69 years.
250
Indeed, the rapid collapse of the USSR in 1991
famously took observers by surprise. The longevity of the Soviet regime is
commonly attributed to the stabilizing benefits of a highly-institutionalized single
political party, the CPSU.
251
Yet despite the regime’s outward stability, intra-ruling elite power-sharing
varied dramatically over the nearly seven decades the CPSU was in power. As
Svolik has argued, variation in elite power-sharing is a structural feature of all
authoritarian regimes.
252
Consequently, the instability of Soviet power-sharing is
not particularly surprising; personalism is a potential problem of afflicting all
authoritarian regimes. However, in addition to structural features common to all
250
Svolik (2012: 21) defines regime ruling coalition spell: “It consists of an uninterrupted
succession in office of politically affiliated authoritarian leaders – typically from the same
government, party, family, or military junta.”
251
Theorists of single-party regime durability in comparative politics note a variety of
mechanisms of single-party rule that potentially enhance authoritarian stability and durability
(patronage, repression, etc.) in comparison to other authoritarian regime types. See, Brownlee,
Jason. Authoritarianism in an Age of Democratization (New York : Cambridge University Press,
2007). Beatrice Magaloni, “Credible Power-Sharing and the Longevity of Authoritarian Rule,”
Comparative Political Studies, vol. 41 (2008), p. 715–41.
252
As Svolik writes (2012:81) writes, “The instability of power-sharing is a consequence of
distinctive, dismal conditions under which authoritarian power-sharing takes place. Authoritarian
elites cannot rely on an independent authority to enforce their agreement about sharing power,
they may use violence to resolve mutual conflicts, and they typically operate under a shroud of
secrecy.”
204
authoritarian regimes, there are also unique aspects of Soviet political history that
contributed to the instability of Soviet power-sharing. As Breslauer writes:
The Soviet political tradition…is dualistic on matters of leadership. On the
one hand, the tenets of democratic centralism, the stress on volunteerism in
Leninist theory, and the tradition of reverting to strong leadership to guide
the “construction of socialism and communism” support personalistic rule.
On the other hand, the Leninist rejection of the Fuehrerprinzip, the
emphasis on collective leadership, the leading role of the party, and the
notion that social forces, not individual leaders, make history, support
oligarchic rule.
253
This dualistic tradition interacted with structural features inherent in all
authoritarian regimes to generate a cyclical pattern of leadership politics in the
Soviet Union: oligarchic arrangements were strongest in the early years of a
paramount leader’s tenure, eventually giving way to an increasing concentration of
personal power, followed by a return to oligarchy as the paramount leader lost
power (often dying in office) and a new leader succeeded him.
254
In other words,
253
George W. Breslauer, Khrushchev and Brezhnev as Leaders: Building Authority in Soviet
Politics (London: George Allen & Unwin, 1982) p. 11.
254
R. Judson Mitchell, Getting to the Top in the USSR: Cyclical Patterns in the Leadership
Succession Process (Stanford: Hoover Institution Press, 1990). See also, Myron Rush, Political
Succession in the USSR. 2d ed. (New York: Columbia University Press, 1968). John H. Hodgson,
205
the General Secretary of the CPSU (or First Secretary from 1952-1966) started his
term relatively weak and balanced by elite political rivals, but as time progressed
the leader was able to consolidate power. By default, then, collective decision-
making, formal or informal, was often strongest in the immediate aftermath of a
succession since the new leader had not yet had sufficient time and opportunity to
consolidate power.
The cyclical instability of Soviet power-sharing was compounded by the
historic absence of the two most important and stabilizing power-sharing
institutions in political regimes: term limits and a formalized power transfer
mechanism. In political regimes that lack term limits and a power transfer
mechanism, each succession can be destabilizing for the ruling elite and the system
as a whole.
255
This is because the cycle of succession (when it will take place) as
well as the outcome (who the potential new leader will be) are both uncertain. The
potential for sustained power struggles among rivals is quite high, as each
succession as marked by confusion and uncertainty. It is for this reason that periods
of power transfer in the Soviet Union were considered “crises” and key tests for the
stability of the Communist regime.
256
Without an institutionalized process for
“The Problem of Succession,” in The Dynamics of Soviet Politics, ed. Paul Cocks, Robert V.
Daniels, Nancy Whittier Heer (Cambridge, Mass: Harvard, 1976), p. 96-115. Seweryn Bialer
Stalin's Successors: Leadership, Stability, and Change in the Soviet Union. 1st paperback ed.
(Cambridge and New York: Cambridge University Press, 1982). Robert Conquest, The Soviet
Succession Problem (Defense Documentation Center no. AD43407, 1964).
255
It can also result in policy stagnation since long tenures of elites stifles policy innovation. See
Mitchell, Getting to the Top, p. 13.
256
Rush, Political Succession in the USSR, p. 72-96.
206
power transfer, succession was a potential free-for-all among rival contenders that
could stretch for several years until decisively resolved.
Moreover, without the critical power-sharing institutions of term limits,
mandatory retirement ages, or a formalized process of leader selection (transfer of
power), the cycle of Soviet succession crises were thus certain to repeat, but at
unknown intervals. The lack of term limits and an institutionalized leader selection
process also meant that the regime was ill-prepared for power transfers when they
unexpectedly did occur. This is due to the interactive effect of lacking both term
limits and routinized leader selection processes: a paramount leader would avoid
designating a successor lest that successor grow powerful enough to “prematurely”
supplant him in office.
257
Without a fixed term of office, not a single Soviet
paramount leader willingly transferred power to a successor.
258
The possibility of a
paramount leader remaining in office indefinitely created a range of perverse
incentives for maintaining power and ensured that the potential for personalism was
a perennial problem throughout Soviet Union’s existence.
In short, the absence of a “transparent, consistently implemented, non-
arbitrary transfer of power mechanism” was the “fundamental defect” at the heart
257
As Uri Ra’anan comments, “No ‘setting sun’ is well advised to risk anointing a successor, i.e. a
‘rising sun,’ since political power will inevitably gravitate to the heir.” See, Flawed Succession:
Russia’s Power Transfer Crises, ed. Uri Ra’anan (Lanham, MD: Lexington Books, 2006), p. xvi.
258
Indeed, Stalin, Brezhnev, Andropov, Chernenko all died in office. Khrushchev was removed in
a party coup. And Gorbachev lost power when the Soviet regime dissolved.
207
of the Soviet regime.
259
Institutionalized mechanisms for power transfer are without
a doubt the most important power-sharing arrangements in any political regime, the
absence of which create unique dilemmas for both the paramount leader and the
ruling elite. Yet power transfer mechanisms, as the name implies, affect only the
transfer of decision-making power, not the amount or scope of decision-making
power concentrated in a political position (i.e., paramount leader). Other
institutional arrangements regulate the amount of power concentrated in a single
political office and whether a single person can hold multiple offices. Thus, the
absence of term limits and power transfer mechanisms may have had the greatest
impact in shaping the cyclical patterns of intra-ruling elite power-sharing in the
Soviet Union, but it is by no means the full story of elite power-sharing.
260
To be sure, the Soviet Union featured many of the specific types of “formal,
deliberative, and decision-making institutions” that Svolik argues facilitate
authoritarian power-sharing in dictatorships: “politburos, councils, [and]
committees that are embedded within authoritarian parties and legislatures.”
261
As
Svolik’s terminology suggests, the Soviet Union was in many ways the innovator
of modern authoritarian power-sharing institutions.
262
Indeed, the strength and
259
Ra’anan, Flawed Succession, xv.
260
Comparatively, as the result of the absence of term limits and power transfer institutions, the
USSR had less robust elite power-sharing than other single party regimes, such as the PRI in
Mexico.
261
Svolik, Politics of Authoritarian Rule, p. 86-87.
262
The USSR is also the starting point for many studies of authoritarian politics. Thus, the study
of Soviet politics has no doubt influenced the field of comparative authoritarianism in myriad
208
durability of the CPSU – and communist regimes more generally – derived from
these innovations.
263
The Soviet Union may never have been able to solve the
transfer of power issue, but the power-sharing institutions embedded in the CPSU
did in fact place constraints on the paramount leader’s accumulation of power at
different times, particularly for Stalin’s successors.
Why and how Stalin was able to slip the shackles of the power-sharing
institutions embedded in the CPSU while his successors were never able to achieve
a similar feat is a central question of this chapter. Collective-rule, of varying
strength, both preceded and succeeded Stalin. Therefore, to understand the unique
process of Stalinization of the CPSU – the declining ability of the Party Congress,
the Central Committee, and the Politburo to constrain the leader– requires an
assessment the CPSU’s institutional composition both before and after Stalin.
Becoming the V ozh d’: The Stalinization of the CPSU
To speak of a Stalinist political “system” is to denote a break with both the
prior Leninist system and the post-Stalin era of Soviet politics on the specific
dimension of the role of the leader within the political system. The roots of
ways. This can create methodological issues when “applying” theories of comparative
authoritarianism to the regime which had major influence on those same theories.
263
As Huntington (1968: 8) wrote, “The real challenge which the communists pose to
modernizing countries is not that they are so good at overthrowing governments (which is easy),
but that they are so good at making governments (which is a far more difficult task). They may not
provide liberty, but they do provide authority; they do create governments that can govern.”
209
Stalinization are found in two trends which were apparent in the earliest period of
Communist rule: centralization and personalization. Both trends affected the
strength, composition, and cohesiveness of political institutions in the new regime.
The centralization trend had two interrelated components. The first was the
shifting of policy decision-making power from the state, primarily the Council of
People’s Commissars (Sovnarkom) to CPSU organs. In the first few years
following the 1917 October revolution the party had yet to attain its fully dominant
institutional position in the regime. As Graeme Gill notes, “Most decisions of
importance during the initial stages of Bolshevik rule were made in Sovnarkom.”
264
The prominence of Sovnarkom in decision-making was a reflection of Lenin’s
influence as chairman, but also due to the relative weakness of the party’s central
executive and administrative organs at the time. The situation began to change with
the restructuring of the party executive organs at the 8
th
Congress in 1919,
particularly with the formalization of the Central Committee Secretariat and the
creation of the Politburo, and further accelerated with Lenin’s withdrawal from
government politics beginning in 1922.
265
Although still retaining considerable
political influence, Sovnarkom – and its successor, renamed the Council of
Ministers in 1946 – would never regain its brief role as the principle policy
264
Graeme Gill, The Origins of the Stalinist Political System (Cambridge and New York:
Cambridge University Press, 2002), p. 55.
265
Ibid., p.55-56. As Myron Rush (1968: 30) writes, “The government’s prestige stemmed chiefly
from Lenin’s position as its head, so that its authority inevitably declined as he withdrew from
affairs.”
210
decision-making body in the regime. The trend toward party dominance was
inaugurated.
The second component of the centralization trend occurred within the CPSU
itself. Power and authority began to flow from the larger, unwieldy Party Congress
to smaller bodies, like the Central Committee (CC) and then the Politburo.
266
Although the Party Congress was formally the supreme institution in the regime,
its increasing size – swelling from 106 delegates at the 7
th
Congress in 1918 to
1,135 delegates at the 10
th
Congress in 1921 – and infrequent (annual) meetings
rendered it ineffective as a policy-making institution.
267
The Central Committee
(CC) was to play a key role in the centralization trend. As Zinoviev argued at the
8
th
Congress, “Fundamental questions of policy, international and domestic, must
be decided by the central committee of our party.”
268
CC members were formally
elected by and accountable to the Party Congress, thus retaining the official
tradition of Congress supremacy, but true power resided with the CC.
However, the membership of the CC itself doubled (from 23 to 46 members)
between the 7
th
and 11
th
Congresses (1918-1922). And although CC meetings were
more regular than Party Congresses, they were still too infrequent to support
266
Graeme Gill, Origins of the Stalinist Political System (Cambridge and New York: Cambridge
University Press, 2002), p. 60.
267
Ibid., 58.
268
Quoted in Jerry F. Hough and Merle Fainsod, How the Soviet Union is Governed (Cambridge:
Harvard University Press, 1979), p. 82-83.
211
effective policy-making.
269
As Gill observes of CC meetings during this timeframe,
“At best, they occurred approximately monthly and rarely lasted longer than one
day, a pattern which was inadequate to enable the CC to maintain close supervision
over the day-to-day running of affairs of the state.”
270
It is for this reason that the
centralization trend continued even further following the creation of the Politburo
in 1919.
271
The Politburo’s organizational advantages for rapid decision-making
were clear: smaller membership and more frequent meetings (averaging twice
weekly after the 10
th
Party Congress).
272
While the CC was given formal
responsibility for selecting Politburo members, the inherent organizational
advantages of the Politburo and the high profile and prestige of its original members
(Lenin, Kamenev, Trotsky, Stalin, and Krestinsky), eventually undermined and
supplanted the CC’s position in the CPSU.
273
By 1922, the Politburo has established
itself as the predominant decision-making body in the Soviet regime.
The second major trend that facilitated the eventual Stalinization of the
CPSU was the lasting influence of Lenin’s personalized authority on the
development of party institutions. As Gill writes:
269
The CC met 81 times between 1918-1922. See Gill, pg. 64.
270
Gill, 63.
271
Hough and Fainsod, p. 83.
272
The Politburo met 253 times between 1919-1922. See Gill, p. 68.
273
Gill, 67.
212
Lenin embodied a notion of authority which was completely at odds with
the notion of regularised authority based upon stable institutional structures.
He legitimised power and authority which was decoupled from institutional
structures, thereby providing a precedent for others to make claims to
authority on that basis and severely hampering the growth of authoritative
institutions…It was, therefore, in the shade of Lenin’s personalised
authority that the early political institutions sought to grow and it was that
shadow which, in significant part, hindered their growth as powerful,
independent institutions.
274
The institutional “fluidity” that marked the early years of the Soviet regime
provided Lenin an opportunity to litigate and appeal policy issues in and across
multiple institutions, utilizing his personal authority to essentially overrule defeats
in some bodies and secure a favorable outcome in another.
275
Such actions by Lenin
weakened the integrity of Soviet political institutions, thereby increasing the
functional need for a strong leader to sustain the efficiency of the system.
276
274
Gill, 108-9.
275
Ibid.
276
Gill (2002:109) notes: “In practice Lenin was the hub of the governmental and administrative
networks. It was his personal role at the top which did much to ensure that the largely
unstructured, and therefore potentially inefficient and conflictual, relationship between the bodies
at the apex of the system did not break down. He provided the lubricant as well as the driving
force for governmental affairs.”
213
Stalinization of the Politburo, 1924-1930
Stalin would masterfully capitalize on both of these structural trends
following Lenin’s death in 1924. Although a full account of Stalin’s consolidation
of power is outside the scope of this chapter, it is important to note that during the
first period of Stalin’s rise the majority of elite conflict continued to take place
within the collective institutional structures of the regime.
277
In other words, the
primary site of factional conflict within the collective leadership – the Politburo –
continued to meet regularly and follow formal procedures, even when dismissing
core members. Similarly, the CC continued to meet regularly as well, at least up
through the 14
th
Congress in 1925.
278
By the middle of the 1920s, however, a
decline in the frequency and regularity of CC meetings was evident.
At about the same time, the composition of CC membership also began to
change. Whereas the bureaucratic constituency of the CC at the 11
th
Congress in
1922 was 47.9 percent state institutions and 34.8 percent full-time party workers,
by the 15
th
Congress in 1927 full-time party worker membership in the CC had
increased to 49.5 percent and the state organ representation had decreased to 28.1
percent.
279
The increased representation for party workers reflected the further
277
A full account of Stalin’s rise and consolidation of power is found in Oleg V. Khlevniuk,
Master of the House: Stalin and His Inner Circle, trans. Nora Seligman Favorov (New Haven and
London: Yale University Press, 2009).
278
The CC met 72 times between 1922-1930. However, the frequency of meetings fell by half
(from 18 to 9) between the 14
th
and 15
th
Congresses. And between the 15
th
and 16
th
Congress, the
CC met only 8 times. See Gill, p. 144.
279
Gill, 148.
214
concentration of power within the party apparatus. Furthermore, the CC
experienced a large increase in the overall number of members in the post-Lenin
era, nearly tripling in size (from 46 to 121 members) between 1922-1927.
280
This
growth in total membership and party representation provided Stalin an opportunity
to increase his support within these organs, thereby securing a factional advantage
over his rivals. Indeed, as the only full member of the Politburo, Orgburo, and
Secretariat, Stalin was ideally placed to monitor personnel issues throughout the
key institutions of the regime.
281
A milestone in Stalin’s rise to personal power
came when the collective leadership institutions were transformed into instruments
of power consolidation. As Gill notes, “Rather than organs of decision-making,
they became arenas within which the opposition was to be pilloried and
defeated.”
282
The opposition remained well-represented in the CC, but with the
newly enlarged membership lending further support to Stalin, the nature of the
intra-party debate shifted; opposition was no longer tolerated.
During this period, Stalin’s defeat of his elite rivals – the Left Opposition
(Trotsky, Kamenev, Zinoviev) and the Right Opposition (Bukharin, Rykov,
Tomsky) – occurred within the context of Soviet collective institutions.
283
It was a
struggle for power, but also a concurrent policy debate over how best to resolve
280
Gill, p. 145.
281
Gill, p. 167.
282
Gill, p. 196.
283
Khlevniuk, Master of the House, p. 1-38.
215
what Preobrazhensky would term the problem of “premature revolution.”
284
In
1921, the Soviet economy was near collapse and Lenin’s New Economic Policy
(NEP) was seen as a necessary strategic retreat from wartime communism. Stalin’s
reputation as a “cynical Machiavellian” derives in part from his apparent flip-flop
on economic policy between 1925 and 1928, namely his views on rapid
industrialization and collectivization versus gradual industrialization and the
continuation of NEP. As Hough and Fainsod write, “He is supposed to have first
allied himself with the right in order to defeat the Left Opposition in the Politburo,
then to have stolen the latter’s program so that he could condemn his erstwhile allies
as a Right Opposition and remove them from the seats of power.”
285
Indeed, after
defeating the Left Opposition, Stalin would eventually enact a more radical forced
collectivization policy to extract capital resources to finance rapid industrialization
than the leaders of the left has ever publicly proposed.
The end result of these maneuvers was, of course, Stalin’s dominance of
Politburo. But what is remarkable – especially in light of what was to come – was
the cautious manner of his approach. Moreover, policy decisions continued to be
issued under the auspices of the collective organs. And importantly, the
oppositionists, although removed from their positions, remained free and alive.
284
Hough and Fainsod, p. 136.
285
Ibid., p. 134.
216
The Era of Terror
By the early 1930s it was clear that Stalin was amassing significant power.
Yet oligarchical tendencies in the regime continued to exert influence. Stalin was
dominant, but his power was not absolute. This is perhaps best reflected in his
inability to secure the death penalty against Riutin in 1932. The Politburo rejected
Stalin’s demand, “thereby affirming the principle that party members (even when
they have been expelled) should not be executed.”
286
Thus, at this stage the
Politburo was still capable of collectively defending the rights of party members
against the wishes of the General Secretary. However, such protection from the
dictator’s whims was not to last much longer.
The Great Terror was the final blow against CPSU power-sharing
institutions under Stalin. It was remarkable in terms of its scope, cruelty, and its
reach into the highest levels of the regime, including the Politburo. Indeed, five
members or candidate members of the Politburo – Rudzutak, Postyshev, Eikhe,
Kosior, and Chubar – were executed, and one – Ordzhonikidze – committed suicide
under pressure.
287
Although Stalin had certainly used mass terror to repress the
general populace – particularly during the campaigns of forced collectivization and
286
Gill, p. 250.
287
Khlevniuk, Master of the House, p. 203-214.
217
the liquidation of the kulaks in late 1920s and early 1930s – the regime’s elite had
been relatively safe from the security organs.
288
The Party’s ability to protect itself from the security apparatus changed in
the aftermath of Kirov’s murder in 1934. It has often been suggested that Stalin was
responsible for the assassination, but notwithstanding his potential involvement in
Kirov’s death, Stalin immediately seized the opportunity to deal a blow to the
opposition by introducing an extraordinary anti-terrorism law, known as the “Law
of 1 December 1934.”
289
Importantly, Stalin sought Party approval for the Law only
after he had already forced it through Presidium of the Central Executive
Committee, demonstrating his increasing ability to work outside formal Party
institutions. The content of the Law of 1 December facilitated the state-sanctioned
violence to come: those accused of terror were denied any procedural protection
during the investigation process, the investigation was to be conducted with utmost
haste, and those found guilty were to be executed immediately.
290
The new law was the literal death knell for effective intra-ruling elite power-
sharing in the Soviet Union under Stalin because it “effectively undermined the
prohibition on the use of the death penalty against party members.”
291
No longer
288
Gill (2002: 191) notes: “It was is in late 1929 that blood was first spilled in intra-party affairs
and a party member was openly arrested for a party offence when Yakov Bliumkin was accused o
treasonable contact with Trotsky and was shot without trial on the authorisation of the OGPU
collegium.”
289
Gill, p. 253.
290
Gill, ibid.
291
Gill, p. 254.
218
were party members insulated from the security forces. As Fainsod commented
about personal dictatorships, “The insecurity of the masses must be supplemented
by the insecurity of the governing elite who surround the dictator.”
292
And the
governing elite surrounding Stalin in the late 1930s were indeed insecure.
Khlevniuk sums it up nicely: “Stalin really could (and often did) suddenly deprive
any Politburo member he chose not only of his job but of his life.”
293
Stalin’s
personal control of the NKVD under Ezhov facilitated the subjugation of the
Politburo and the CC.
294
Following the Great Terror, the Soviet regime had reached the stage of
“mature Stalinism.” As Bialer writes, “Stalin’s domination of the top decision-
making bodies was absolute. As a matter of fact, it is difficult under mature
Stalinism to speak about collective decision-making bodies at all.”
295
Although
policy decisions continued to be issued in the name of the CC, the CC was only
elected twice during the stage of mature Stalinism, at the 18
th
Congress in 1939 and
the 19
th
Congress in 1952. Moreover, the CC only met six times in plenary session
from 1938 until Stalin’s death in 1953.
296
The Stalinized Politburo continued to
meet, often informally, but it was an institution that was thoroughly subjugated to
the whims of the ruler.
292
Merle Fainsod, How Russia is Ruled (Cambridge,. Harvard University Press, 1953), p. 441.
293
Khlevniuk, Master of the House, p. 223.
294
Gill, 297.
295
Bialer, Stalin’s Successors, p. 32.
296
Ibid.
219
Power-Sharing Post-World War II
Stalin’s personal dictatorship would persist throughout the Second World
War. In fact, during the war Stalin further consolidated power through the creation
of the State Defense Committee (Gosudarstvennyi Komitet Oborony, or GKO) in
1941, which was invested with full plenipotentiary power to coordinate the Soviet
regime’s war effort. By the end of the war Stalin held five leadership positions
within the regime: Chair of the State Defense Committee (GKO), Chair of the
Council of Ministers, Minister of Defense, General Secretary of the CPSU Central
Committee, and Supreme Commander in Chief of the Soviet Armed Forces.
297
However, the workload throughout the war was such that Stalin had little choice
but to delegate some degree of responsibility to his colleagues in order to keep the
war effort functioning. Thus Stalin’s deputies had come to “resemble more the
semiautonomous leaders of the early 1930s than the cowed and submissive
Politburo leaders of the first post-purge years.”
298
Stalin’s primary goal in the
immediate post-war years was to reverse this trend of limited autonomy and
reestablish total control over his key deputies.
Rather than ruling through the full Politburo, Stalin ruled primarily through
informal “ruling groups” in the post-war era. In the immediate post-war years, the
297
See fn# 1, pg. 176 in Yoram Gorlizki and Oleg Khlevniuk, Cold Peace: Stalin and the Soviet
Ruling Circle, 1945-1953 (Oxford: Oxford University Press, 2004).
298
Gorlizki and Khlevniuk, Cold Peace, p. 17.
220
ruling group was a “quintet” consisting of Stalin, plus Molotov, Mikoyan,
Malenkov, and Beria. With the eventual inclusion of Khrushchev, Voznesenski,
Kaganovich, and Bulganin, the ruling group would swell to a novenary for a
time.
299
The lack of formal rules or membership regulations provided Stalin the
flexibility to personally appoint or remove members from the ruling group
according to his whims.
300
Indeed, Stalin exploited the ability to demote and
remove his colleagues from the ruling group as a key control measure to regulate
their autonomy and influence. The original quintet were members of both the
Politburo and the GKO, until its dissolution in 1945. To reestablish his personal
control, Stalin launched a series of attacks against each member of the quintet,
essentially reminding them of who was boss.
301
Although the threat of physical
repression still loomed large, only one member of the ruling group – Voznesenski
– was executed in the post-war period.
302
Furthermore, the decline of the Central Committee as a functional power-
sharing institution continued apace. In 1947, Stalin “turned the Central Committee
into a purely passive rubber stamp for Politburo decisions.”
303
The CC did not
convene during the five year period between 1947-1952. Instead, the CC was
reduced to voting by correspondence, which essentially prevented any debate or
299
Ibid., p. 49.
300
Ibid.
301
Ibid. p. 17-44.
302
Ibid. p. 79-95.
303
Gorlizki and Khlevniuk, p. 51.
221
discussion, and simply allowed for a favorable vote in support of a predetermined
Politburo decision. And even correspondence votes were conducted only
infrequently, with two in 1946, two in 1947, and four in 1948.
304
As is clear, in the
post-war era, the Politburo’s formal accountability to the CC was basically a total
farce.
Throughout the post-war period, Stalin also continued to carefully maintain
personal control of the security services. As Gorlizki and Khlevniuk explain,
“Stalin’s direct control of the secret police assumed particular importance in the
leader’s old age. Although quite willing to excuse himself from secondary
commitments, Stalin would not let the direction of the security forces out of his
clutches for a moment.”
305
His paranoia seemed to increase with his age and
infirmity as well.. It is no surprise that in 1950, he reportedly had listening devices
installed in the apartments of Molotov and Mikoyan.
306
And even in his final years
of life, Stalin continued to obsess over secret police reports: “sitting, secluded, for
hours, poring over the transcripts of interrogations with ‘wreckers’ and ‘doctor-
murderers’ would prove, in his final months, to be one of Stalin’s favorite
pastimes.”
307
Old habits, indeed, die hard.
304
Ibid.
305
Gorlizki and Khlevniuk, p. 113.
306
Ibid., p. 7.
307
Gorlizki and Khlevniuk, p. 7-8.
222
Stalin as Personalist (Boss) Dictator
In terms of the personalist index discussed in chapter four, the brief review
above indicates that Stalin was quite clearly a civilian personalist dictator (boss) at
the time the Soviet Union crossed the nuclear threshold. He personally controlled
the appointment and removal from the Soviet regime’s highest decision-making
bodies. He transformed potential power-sharing institutions into mere rubber
stamps. He personally controlled the security forces and secret police, directly
supervising the purges of his elite rivals and colleagues. During the height of the
Great Terror, he murdered many of his senior elite rivals, and caused the deaths of
many thousands of dissident and loyal CPSU members. He personally controlled
the fates of senior military officers, decimating the high command during the Great
Terror. He occupied multiple leadership positions simultaneously. In sum, there can
be no doubt that Stalin was a personalist dictator.
Civil-Military Relations under Stalin, 1945-1953
The second core question in this chapter is the relationship between the
breakdown of intra-ruling elite power-sharing institutions and civil-military
arrangements. As we have seen, by 1936, Stalin – through a series of intrigues,
purges, and institutional reforms – had successfully subverted the collective
decision-making institutions of the CPSU and concentrated an immense amount of
political power within his sole person. According to the framework developed in
223
chapter four, the structural conditions were thus present for Stalin to employ
despotic control measures. As such, I will evaluate Stalin’s civil-military
arrangements on the four categories specified in the previous chapter: promotions
and assignments, command authority, monitoring mechanisms, and punishment.
Prior to evaluating the despotic nature of Stalin’s civil-military relations, I will
provide a little historical context with a brief sketch of Soviet civil-military
relations.
From a broad historical view, civil-military relations under Stalin were
marked by the dual legacies of the vicious pre-war purges during the Great Terror
and the subsequent challenge and victory over the Nazis in the Second World
War.
308
The impact of the Great Terror on the CPSU has been discussed already.
The military purge component of the Great Terror was similarly devastating. To be
sure, the Soviet military had faced purges prior to the Great Terror, losing 5 percent
of military communists in the 1929 purge and 4.3 percent in the 1933 purge. In
contrast, the figures for the Party as a whole were 11.7 percent and 17 percent,
308
Roman Kolkowicz, The Soviet Military and the Communist Party (Princeton: Princeton
University Press, 1967). Timothy J. Colton, “Perspectives on Civil-Military Relations in the
Soviet Union,” in Soldiers and the Soviet State: Civil-Military Relations from Brezhnev to
Gorbachev, eds. Timothy J. Colton and Thane Gustafson (Princeton: Princeton University Press,
1990); Timothy J. Colton, Commissars, Commanders, and Civilian Authority: The Structure of
Soviet Military Politics (Cambridge and London: Harvard university Press, 1979); William E.
Odom, The Collapse of the Soviet Military (New Haven and London: Yale University Press,
1998); Amos Perlmutter and William M. LeoGrande, “The Party in Uniform: Toward a theory of
Civil-Military Relations in Communist Political Systems,” The American Political Science
Review, vol. 76, no.4 (December 1982), p. 778-789.
224
respectively.
309
But these relatively mild military purges would give little notice of
the horror to come, especially since punishment during these purges did not include
execution.
310
The activities involved in the Great Purge of 1937 cut across all four of the
analytical assessment categories (promotions and assignments, command authority,
monitoring mechanisms, and punishment). As such, it is best understood it is a
watershed event in Soviet civil-military relations under Stalin. Moreover, the Great
Purge undoubtedly represents the nadir of civil-military relations in the history of
the Soviet Union. In this sense, it is representative of Stalin’s despotic control of
the military, but also a single distinct phase in the cycle of what Kolkowicz
describes as Stalin’s “balancing act”:
Stalin developed the technique…in which every concession to the military’s
professional and institutional interests (broader commanding authority,
professional rank and status, economic and social privileges, etc.) was
accompanied by a tightening of political and police controls. Only during
periods when Stalin had acute need of the military’s expertise and good will
did he forgo the ‘balancing’ of concessions by stricter controls. However,
309
Robert Conquest, The Great Terror: A Reassessment (New York and Oxford: Oxford
University Press, 1990), p. 185.
310
Brian D. Taylor, Politics and the Russian Army: Civil-Military Relations, 1689-2000
(Cambridge: Cambridge University Press, 2003), p. 155.
225
these ephemeral moments of relaxation were usually followed by periods of
renewed restraint, if not outright coercion, of the military.
311
Indeed, prior to the great Purge the military had been steadily gaining in prestige
and professionalism throughout the 1930s. In March 1934, unitary command
authority (edinoachaliye) had been restored, meaning the principle of shared
command with political commissars was (temporarily) abolished.
312
In September
1935, new personnel ranks were introduced for officers, including Marshal of the
Soviet Union. And the personal security of commanding officers was strengthened
by the 1935 Statue which stipulated that they could not be arrested by civil organs
without prior authorization from the Commissar of Defense.
313
These protections
would prove to be short-lived.
It is impossible to overstate the damage the Great Terror caused in the Red
Army beginning in 1937.
314
The losses at the highest levels of command were
simply astonishing. Of the five Marshals of the Soviet Union, three were executed
(Tukhachevskiy, Blyukher, and Yegorov). Also killed were three of four
commanders first rank (Yakir, Uborevich, and I.P. Belov), and all nine army
311
Roman Kolkowicz, The Soviet Military and the Communist Party, p.49.
312
Commissars would be re-instated in 1937 and again in 1941.
313
Roman Kolkowicz, The Soviet Military and the Communist Party, p. 54-55.
314
Peter Whitewood, The Red Army and the Great Terror: Stalin’s Purge of the Soviet Military
(Lawrence, KS: University Press of Kansas, 2015). See also, Roger R. Reese, “The Red Army and
the Great Purges,” in Stalinist Terror: New Perspectives, eds. J. Arch Getty and Roberta T.
Manning (Cambridge: Cambridge University Press, 1993), p.198-214.
226
commanders second rank.
315
Furthermore, all military district commanders were
removed from their posts. Out of the 85 members of the country’s top military body
– the Military Soviet – 76 were targeted for repression.
316
Stalin’s purpose in
destroying the Red Army’s “best brains and leading personalities” was
unmistakable. As Erickson writes, “In liquidating the most independent section of
the high command, Stalin rid himself of the last potential source of leadership
which could rival his own.”
317
In other words, by early 1937 Stalin had already
consolidated his power over his political rivals in the Party. All that was left was
the Red Army. It was the final step in transforming the Soviet regime into a personal
dictatorship.
The advent of World War II created a new phase in the Stalinist civil-
military cycle. A full assessment of Soviet civil-military relations during the
Second World War is outside the scope of this chapter. However, it should be noted
that at various times during the war Stalin reduced political controls on the military,
and by necessity when facing defeat, granted somewhat greater autonomy to his
commanders. But, as Rice notes, “there was no doubt that everything was the
preserve of the General Secretary.” Stalin was not shy in asserting his authority and
exercising his “personal power through the demoralized General Staff. Calling its
315
Taylor, Politics and the Russian Army, p. 163-164.
316
Taylor, Politics and the Russian Army, p. 154.
317
John Erickson, The Soviet High Command: A Military Political History, 1918-1941, p. 465.
227
members in to report to him, berating them and personally altering their plans.”
318
Stalin assumed total and direct responsibility for strategic leadership during the
war. He held all relevant pinnacle command positions, including Chairman of
Stavka (Supreme Military Headquarters).
319
After the war, Stalin resumed his normal pattern of tightening political
control once a crisis had passed. Garthoff has described Soviet civil-military
relations in the post-war era as the “years of Stalinist suppression.”
320
Indeed, Stalin
would arrogate to himself the full glory of the Soviet victory over Germany. The
role of the Red Army in defeating Germany was obscured, and prominent
commanders were demoted, some even arrested. As Kolkowicz writes, “The
remaining years of Stalin’s rule were years of tribulation for the victorious officers
318
Condoleezza Rice, “The Party, The Military, and Decision Authority in the Soviet Union,”
World Politics vol. 40, no. 1 (Oct. 1987), p. 59-60.
319
Timothy J. Colton, Commissars, Commanders, and Civilian Authority: The Structure of Soviet
Military Politics (Cambridge and London: Harvard university Press, 1979), p. 158.
320
Garthoff offers a particularly perceptive assessment: “The history of civil-military relations in
the postwar Stalin period offers little of significance. Perhaps we should say that there was, on the
whole, little such history. For the main point is that under the personal dictatorship of Stalin the
military had been reduced to administrators of the military bureaucracy. They were subjected to
the autocrat’s own method of periodic juggling of senior personnel to avoid the creation of the
kind of cliques or subdomains which perforce he had had to allow during the war. Although some
of the military leaders were personally suppressed in the early postwar years, the period from
1946-1953 especially deserves to be termed the period of Stalinist suppression because the
military were not only cut off from politics but were even denied initiative within their own realm
of competence.” See, Raymond L Garthoff, “The Marshals and the Party: Soviet Civil-Military
Relations in the Postwar Period,” in Total War and Cold War: Problems in Civilian Control of the
Military, ed. Harry Coles (Columbus, OH: Ohio State University Press, 1962), p. 248-249.
228
and men of the Soviet military.”
321
Victory in war had not brought them respite
from the Stalin’s suspicions.
The nature of Stalin’s despotic control is as follows:
Promotions and Assignments: Stalin retained full control over promotions
and assignments. There were essentially no limits on Stalin’s appointment power.
In fact, as Khrushchev relates in a funny anecdote, Stalin even promoted his cook
to the rank of major general.
322
Furthermore, Stalin maintained the power to demote
even the most respected and accomplished military leaders, as he demonstrated in
1949 when he removed Zhukov as Commander-in-Chief of the Ground Forces and
essentially exiled to the Odessa Military District.
323
Command Authority: Stalin placed significant restrictions on command
authority. Dual command under the commissar model was introduced and re-
introduced at various intervals according to Stalin’s threat assessment. Commissars
were re-introduced on May, 8 1937, then revoked on August 12, 1941, reinstated
again on July 16, 1941, and then finally replaced by one-man (unified) command
on October 9, 1941.
324
Even after edinoachaliye was restored, political officers
from the Main Political Administration (MPA) continued to be embedded within
321
Kolkowicz, The Soviet Military and the Communist Party, p. 71.
322
As Khrushchev recalls, “There was a Caucasian chef who made shashlik for Stalin. He had
been made a major general. Every time I came to Moscow, I saw this chef had more and more
ribbons and medals, apparently in recognition of his skill at cooking shashlik.” Khrushchev
Remembers, trans. Strobe Talbott, p. 311.
323
Kolkowicz, The Soviet Military and the Communist Party, p. 72.
324
Colton, Commissars, Commanders, and Civilian Authority, p. 14.
229
the units at most levels of the command hierarchy. Furthermore, command
authority was extremely centralized in Moscow. Senior battle commanders had to
follow operational orders to the very letter, even when such orders were useless or
downright stupid.
325
Monitoring Mechanisms: the most intrusive monitoring mechanisms were
utilized under Stalin. The NKVD routinely spied on military officers, even
including political officers in the Main Political Administration (MPA). Officers
were interrogated and tortured. As Erickson observes, the NKVD was the Red
Army’s “inveterate and real enemy,” and even during wartime “the organs of
surveillance and repression operated according to their original purpose.”
326
Punishment: Under Stalin, military officers were subject to the most severe
and arbitrary punishments. The murder of the High Command in the Great Purge
of 1937-38 is obviously the most notable example, but executions remained
common even throughout the war. The Western Front commander, General of the
Army Pavlov, his staff, and the North-Western Front staff were executed in 1941.
327
Indeed, when General Petrov learned he had been promoted to a front command,
he quipped, “So now they are going to shoot me, too.”
328
Civil-military relations
under Stalin was a life or death struggle for Soviet officers.
325
Erickson, The Soviet High Command, p. 666-667.
326
Ibid., p. 668.
327
Erickson, p. 626.
328
Ibid.
230
In sum, civil-military relations under Stalin satisfy all the classification
requirements to render a despotic coding. Or as Colton has remarked, “individual
officers lived in fear of Stalin’s wrath and the midnight knock on the door by the
secret police.”
329
Although the excesses associated with the massacres of 1937
would never repeat on that scale again, civil-military relations remained despotic
until the end of Stalin’s life.
The First Nuclear Dictator: Nuclear Posture under Stalin, 1949-1953
The previous sections of this chapter have established that Stalin was a
personalist dictator par excellence who adopted despotic civil-military
arrangements in the late 1930s which persisted through the WWII and until his
death. Per the theory of internal political constraints on nuclear posture that I
developed in chapter three, as a personalist dictator, Stalin should have confronted
severe constraints on each dimension of nuclear posture: capabilities, doctrine, and
management. Evaluating Soviet nuclear posture under Stalin is thus the final task
in this chapter.
Capabilities: How did Stalin’s personalist rule affect the development of Soviet
nuclear capabilities?
329
Timothy J. Colton, “Perspectives on Civil-Military Relations in the Soviet Union,” in Soldiers
and the Soviet State: Civil-Military Relations from Brezhnev to Gorbachev, eds. Timothy J. Colton
and Thane Gustafson (Princeton: Princeton University Press, 1990), p. 20.
231
When the Soviet Union successfully tested its first nuclear device (RDS-1)
on August 29, 1949, Stalin was approaching the twilight of his three decades of
personal rule. The Soviet Union has passed through the crucible of the Second
World War, emerging victorious, but Stalin was a changed man. The heavy
workload during the war had taken a heavy toll on his health and he was suffering
from a cardiac condition and a variety of other ailments.
330
To preserve his health,
his summer vacations to the Black Sea stretched longer than ever before. In 1950
and 1951 he spent nearly five months at his dacha in Abkhazia.
331
The Soviet Union’s entry into the “nuclear club” and its first years as a
nuclear power coincided with Stalin’s personal decline. In this sense, Stalin’s actual
experience leading a nuclear weapon state is quite modest: merely four years.
Furthermore, Soviet nuclear capabilities during Stalin’s rule were similarly modest.
As Khrushchev noted, “Under Stalin we had no means of delivery. We had no long-
range bombers capable of reaching the united States, nor did we have long-range
rockets. All we had was short-range rockets.” And in terms of actual bombs, he
adds: “We had only just developed the mechanism and had a negligible number of
finished bombs.”
332
Reporting indicates that a pilot series of five RDS-1 weapons
330
Robert Service, Stalin: A Biography (Cambridge, Mass: Harvard University Press, 2004), p.
491, 571-580. For a list of Stalin’s various ailments, see Montefiore p. 614.
331
Service, Stalin, p. 572.
332
Quoted in David Holloway, Stalin and the Bomb: The Soviet Union and Atomic Energy 1939-
1956, p 270.
232
was completed in 1950.
333
And CIA reporting estimated that the Soviet Union
would possess between 25-50 nuclear bombs available in 1953, but this estimate is
likely high given the fact that the first serial production bombs were delivered in
1953.
334
Moreover, only three nuclear tests were conducted in Stalin’s lifetime:
RDS1 on August 29, 1949, RDS-2 on September, 24, 1951, and RDS3 on October
18, 1951.
335
Therefore, to speak of nuclear posture under Stalin is a bit of a misnomer;
it is premature. Under Stalin the Soviet nuclear arsenal remained in its infancy.
Nevertheless, insights can be drawn from Stalin’s personalized decision-making
approach to build the bomb in the first place. One question of particular importance
is how did the personalist nature of Stalin’s regime affect the pace of Soviet
technological progress?
There is little doubt that the intense suspicion and secrecy that pervaded the
Stalinist regime was deleterious to the bomb development effort. Stalin distrusted
Soviet scientists and engineers to such a degree that he discounted their innovative
proposals unless the technology had been verified by Western experience.
336
As
Holloway writes, “Stalin regarded scientists and engineers with suspicion, fearing
that they might be wreckers or saboteurs, and claimed for himself the right to say
333
Podvig ed., Russian Strategic Nuclear Forces, p. 2.
334
Holloway, Stalin and the Bomb, p. 322.
335
Podvig, Russian Strategic Nuclear Forces, p. 485.
336
Holloway, Stalin and the Bomb, p. 365.
233
what constituted valid science.”
337
It is for this reason that Stalin and Beria
originally discounted intelligence reports of the American bomb program.
338
A similar dynamic applied to other advanced technologies, such as the
Soviet programs for radar, rockets, and jet propulsion. In each case, political
authorities had quashed promising research and development programs because
they did not trust science and scientists. These projects were resumed only after
Western countries had demonstrated the viability of these technologies.
339
Of
course, this ensured that Soviet technology programs were always playing “catch
up” to their Western counterparts.
Beyond the bomb itself – meaning the physics package – Stalin’s purges
also had a direct negative effect on the development of delivery vehicles and
supporting technology. As Holloway writes, “Missile development suffered greatly
when leading figures – including Tukhachevski – were arrested and shot in the
Great Purge.”
340
Other notable weapons designers – such as Korolev and Glushko
– escaped death, but were arrested and imprisoned for a time. Needless to say,
technological progress is likely to be slowed down when key scientists and
engineers are in prison camps. Of course, despite all of the political meddling and
interference, the Soviet Union under Stalin was able to build and successfully test
337
Ibid., p. 366.
338
Ibid., p. 115.
339
Holloway, Stalin and the Bomb, p. 145
340
Holloway, Stalin and the Bomb, p. 146.
234
atomic weapons in just four years after the initiation of the all-out effort. Thus it is
difficult to estimate the true impact of the Stalinist system on Soviet nuclear
development. To be sure, without the clear evidence of the possibility of nuclear
weapons following Hiroshima and Nagasaki, there is some doubt whether the
suspicious Stalin would have trusted Soviet scientists sufficiently to devote the
tremendous economic and manpower resources required to build the bomb.
Indeed, on the one hand, Soviet success in building the bomb should not be
taken as a given. As Hymans has argued, not all of the states which have seriously
attempted to develop nuclear weapons have been successful in that effort, with
roughly 40% failing.
341
Importantly, the main reason for proliferation failure,
according to Hymans, is “political meddling” from a top leadership that faces few
constraints on its freedom of action. Hymans writes:
When state institutions are neo-patrimonial, they cannot prevent short-
sighted meddling in the nuclear program by the top leadership – and
therefore meddling happens, and scientific and technical workers lose their
autonomy. As a result, nuclear weapons projects become long, hard slogs.
By contrast, when state institutions are Weberian legal-rational, they can
341
There are 17 historical cases of nuclear weapons “pursuit,” seven of which failed. See, Jacques
E. C. Hymans, Achieving Nuclear Ambitions: Scientists, Politicians, and Proliferation (New
York: Cambridge University Press, 2012), p. 3-5. See also, Malfrid Braut-Hegghammer, Unclear
Physics: Why Iraq and Libya failed to Build Nuclear Weapons (Ithaca and London: Cornell
University Press, 2016).
235
block top leaders from engaging in such short-sighted actions that
undermine their scientific and technical workers’ autonomy.
342
In other words, the lack of constraints on top leaders’ ability to interfere in a nuclear
program dooms that program to inefficiency and potential failure.
343
342
Ibid., p. 26.
343
Hymans uses the neopatrimonialism concept in a curious way. First, Hymans (2012:77)
goes to great lengths to “draw a clear distinction between the concept of state institutionalization
and of regime type.” As he writes, “Both neo-patrimonialism and Weberian legal-rationalism are
compatible with a wide variety of regime types, both democratic and non-democratic.” But this is a
curious assertion given that the concept of a neo-patrimonialism has a long history within
comparative politics as a subtype of authoritarian regime. It is certainly true that some authoritarian
regimes are far better institutionalized than others, with some exhibiting legal-rational tendencies.
The reverse contention that neo-patrimonialism is compatible with democracy, however, is not
accurate unless one resorts to the various definitions of “diminished” or “flawed” democracy. And
even in hybrid regimes, once neo-patrimonialism reaches a certain level of intensity it assuredly
harms rule of law and is therefore antithetical to core dimensions of democracy. There is, however,
no inherent conceptual conflict between neo-patrimonialism and autocracy. Neo-patrimonialism can
flourish unfettered in dictatorships.
In one sense, the underlying purpose of Hymans’ truncated discussion of regime type is to
simply warn against conflating the concept of state institutionalization with democracy, reminding
us that although there is undoubtedly an empirical correlation between democratic regime type and
nuclear weapons project efficiency, correlation is not causation. Therefore, he argues, democratic
regimes should not have any inherent advantages in pursuing the bomb. In other words, all things
equal, a well-institutionalized authoritarian regime should have the same baseline nuclear weapons
program efficiency as a well-institutionalized democracy. But as he notes, “It is true that as an
empirical matter longstanding liberal democracies have generally been grounded solidly, if often
unhappily, in strong Weberian legal-rational state foundations; whereas other regime types are
empirically less likely to have such strong institutional foundations.” It is left unexplored why this
may be so. But if democracies have an inherent advantage in developing or sustaining legal-rational
institutions, then by Hymans’ own logic democracies have an advantage in nuclear weapons
program efficiency.
Second, the indicators for neo-patrimonialism he cites are all better understood as
consequences of the concentration of power. If “political meddling” from the top leadership is the
root cause of decreasing efficiency and proliferation failure, then political regimes with institutions
circumscribing the leader’s discretion are at a distinct advantage. Likewise, personalist regimes are
at a tremendous disadvantage. In other words, it is not the use of personal networks, corruption,
nepotism, or personality cult that causes inefficiency, it is the ability of the paramount leader to
arbitrarily reorganize regime institutions, redraw bureaucratic boundaries at a whim, and personally
intervene. This leads to the conclusion that it is not the level of state institutionalism, per se, that
increases program efficiency and facilitates proliferation success. It is the presence of a very
particular set of political power-sharing institutions that constrain the ability of the paramount leader
to concentrate power in his or her person.
236
On the other hand, the Stalinist system had distinct “advantages” in terms
of marshalling resources – both human and material – for a specified objective. If
a personalist dictator is willing to “pay any cost necessary” to achieve their nuclear
ambitions, the entire regime can be mobilized in pursuit of the goal. To some extent,
this dynamic was visible in Stalin’s rapid industrialization program in the late 1920s
and 1930s. Despite its massive toll in human lives, the industrialization program
was “spectacularly successful in terms of its military consequences.”
344
Rapid
industrialization enabled the Soviet Union to militarily withstand the German
onslaught in World War II. The same principle applied to the Soviet nuclear
program.
This raises the issue of the comparative advantage of an extensive versus
and intensive nuclear program. Stalin may have been successful in building the
bomb because the Soviet nuclear program was an extensive, whole-of-regime
effort. The program may have been inefficient in terms of utilization of allocated
resources, but with Stalin willing to devote all available regime resources to project,
efficiency was not necessary for success. Indeed, Stalin told Kurchatov, “it is not
worth engaging in small-scale work, but necessary to conduct the work broadly,
with Russian scope, that in that connection the broadest all-round help would be
provided.” And directly to the point, Kurchatov recalls, “Comrade Stalin said that
344
Hough and Fainsod, p. 155.
237
it was not necessary to seek cheaper paths…the project should proceed without
regard to cost.”
345
Stalin wanted his bomb fast and at any cost.
Nevertheless, the enduring trends of the Stalin era persisted throughout the
Cold War. The Soviet Union nearly always lagged behind the U.S. in terms of
weapons innovation. As Evangelista notes, “the highly centralized Soviet system
inhibits incentive and innovation.”
346
As the Cold War progressed, the “action-
reaction” dynamic became more pronounced, with the Soviet Union responding to
the “external stimuli” of U.S. weapons development.
347
And it became more
evident as time progressed that the ever-widening technological gap between the
two countries was directly the result of Soviet political structures.
Management: How did Stalin’s personalism affect nuclear command and control?
As discussed above, Soviet nuclear capabilities were extremely modest
during Stalin’s lifetime. There is no evidence of the small nuclear arsenal of air-
dropped (gravity) bombs ever being put on alert status under Stalin. Therefore,
there is little information on nuclear command-and-control practices under Stalin.
However, it is known that the few nuclear weapons that did exist under Stalin were
kept in civilian custody by the Ministry of Medium Machine Building until late
345
Quoted in Holloway, Stalin and the Bomb, p.148. Emphasis mine.
346
Mathew Evangelista, Innovation and the Arms Race: How the United States and the Soviet
Union Develop New Military Technologies (Ithaca and London: Cornell University Press, 1988),
p. 29.
347
Ibid., p. 73-82.
238
1954. Consistent with Stalin’s use of the internal security forces to counterbalance
the military, KGB troops were tasked with guarding the weapons. Nuclear weapons
were not transferred into the Ministry of Defense custody until the late 1950s.
348
In fact, political control over Soviet nuclear weapons was maintained by the
physical separation of warheads and delivery vehicles until the mid-1960s. This
was further strengthened by the practice of giving the KGB custody over the
warheads/bombs and the military custody over the delivery vehicles.
349
Importantly, this practice was maintained in the face of an ever increasing US first-
strike ability. Furthermore, a similar custody arrangement applied to Soviet tactical
nuclear weapons as well, once those weapons were developed in the late 1950s.
350
Given that Stalin prized centralized control in all areas, it is quite likely that had he
lived longer into the nuclear era an even more centralized management scheme
would have been employed.
It is perhaps worth noting, as a point of comparison, how the early U.S.
nuclear arsenal was managed. The U.S. nuclear weapon program was a military-
run program, under the auspices of the Manhattan Engineering District (MED). The
MED was led by a military officer, General Leslie Groves. As such, Feaver writes
that “essentially the military had custody of nuclear weapons. The initial pattern of
348
Holloway, Stalin and the Bomb, p.328.
349
Stephen M. Meyer, “Soviet Nuclear Operations,” in Managing Nuclear Operations, eds.
Ashton B. Carter, John D. Steinbruner, and Charles A. Zraket (Washington, D C.: The Brookings
Institution, 1987), p. 487.
350
Ibid., p. 489.
239
civilian control, then, was delegative.”
351
However, with the passage of the Atomic
Energy Act in 1946, civilian custody of nuclear weapons was established. All
nuclear weapons were transferred to the civilian Atomic Energy Commission, thus
inaugurating assertive civilian control of the U.S. nuclear arsenal.
The U.S. military fought against assertive civilian control from the onset of
the policy.
352
In April 1951, the military won a small victory in the bureaucratic
battle over nuclear custody when Truman authorized the transfer of nine complete
weapons to the military to be station in Guam.
353
The reason for the policy change
was twofold: the Soviet Union tested its first nuclear device in August 1949 and the
Korean war commenced in June 1950. Thus, U.S. arsenal management procedures
were responding directly to perceived changes in the security environment. By
1956, military custody of U.S. nuclear weapons of all yields, including
thermonuclear weapons, was the new norm. Nuclear weapons were widely
dispersed and deployed around the globe, with use-authority pre-delegated to
specific military commanders.
354
The primary difference between U.S. and Soviet nuclear management
procedures is that U.S. nuclear policy rapidly evolved in anticipation of the Soviet
351
Peter D. Feaver, Guarding the Guardians: Civilian Control of Nuclear Weapons in the United
States (Ithaca and London: Cornell University Press, 1992), p. 89.
352
Ibid., p. 110-127.
353
Ibid. p. 138.
354
Ibid., p. 152, 168-171.
240
nuclear threat, often grossly overestimating Soviet capabilities.
355
In contrast,
Soviet nuclear weapons management policy evolved much more slowly, retarded
by a legacy of pathological civil-military relations. Simply put, the U.S. had the
luxury of adopting a highly delegative nuclear management arrangement whereas
the Soviet Union, from Stalin onward, was forced to accept extended periods of
severe vulnerability.
Doctrine: How did Stalin’s personalism affect Soviet nuclear strategic thought?
The final dimension of nuclear posture – doctrine or employment mode – is
the domain in which it is easiest to strong conclusions regarding Stalin’s impact on
Soviet nuclear development. This is because, as I argued in chapter three, the
concentration of power in a single leader is likely to introduce bias into the
development of military doctrine. A personalist dictator can exert tremendous
influence over doctrinal discussions by restricting pertinent information about
nuclear weapons from regime elites, curtailing professional military debate, and
enforcing the acceptance of their preferred doctrine. The result is likely to be a
convoluted synthesis emphasizing the importance of possessing nuclear weapons
as a symbol of regime strength and a continued diminishment of the strategic
challenges of the nuclear revolution.
355
The most famous overestimate is the perceived “missile gap” in the late 1950s. See, Desmond
Ball, Politics and Force Levels: The Strategic Missile Program of the Kennedy Administration
(Berkeley and Los Angeles: University of California Press, 1980).
241
Indeed, this is exactly what occurred under Stalin. The total absence of any
discussion of nuclear doctrine is the central theme of Stalin’s era. Stalin forbid
Soviet military academies from discussing nuclear weapons. Instead, he forced
Soviet officers to focus on the lessons of WWII and his “permanent operating
factors.”
356
As Scott and Scott note, “In spite of the development of nuclear
weapons, no discussion on doctrine was permitted during Stalin’s lifetime, and,
indeed, he disliked the term.”
357
This amounted to an intellectual straightjacket for
Soviet military thinkers, retarding Soviet doctrinal development by many years.
What is truly astounding is the immediate shift in military thinking upon
Stalin’s death. As Meyer notes, “Contentious issues in Soviet military doctrine and
military science, particularly issues related to nuclear weapons, began to receive
‘open’ and critical treatment.”
358
This was a direct result of the changing nature of
the regime and a fundamental shift in civil-military relations. Military officers
began to develop and expound doctrinal concepts, and, importantly, engage with
the political leadership on questions of nuclear use. The nuclear revolution had
finally arrived in the Soviet Union.
359
356
Matthew Evangelista, Innovation and the Arms Race (Ithaca and London: Cornell University
Press, 1988), p. 157.
357
Harriet Fast Scott and William F. Scott, The Armed Forces of the Soviet Union, 2
nd
edition
(Boulder, CO: Westview Press, 1982), p. 38.
358
Stephen M. Meyer, “Soviet Nuclear Theater Forces,” p. 12.
359
Garthoff charts this doctrinal revolution. See, Garhoff, Deterrence and the Revolution in Soviet
Military Doctrine (Washington D.C: The Brookings Institution, 1990).
242
Conclusion
How, then, did intra-ruling elite power-sharing and civil-military relations
affect the evolution of Soviet nuclear posture? In this chapter I have argued that the
history of the Soviet Union’s nuclear posture evolution follows the outlines of the
theory I proposed in chapter three. When the Soviet Union first crossed the nuclear
threshold in 1949, the political power-sharing institutions of the Communist Party
of the Soviet Union (CPSU) were in a state of decay. Over the course of two
decades through complicated intrigues, brutal purges, and reorganizations of the
CPSU administrative organs, Stalin systematically reduced the already limited
power-sharing institutions of the CPSU into instruments of his personal rule.
Importantly, with Stalin’s ascendance and the subversion of CPSU power-sharing
institutions in the late 1930s the dictator’s relationship with the Red Army
deteriorated rapidly and civil-military relations turned highly despotic. The Second
World War offered a brief interlude from Stalin’s pernicious purges, but after the
war Stalin reverted to despotic measures to control the Soviet military. The effect
of Stalin’s despotic relationship with the Soviet military on nuclear operations was
severe. Open doctrinal debate within the military was nonexistent, constrained by
secrecy and the need to contextualize all military thought with Stalin’s “permanent
operating factors.” In short, during Stalin’s tenure, the nuclear revolution could not
be debated by senior officers let alone embraced.
243
Stalin’s death in 1953 heralded a new era for intra-elite power-sharing and
civil-military relations. In the post-Stalin era power-sharing institutions embedded
within the CPSU regained regular function, leading to opportunities for increased
military autonomy. The CPSU returned to “collective rule” under Khrushchev. And
though Khrushchev consolidated his position as paramount leader (First Secretary
of CPSU) in 1957 – following his defeat of the Anti-Party Group led by Malenkov,
Molotov, and Kaganovich – he never attained the concentration of power that Stalin
did. There was no return to terror. Power-sharing institutions met regularly and,
ironically, functioned as intended when in 1964 Khrushchev was removed by elite
rivals within the Politburo. The coup marked a high-point in power-sharing,
demonstrating that Soviet power-sharing institutions in the CPSU had properly
served their information and monitoring function.
As one scholar has described it, the post-Stalin era was “shaped by a shifting
mix of collective and personal leadership. Khrushchev came to be less constrained
by the collective than Brezhnev, but within his administration, Brezhnev was more
constrained at some points than at others. In the early years (late 1960s), collective
restraints appear to have been stronger than in the later years (1970s).”
360
Importantly, neither Khrushchev nor Brezhnev reached the extreme concentrations
of power that marked Stalin’s rule. Although both wielded power at the pinnacle of
360
George W. Breslauer, Khrushchev and Brezhnev as Leaders: Building Authority in Soviet
Politics (London: George Allen Unwin, 1982), p. 17.
244
the Soviet political system, neither Khrushchev nor Brezhnev used terror as a tool
of power consolidation against their senior Politburo colleagues or military leaders.
The physical security of the ruling elite was ensured. As the discussion of
personalism in chapter four revealed, violence against rival elites is the defining
indicator and necessary requirement of a personalist regime.
361
The return to collective rule under Khrushchev and Brezhnev created a more
permissive environment for nuclear autonomy. By avoiding the excesses of Stalin’s
despotic rule, his successors were able to grant the Soviet military considerable
professional autonomy. Civil-military relations varied in the intensity of
assertiveness throughout Khrushchev and Brezhnev’s tenure, never returning to the
despotic conditions under Stalin nor ever reaching the delegative arrangements of
its democratic rivals.
362
As such, the Soviet Union was able to more fully embrace
the nuclear revolution after Stalin’s death in 1953. And once free from the internal
constraints of personalism, the Soviet regime developed the largest and most
technologically diverse nuclear arsenal of any authoritarian regime, including a full
strategic nuclear triad of air, sea, and missile delivery capabilities, tactical
(battlefield) nuclear weapons, robust command-and-control infrastructure, and
eventually even space based early-warning systems.
361
The violence must be directed to elites, not just the masses. Analysts have often have focused
on the mass killings. But it is the violence against senior elites that is hallmark of personalism.
362
Nor did civil-military relations in the USSR ever come close to delegative.
245
Chapter Six: Conclusion and Alternative Nuclear Futures
In this dissertation I have sought to understand the role of political regime
type in nuclear posture formation. One of the great debates in nuclear security
studies, and international relations more broadly, is whether the sources of a state’s
nuclear behavior are purely internal (domestic politics) or purely external
(international security environment). The answer is undoubtedly an amalgam of the
two since external security threats are refracted through domestic political
institutions. But that is an unsatisfying response insofar as it provides no
mechanism to adjudicate between conflicting imperatives (internal vs. external).
Put differently, the question is when are internal political imperatives likely to
trump external security imperatives? The answer I propose in this study is only in
a very specific type of authoritarian regime: a personalist regime. This is because
the extreme concentration of political power in a single leader enables the dictator
to organize the regime for the single purpose of staying in power.
The theoretical innovation I offer is to conceptualize the dictator’s extreme
concentration of power as a paradox. The process of concentrating and
consolidating such extreme power requires the subversion and degeneration of the
political institutions that facilitate the sharing of information – such as alternate
threat assessments – among regime elites. And once these authoritarian power-
sharing institutions have failed, it becomes dangerous, often deadly, to dissent or
246
disagree with the dictator. In such conditions, the dictator can restrict information
and enforce consensus views that essentially reduce structural pressures emanating
from the international system.
In chapter two I conducted an extensive critical review of the literature on
nuclear behavior and political regime type, starting with the legacy of the Cold War
and followed by the optimism-pessimism debate that raged in the early 1990s. Then
I analyzed the debate on the determinants of nuclear posture in the second nuclear
age. I found that the widespread focus on civil-military relations as the unit-level
variable of choice was theoretically useful, but incomplete because civil-military
relations was often considered in isolation from political regime type.
In chapter three, I argued that all political regimes feature three observable
power-sharing relationships of significant importance: elite-mass, intra-ruling elite,
and civil-military relations. I stressed that these power-sharing relationships form
the core of existing definitions of political regime types and sub-types. In simplest
terms, political regimes that feature all three functional power-sharing relationships
are considered democratic. Political regimes that lack elite-mass power-sharing
arrangements are considered authoritarian. Authoritarian regimes can be further
classified by the existence or absence of intra-ruling elite power-sharing
arrangements. Authoritarian regimes that lack intra-ruling elite power-sharing
arrangements, whether the ruling elite is civilian or military, are at risk of
247
degenerating into personalist regimes – defined by the extreme concentration of
power in a single leader.
I extended these insights into the study of nuclear posture by constructing a
theory of domestic political constraints on nuclear posture formation in
authoritarian regimes. The theory postulated that authoritarian regimes lacking
functional intra-ruling elite power-sharing institutions confront unique internal
political constraints in developing, deploying, and managing nuclear weapon
capabilities required for first-use asymmetric or full-spectrum escalation postures.
This is because in the absence of authoritarian power-sharing institutions a dictator
can acquire an extreme concentration of political power, which introduces
significant distortions into multiple domains of the regime, including civil-military
relations, strategic assessment, and technological innovation. These distortions
negatively affect each of the three core dimensions of nuclear posture: capabilities,
doctrine, and management.
In chapter four, I focused on current conceptualizations of the personalist
regime concept, developing a minimalist definition of the personalist regime
concept that can be utilized as the independent variable in future empirical tests of
the theory. I also developed a suitable measurement index for the minimalist
conceptualization of personalism to aid in political regime classification schemes.
Furthermore, I developed a four part typology of civil-military relations that can be
used as the intervening variable in the analysis of the relationship between intra-
248
ruling elite power sharing institutions and nuclear posture constraints. The typology
is constructed by combining different values on the institutional identity of who
exercises political control of the military in a regime – the institutional identity of
the political elite – along with the specific methods and mechanisms of political
control. In other words, the who and the how of political control of the military are
evaluated in tandem with disaggregated authoritarian regime type.
In chapter five, I illustrated the core hypotheses of my theory through a case
study of the nuclear history of the Soviet Union, the world’s first nuclear
authoritarian regime. I traced the connection between power-sharing institutions,
civil-military relations, and nuclear posture in the Soviet Union under the first
nuclear dictator: Joseph Stalin (1949-1953). I evaluated the functionality of power-
sharing institutions within the Communist Party of the Soviet Union (CPSU) and
the chronology of power concentration and consolidation. The findings provide
initial support to my theory of internal political constraints on nuclear posture in
authoritarian regimes. However, the brief duration of Stalin’s nuclear tenure – four
years – and the limited Soviet nuclear capabilities at the time of his death mean that
the case study is best considered an illustrative case than a true empirical test of my
theory.
249
Regime Type and Nuclear Posture: Empirical Patterns
As a final task of theory development in this study, I have organized existing
nuclear states according to their disaggregated regime type and nuclear posture
categories. This exercise reveals an empirical pattern that lends further support to
the theoretical framework offered in the dissertation. Table 6.1 displays the
variation in empirical outcomes by regime type.
250
Table 6.1 Regime Type and Nuclear Posture
Catalytic
363
Assured Retaliation Asymmetric Escalation
Full-Spectrum
Escalation/LoW
(Counter-Force)
Democracy Israel (1969-1991)
India (1998- present)
UK (1991- present)
Israel (1991- present)
US (1945-1970)
364
France (1960-present)
365
UK (1952-1991)
Pakistan (2007-present)
366
US (1970-present)
367
Junta
Pakistan (1990-
98)
368
Pakistan (1999-2007)
Machine S. Africa (1979-91) China (1976- present) USSR (1957-1982)*
369
USSR (1982-
1991)
370
Boss
USSR [Stalin] (1949-53)
China [Mao] (1964-76)
Strongman
363
Narang includes catalytic as a nuclear posture although I have excluded it from my analysis
because I view it as a proliferation strategy rather than a weaponized posture. I include it in the
table for the sake of comprehensiveness.
364
US first deployed tactical nuclear weapons to Europe in the spring of 1952. Evangelista,
Innovation and the Arms Race, p. 152. However, since the US used nuclear weapons in an
asymmetric role at Hiroshima and Nagasaki I’ve opted to code US posture as asymmetric from the
date of those attacks.
365
France’s AN-52 variable yield (6-8kt or 25kt) warhead entered into service 1972.
366
Coding Pakistan as a democracy is a contentious coding. It is unclear whether the elected
government has in fact established civilian control of the military. In any event, Pakistan’s civil-
military relations remain highly delegative, if not praetorian, and therefore generate the same
theoretical prediction for an asymmetric escalation posture.
367
US first deployed space-based early-warning satellites in 1970.
368
Some analysts may object to categorizing Pakistan as an authoritarian regime at the time of the
nuclear 1998 tests, meaning prior to the Musharraf coup in 1999. I find it reasonable to include
Pakistan in this group because its post-nuclear test posture decisions were made when Pakistan’s
regime type was a junta, in other words an unmistakable military dictatorship.
369
Evangelista, Innovation and the Arms Race notes 1957 as year of deployment for various
Soviet tactical nuclear systems.
370
USSR integrated space-based early-warning satellites with nuclear command and control in
1982, see Blair, Logic of Accidental Nuclear War. The Soviet system was never truly complete or
functional. High rate of technical failure. Not full coverage.
251
One pattern that emerges is that democracies and juntas have tended to
adopt asymmetric escalation postures (US, UK, France, and Pakistan). This is not
surprising since I hypothesize these regimes would face few internal constraints on
nuclear posture development. The sole remaining democratic nuclear state – India
– has not adopted an asymmetric escalation posture as of yet. This may be changing
though as India has begun to seriously discuss the development of tactical nuclear
weapons.
371
As my theory would expect, India’s power-sharing institutions could
support the development of delegative civil-military relations required for a future
transition to an asymmetric escalation posture.
Another trend is that civilian authoritarian regimes (machines) have
demonstrated variance in nuclear posture outcomes. Yet the Soviet Union post-
Stalin is the only civilian authoritarian regime to adopt an asymmetric escalation
posture. Importantly, this posture selection only occurred after Stalin had passed
from the scene and intra-ruling elite power-sharing institutions were reconstituted.
Stalin’s death heralded a dramatic change in the Soviet political system, freeing the
military from despotic control measures and permitting a renaissance of
professional military doctrinal debates. Under Stalin’s successors, especially
Brezhnev, the Soviet military experienced a “golden age” of civil-military relations
371
Toby Dalton and George Perkovich, “India’s Nuclear Options and Escalation Dominance,”
Carnegie Endowment for International Peace (May 2016). Available at:
http://carnegieendowment.org/files/CP_273_India_Nuclear_Final.pdf. See also, Vipin Narang,
“Five Myths About India’s Nuclear Posture,” Washington Quarterly vol. 36 no. 3 (Summer 2013),
p. 143-157.
252
as a result of the stabilized oligarchy. The end result of the process of political
development and stabilization was the most expansive nuclear posture of any
authoritarian regime to date.
The story of China’s nuclear posture, interestingly, is continuity across
decades and different leaders.
372
The transition from a personalist (boss) regime
under Mao to a non-personalist (machine) regime under Deng did not trigger a
wholesale reevaluation of China’s nuclear posture as it did in the Soviet Union after
Stalin’s death. This is perhaps partly due to the different leadership roles played by
Stalin and Mao in their respective regimes. Although both Stalin and Mao were
personalist dictators, Mao was also the revolutionary founder, chief ideologue, and
father of Chinese Communism. In this sense, Mao played the combined role of both
Lenin and Stalin. Whereas Stalin could be rapidly and harshly denounced after his
death by pointing to his “deviations” from Leninist thought, the same maneuver
could not be used to denounce Mao.
373
As such, it was a far more sensitive task to
denounce Mao following his death, because as Deng understood, “discrediting
Comrade Mao Zedong…would mean discrediting our Party and state.”
374
Mao’s
372
John Wilson Lewis and Xue Litai, Imagined Enemies: China Prepares for Uncertain War
(Stanford: Stanford University Press, 2006). John Wilson Lewis and Xue Litai, China’s Strategic
Seapower: The Politics of Force Modernization in the Nuclear Age (Stanford: Stanford University
Press, 1994). John Wilson Lewis and Xue Litai, China Builds the Bomb (Stanford: Stanford
University Press, 1988).
373
See text of “Khrushchev’s Secret Speech,” in Khrushchev Remembers, trans. Strobe Talbott, p.
559-618.
374
Roderick MacFarqhar and Michael Schoenals, Mao’s Last Revolution (Cambridge: Harvard
University Press, 2006), p. 456-458.
253
thought – including his views on nuclear weapons – would continue to loom large
even after his death.
Furthermore, the tumult of the Cultural Revolution created a lost decade in
Chinese nuclear development. The political turmoil reached into the nuclear
program, halting the work on critical delivery vehicles. Furthermore, the People’s
Liberation Army (PLA) was deeply involved in politics throughout the Cultural
Revolution, restoring domestic order, and even administrating large swaths of the
country. As such, the PLA had little opportunity or inclination to evaluate the
strategic implications of the nuclear revolution. Indeed, serious discussion of
nuclear strategy and doctrine in the PLA did not until the mid-1980s, nearly a
decade after Mao had passed from the scene.
375
Implications for Future Research
As is the case with all research on nuclear weapon states, the inferences of
this study have been limited by the small number of cases in the universe. The
pervasive “small N” problem has actually been intentionally exacerbated by using
a disaggregated concept of regime type to further reduce the universe of cases from
four existing authoritarian nuclear regimes (USSR/Russia, China, Pakistan, and
375
M. Taylor Fravel and Evan S. Medeiros, “China’s Search for Assured Retaliation: The
Evolution of Chinese Nuclear Strategy and Force Structure,” International Security, vol. 35, no. 2
(Fall 2010), p. 48-87.
254
North Korea) to only those nuclear regimes that have featured personalist dictators,
namely the USSR, China, and North Korea. Despite the methodological challenges
involved, the focus on the nuclear posture of personalist dictators is warranted for
two reasons. First, personalist dictators have been shown to be more likely to pursue
nuclear weapons.
376
Second, personalist dictators have been theorized to be the
most likely to initiate interstate conflict.
377
Therefore, understanding how the most
violent leaders in the world adapt to the nuclear revolution is important for both
theory and policy.
To that purpose, future research should continue to explore the question of
internal political constraints on nuclear posture in personalist regimes. And, in fact,
if the North Korean regime under Kim Jong-un continues to advance its nuclear
capability, it will soon present itself as the ideal “crucial” case for empirically
testing my theory of internal political constraints in personalist regimes. Whereas,
prior nuclear personalist dictators – Stalin and Mao – led states with large
conventional forces and considerable strategic depth, partially mitigating the need
for an asymmetric escalation posture – the DPRK under Kim Jong-un is in a far
more severe security environment. Thus, external threat and internal constraints
may come into direct conflict in Kim’s regime to an extent that has not occurred in
376
Christopher Way and Jessica L. P. Weeks, “Making it Personal: Regime Type and Nuclear
Proliferation,” American Journal of Political Science Volume 58, No. 3 (2014): 705-719.
377
Jessica Weeks, “Strongmen and Straw Men: Authoritarian Regimes and the Initiation of
International Conflict,” American Political Science Review Vol. 106, No. 2 (2012): 326- 347.
255
the nuclear era to date. As such, the North Korean case could help answer the
question of when internal political imperatives are likely to trump external security
imperatives.
Another avenue of potential research on personalism and nuclear posture is
what I will term “inherited arsenals.” Although personalist dictators have obviously
led their countries across the nuclear threshold, no regime has degenerated back
into personalism after acquiring nuclear weapons. In other words, no new
personalist dictators have arisen in an existing nuclear weapon state. However, with
the rise of Putin in Russia and Xi in China, this trend may be changing. Both leaders
have inherited nuclear arsenals and both are consolidating power. It is outside the
scope of this study to evaluate whether Putin and Xi have indeed become
personalist dictators, but the possibility exists that they will do so. And since both
Russia and China are currently engaged in substantial nuclear modernization
projects, the dual trend of personalization and nuclear modernization is worthy of
attention. Nuclear personalists may not be a thing of the past.
256
Bibliography
Aguero, Felipe. Soldiers, Civilians, and Democracy: Post-Franco Spain in
Comparative Perspective (Baltimore and London: The Johns Hopkins
University Press, 1995).
Albright, David. “South Africa and the Affordable Bomb,” Bulletin of the Atomic
Scientists July/August 1994.
Alexander, Brian and Alistair Millar, eds. Tactical Nuclear Weapons: Emergent
Threats in an Evolving Security Environment (Washington D.C.: Brassey’s
Inc., 2003).
Art, David. “What Do We Know About Authoritarianism After Ten Years?”
Comparative Politics vol. 44, no. 3 (April 2012).
Avant, Deborah D. “The Institutional Sources of Military Doctrine: Hegemons
and Peripheral Wars,” International Studies Quarterly, vol. 37, no. 4 (1993).
Ball, Desmond. Politics and Force Levels: The Strategic Missile Program of the
Kennedy Administration (Berkeley and Los Angeles: University of California
Press, 1980).
Bialer, Seweryn. Stalin's Successors: Leadership, Stability, and Change in the
Soviet Union. 1st paperback ed. (Cambridge and New York: Cambridge
University Press, 1982).
Biddle, Stephen and Robert Zirkle, “Technology, Civil-Military Relations, and
Warfare in the Developing World,” The Journal of Strategic Studies, vol.
19, no. 2 (June 1996).
Blair, Bruce. The Logic of Accidental Nuclear War (Washington, D.C.: The
Brookings Institution, 1993).
Bracken, Paul. The Second Nuclear Age: Strategy, Danger, and the New Power
Politics (New York: Times Books, 2012).
________. The Command and Control of Nuclear Weapons (New Haven and
London: Yale University Press, 1983).
Bratton, Michael and Nicolas van de Walle. Democratic Experiments in Africa:
Regime Transitions in Comparative Perspective (Cambridge and New
York: Cambridge University Press, 1997).
257
________. “Neopatrimonial Regimes and Political Transitions in Africa” World
Politics vol. 46, no. 4 (July 1994).
Braut-Hegghammer, Malfrid. Unclear Physics: Why Iraq and Libya failed to
Build Nuclear Weapons (Ithaca and London: Cornell University Press,
2016).
Breslauer, George W. Khrushchev and Brezhnev as Leaders: Building Authority
in Soviet Politics (London: George Allen Unwin, 1982).
Brooker, Paul. Non-Democratic Regimes 2
nd
ed. (London: Palgrave Macmillan,
2009).
Brooks, Risa A. Shaping Strategy: The Civil-Military Politics of Strategic
Assessment (Princeton: Princeton University Press, 2008).
Brownlee, Jason. Authoritarianism in an Age of Democratization (New York :
Cambridge University Press, 2007).
Bundy, McGeorge. “Existential Deterrence and Its Consequences,” in Douglas
MacLean, ed., The Security Gamble: Deterrence Dilemmas in the Nuclear
Age (Totowa, N.J.: Rowman and Allanheld, 1984).
Chehabi, H.E. and Juan J. Linz, eds. Sultanistic Regimes (Baltimore and London:
The Johns Hopkins University Press, 1998).
Cochran, Thomas B., Robert S. Norris, and Oleg A. Bukharin. Making the
Russian Bomb: From Stalin to Yeltsin (Boulder: Westview Press, 1995).
Cocks, Paul and Robert V. Daniels, Nancy Whittier Heer, eds. The Dynamics of
Soviet Politics (Cambridge: Harvard, 1976).
Colton, Timothy J. “Perspectives on Civil-Military Relations in the Soviet
Union,” in Soldiers and the Soviet State: Civil-Military Relations from
Brezhnev to Gorbachev, eds. Timothy J. Colton and Thane Gustafson
(Princeton: Princeton University Press, 1990).
________. Commissars, Commanders, and Civilian Authority: The Structure of
Soviet Military Politics (Cambridge and London: Harvard university
Press, 1979).
Cohen, Avner. “Israel: A Sui Generis Proliferator,” in Muthia Alagappa, ed., The
Long Shadow: Nuclear Weapons and Security in 21st Century Asia
(Stanford, CA: Stanford University Press, 2008).
258
Cohen, Avner and Marvin Miller. “Bringing Israel's Bomb Out of the Basement:
Has Nuclear Ambiguity Outlived Its Shelf Life?” Foreign Affairs vol. 89,
no. 5 (September/October 2010).
Conquest, Robert . The Great Terror: A Reassessment (New York and Oxford:
Oxford University Press, 1990).
________. The Great Terror: Stalin’s Purge of the Thirties (New York: The
Macmillan Company, 1968).
________. The Soviet Succession Problem (Defense Documentation Center no.
AD43407, 1964).
________. Power and Policy in the U.S.S. R.: The Study of Soviet Dynasties (New
York: St. Martin’s Press, 1961).
Dahl, Robert A. Democracy and its Critics (New Haven and London: Yale
University Press, 1989).
________. Controlling Nuclear Weapons: Democracy Versus Guardianship
(Syracuse, N.Y.: Syracuse University Press, 1985).
Dalton, Toby and George Perkovich, ”India’s Nuclear Options and Escalation
Dominance,” Carnegie Endowment for International Peace (May 2016).
Desch, Michael C. Civilian Control of the Military: The Changing Security
Environment (Baltimore and London: The Johns Hopkins University
Press, 1999).
Erickson, John. The Soviet High Command: A Military Political History, 1918-
1941 (Boulder and London: Westview Press, 1984).
Evangelista, Matthew. Innovation and the Arms Race: How the United States and
the Soviet Union Develop New Military Technologies (Ithaca and London:
Cornell University Press, 1988)
Fainsod, Merle. How Russia is Ruled (Cambridge,. Harvard University Press,
1953).
Feaver, Peter D. Armed Servants: Agency, Oversight, and Civil-Military Relations
(Cambridge, MA: Harvard University Press, 2003).
________. “Neooptimists and the Enduring Problem of Nuclear Proliferation,”
Security Studies vol. 6, no. 4 (Summer 1997).
________. “Optimists, Pessimists, and Theories of Nuclear Proliferation
Management,” Security Studies vol. 4, no. 4 (Summer 1995).
259
________. “Proliferation Optimism and Theories of Nuclear Operations,”
Security Studies vol. 2, no. 3-4 (1993): 159-191.
________. “Command and Control in Emerging Nuclear Nations,” International
Security vol. 17, no. 3 (1993): 160-187.
________. Guarding the Guardians: Civilian Control of Nuclear Weapons in the
United States (Ithaca and London: Cornell University Press, 1992).
Finer, S.E. The Man on Horseback: The Role of the Military in Politics (London
and Dunmow: Pall Mall Press, 1962).
Frantz, Erica and Natasha Ezrow. The Politics of Dictatorship (Boulder and
London: Lynne Reinner Publishers, 2011).
Fravel, M. Taylor and Evan S. Medeiros, “China’s Search for Assured
Retaliation: The Evolution of Chinese Nuclear Strategy and Force
Structure,” International Security, vol. 35, no. 2 (Fall 2010).
Freedman, Lawrence. The Evolution Nuclear Strategy, 3rd ed. (London: Palgrave
Macmillan, 2003).
Fuhrmann, Matthew. “Spreading Temptation: Proliferation and Peaceful Nuclear
Cooperation Agreements,” International Security vol. 34, no. 1 (2009).
Gaddis, John Lewis. “International Relations Theory and the End of the Cold
War,” International Security 17, no. 3 (Winter 1992/93): 5-58.
________. “The Long Peace,” International Security vol. 10, no. 4 (Spring 1986).
Gandhi, Jennifer. Political Institutions under Dictatorship (New York: Cambridge
University Press, 2008).
Gandhi, Jennifer and Ellen Lust-Okar. “Elections Under Authoritarianism,”
Annual Review of Political Science vol. 12, no. 1 (2009).
Gandhi, Jennifer and Adam Przeworski, “Authoritarian Institutions and the
Survival of Autocrats,” Comparative Political Studies vol. 40, no. 11
(2007).
Garthoff, Raymond L. Deterrence and the Revolution in Soviet Military Doctrine
(Washington D.C: The Brookings Institution, 1990).
________. “A Garthoff-Pipes Debate on Soviet Strategic Doctrine,” Strategic
Review, vol. 10, no. 4 (Fall 1982), p. 36-83.
________. “Mutual Deterrence and Strategic Arms Limitation in Soviet Policy,”
International Security, vol. 3, no. 1 (Summer 1978), p. 112-47.
260
________. “The Marshals and the Party: Soviet Civil-Military Relations in the
Postwar Period,” in Total War and Cold War: Problems in Civilian
Control of the Military, ed. Harry Coles (Columbus, OH: Ohio State
University Press, 1962)
Gause, Ken E. North Korean House of Cards: Leadership Dynamics under Kim
Jong-un (Washington D.C.: Committee for Human Rights, 2015).
Gavin, Francis J. “Politics, History, and the Ivory Tower-Policy Gap in the
Nuclear Proliferation Debate,” Journal of Strategic Studies 35, No. 4
(2012).
________. “Same As It Ever Was: Nuclear Alarmism, Proliferation, and the Cold
War,” International Security vol. 34, no. 3 (Winter 2009/10).
________. “Blasts from the Past: Proliferation Lessons from the 1960s,”
International Security vol. 29, no. 3 (Winter 2004/2005).
Gavin, Francis J. and James B. Steinberg. “Mind the Gap: Why Policymakers
and Scholars Ignore Each Other, and What Should be Done About It,”
Carnegie Reporter 6, no. 4 (Spring 2002).
Geddes, Barbara. Paradigms and Sandcastles: Theory Building and Research
Design in Comparative Politics (Ann Arbor, MI: University of Michigan
Press, 2003).
________. “What Do We Know About Democratization After Twenty Years?”
Annual Review of Political Science 2 (1999).
________. “Authoritarian Breakdown: Empirical Test of a Game-Theoretic
Argument,” Working Paper (1999).
George, Alexander L. and Andrew Bennet, Case Studies and Theory
Development in the Social Sciences Cambridge (BCSIA Studies in
International Security, MIT Press, 2005).
Gelman, Harry. The Brezhnev Politburo and the Decline of Détente (Ithaca and
London: Cornell University Press, 1984).
Gerring, John Case Study Research: Principles and Practices (New York:
Cambridge University Press, 2007).
Gill, Graeme. The Origins of the Stalinist Political System (Cambridge and New
York: Cambridge University Press, 2002).
261
Goldstein, Avery. Deterrence and Security in the 21st Century: China, Britain,
France, and the Enduring Legacy of the Nuclear Revolution (Stanford,
CA: Stanford University Press, 2000).
Gorlizki, Yoram and Oleg Khlevniuk, Cold Peace: Stalin and the Soviet Ruling
Circle, 1945-1953 (Oxford: Oxford University Press, 2004).
Gourevitch, Peter. “The Second Image Reversed: The International Sources of
Domestic Politics,” International Organization, vol. 32, no. 4 (Autumn
1978).
Gray, Colin. "National Styles in Strategy: The American Example," International
Security, vol. 6, no. 2 (Fall 1981);
________. Nuclear Strategy and National Style (Lanham, Md.: Hamilton Press,
1986).
Hamilton, Eric J. “International Politics and Domestic Institutional Change: The
Rise of Executive War-making Autonomy in the United States,” Ph.D.
Dissertation, University of Southern California, Los Angeles, 2015.
Hadenius, Axel and Jan Toerell, “Pathways from Authoritarianism,” Journal of
Democracy vol. 18, no. 1 (2007).
Hodgson, John H. “The Problem of Succession,” in The Dynamics of Soviet
Politics, ed. Paul Cocks, Robert V. Daniels, Nancy Whittier Heer
(Cambridge, Mass: Harvard, 1976).
Holloway, David. Stalin and the Bomb: The Soviet Union and Atomic Energy
1939-1956. (New Haven and London: Yale University Press, 1994).
________. The Soviet Union and the Arms Race (New Haven and London: Yale
University Press, 1983).
Horowitz, Michael. The Diffusion of Military Power: Causes and Consequences
for International Politics (Princeton: Princeton University Press, 2010).
Hough, Jerry F. and Merle Fainsod, How the Soviet Union is Governed
(Cambridge: Harvard University Press, 1979).
Huntington, Samuel P. “ Reforming Civil-Military Relations." Journal of
Democracy, vol. 6 no. 4, 1995.
________. The Third Wave: Democratization in the Late Twentieth Century
(Norman, OK: University of Oklahoma Press, 1991).
262
________. Political Order in Changing Societies (New Haven and London: Yale
University Press, 1968).
________. The Solider and State: The Theory and Politics of Civil-Military
Relations (New York: Vintage Books, 1957).
Hymans, Jacques E. C. Achieving Nuclear Ambitions: Scientists, Politicians, and
Proliferation (Cambridge: Cambridge University Press, 2012).
________. “When Does a State Become a Nuclear Weapons State? An Exercise
in Measurement Validation,” The Nonproliferation Review vol. 17, no. 1
(2010).
________. The Psychology of Nuclear Proliferation: Identity, Emotions, and
Foreign Policy (New York: Cambridge University Press, 2006).
Hymans, Jacques E. C. and Matthew S. Gratias, “Iran and the Nuclear Threshold:
Where is the Line?” The Nonproliferation Review vol. 20, no. 1 (2013).
Jervis, Robert. The Meaning of the Nuclear Revolution: Statecraft and the
Prospect of Armageddon (Ithaca and London: Cornell University Press,
1989).
Jervis, Robert, Richard Ned Lebow, and Janice Stein, eds., Psychology and
Deterrence (Baltimore: Johns Hopkins University Press, 1985).
Jo, Dong-Joon and Erik Gartzke. “Determinants of Nuclear Weapons
Proliferation,” The Journal of Conflict Resolution vol. 51, no. 1 (2007):
167-194).
Joffe, Ellis. Party and Army: Professionalism and Political Control in the
Chinese Officer Corp, 1949-1964. (Harvard: East Asian Monographs,
1967).
Johnston, Alastair Iain. “Thinking about Strategic Culture,” International Security
vol. 19, no. 4 (1995).
Joyner, Daniel H. “The Implications of the Proliferation of Weapons of Mass
Destruction for the Prohibition on the Use of Force,” in Marc Weller, ed.
The Oxford Handbook of the Use of Force in International Law (Oxford:
Oxford University Press, 2015).
Karl, David J. “Proliferation Optimism and Pessimism Revisited,” Journal of
Strategic Studies vol. 34, no. 4 (2011).
________. “Proliferation Pessimism and Emerging Nuclear Powers,”
International Security vol. 21, no. 3 (Winter 1996/97).
263
Kier, Elizabeth. Imagining War: French and British Military Doctrine between
the Wars (Princeton: Princeton University Press, 1997).
Khlevniuk, Oleg V. Stalin: A New Biography of a Dictator, trans. Nora Seligman
Favorov (New Haven and London: Yale University Press, 2015).
________. Master of the House: Stalin and His Inner Circle, trans. Nora
Seligman Favorov (New Haven and London: Yale University Press,
2009).
Khrushchev, Nikita S. Khrushchev Remembers, trans.and ed. Strobe Talbott
(Boston: Little, Brown & Company, 1970).
Kissinger, Henry A. Nuclear Weapons and Foreign Policy (New York: Harper &
Brother, 1957).
Kolkowicz, Roman The Soviet Military and the Communist Party (Princeton:
Princeton University Press, 1967).
Knopf, Jeffrey W. “Recasting the Proliferation Optimism-Pessimism Debate,”
Security Studies 12, no. 1 (2002).
Kroenig, Matthew. “Time to Attack Iran: Why a Strike is the Least Bad Option,”
Foreign Affairs vol. 91, no. 1 (January/February 2012).
________. “Exporting the Bomb: Sensitive Nuclear Assistance and Proliferation,”
American Political Science Review vol. 103, no. 1 (2009).
Lai, Brian and Dan Slater. “Institutions of the Offensive: Domestic Sources of
Dispute Initiation in Authoritarian Regimes, 1950-1992,” American
Journal of Political Science, vol. 50, no. 1 (January 2006).
Lavoy, Peter R., Scott D. Sagan, and James J. Wirtz, eds., Planning the
Unthinkable: How New Powers Will Use Nuclear Biological, and
Chemical Weapons (Ithaca and London: Cornell University Press, 2000).
Lavoy, Peter R. “The Strategic Consequences of Nuclear Proliferation,” Security
Studies 4, no. 4 (Summer 1995).
Lewis, John Wilson and Xue Litai, Imagined Enemies: China Prepares for
Uncertain War (Stanford: Stanford University Press, 2006).
________. China Builds the Bomb (Stanford: Stanford University Press, 1988).
________. China’s Strategic Seapower: The Politics of Force Modernization in
the Nuclear Age (Stanford: Stanford University Press, 1994).
Li, Nan, ed. Chinese Civil-Military Relations: The Transformation of the People’s
264
Linz, Juan J. and Alfred Stepan. Problems of Democratic Transition and
Consolidation: Southern Europe, South America, and Post-Communist
Europe (Baltimore and London: The Johns Hopkins University Press,
1996).
Lobell, Steven E., Norrin M. Ripsman, Jeffrey W. Taliaferro, eds., Neoclassical
Realism, The State, and Foreign Policy (Cambridge: Cambridge
University Press, 2009).
Lord, Carnes. "American Strategic Culture," Comparative Strategy, Vol. 5, No. 3
(1985).
MacFarqhar, Roderick, ed. The Politics of China, 2
nd
ed. (Cambridge: Cambridge
University Press, 1997).
MacFarqhar, Roderick and Michael Schoenals. Mao’s Last Revolution
(Cambridge: Harvard University Press, 2006).
Magaloni, Beatrice. “Credible Power-Sharing and the Longevity of Authoritarian
Rule,” Comparative Political Studies, vol. 41 (2008).
Mawdsley, Evan and Stephen White. The Soviet Elite from Lenin to Gorbachev:
The Central Committee and its Members, 1917-1991 (Oxford: Oxford
University Press, 2000).
Mearsheimer, John J. “Back to the Future: Instability in Europe After the Cold
War,” International Security 15, no. 1 (Summer 1990).
Meyer, Stephen M. “Soviet Nuclear Operations,” in Managing Nuclear
Operations, eds. Ashton B. Carter, John D. Steinbruner, and Charles A.
Zraket (Washington, D C.: The Brookings Institution, 1987).
________. Soviet Theatre Nuclear Forces, Adelphi Papers no. 187 (London :
International Institute for Strategic Studies, 1983-84).
Mitchell, R. Judson. Getting to the Top in the USSR: Cyclical Patterns in the
Leadership Succession Process (Stanford: Hoover Institution Press, 1990).
Montefiore, Simon Sebag. Stalin: The Court of the Red Tsar (New York: Alfred
A. Knopf, 2004).
Monteiro, Nuno P. and Alexandre Debs. “The Strategic Logic of Nuclear
Proliferation,” International Security 39, Vol. 2 (2014).
Morgan, Patrick M. Deterrence: A Conceptual Analysis (Beverly Hills, CA: Sage
Publications, 1983) and Deterrence Now (Cambridge: Cambridge
University Press, 2003.
265
Narang, Vipin. Nuclear Strategy in the Modern Era: Regional Powers and
International Conflict (Princeton and Oxford: Princeton University Press,
2014).
________. “Five Myths About India’s Nuclear Posture,” Washington Quarterly
vol. 36 no. 3 (Summer 2013).
________. “What Does it Take to Deter? Regional Power Nuclear Postures and
International Conflict,” Journal of Conflict Resolution, vol. 57, no. 3 (June
2013).
________. “Posturing for Peace?: Pakistan’s Nuclear Postures and South Asian
Stability,” International Security vol. 34, no. 3 (Winter 2009/10).
Odom, William E. The Collapse of the Soviet Military (New Haven and London:
Yale University Press, 1998).
Parrish, Michael. The Lesser Terror : Soviet State Security, 1939-1953.
(Westport, Conn. : Praeger, 1996).
Paul, T.V., Patrick M. Morgan, and James J. Wirtz, eds. Complex Deterrence:
Strategy in the Global Age (Chicago and London: Chicago University
Press, 2009).
Perlmutter, Amos and William M. LeoGrande. “The Party in Uniform: Toward a
theory of Civil-Military Relations in Communist Political Systems,” The
American Political Science Review, vol. 76, no.4 (December 1982).
Pipes, Richard. "Why the Soviet Union Thinks It Could Fight and Win a Nuclear
War," Commentary, vol. 64, no. 1 (July 1977).
Podvig, Pavel ed. Russian Strategic Nuclear Forces (Cambridge, Mass: the MIT
Press, 2001).
Pollack, Kenneth M. Unthinkable: Iran, the Bomb, and American Strategy (New
York: Simon & Schuster, 2013).
Posen, Barry R. The Sources of Military Doctrine: France, Britain, and Germany
Between the World Wars (Ithaca and London: Cornell University Press,
1984).
Potter, William C. and Jeffrey Lewis. “Cheap and Dirty Bombs: Could These
Creepy Chest Packs be North Korea’s Way of Threatening Radiological
War?” Foreign Policy. Available at:
http://foreignpolicy.com/2014/02/17/cheap-and-dirty-bombs/
266
Ra’anan, Uri. Flawed Succession: Russia’s Power Transfer Crises (Lanham, MD:
Lexington Books, 2006).
Reese, Roger R. “The Red Army and the Great Purges,” in Stalinist Terror: New
Perspectives, eds. J. Arch Getty and Roberta T. Manning (Cambridge:
Cambridge University Press, 1993).
Rice, Condoleeza. “The Party, The Military, and Decision Authority in the Soviet
Union,” World Politics vol. 40, no. 1 (Oct. 1987).
Roberts, Geoffrey. Stalin’s General: The Life of Zhukov (New York: Random
House, 2012).
Rose, Gideon. “Neoclassical Realism and Theories of Foreign Policy,” World
Politics 51 (October 1998).
Rush, Myron. Political Succession in the USSR. 2d ed. (New York: Columbia
University Press, 1968).
________. The Rise of Khrushchev (Washington, D.C.: Public Affairs Press,
1958).
Sagan, Scott D. ed., Inside Nuclear South Asia (Stanford: Stanford University
Press, 2009).
________. “The Origin of Military Doctrine and Command and Control
Systems,” in Peter R. Lavoy, Scott D. Sagan, and James J. Wirtz, eds.,
Planning the Unthinkable: How New Powers Will Use Nuclear Biological,
and Chemical Weapons (Ithaca and London: Cornell University Press,
2000).
________. “Why Do States Build Nuclear Weapons?: Three Models in Search of
a Bomb,” International Security vol. 21, no. 3 (1996).
________. The Limits of Safety: Organization, Accidents, and Nuclear Weapons
(Princeton: Princeton University Press, 1995).
________. ed., Civil-Military Relations and Nuclear Weapons (Palo Alto, CA:
Center for International Security and Cooperation, 1994).
Sagan, Scott D. and Kenneth N. Waltz, The Spread of Nuclear Weapons: An
Enduring Debate (New York and London: W.W. Norton & Company,
2013).
Sasikumar, Karthika and Christopher Way. “Testing Theories of Nuclear
Proliferation: The Case of South Asia,” in Inside Nuclear South Asia (Palo
Alto: Stanford University Press, 2009).
267
Schapiro, Leonard, and John W. Lewis. “The Roles of the Monolithic Party Under
the Totalitarian Leader.” The China Quarterly vol. 40, no. 40 (1969).
Schweller, Randall L. “Bandwagoning for Profit: Bringing the Revisionist State
Back In,” International Security, vol. 19, no. 1 (Summer 1994).
Scott, Harriet Fast and William F. Scott, The Armed Forces of the Soviet Union,
2nd edition (Boulder, CO: Westview Press, 1982).
Seng, Jordan. “Optimism in the Balance: A Response to Peter Feaver,” Security
Studies, vol. 6, no. 4 (1997).
________. “Less is More: Command and Control Advantages of Minor Nuclear
States,” Security Studies vol. 6, no. 4 (Summer 1997).
Service, Robert. Stalin: A Biography (Cambridge, Mass: Harvard University
Press, 2004)
Sescher, Todd S. “Are Soldiers Less War-prone than Statesmen?” Journal of
Conflict Resolution Vol. 48, No. 5 (2004).
Singh, Sonali and Christopher Way. “The Correlates of Nuclear Proliferation,”
Journal of Conflict Resolution vol. 48, no. 6 (2004).
Slater, Dan. “Iron Cage in and Iron Fist: Authoritarian Institutions and the
Personalization of Power in Malaysia,” Comparative Politics vol. 36, no. 1
(October 2003).
Snyder, Richard. “Paths out of Sultanistic Regimes: Combining Structural and
Voluntarist Perspectives” in H.E. Chehabi and Juan J. Linz, eds.
Sultanistic Regimes (Baltimore and London: The Johns Hopkins
University Press, 1998).
Solingen, Etel. Nuclear Logics: Contrasting Paths in East Asia and the Middle
East (Princeton: Princeton University Press, 2007).
Staniland, Paul. “Explaining Civil-Military Relations in Complex Political
Environments: India and Pakistan in Comparative Perspective,” Security
Studies, vol 17, no. 2 (2008).
Stepan, Alfred. Rethinking Military Politics: Brazil and the Southern Cone
(Princeton, New Jersey: Princeton University Press, 1988).
Svolik, Milan W. The Politics of Authoritarian Rule (New York: Cambridge
University Press, 2012).
268
Roth, Guenter “Personal Rulership, Patrimonialism, and Empire-Building in the
New States,” World Politics 20 (January 1968).
Talmadge, Caitlin. The Dictator’s Army: Battlefield Effectiveness in Authoritarian
Regimes (Ithaca and London: Cornell University Press, 2015).
________. “The Puzzle of Personalist Performance: Iraqi Battlefield
Effectiveness in the Iran-Iraq War,” Security Studies, vol. 22, no. 2,
(2013).
Taubman, William. Khrushchev: The Man and His Era (New York and London:
W.W. Norton & Company, 2003).
Taylor, Brian D. Politics and the Russian Army: Civil-Military Relations, 1689-
2000 (Cambridge: Cambridge University Press, 2003).
Thayer, Bradley A. “The Risk of Nuclear Inadvertence: A Review Essay,”
Trachtenberg, Marc. A Constructed Peace: The Making of the European
Settlement 1945-1963 (New Jersey: Princeton University Press, 1999).
Vogel, Ezra F. Deng Xiaoping and the Transformation of China (Cambridge and
London: Harvard University Press, 2011).
Walt, Stephen M. The Origins of Alliances (Ithaca, N.Y.: Cornell University
Press, 1987).
Waltz, Kenneth N. The Spread of Nuclear Weapons: More may be Better,
Adelphi Papers no. 171 (London: International Institute for Strategic
Studies, 1981).
________. Theory of International Politics (Reading, Mass.: Addison-Wesley,
1979).
Way, Christopher and Jessica L. P. Weeks. “Making it Personal: Regime Type
and Nuclear Proliferation,” American Journal of Political Science vol. 58,
no. 3 (2014).
Weber, Max. Economy and Society: An Outline of Interpretive Sociology, ed.
Guenter Roth and Claus Wittich (Berkeley: University of California Press,
1978).
Weeks, Jessica L. P. Dictators at War and Peace (Ithaca and London: Cornell
University Press, 2014).
269
________. “Strongmen and Straw Men: Authoritarian Regimes and the Initiation
of International Conflict,” American Political Science Review vol. 106, no.
2 (2012).
________. “Autocratic Audience Costs: Regime Type and Signaling Resolve,”
International Organization vol. 62, no. 1 (Winter, 2008).
Welch, Claude E. Jr. ed. Civilian Control of the Military: Theory and Cases from
Developing Countries (Albany: State University of New York Press,
1976).
Whitewood, Peter. The Red Army and the Great Terror: Stalin’s Purge of the
Soviet Military (Lawrence, KS: University Press of Kansas, 2015).
Wohlstetter, Albert. “The Delicate Balance of Terror,” Foreign Affairs 37, no. 2
(January 1959).
Yoshihara, Toshi and James R. Holmes, eds. Strategy in the Second Nuclear Age:
Power, Ambition, and the Ultimate Weapon (Washington D.C.:
Georgetown University Press, 2012).
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Lesser evils? WMD pursuit beyond nuclear weapons
PDF
Pathways to nuclear weapon reversal: exploring mechanisms and understanding non-proliferation policy
PDF
Power, profits, and politics: energy security and cooperation in Eurasia
PDF
Regional trade association and civil society: an analysis of contingent interactions between regime-level and organizational-level variables
PDF
When bullets are not enough: determinants of chemical weapons usage and abstention
PDF
Interdependency of the water-oil-nuclear industries in the Persian/Arabian Gulf: understanding risk and improving prevention and preparation of disasters
PDF
Like father, like son? A succession-based explanation for conflict initiation by authoritarian regimes
PDF
The social making of authoritarian environmentalism: protest-litigation nexus and policy changes in China
PDF
Shaking the foundations of violent civil conflict: institutions, disasters, and the political economies of state-rebel interaction
PDF
Thawing rivalries and fading friendships: a multi-method approach to rapprochement and alienation
PDF
Identity salience and political violence: centering social identity in the study of intrastate conflict
PDF
A two-level analysis of foreign policy decision making: an empirical investigation of the case of China-Taiwan
PDF
When passions run high: emotions and the communication of intentions in face-to-face diplomacy
PDF
The political economy of implementation: intellectual property rights protection across the world
PDF
Perspectives on state capacity and the political geography of conflict
PDF
Developing countries in the digital era: state and business interactions for industrial upgrading
PDF
Carnival in the birdcage: a study of the film and media industries' reforms in China after 1978 and its soft power
PDF
The allocation and accessibility of aid: the role of recipient government agency in dictating donor decisions
PDF
Legitimizing self-determination: advancing the sovereignty of separatist movements
PDF
Supply and demand of economic hierarchy: the Northeast Asia case
Asset Metadata
Creator
Gratias, Matthew S.
(author)
Core Title
The paradox of personalism: internal political constraints on nuclear posture in authoritarian regimes
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Political Science and International Relations
Publication Date
02/10/2019
Defense Date
01/25/2017
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
authoritarian political regimes,civil-military relations,nuclear posture,nuclear weapons,OAI-PMH Harvest,personalist regimes
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
James, Patrick (
committee chair
), English, Robert (
committee member
), Meshkati, Najmedin (
committee member
)
Creator Email
gratias@usc.edu,matthewgratias@yahoo.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-334250
Unique identifier
UC11255889
Identifier
etd-GratiasMat-5037.pdf (filename),usctheses-c40-334250 (legacy record id)
Legacy Identifier
etd-GratiasMat-5037.pdf
Dmrecord
334250
Document Type
Dissertation
Rights
Gratias, Matthew S.
Type
texts
Source
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 a...
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
Tags
authoritarian political regimes
civil-military relations
nuclear posture
nuclear weapons
personalist regimes