Citation
Central Asian repression and Western influence

Material Information

Title:
Central Asian repression and Western influence
Creator:
Goudge, Sara L. ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (86 pages). : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of Arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Political Science, CU Denver
Degree Disciplines:
Political science

Subjects

Subjects / Keywords:
Foreign relations -- United States -- Asia, Central ( lcsh )
Foreign relations -- Asia, Central -- United States ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
Repression, used by dictators as a tool to control the populous of a nation, is a defining feature of authoritarian regimes. Often the most concerning response, the used of coercion is centrally important to a greater understanding of how dictators maintain stability. Emerging theories have proposed that autocracies may be influenced by the West toward democratic transition, thus away from repressive autocratic behavior. For both intellectual and ethical reasons it is important to know when a regime will use repression and how the West may be able to modify these reposes. This study explores when the regimes of Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan have used violent forms of coercion or restricted the rights and freedoms of their people in order to maintain stability, and if ties to Western nations influenced this behavior. Looking at problems that arise vertically (from the people), horizontally (from elite division), or socio-economically (from economic decline or rising inflation), this study explores how these authoritarian regimes respond to a variety of challenges, and how Western linkages alter these responses. Using interaction terms to test the conditional relationship between challenges to regime stability and Western linkages, this study explores how ties with the West may be able to influence coercive responses in these authoritarian regimes. The following analysis finds that these regimes view rising inflation as a greater threat to their stability, resulting in an increased use of violence against their people, but interestingly, when challenged during periods of economic decline these dictators expand freedoms and rights. Additionally, when these regimes increase their ties to the West they utilize more difficult to distinguish forms of repression and are more apt to restrict the rights of their people. Finally, the following analysis shows that whether a challenge is socio-economic or domestic alters the influence of Western linkages, and that in some cases the growth of these ties may be viewed by these regimes as liabilities to their rule.
Thesis:
Thesis (M.A.)--University of Colorado Denver. Political science
Bibliography:
Includes bibliographic references.
System Details:
System requirements: Adobe Reader.
General Note:
Department of Political Science
Statement of Responsibility:
by Sara L. Goudge.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
891657113 ( OCLC )
ocn891657113

Downloads

This item has the following downloads:


Full Text
CENTRAL ASIAN REPRESSION
AND WESTERN INFLUENCE
by
SARA L. GOUDGE
B. A.University of Colorado at Denver2010
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Arts
Political Science
2013


2013
SARA GOUDGE
ALL RIGHTS RESERVED


This thesis for the Master of Arts degree by
Sara L. Goudge
has been approved for the
Political Science Program
by
Christoph H. Stefes, Chair
Michael Berry
Thorsten Spehn
October 8th 2013


GoudgeSara L. (M.A.Political Science)
Central Asian Repression and Western Influence
Thesis directed by Associate Professor Christoph H. Stefes
ABSTRACT
Repression, used by dictators as a tool to control the populous of a nation, is a
defining feature of authoritarian regimes. Often the most concerning response, the use of
coercion is centrally important to a greater understanding of how dictators maintain
stability. Emerging theories have proposed that autocracies may be influenced the West
toward democratic transition, thus away from repressive autocratic behavior. For both
intellectual and ethical reasons, it is important to know when a regime will use repression
and how the West may be able to modify these responses. This study explores when the
regimes of Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan have used violent forms
of coercion or restricted the rights and freedoms of their people in order to maintain
stability, and if ties to Western nations influenced this behavior. Looking at problems that
arise vertically (from the people), horizontally (from elite divisions), or socio-
economically (from economic decline or rising inflation), this study explores how these
authoritarian regimes respond to a variety of challenges, and how Western linkages alter
these responses. Using interaction terms to test the conditional relationship between
challenges to regime stability and Western linkages, this study explores how ties with the
West may be able to influence coercive responses in these authoritarian regimes. The
following analysis finds that these regimes view rising inflation as a greater threat to their
stability, resulting in an increased use of violence against their people, but interestingly,
when challenged during periods of economic decline these dictators expand freedoms and
iii


rights. Additionally, when these regimes increase their ties to the West they utilize more
difficult to distinguish forms of repression and are more apt to restrict the rights of their
people. Finally, the following analysis shows that whether a challenge is socio-economic
or domestic alters the influence of Western linkages, and that in some cases the growth of
these ties may be viewed by these regimes as liabilities to their rule.
The form and content of this abstract are approved. I recommend its publication.
Approved: Christoph H. Stefes
IV


ACKNOWLEDGEMENTS
Though there have been many individuals that contributed to this thesis, I would
like to first thank all of the wonderful professors in the Political Science Department at
the University of Colorado at Denver. Specifically, I must thank Christoph Stefes and
Michael Berry for all of their personal assistance with this project. Thank you both for all
of your insight and direction, because this would have been impossible without both of
you. I would also like to thank James Myers for his astute suggestions and his extremely
detailed editing technique, and though we may disagree on stylistic issues, I still
occasionally respect your opinion. And finally, I must thank my parents, Linda and
Chuck Goudge, for all of their support throughout the years, because without them none
of this would have been possible.


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION.................................................1
II. REVIEW OF THE LITURATURE.................................... 6
III. HYPOTHESES, MEASUREMENT, AND MODELS.........................16
Hypotheses..................................................16
Measurement.................................................18
Models......................................................27
IV. RESULTS.....................................................28
Model1:Coercion and Horizontal, Vertical and Socio-Economic
Challenges..................................................28
Model 2: Inflation, Western Linkages, and Repression........35
Model 3: Economic Decline, Western Linkages, and Repression.37
Model 4: Terrorism, Western Linkages, and Repression........42
Model 5: Elite Defection, Western Linkages, and Repression..46
Model 6: Protests, Western Linkages, and Repression.........49
V. DISCUSSION..................................................53
The First Model.............................................53
Geopolitical Linkages.......................................59
Without Challenges, Low Intensity Coercion Increases........60
Socio-Economic vs. Domestic Challenges......................62
VI. CONCLUSION..................................................68
REFERENCES..............................................................72
vi


APPENDIX
A. Document Analysis Coding..................................75
B. Variable Ranges...........................................76
vii


LIST OF TABLES
Tables
1. Model1 Coercion and Challenges..........................................30
2. Inflation and Coercion..................................................33
3. Economic Decline and Coercion...........................................34
4. Model 2 Inflation, Western Linkage and Repression.......................35
5. Model 3 Economic Decline, Western Linkages, and Repression..............38
6. Inflation, Western Linkages, and Repression 1996-2010.................. 40
7. Economic Decline, Western Linkages, Repression 1996-2010............... 41
8. Terrorism, Western Linkages, and Repression.............................43
9. Elite Defection, Western Linkages, and Repression.......................47
10. Protests, Western Linkages, and Repression.............................49
11. Models 2-6 Linkage Interaction Terms...................................51
12. Model1 Expected and Directional Results................................54
13. Socio-Economic Hypotheses 1993-2010 vs. 1996-2010..................... 56
14. Linkages Increase Low Intensity Repression when Challenges are Absent... 60
viii


Figures
LIST OF FIGURES
1.Socio-Economic and Domestic Challenges with Linkages
62
IX


CHAPTER I
INTRODUCTION
"We never expected they would shoot at women and children. I've got three children, I
was trying to cover them up. I was crying: 'Take my life, don't shoot my children.'" (BBC
17 May2005).
After several days of peaceful antigovernment protest, on the evening of May 13,
2005 the military began openly firing on the crowds that had gathered in Andijon,
Uzbekistan (Radio Free Europe/Radio Liberty 2010Most accounts conclude that these
demonstrators went to the streets to protest the injustice, harsh repression and poverty
they were experiencing under the mle of Islam Karimov (Radio Free Europe/ Radio
Liberty 2010). On that day it is estimated that somewhere between several hundred and
nearly 1,000 people lost their lives (Radio Free Europe/ Radio Liberty 2010). Islam
Karimovs regime has not only argued that the casualties were much lessbut that this
harsh response was necessary because these were not protesters, but religious extremists
(Radio Free Europe/ Radio Liberty 2010). What actually happened during the Andijon
Massacre is still somewhat unclear, but what is known is that this action was a striking
example of an authoritarian regime reacting with bmtal violence when challenged by
popular protests.
Although it may seem that the choice to fire on peaceful protesters was an
overreaction, this behavior is not only common in authoritarian regimes, but can at times
be considered acceptable. Knowing when these regimes resort to violence is not only of
importance for a greater understanding of political science, but also because this
knowledge could be the basis for informed and responsible foreign policy around the
1


globe. The following study hopes to provide insight into when these regimes react with
repression and how ties with Western nations may be able to influence these leaders away
from using coercion to control their people.
The study that follows utilizes a theoretical framework that recognizes the
dynamic nature of authoritarian regimes, rather than focusing on the possibility for
democratization. Using the cases of Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan,
this research hopes to avoid the misconception that these regimes are merely failed or
semi-democracies. Much of the research conducted on this region has focused primarily
on the state, the culture, or the possibilities of democratic transition (Liu 2011, Spechler
2007, Hale 2011). Though work in this region has been preformed in many different
academic disciplines, each with their own important contributions, political science
studies have mainly focused on the process of democratization (Petrie 2005, Juraev 2008,
Spechler 2007). Although this research is important, it makes the assumption that these
nations are working toward greater liberalization or expanded human rights (as if this
were the natural and expected progression). By utilizing theories that explore the
dynamic nature and durability of authoritarian regimes, the following research will
examine these nations through the lens of understanding authoritarian forms of
governance, rather than imperfect democracies.
The following study is interested in determining when theses Central Asian
nations react not only with violence against their people, but also by restricting the rights
of their citizens more generally. The research that follows will explore the challenges that
have arisen in the selected cases from 1993-2010. By looking at challenges that may
arise in these nations, such as socio-economic problems, societal instability, or elite
2


divisions, this work hopes to contribute to a nuanced understanding what kinds of
problems elicit repressive responses from these regimes and how the West might assuage
these responses.
The two central research questions that have guided this research are: under what
circumstances do the authoritarian regimes in Central Asia use repression or coercion to
control their people, and what types of Western linkages deter these regimes from using
these techniques? The measurement for the level of coercion in these nations is based off
the ratings provided by the Cingranelli-Richards (CIRI) Human Rights Dataset. Although
it can be difficult to find information from this region of the world, this study will use
data collected from the World Bank, the US Department of Homeland Security, US
Census, and the United Nations Economic and Social Commission for Asia and the
Pacific (UNESCAP). For the concepts that are more difficult to measure, or where
reliable information was unavailable, the Keesing5s World News Reports are used and a
process of document analysis is performed. This research uses a time-series panel design
that will test not only the use of coercion when these regimes face challenges, but also
how ties to Western nations interact with these responses.
The following study focuses on the nations of Kazakhstan, Kyrgyzstan, Tajikistan
and Uzbekistan, but omits Afghanistan and Turkmenistan. Although it would have been
optimal to use all of the nations in Central Asia in order to make broad assertions about
this region, these four were selected because they are each interesting cases individually.
This research hopes to explore Western influence in the region, thus Afghanistan has
been omitted because the Western led war being conducted in this nation changes the
dynamic that hopes to be explored in this research. Because this nation was invaded by
3


the US and NATO, this use of force does not allow for this to be a possible case in which
Western influence is examined, since for the last 12 years it has been subject to only
Western intervention. Turkmenistan was omitted because the data from this nation are
even scarcer than the data that are available for the other countries in the region.
Kazakhstan was selected because it is the largest nation in the region, which also
possesses the greatest oil wealth. To examine the range of authoritarian practices,
Kyrgyzstan was selected because it is arguably the most democratic in the region, and
conversely Uzbekistan was selected because it falls on the more repressive end of the
authoritarian continuum. And finally, Tajikistan was chosen because the nation
experienced a civil war after independence. The limited scope of this study did not allow
for all of the cases in Central Asia to be examined, but those that were selected each in
their own way expose the complicated and diverse nature of the nations in this region.
Using theories of authoritarianism, guided by an understanding of the dynamic
nature of these types of regimes, this study produces several major findings. First of all,
this research shows that when these regimes are experiencing economic downturns, they
surprisingly, respond by broadening the rights and freedoms of their people.
Theoretically, it is predicted that authoritarian regimes, when faced with economic
hardships will increase repression in order to maintain stability, rather than expanding the
rights of the citizen. This finding shows that these regimes react contrary to anticipated
expectations. Additionally, during periods of increased inflation from 1996-2010 these
regimes tend to increase their use of violent repression. This study also finds that these
regimes are more likely to restrict the rights of their people when there are greater ties to
the West. And finally the following research shows that the type of problem that arises in
4


these nations, whether socio-economic or domestic, can dramatically change the
influence of ties to Western nations. Though these findings are ultimately important
contributions to the research on these types of regimes, as well as Central Asia, they
highlight how dynamic, yet responsive dictatorships can be. The study that follows will
show that understanding the behavior of authoritarian regimes is complex and that the
power and influence of the West in changing the coercive behavior of these regimes is
dependent on the type of problems that the leaderships face in these nations.
5


CHAPTER II
REVIEW OF THE LITURATURE
When the Soviet Union collapsed the initial reaction was one of hope for the
future of democracy around the globe. Many political scientists claimed that the
democratic model had succeeded in conquering all other political systems, but as the dust
finally settled it became clear that although many of the newly independent nations had
pursued democratic institutions, these arrangements were merely facades for the
increasingly authoritarian characteristics of these regimes (Schedler 2002, Levitsky and
Way 2010). Because these governments were unlike the traditional democracies, or
outright autocracies, theorists were forced to address the possibility that many of these
new nations were not just semi-democracies, but had become a new form of
authoritarianism (Gilbert and Mohseni 2011). In facing the emergence of systems that
appear to have the democratic prerequisites, but are very clearly not democracies
theorists have proceeded to reconsider previous understandings of these types of regimes.
This has led to a broader understanding of the complexity of authoritarian governance
rather than just viewing them as imperfect forms of democracy (Brooker 2009, Levitsky
and Way 2002). Recognizing that authoritarian regimes are their own distinct form of
governance, and that these systems do not necessarily lead to democracy, studies have
begun examining the durability and responsiveness of these regimes and how they
respond to challenges. Theorists have proposed that challenges to a regime can emerge
from within the governing elite, from the people, or the international community
(Gerschewski 2013, Levitsky and Way 2010). Although there are many ways that a
regime can respond to these challenges, repression is not only one of the most visible, but
6


one of the most definitive authoritarian reactions. By looking at the work of multiple
theorists that explore the plethora of challenges that face authoritarian regimes and the
importance of understanding when a regime uses repressive or coercive techniques, it
will become clear that the dynamic nature of these regimes must be further explored.
Regimes can face challenges that come from vertical, horizontal, and
international sources. More specifically, a regime can fall victim to cleavages within the
ruling elitewhich can be considered a horizontal problem. Problems may also arise
vertically originating from the populous. And finally, the pressure of international actors
to democratize has also been theorized to influence the behavior of regimes or destabilize
them and possibly dislodge a dictator. The central concern of a regime is to maintain its
power. To do this leadership must be able address specific problems.
One of the most important challenges can come from within the regimes elite.
Gerschewski (2013) proposes a theory of three pillars within an authoritarian regime that
can predict the stability of the government in power. These pillars include: repression,
cooptation, and legitimation (Gerschewski 2013). He defines cooptation uas the capacity
to tie strategically-relevant actors (or group of actors) to the regime eliteGerschewski
2013, 22). It is centrally important for the leadership to continue to maintain the support
of other powerful people within the ailing elite, because these actors have their own
power base, whether it may be financial, tribal, or regional (just to name a few) and can
easily defect from the ailing party and join the opposition.
Gandhi and Przeworski (2007) explain that there are multiple levels to the elite
power structurewhich they call trenches. The first trench is made up of the central
members of a leaders administrationsuch as those that make up [c]onsultive councils
7


juntasand political bureaus (Gandhi and Przeworski 2007,1289). The second trench is
comprised of ua legislature that encapsulates some opposition, a party that mobilizes
popular support for the dictator, or even multiple parties^ (Gandhi and Przeworski 2007,
1293). The cooperation from these political elite is thought to enhance the durability of
the authoritarian leader. These theories suggest that some of the greatest challenges to a
regime can be horizontal.A dictator must be concerned with the leadership that is closest
to them, and the opposition within the governing bodies. The complexity of this situation
requires a leader to maintain many relationships and ward off challenges from these
groups and individuals.
Vertical challenges can also come in many forms, and can also require some
astute maneuvering for the regime to combat these problems. Gerschewskis (2013) pillar
of legitimation can be used to explain challenges that arise from the societys discontent
with the leadership, as well as the socio-economic factors that may destabilize a regime.
He explainsLegitimation seeks to guarantee active consentcompliance with the rules
passive obedienceor mere toleration within the population (Gerschewski 2013,18). He
breaks up this concept into the two distinct categories of diffuse and specific support
(Gerschewski 2013, 20). Specific support is gained when a regime is able to fulfill
demands, such as guaranteeing security and improved economic conditions (Gerschewski
2013, 20). Diffuse support refers to more general ideas about a regime, and includes
political ideologies promoted, national identity and leadership (Gerschewski 2013, 20).
Problems that arise from below generally challenge the legitimacy of a regime.
One of the most important vertical challenges to authority can come in the form of mass
protests, which are generally a sign that a regime has lost legitimacy. These types of
8


challenges can appear instantaneous and remove a leader very quickly (Kuran 1991).
Although protests are the most studied of the various vertical challenges, terrorism can
also be considered a vertical challenge, though generally from a smaller faction in the
society. Using Gerschewski5s (2013) concept of legitimation, both of these challenges are
symptoms of a regime losing support from the people (20). Although terrorist acts are
generally committed by a small minority of people in society, they are by definition
intended to destabilize a nation and challenge the regime in power. The final type of
vertical challenge, that is particularly important in Central Asia, are instances of ethnic
violence. Though this factor does not specifically challenge the central governments
policiesit does challenge a regimes ability to ensure the security of the people in
society.
It has also been theorized that a regime can face challenges when there are
dramatic economic fluctuations. These economic shocks are not in origin vertical, but
they can manifest themselves into problems that challenge the regime from below
(Aleman and Yang 2011). In other words, though the economic problems can directly
disrupt the governmentthe major concern is that it will promote social unrest. This idea
fits nicely into Gerschewskis (2013) definition of specific support that translates into a
stable pillar of legitimation. If a nation is facing economic hardshipsand the regime is
unable to remedy these problems, the pain will be felt by those at the bottom, which can
translate into a loss of specific support, and thus legitimation (Gerschewski 2013).
Economic challenges can arise from decreases in economic performance as well
as increases in inflation. Studies have found that there is a connection between economic
growth and development, which can decrease the human rights abuses and promote
9


democratization in Central Asia and other post-Soviet cases (Spechler 2007, Neundorf
2010). In the same strand of logic that economic growth would promote democratization,
the inverse, that economic decline could induce a government to become more repressive
and authoritarianmight also be true.
Gasiorowski (1995) finds that when regimes are faced with economic crises there
is an increased likelihood for regime breakdown. He finds that there are differences in the
effect of specific types of crises depending on the time period in which they occur
(Gasiorowski 1995). The author explains:
We can conclude from these findings that economic crises do not simply
undermine the legitimacy of whatever type of regime currently exists in a country,
thus triggering regime change in either direction, as argued by several authors.
However, we must also conclude that while inflationary crises inhibited
democratization in both contexts in a complementary manner in the 19505s and
1960s [...] they may actually have facilitated democratic transitions in the late
19805s; and recessionary crises did not have this kind of complementary effect
throughout 1950-1989.
-Gasiorowski 1995, 892
This study shows that the impact of inflation has different effects depending on when the
challenge to the regime occurs. It is clear from these findings that inflation and economic
decline are important challenges that may lead to regime breakdown and should be
centrally important in the study of authoritarian regimes.
Though a regime can encounter problems from multiple levels within society, it
may also be challenged by the international community. Challenges that come from
international organizations, such as the UN, generally, come in the form of pressure to
democratize. Within the theory of competitive authoritarianism proposed by Levitsky and
Way (2010), the importance of Western influence on democratic policies and the
10


possibility of democratic regime change is explored. These theorists explain that
international pressure may influence regimes though both linkages and leverage exercised
by Western powers, particularly the United States (Levitsky and Way 2010). Leverage is
defined as the level of vulnerability of a regime to pressures from the international
community to democratize, which can come in the form of "positive conditionality (for
example, EU membership), punitive sanctions (aid withdrawal, trade sanctions),
diplomatic persuasionand military forceLevitsky and Way 2006, 382). The influence
of these forms of leverage may be diminished by the size of the country that the West is
attempting to influence, for instance, a large nation with many resources would be less
vulnerable to these types of pressures in comparison to a smaller nation (Levitsky and
Way 2006, 383). Additionally, the authors explain that there are two other limitations that
may negate the influence of Western leverage, which are: an alternative regional power
that can offer support to the regime that does not push democratization (such as Russia or
China in the case of Central Asia), or there may be other Western policy objectives that
outweigh the importance of democratization of the nation (Levitsky and Way 2006, 383).
The authors explain that leverage can increase the costs to the regime to commit
authoritarian abuses, but that there are limitations to using leverage to change the
behavior of authoritarian regimes (Levitsky and Way 2006, 383).
The authors explain that linkages "can be defined as the density of ties and cross-
border flows between a particular country and the U.S, the EU and western-dominated
multilateral institutionsLevitsky and Way 2006, 383). These ties can be broken up into
five dimensions that include: economic, geopolitical, social, communication, and through
transnational civil society (Levitsky and Way 2006). These types of linkages apply
11


pressure on authoritarian regimes because when a government commits an abuse there is
a heightened chance of international salience and thus an increased chance that Western
governments may act to combat these abuses (Levitsky and Way 2006, 384). The authors
explain that linkages help to make leverage more effective because they increase the
boomerang effects that can draw attention to even minor abuseswhich increases the
costs to the regime for acting outside of the expected norms promoted by the West
(Levitsky and Way 2006, 386). These types of bonds change the expectations from
society of the government and also make the use of coercion or repression less likely.
This theory indicates that countries that have higher linkages and leverage will be
pressured to democratize. The authors emphasize that without linkages the influence of
leverage is lower, if not entirely ineffective. Because linkages are so important to the
successful use of leverage, these types of bonds are of central importance in
understanding how to change the repressive or coercive behavior of authoritarian
regimes.
As the previous discussion has shown there are many types of challenges that
authoritarian regimes may face and although there are many ways that a government may
respond to these problems one of the most visible and significant is through their use of
repression to maintain power.
One of the defining characteristics of authoritarian regimes is their use of
repression or coercion (Gerschewski 2013, 21). Though these types of governments have
other tools that can be used to maintain their powera regimes use of coercive
techniques helps to differentiate this type of governance from more democratic systems.
One of the most important ways an authoritarian regime can respond to challenges and
12


maintain stability is through the use of repression and the ability to use the coercive
capability of the security apparatuses, which has been found to be centrally important to
the stability and success of a regime (Gerschewski 2013, Beilin 2012). The use of
coercion is not a constant factor though; rather it increases and decreases in response to
challenges that face the regime in power.
Levitsky and Way (2010) explain that conceptually, repression or coercion is not
only limited to the level of visible repression, but also includes more subtle techniques.
These authors propose that there are two types of coercion: high intensity and low
intensity (Levitsky and Way 2010, 57-58). High intensity coercion, they explain, is more
visible because the regime may target large groups, important people, or institutions
(Levitsky and Way 2010, 57). Specificallyhigh-intensity coercion could include
campaigns of violence against opposition parties (e.g. Cambodia and Zimbabwe),
imprisonment (e.g. Malaysia and Russia), attempted assassination of major opposition
leaders (e.g. Belarus and Ukraine)and high-profile assaults on democratic institutions
such as parliament (e.g. Russia in 1993)Levitsky and Way 2010, 57-58). Low-intensity
coercion is much less visible and can take multiple forms, such as surveillance,
harassment, denial of public services, or employment (Levitsky and Way 2010, 58). This
type of coercion can take many forms, but the use of this type or repression is always to
accomplish the same goal: to diminish the power of the possible opposition to such a
degree that they no longer are able to pose a serious challenge (Levitsky and Way 2010,
58). By creating these conceptual categories Levitsky and Way (2010) bring attention to
the fact that these regimes have many tools that they can use in response to possible
challenges that may threaten their regime.
13


There are also limitations to the ability of a regime to use some forms of
repression. Johnston (2011) explains dictators can respond to protests with repression, but
that increased brutality canin some casesembolden the protesters. It is thought that low
levels of repression or high levels will both result in an increase in protest, and that a
government needs to find a level of repression that is neither too tough or too lenient to
quell social movements (Johnston 2011: 108-109). Attempting to find this line of
proportionality is referred to as the dictators dilemmaJohnston 2011Francisco
2005). Repression, particularly high-intensity or more visible types of coercion, also
carry risks in the eyes of the West. Regimes may also be influenced by international
actors to decrease the use of repression. Because there are many concerns that surround
the use of coercion, whether it is in the form of an increased reaction from the people, or
from the West, it is clear that there are many concerns that must be weighed by a leader
or regime when faced with the many types of problems that may arise.
Though repression is not the only technique that a regime can choose to use, how
and when they choose to use coercion to elicit desired responses is arguably the most
valuable to understand for political scientists and for the discussion international foreign
policy. This can be said because knowing when a regime may respond with high-intensity
or low-intensity coercion could help to guide policies that may influence democratization
and at the very least could limit morally negligent policies. A regime can use repression
or coercive techniques whenever it chooses, but some have recognized that there are
different challenges from different sources that may destabilize a regime. By recognizing
multilevel challenges and particular forms of repression that a regime may use to combat
them, the following research will examine how these pressures interact and whether
14


Western linkages are able to mediate the coercive responses from these regimes in
Central Asia.
15


CHAPTER III
HYPOTHESES, MEASUREMENT, AND MODELS
Hypotheses
There are many ways that an authoritarian regime can be challenged, but how the
government responds to these problems speaks to the dynamic nature of these regimes
and the necessity for scholars to understand their behavior. One of the most important
reactions that a regime may have to particular issues is to repress its people, through
either high intensity or low intensity forms of coercion. To explore the relationship
between the level and type of coercion implemented by a regime, this study will test
various types of challenges and the specific actions taken by the governments in Central
Asia. By applying the previously discussed theories to the cases of Uzbekistan,
Tajikistan, Kyrgyzstan and Kazakhstan, the relationship between coercion or repression
and vertical, horizontal, and international challenges will be explored.
The discussion presented in the literature review lends itself to four different
hypotheses.
HYPOTHESIS 1:Societal challenges to a regime will lead to an increase in the
level of coercion.
HYPOTHESIS 2: Pronounced economic downturns will cause an increase in the
level of coercion.
HYPOTHESIS 3: Rises in inflation will result in an increased level of coercion.
HYPOTHESIS 4: Increasing divisions within the mling elite will challenge the
stability of the current regime and will result in increased level of coercion.
HYPOTHESIS 5: The effect of rising inflation on the use of coercion is expected
16


to be negatively conditioned by the presence of Western linkages.
HYPOTHESIS 6: Western linkages will negatively condition the use of coercion
when a regime is faced with economic decline.
HYPOTHESIS 7: When a regime is faced with the challenge of terrorist acts
linkages to the West will negatively condition the use of coercion.
HYPOTHEISIS 8: The use of coercion when a regime is challenged with elite
defection will be negatively conditioned by the presence of Western linkages.
HYPOTHESIS 9: Western linkages, when a regime is challenged by protests,
will negatively condition the use of coercion.
For this study the dependent variable will be the level of coercion or repression
that a regime implements, which will be broken up into high intensity and low intensity
forms of coercion. The first hypothesis will test the types of vertical challenges discussed
earlier. These types of problems will come from the people and can challenge the
stability of the regime. The second and third hypotheses test the economic challenges that
can manifest themselves into vertical challenges. And the fourth hypothesis will examine
the horizontal challenges that a regime may face that come from the ailing elite within
the regime or the opposition parties or leaders. Regarding the international dimension, as
the previous chapter briefly discussed, these types of challenges to a regime are not
direct, instead they have subtle effects that alter the expectations and, theoretically, would
mediate a regimes use of repression or coercion. Because this is not a direct relationship
these factors will be tested for their conditional effects in the responses of these regimes.
These conditional relationships will be tested using the fifth thru ninth hypotheses.
17


Measurement
This study will cover Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan during
the period from 1993-2010. These cases and years produce a sample size of 72. Though
all of these countries became independent from the Soviet Union in the final days of
19911992 was omitted because the turmoil during this year led to a lack of reliable data.
The use of repression or coercion will be the dependent variable in this study.
And in light of the discussion presented by Levitsky and Way (2010), this concept must
be divided into two different types of coercion, high intensity and low intensity. This
keeps these two forms of coercion conceptually separate, which is useful because in some
instances a regime may use one form and not the other. This distinction is important for a
broader understanding of the circumstances in which these regimes use particular types of
repression. For the purpose of this study high intensity coercion is considered the more
violent and visible form of coercion, whereas low intensity involves more subtle
techniques that remove the liberties and rights of the people in order to minimize their
possible oppositional power. The dependent variables are based off of the ratings
produced in the 'CIRI Human Rights Data set5 (Cingranelli and Richards 2010). This
dataset looks at nations using various factors that contribute to the level of human rights
in the nation. These ratings are scaled from 0 to 2 for each component, with 0 indicating
higher levels of the repressive technique being used and 2 being little to no instances.
These ratings give a higher value to the lower levels of repression. Because the direction
of these ratings is negative, for the purpose of this study the ratings that make up the two
dependent variables have been reversed by multiplying them by -1,thus making a higher
rating indicate a higher level of repression or coercion.
18


The measure for high intensity coercion is constructed from the ratings for
extrajudicial killings and torture as presented in the dataset. The measure for low
intensity coercion is comprised of the ratings for freedom of assembly and association,
foreign movement, domestic movement, speech, electoral self-determination, religion,
workers rightsand the independence of the judiciary. Because there are a different
number of components included in each of the dependent variables when they are
combined (resulting in a high intensity range of 0-4 and low intensity range of 0-16) the
coefficients produced should not be directly compared, but rather understood with regard
to their specific range.
The first and the fourth hypotheses selected for this study create independent
variables that lend themselves to a panel design because they have many components that
should be included. The first hypothesis predicts a relationship between coercion and
problems that arise in society, which is a fairly broad statement that requires a multiple
component understanding of these types of challenges. For this study, problems that may
arise in society that can challenge the stability and legitimacy of a regime include
protests, terrorist acts and ethnic violence. The previous chapter discussed the important
role that protests can have in removing dictatorships and disrupting a regimeso it is of
paramount importance that this type of challenge be tested in relationship to the possible
types of repressive responses that may come from the regime in power. Though terrorist
actions have not been studied to the same extent as the influence of protests, the same
rationale seems to apply: discontent within society can spread to the point that the regime
no longer can claim it is able to represent the people or protect them. And finally,
instances of ethnic violence can be seen as problems where a regime, as Gerschewski
19


(2013) might put it, has lost legitimacy because they are unable to keep up their promises
to ensure the security of the people.
The fourth hypothesis proposes that there is a relationship between elite divisions
and the level of repression used by the government. As the previous discussion showed,
some regimes may face challenges that can come from people within the government that
are close to the leader, and other horizontal challenges that may come from the
opposition that maintains some level of power (Gandhi and Przeworski 2007). It is
particularly difficult to measure the level of division that can exist at a given time within
a regime, mainly because these types of problems are not nearly as visible as societal or
vertical challenges. That being said, there are some events that could signal these types of
divisions. For the purpose of this study there are five observable events that may occur
which might be signs that there are broadening elite divisions occurring. These include:
arresting or charging individuals within the government or the opposition, violent acts
perpetrated against members of the elite or the opposition (these can be with or without
ties to the government), members of the elite publicly criticizing the regime, high ranking
members of the government defecting, or the creation of new parties or opposition blocs.
The first four indicate that there are problems within these groups that may be
challenging the regimes control. The final component is slightly different because it
measures the growth of opposition voices within the nation, which could, given enough
power, become enticing places to defect to for members of the elite if the regime appears
to be weakening.
Measuring concepts such as the levels of protest, terrorism, or elite divisions is
primarily difficult because of the lack of available research and data from these Central
20


Asian nations. That being said a process of document analysis was selected for the
components for the first and the fourth hypotheses. Using Keesings World News
archive, each of the countries names were searched and all relevant returns were read and
coded according to their content regarding the component of the variables that they
included (Keesings Record of World Events 2010). The totals were then collected for
each year and weighted on a 0-4 scale, with 0 indicating that there were no instances of
the event, whether it be a protest or an instance of elite criticism, occurring and 4
meaning that they were comparatively frequent. This technique of document analysis was
used for each of the components of the first and the fourth hypothesis. Although
document analysis is not always the most reliable way to generate measurements for
concepts, it was really the only way to examine these types of events in these countries.
There are two reasons that using this form of document analysis in this case was a
better option than all other available choices. One of the major issues with researching
these nations is that there is a very limited pool of data. Many of the indicators used in
political science are just not available consistently for these countries. The second, and
most important reason for choosing document analysis is that it allows for concepts that
are difficult to measure to be examined, such as the components that might indicate elite
divisions. It would be difficult and possibly fraught with mistakes to measure the
divisions within the elite in another way than through the possible signs that might
indicate that there are fractures emerging within the ailing class in these nations. It is
particularly difficult to measure these divisions without tracing the paths of all of the elite
in each of these cases. Other studies have tried to follow the elite movement, but this is a
particularly difficult process in nations, such as those selected for this research, because
21


even though a member of the ailing elite might leave the government this does not
necessarily mean that they have lost favor with the regime, often it is quite the contrary.
Jonathan Murphy (2006), in his study of the elite in Kazakhstan from 1989-2002, finds
that although many of the elite left the government for positions within the private sector,
they eventually returned, indicating that their departure was more of a reshuffle because
they had earned favor with the regime, rather then an indication that there were elite
divisions. The findings presented in this piece point out how difficult it is to trace the
behavior of the elite and that even though a member of the ailing class might leave the
government this does not necessarily mean that this is a sign of elite division. It would be
impossible to follow all of the members of the elite for the two decades at issue in this
study, not only because it would be an enormous task, but because the available
information about these individuals is very limited and could lead to many errors in the
conclusions.
The second and third hypotheses, the influence of economic decline and inflation
in relation to the use of coercion, are much easier concepts to measure. To measure the
economic decline in the second hypothesis the gross national income ($US per capita)
was used from the database created by the United Nations Economic and Social
Commission for Asia and the Pacific (UNESCAP). The thought behind this hypothesis is
that the greater the impact on the people as their economic wellbeing diminishes, the
greater the challenge will be to the legitimacy of the government, to which the regime
must respond. Thus the magnitude of the economic decline is important. To capture this
the percent change from the year previous was produced and these figures were lagged
one year to allow for the impact of the economic downturn to reach society. In order to
22


measure the level of decline and to make the coefficients more easily interpretable these
data was reversed by multiplying all of the figures by -1.The third hypothesis, regarding
inflationis based on the same logic discussed in the literature review. To measure this
conceptthe indicator from the World Bank for inflation as GDP deflator (annual %) was
used (World Bank). These data were also lagged one year, as to be able to measure the
societal pressure, rather than the immediate reaction.
As the previous chapter explained, challenges to an authoritarian regime can also
come from the international community through the emergence of Western Linkages. As
Levitsky and Way (2010) explain the level of linkages and leverage experienced by a
nation has an impact on the nations behavior and the pressure to democratize. Leverage
and linkage work together, but linkages are centrally important. Though both of these
types of influence could be examined, this study will concentrate on the influence of
linkages. This was selected because the theories presented by Levitsky and Way (2010)
indicate that leverage without linkage is ineffective. That being said even if there were
years within these nations in which leverage was high, without linkage the leverage
would theoretically be futile.
Because authoritarian regimes are not constantly repressive, rather they are
repressive because they are challenged, this study will explore whether international
linkages may play a meditating role in the behavior of an authoritarian regime. Although
the relationship between coercion and linkages could be tested directly, this work
recognizes that these regimes responsive, rather than maintaining a static level of
repression, thus the influence of Western linkages is thought to condition the responses to
specific types of challenges. In other words a regime may have low linkages and low
23


repression, but also no challenges to the regime, which does not prove that this theory
presented by Levitsky and Way (2006) is incorrect, rather it shows that there must be
challenges to a regime to actually test the influence of the Western linkages. For that
reason this study will apply the theory that linkages have a conditioning effect on the
behavior of authoritarian regimes. Using this concept as an interaction term, rather than
an independent variable allows for the conditional effects of increased linkages with the
West to be explored. The use of interaction terms have been argued by many, that when
used correctly, these types of models can explain more complicated relationships
(Jaccard, Turrisi, and Wan 1990, Brambor, Clark, and Golder 2006). In order to make the
coefficients more meaningful in relation to each other, each of variables and interaction
terms used in the all of the models, excluding the first, have been centered (Jaccard,
Turrisiand Wan 1990, BramborClarkand Golder 2006). Hypotheses five thru nine
show the five challenges that have been selected in this study to test the influence of
Western linkages, which include: inflation, economic decline, terrorism, elite defection,
and protests. Unfortunately, all of the other types of challenge components could not be
tested in this matter because the study size limited the ability for such a lengthy process.
Using the five different types of linkages explained by Levitsky and Way (2006)
this study will also explore what types of linkages produce conditioned responses in the
use of coercion from these regimes. Though these measures could be combined to make a
composite score for linkages, keeping them separate helps to keep the results distinct and
find a possible hierarchy in types of linkages. The first type of linkage explored is
economic. To measure this concept, this study uses the exports and imports between each
nation and the United States, which is obtained through the US census website (United
24


States Census). The exports and imports were combined to show the amount of
commerce, without focusing on one of the particular aspects of it. The theory proposed
by Levitsky and Way (2006) does not focus on something like economic dependency, but
rather just the amount of interaction, so it is coherent to combine the imports and exports
within the theoretical framework.
The second linkage, geopolitical ties, is a much more difficult concept to measure.
These types of ties could be measured using treaties or agreements that the nations of
focus have decided to enter into with Western powers. Unfortunately, the data regarding
these types of interactions is difficult to find. What are available though are the dates that
these nations became signatories to United Nations treaties (UNTS Treaty Collection).
Though the UN is not rnn entirely by the West, it does promote more democratizing
ideals. By going through the dates on which the nations in this study became signatories
to different treaties within the UN, though not the perfect measure for geopolitical ties,
does capture the years in which these nations were attempting to appear more open to
participation in the international community. Becoming signatories to these treaties does
not necessarily prove that these nations are trying to democratize, but it does indicate that
they were interested in appearing to be cooperating with the international community.
The annual figures created to measure geopolitical linkage were obtained by doing a
participant search with the 'all matching5 specification selected (UNTS Treaty
Collection). The returns generated were reviewed and each treaty and the year in which
the nation became a signatory were collected. While this may not be the best measure for
geopolitical linkages, with the limitations to the available data, this is the closest and best
measure available for these nations.
25


The third type of linkage discussed by Levitsky and Way (2010) is social, that is
the flow of people across borders[which] includes migrationtourismrefugeesand
diaspora communities, as well as elite education in the West (Levitsky and Way 2006,
383). To capture this type of linkage this study uses the figures for immigrants admitted
to the United States by region and country of birth from the United States Department of
Homeland Securitys yearbooks of statistics (DHS).
The fourth linkage, communication, fortunately, is a much easier concept to
measure. It would be optimal for the data regarding telephone and Internet access to be
used, but the data for Internet users is incomplete, so it has been omitted (World Bank).
There were consistent figures available though for cell phones and telephones per 100
people from the World Bank (World Bank). To create a single annual figure the data was
combined resulting in a value for cell phones and telephone lines per 100 people in each
of these nations.
The final type of linkage, transnational civil society, is also one of the more
difficult concepts to measure. Levitsky and Way (2006) explain that this concept includes
local ties to western-based nongovernmental organizationsreligious groupsand party
organizations (384). To measure this concept this study uses the World Bank data on net
official development assistance and official aid received (current $US) (World Bank).
These forms of aid that are received from outside governments are generally for the
purposes of promoting democratization programs on the ground in these nations. Though
tms is not the perfect measure it is similar to the concept presented by Levitsky and Way
(2006), while not being a direct measure for aid, which they would consider a type of
leverageas discussed in the literature review (Levitsky and Way 2006).
26


Though some of these measures may not fit the concepts perfectly, they are the
best options because the data from this region are so limited. The following discussion
will explain the specific models that were used in this research.
Models
This study, because each of the concepts covered in the hypotheses have multiple
components, will utilize a panel design. To explore the relationship between the levels of
coercion and the different types of challenges that a regime may face, the first model will
compare the two dependent variables, high and low intensity coercion, and each of the
components that for the first four hypotheses. The first model will test the direct
relationship between the dependent variables and the levels of terrorism, ethnic violence,
protest, economic downturn, inflation, arrests, violence, public criticism, new opposition,
and defection. The next five models will test the fifth thru ninth hypotheses using
interaction terms for the level of Western linkage during periods of inflation, economic
downturn, terrorism, defection and protest. Each of the models that include the
interaction terms all of the variables will be centered and will test both of the dependent
variables.
27


CHAPTER IV
RESULTS
This chapter will review the models used in this research that directly test the
relationship between the dependent variables of coercion and the nine hypotheses and
also will explore the conditioning relationships between Western linkages and various
challenges. These results will expose some of the possible patterns between the models
that will then be discussed in the following chapter. Although many of the findings in
these models do not appear to conform to the expectations, they do expose some
interesting trends that will become apparent in the following review. From this discussion
it will become clear that the type of challenge that a regime faces changes how Western
linkages influence the behavior of these regimes in Central Asia. Dependent on whether
the challenge is socio-economic or from the people or the elite can change how particular
types of Western linkage influence the behavior of these regimes. Additionally, although
this research is primarily interested in exploring how Western linkages condition the
coercive responses of these regimes when faced with challenges to legitimacy, these
models will show that in many cases Western linkages alone help to increase the use of
low intensity coercion.
Model1:Coercion and Horizontal, Vertical and Socio-Economic Challenges
The first model shows that specific types of challenges can result in different
types of behavior from the regimes of Kazakhstan, Kyrgyzstan, Tajikistan and
Uzbekistan. The first hypothesis, that challenges that arise from society will result in an
increase in the use of coercion from these regimes, produces some interesting results. The
findings show that when there are instances of terrorism and ethnic violence the
28


government will increase violent forms of coercion. These results estimate that for every
three terrorist acts or instances of large-scale ethnic violence in these nations there is an
expected one-point increase in the level of high intensity coercion, which on a -4-0 scale
is dramatic. The relationship between the instances of protest and high intensity coercion,
is very close to being statistically significant (.12), but has a negative impact on the use of
this more violent form of repression from these regimes. These findings show that when
these regimes face terrorist acts or there is an outbreak of ethnic violence that they will
respond with an increase in the use of high intensity coercion, but when there are protests
this form of repression decreases. Though this is not a significant relationship, it is
interesting to note that the type of vertical challenge produces different reactions from
these regimes.
Regarding low intensity coercion, the model results provide no support for the
first hypothesis. This indicates that there does not appear to be a direct relationship
among protests, terrorism, and ethnic violence and low intensity coercion. Though it
might be assumed that there would be a restriction of the freedoms included in low
intensity repression, such as freedom of the press or speech, in the face of vertical
challenges this does not appear to be the case in these Central Asian nations.
No evidence is found in support of the second hypothesis, that economic
downturn will increase the use of coercion, as it pertains to the use of high intensity
coercion. However, a strong relationship with low intensity coercion is produced, though
it is not in the predicted direction. The negative coefficient indicates that as the economy
declines the use of low intensity coercion actually decreases. It was hypothesized that
when a regime is challenged with the instability created by economic decline this
29


problem could translate into dissatisfaction from the people and in response to this type
of challenge a regime will use repression to maintain control over the population. The
results from this model indicate that these regimes react oppositely and actually decrease
the level of low intensity coercion in the face of economic decline.
Table 1:Model1 Coercion and Challenges
High Intensity Coercion Low Intensity Coercion
HI: Protest -.325 (.21) -.010(40)
HI: Terrorism .345*** (.05) .124 (.18)
HI: Ethnic 370** (.18) -.069 (.59)

H2: Economic Decline -.064 (.62) -2.876*** (.06)

H3: Inflation -***(00) -2* (00)

H4: Arrests -.045 (.14) -569** (.28)
H4: Violence .257 (.22) -.323 (.22)
H4: Public Criticism -.143 (.13) -.178 (.28)
H4: New Opposition -.1178 (.15) .109 (.17)
H4: Defection -154* (.08) -.388 (.34)

N 72 72
R-squared 0.367 0.301
*p<.10
**p<.05
***p<01
The third hypothesis, which proposes that the challenge of inflation, because it
may destabilize society, will increase the use of repression is also tested. The results
indicate that there is a relationship between inflation and levels of coercion, but one that
30


is negative. Both dependent variables are statistically significant, but indicate that when
inflation increases these regimes actually decrease the use of violent repression and low
intensity coercion. Though it was theorized that as inflation increases there will be a
corresponding increase in the possible vertical challenges in a nation, and thus the regime
will choose to implement coercive techniques to maintain stability, this does not appear
to be the case.
The fourth hypothesis, which states that elite divisions will lead to increased
levels of repression or coercion, also did not produce the expected relationships. Most of
the components used to measure the level of elite division did not produce significant
findings, with two exceptions. Surprisingly, this model shows that when defection of the
elite increases there is actually a decrease in the use of violent forms of repression. For
instance in 2004 Kyrgyzstan experienced a very high level of defection, but there was
little violent repression. Even though important members of the ailing party, such as
Kurmanbek Bakiev, left the governing party and established new voting blocs to oppose
the current regime of Askar Akayev, this did not translate into an increase in the use of
more visible forms of repression in Kyrgyzstan. These defections in 2004 show the elite
movement away from Akayev5 s regime, who was ultimately overthrown the next year
and replaced by a government headed by the defector Bakiev. Though it would be
expected that when a regime is losing legitimacy to such a degree that important elite
leaders are defecting, the regime would use both forms of coercion to maintain power,
this model and example find the opposite relationship.
Additionally, this model also finds that low intensity coercion also decreases with
arrests of elite and opposition. This shows that as arrests of leadership and opposition
31


increase there will be a decrease in the use of the non-violent forms of coercion, such as
restricting freedoms of speech, press, or the judiciary as a few examples. These findings
indicate that there is not the expected reaction from these regimes in their use of coercion
when faced with these types of elite divisions.
In sum, this model shows that there are specific types of challenges like terrorist
acts and ethnic violence that do result in the hypothesized reactions from these regimes,
but the other variables show that challenges such as economic declines, rising inflation
and elite divisions, do not lead to increased levels of coercion. One of the most surprising
findings in this model was with the socio-economic variables of inflation and economic
decline, which show that when challenged by these types of problems, these regimes will
actually move away from violent repression and broaden liberties and freedoms, in most
cases. However, it is possible that these contrary findings are being skewed by the first
few years of independence in these nations.
After the collapse of the Soviet Union the newly independent nations in Central
Asia experienced economic instability and inflation at staggering levels. For instance
within the first few years of independence the Kyrgyz people experienced a drop in their
per capita GDP of 20 percent in 1994 compared to the previous year, while the Tajik
people saw almost a 30 percent drop in 1992. During this time there was an intense push
from the West for these nations to transition from communism to democracy. This
pressure in the face of staggering socio-economic challenges may explain the negative
findings for the first model. As the work by Gasiorowski (1995) discussed, timing
matters when understanding the effects of economic challenges, and this period in the
history of these nations may be unique, which may help to explain these negative
32


findings. To explore this possible explanation the first three years were dropped from the
model to test if these years were skewing the data in the negative direction.
Table 2: Inflation and Coercion
1993-2010 1996-2010
High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion
Inflation -.001(.00) -***( 003***( -4(

N 72 72 56 56
R-Squared .051 .132 .072 001
*=p<.l
**=p< .05
***=p< .01
****=p< .001
By removing the first three years the significant relationship between inflation
and for low intensity coercion does not hold, but the expected relationship between rising
inflation and an increase in high intensity coercion emerges. This shows that from 1996-
2010 when these nations are experiencing periods of high inflation that the regimes do
tend to resort to violent forms of repression in order to maintain control. This indicates
that the first three years were, to some degree, skewing the data in the case of inflation.
Although these years being removed exposed the expected relationship for inflation, the
removal of these years actually indicates the robustness of these findings for economic
decline.
As the third table shows, even with the removal of the first three years the
negative and significant findings that show that economic decline actually decreases the
use of low intensity coercion, though lower is still significant. Unfortunately these
findings together do not help to explain the results completely; rather they only further
33


indicate the complicated nature of the relationship between these types of challenges and
the regimes use of coercion. Althoughthe inflation findingswith the removal of the
first years of independence, do produce the predicted relationship, the economic decline
results are still difficult to explain. To make broad assertions about the possible meaning
of these findings would require further research in Central Asia and the passage of more
time to test whether these findings persist.
Table 3: Economic Decline and Coercion
1993-2010 1996-2010
High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion
Economic Decline -.345 (1.10) -3.410*** (1.29) -.662 (.66) -1.805** (.99)

N 72 72 56 56
R-Squared .002 .083 008 .012
*=p<.l
**=p< .05
***=p< .01
****=p< .001
To provide another perspective, the following models incorporate interaction
terms to test how linkages to the West can influence the behaviors of these regimes. As
discussed in the literature review, it is thought that when nations increase their linkages
with the West, there will be a normative pressure to reduce the use of repression. These
models will explore whether these nations in Central Asia conform to these theoretical
expectations. The second model tests both of the dependent variables against increased
levels of inflation with the five linkage components that include the levels of economic,
geopolitical, social, communication, and civil society ties to the West.
34


Model 2: Inflation, Western Linkages, and Repression
The second model, which tests the fifth hypothesis, indicates that there is a
conditional effect of Western linkages to a regimes use of coercion when faced with the
challenge of rising inflation. All of the p-values for the interaction terms, excluding
economic linkages for both dependent variables and communication linkages for low
intensity coercion, are significant. What is interesting about these findings though, is that
they each have a negative conditional effect on the dependent variable, except for
geopolitical linkages.
Table 4: Model 2 Inflation, Western Linkage and Repression
Impact of Inflation on Coercion with Linkages (Model 2)
High Intensity Coercion Low Intensity Coercion
H3: Inflation -.015*** (.00) - 8***(.00)
Link: Economic .001** (.00) (X)0 (00)
Link: Geopolitical -.028*** (.01) -.043 (.04)
Link: Social '(X)0 (00) ***(00)
Link: Communications -0.179*** (.01) -.058 (.06)
Link: Civil Society '(X)0 (00) (X)0 (00)

Inflation*Economic (X)0 ( '(X)0 (00)
Inflation*Geopolitical . 0***(.00) . 0***(.00)
Inflation*Social -00***(.00) -00***(.00)
Inflation*Communication -***(00) '(X)0 (00)
Inflation*Civil Society -00***(.00) -0* (00)

N 72 72
R-squared 0.356 0.538
*p<.10
**p<.05
***p<01
35


Interpreting the interaction terms, there are three expected and one unexpected
finding for the conditional influence of Western linkages and the use of coercion during
periods of inflation. This model shows that the interaction between inflationsocial and
civil society linkages helps to decrease both forms of coercion in these nations.
Additionally, the results show that communication linkages help to decrease the use of
violent repression while inflation increases. Each of these interaction terms shows a
negative relationship to the dependent variables, which is what was expected. In other
words, when there are communication, social, and civil society linkages their interaction
with inflation puts pressure on these regimes, to which they will respond with less violent
repression and a loosing of restrictions on personal freedoms.
Looking at the Kazakhstan in 1993 confirms these findings. In this year, this
nation was experiencing extraordinary levels of inflation, though accompanied a high
amount of civil society linkages, which together produced no instances or high intensity
coercion. This example indicates the conditioning effect of the civil society linkages, that
the above table shows, that helped to deter the regime of Nursultan Nazarbayev away
from more violent forms of repression.
The unexpected finding in this model is the positive impact of geopolitical
linkages on the use of coercion. This finding indicates that when a regime has increased
levels of geopolitical linkages and is facing rising inflation the leadership will increase
the use of both subtle and violent forms of repression
This model also identifies a significant relationship between the inflation variable
and both types of coercion in the absence of linkages. These findings show that when
36


there is only inflation, and Western linkages are missing, these regimes will move away
from using both types of coercion.
Though this study is interested in the interaction between Western linkages when
these regimes are faced with challenges, these models also automatically measure the
influence of these ties when the specific challenges are absent. Even though it seems
contrary to the intent of this study to include a discussion of these findings, because they
are so often significant within these models and they also produce interesting trends, they
merit discussion. The coefficient for geopolitical and communication linkages in the case
of high intensity coercion also shows that these types of linkages when inflation is zero
will decrease violent repression in these countries. Of course this is the expected
relationship, because as discussed before it is thought that as linkages increase, violent
forms of repression will become too costly for these regimes to use. Conversely, this
model finds that when there is no inflation economic linkages increase the use of violent
forms of repression. Additionally, this model finds that when inflation is zero, social
linkages increase the use of low intensity forms of coercion. These findings show that
there are some types of linkages that reduce the use of the different types of coercion, but
others that do not. The following models will continue to expose the complicated nature
of these relationships.
Model 3: Economic Decline, Western Linkages, and Repression
The sixth hypothesis proposes that when a regime is faced with economic decline
that Western linkages will decrease the use of repression in these regimes. The third
model tests the influence of these linkages on the use of both forms of coercion when a
regime is faced with economic decline. The coefficients for the variables show an
37


unexpected relationship for most of the linkages when a regime is not experiencing a
period of economic hardship. This model shows that when these regimes are not
experiencing periods of decreased economic performance, economic, social and civil
society linkages increase the use of low intensity coercion. Interestingly, this model
seems to indicate that the less visible forms of coercion may be preferable when Western
linkages increase, even when the regime is not experiencing any problems.
Table 5: Model 3 Economic Decline, Western Linkages, and Repression
Impact of Economic Decline on Coercion with Linkages (Model 3)
High Intensity Coercion Low Intensity Coercion
H2: Economic Decline -1.125 (.82) -1.003 (1.06)
Link: Economic -0 ( ***(00)
Link: Geopolitical 030*** (.01) -077* (.04)
Link: Social '(X)0 (00) ***(00)
Link: Communications -.002 (.02) -.026 (.03)
Link: Civil Society '(X)0 (00) 0* (00)

Econdown*Economic .002 (.00) .004 (.00)
Econdown*Geopolitical 261** (.11) .150 (.33)
Econdown*Social -.001(.00) '(X)0 (00)
Econdown*Communication -.055* (.03) -.156 (.11)
Econdown*Civil Society . 0***(.00) (X)0 (00)

N 72 72
R-squared 0.217 0.488
*p<.10
**p<.05
***p<01
38


Additionally, this model finds that when the economy is not declining,
geopolitical linkages increase the use of high intensity coercion. The only expected
finding in the linkage variables shows that when there is no economic decline
geopolitical linkages will decrease the use of the more subtle forms of repression in these
countries. Like the model that tested inflation, this model shows that when the economies
in these nations are not experiencing economic hardships, economic, social, and civil
society linkages actually appear to increase restrictions on civil liberties and freedoms,
which as the following discussion will show is a pattern between each of the models that
explore Western linkages.
In this model, only one interaction term has a significant coefficient estimate for
the dependent variable of high intensity coercion. Communication linkages in periods of
economic decline condition the reaction of these regimes to decrease their use of violent
repression. Surprisingly, there are significant, but positive findings for the conditional
effect of geopolitical and civil society linkages on the use of high intensity coercion
though. These results show that economic decline interacts with geopolitical and civil
society linkages in a way that actually increases the use of the more brutal forms of
repression. Interestingly, these findings show that in some cases certain forms of linkage
may actually increase the chances that these regimes will use coercion. Altogether these
findings show that when these regimes are experiencing economic hardships
communication linkages will decrease the use of high intensity coercion, but that civil
society and geopolitical linkages can increase violent repression.
Because the second and third models examined the socio-economic variables that
produce such contrary results in the first model, the same process of dropping the first
39


three years was conducted to see how the omission of these turbulent years changes the
results. As the following discussion will show, dropping the first years of independence
in these nations helps to expose some more interesting trends in the data.
Table 6: Inflation, Western Linkages, and Repression 1996-2010
1993-2010 1996-2010
High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion
Inflation -.015*** (.00) - 8***(.00) -(01) -.008 (.01)

Link: Economic .001** (.00) 03 (00) .008 (.01) .020 (.01)
Link: Geopolitical -.028*** (.01) -.043 (.04) -.242 (.18) -875*** (.26)
Link: Social '(X)0 (00) . 0***(.00) -00 ( -.004 (.00)
Link: Communications -179*** (.01) -.058 (.06) -.396** (.19) -.684 (.50)
Link: Civil Society '(X)0 (00) (X)0 (00) (X)0**(.00) (X)0**(.00)

Inflation*Economic (X)0 (00) '(X)0 (00) (X)0 (00) (X)0 (
Inflation*Geopolitical . 0***(.00) . 0***(.00) -.001(.00) -.005** (.00)
Inflation*Social -00***(.00) -00***(.00) '(X)0 (00) '(X)0**(.00)
Inflation*Communication -***(00) '(X)0 (00) -.002** (.00) -.004 (.00)
Inflation*Civil Society -00***(.00) -0* (00) (X)0**(.00) (X)0**(.00)

N 72 72 56 56
R-Squared 0.356 0.538 .246 .592
*=p<.l
**=p< .05
***=p< .01
****=p< .001
40


Table 7: Economic Decline, Western Linkages, and Repression 1996-2010
1993-2010 1996-2010
High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion
Economic Decline -1.043 (.83) .848 (.99) -1.452* (.88) .885 (2.44)

Link: Economic (X)0 (00) ***(00) .003 (.00) .001(.00)
Link: Geopolitical -.030*** (.00) -077* (.04) .027* (.02) -.032 (.08)
Link: Social (X)0 (00) ***(00) '(X)0 (00) 001*** (
Link: Communications .002 (.00) -.026 (.03) -.025 (.03) .008 (.04)
Link: Civil Society (X)0 (00) 00*(00) (X)0 (00) '(X)0 (00)

Econdown*Economic .002 (.00) -.004 (.00) *(00) (X)0 (00)
Econdown*Geopolitical 261** (.11) -.150 (.33) .127 (.12) .529 (.61)
Econdown*Social -.001(.00) -.001(.00) -.002 (.00) .002 (.00)
Econdown*Communicati on -055* (.03) .156 (.11) -096** (.04) -.076 (.15)
Econdown*Civil Society . 0***(.00) '(X)0 (00) (X)0**(.00) (X)0 (00)

N 72 72 56 56
R-Squared 0.217 0.488 .012 .518
*=p<.l
**=p< .05
***=p< .01
****=p< .001
When the first three years are dropped the negative relationship between civil
society linkages and inflation on high and low intensity coercion reverses, but remains
positive in the case of economic decline. In other words, the 1996-2010 model shows that
civil society linkages interact with the challenge of inflation and actually increase the use
of repression in these nations. The removal of these years also confirms that the presence
of communication linkages negatively conditions the violent responses of these regimes.
41


This shows that when there are socio-economic challenges, communication linkages help
to decrease the use of coercion. The models for 1996-2010 also show a positive
conditional relationship between economic decline and economic linkages with high
intensity coercion. This means that when there are economic linkages and these
economies are declining, this model predicts that these regimes will increase their use of
more bmtal forms of repression. From all of this, it appears that socio-economic
challenges increase the use of coercion when there are civil society linkages, but decrease
the use of high intensity coercion with communication and social linkages.
These findings seem to indicate that the type of challenge changes the effect of
Western linkages, considering these two models share several important similarities in
their results. The following discussion will show that the models that test the socio-
economic factors that can destabilize a regime, inflation and economic decline, produce
completely opposite effects to those that test domestic challenges, such as terrorist acts,
defection and protest. There seem to be distinct differences between how a regime
responds while in periods socio-economic challenge that changes the effect of specific
types of Western linkages. The findings from the following models show that the type of
challenge, whether socio-economic or domestic, changes which types of linkage increase
or decrease the use of repression in these cases.
Model 4: Terrorism, Western Linkages, and Repression
Based on the seventh hypothesis, the fourth model tests the conditional effects of
Western linkages on the use or high intensity and low intensity coercion when the
regimes in Central Asia have been challenged by acts of terrorism. This model shows
that when faced with terrorist acts, without linkages, these regimes respond with the use
42


of violent forms of repression. This finding is to be expected, but many of the others
produced by this model are not.
Table 8: Terrorism, Western Linkages, and Repression
Impact of Terrorist Acts on Coercion with Linkages (Model 4)
High Intensity Coercion Low Intensity Coercion
HI: Terrorism 575*** (.11) .189 (.45)
Link: Economic (X)0 (00) .001** (.00)
Link: Geopolitical -021* (.01) -.032 (.07)
Link: Social (X)0 (00) ***(00)
Link: Communications .016 (.02) -.018 (.04)
Link: Civil Society '(X)0 (00) (X)0 (00)

Terrorism*Economic -.001(.00) (X)0 (00)
Terrorism*Geopolitical -.010 (.00) 091* (.06)
Terrorism*Social (X)0**(.00) '(X)0**(.00)
Terrorism*Communication .032*** (.01) -.005 (.05)
Terrorism*Civil Society '(X)0**(.00) '(X)0 (00)

N 72 72
R-squared 0.383 0.514
*p<.10
**p<.05
***p<01
The coefficient estimates for economic and social linkages when tested against
low intensity coercion are both positive and significant, which indicate that if there are no
terrorist acts these linkages actually have a positive impact on the use of subtle forms of
repression. This is similar to positive relationship found for low intensity coercion in the
43


previous models, with the exception of geopolitical linkages in the case of economic
decline. This model also shows that geopolitical linkages, when there are no instances of
terrorism, decrease the use of violent repression.
The interaction terms in this model produce more surprising findings. Out of the
five significant interactions using both of the dependent variables, three of these
conditional effects are in the positive direction, which is not what the theories would
predict. In the case of high intensity coercion a positive conditional relationship exists
between social linkages and terrorism, as well as between communication and terrorism.
This means that in these Central Asian nations, when these regimes are challenged
through terrorist acts, communication and social linkages actually increases the use of the
more violent a visible forms of coercion. The relationship here is positive, unlike the
previous models that looked at inflation and economic decline, which found that social
and communication linkages decrease the use of repression. This discrepancy can be
explained though.
In the context of the US-led War on Terror it may be expectable for these
regimes to respond by any means necessary when they are able to convincingly argue that
the challenge is from a terrorist group. Of course in many cases these regimes quickly
defend their more bmtal reactions on fighting terrorism, which may be a reason why
some Western linkages actually positively affect the use of coercion. In these cases it
may be seen as an acceptable excuse to the West.
Additionally, in the case of low intensity coercion there exists a positive
conditional relationship between terrorism and geopolitical ties. This indicates that this
form of linkage actually increases the likelihood of more subtle forms of coercion, such
44


as the restriction of freedoms, when a regime is faced with increased terrorism. Possible
explanations for all of these unexpected findings will be discussed in the following
chapter.
There are two interaction terms though, that produce the expected negative
conditional effects. This model finds that civil society linkages decrease the use of high
intensity forms of repression when these regimes are experiencing increased acts of
terrorism. This shows that when there is an increase in civil society linkages, such as
funding for democratization programs or NGOs on the ground, that even when a regime
faces terrorist actions these linkages help to decrease the use of violent repression. What
is important to understand about these findings is that with low levels of civil society
linkages these regimes still tend to continue to resort to high intensity forms of coercion.
For instance, in Tajikistan in 1997 there was great social unrest that was exhibited
through the high level of terrorist actions that year, including the kidnapping of 15 Red
Cross workers (Keesings41Feb41500). This year also experienced an increased level
of instances of torture and extrajudicial killings. This relationship is to be expected,
particularly with the knowledge that this year was also accompanied by very low,
comparatively, levels of civil society linkages.
The final expected finding for this model is that it shows, that when faced with
terrorist acts, social linkages can decrease the use of low intensity coercion. This
indicates that increases in social linkages, such as heightened levels of migration to the
US, will decrease the use of less intense forms of repression when a regime is challenged
by acts of terrorism.
45


This model, compared to the inflation and economic decline linkage models,
produced similar findings in that they each seem to show that linkages, when challenges
are absent, seem to increase the use of low intensity coercion. The findings contained in
this model though, stand apart from the previous models in that it shows that there is an
increase in the use of violent forms of repression when communication linkages and
terrorism are present, while the socio-economic models found that these linkages
decreased the use of coercion. This indicates that these types of linkages have the
opposite effect when these regimes are challenged by terrorist acts.
Model 5: Elite Defection, Western Linkages, and Repression
The eighth hypothesis predicts that repression will decrease in the face of elite
defection when there are increased linkages to Western nations. The fifth model tests the
conditional effect of these ties on the use of both types of coercion when the regime is
challenged horizontally by elites defecting from the leaders party. This model identifies
multiple factors that can influence these regimes use of coercionwhich further indicate
the complicated relationship that exists between different forms of linkages and how
these relationships translate into a regimes use of coercion. One of the most surprising
findings is that defection has a positive effect on the use of high intensity coercion, but a
negative impact on the use of low intensity forms of coercion when there are no linkages.
This shows that when these regimes are challenged in this way, their response is to use
more violent forms of repression rather than more subtle forms that restrict the freedoms
of the people. In fact when these regimes are experiencing an increase in elite defection
they increase their use of violent and visible repression while decreasing the use of low
intensity. Additionally, this model finds that in years when there is no elite defection,
46


economic, social, and civil society linkages actually increase the use of low intensity
coercion, which is similar to the previous linkage models. Just as in the second and fourth
models, the fifth model finds that geopolitical ties decrease the use of high intensity
coercion when defection is zero.
Table 9: Elite Defection, Western Linkages, and Repression
Impact of Defection on Coercion with Linkages (Model 5)
High Intensity Coercion Low Intensity Coercion
H4: Defection 241*** (.02) -536*** (.14)
Link: Economic (X)0 (00) .001** (.00)
Link: Geopolitical -056*** (.02) -.039 (.03)
Link: Social (X)0 (00) ***(00)
Link: Communications .006 (.02) -.0317 (.02)
Link: Civil Society '(X)0 (00) . 0***(.00)

Defect*Economic '(X)0 (00) '(X)0 (00)
Defect*Geopolitical -043*** (.01) .125*** (.02)
Defect*Social '(X)0 (00) '(X)0 (00)
Defect*Communication 030*** (.01) -.004 (.00)
Defect*Civil Society -00***(00) (X)0 (00)

N 72 72
R-squared 0.174 0.518
*p<.10
**p<.05
***p<01
47


The interaction terms show several unexpected positive conditional relationships.
Similar to the other interaction models discussed, these findings show that positive
conditional effect of geopolitical linkages and defection on the use of both forms of
repression. The previous models show, that geopolitical linkages actual increase the use
both types of coercion, in most cases, when confronted by the various challenges.
Additionally, this model identifies a positive conditional relationship with
communication and defection and the use or high intensity coercion. This means when
experiencing increasing levels of defection, higher communication ties will actually
increase the use of violent repressive techniques. Just like the model that looked at how
regimes react to terrorism, this model shows Western communication linkages, when
faced with defection, actually increases the use of violent repression. These findings
indicate that, contrary to what was expected, communication linkages may be viewed by
these regimes as liabilities to which, when challenged domestically, these Central Asian
nations react violently.
In this model there is one interaction term with a significant and negative effect.
This model finds that the interaction between defection and civil society linkages
decreases the use of high intensity coercion. Similar to the model that tested terrorism,
though it is the opposite when compared to the socio-economic models after 1996. These
findings suggest an emerging trend between the models: one that indicates that socio-
economic challenges and domestic challenges, such as terrorism or in this case defection,
have dramatically different effects on the influence of Western linkages. The other
important trend in all of theses linkage models discussed thus far show that linkages
when the specific challenges are absent increase the use of lower intensity forms of
48


repression. Although this study intended to explore how Western linkages condition the
repressive responses of these regimes when they are challenged, because these models
also inadvertently tests the lack of challenge as well, these findings must also be
discussed particularly because they expose some interesting trends.
Model 6: Protests, Western Linkages, and Repression
The final hypothesis in this study states that Western linkages when regimes are
faced with protests will decrease the use of repression. The sixth and final model tests
tms idea, but finds that only economic linkages produce the predicted decrease in the use
of coercion.
Table 10: Protests, Western Linkages, and Repression
Impact of Protest on Coercion with Linkages (Model 6)
High Intensity Coercion Low Intensity Coercion
HI: Protest -.263 (.21) .060 (.20)
Link: Economic (X)0 (00) (X)0 (00)
Link: Geopolitical -.022 (.02) -.0356 (.02)
Link: Social (X)0 (00) ***(00)
Link: Communications .006 (.02) -.005 (.02)
Link: Civil Society '(X)0 (00) (X)0**(.00)

Protest*Economic -.001(.00) -***(00)
Protest*Geopolitical .016 (.07) .114*** (.03)
Protest*Social . 0***(.00) ***(00)
Protest*Communi cation -.003 (.01) .004 (.01)
Protest*Civil Society (X)0 (00) '(X)0 (00)

N 72 72
R-squared 0.284 0.566
*p<.10
**p<.05
***p<01
49


This model finds two significant linkages when tested against the dependent
variable of low intensity coercion. According to the models estimateswhen there are no
protests civil society and social linkages increase the use of less visible forms of
coercion. This means that when there are no vertical challenges, that come in the form of
popular protests, these types of linkages tend to result in a restriction of freedoms and
liberties of the people. This model, just like all of the other linkage models, for the most
part, shows that increased linkages in these nations increase the use of low intensity
coercion.
This model also produces four significant interaction terms, three of which impact
the use of low intensity repression and one that influences the use of high intensity
coercion. There is a positive conditional effect on the use of low intensity coercion with
the interaction between protest and geopolitical ties. Which means that when there are
periods of increased protest, higher levels of geopolitical ties in these nations will
increase the use of less violent forms of repression. Similarly, social linkages together
with protests also increase the use of both violent and subtle forms of repression in these
nations. This model, interestingly, finds a negative conditional effect of economic
linkages and protest on the use of low intensity coercion. Unlike any of the previous
models, this shows that when faced with protests, economic linkages actually deter these
regimes from restricting the rights of the people through low intensity repression. The
increased use of coercion when social linkages and protest work together also mimics the
findings in the models that tested terrorism and defection.
50


Table 11:Models 2-6 Linkage Interaction Terms
Linkage Interaction Terms
High Intensity
Coercion
Challenge*
Economic
Challenge*
Social
Challenge*
Communication
Challenge*
Civil Society
Low Intensity
Coercion
Model 2 Model 3
(1996-2010) (1996-2010)
Challenge*
Economic
Challenge*
Social -
Challenge*
Communication
Challenge*
Civil Society
Model 4
Expected
Results
Unexpected
Results
No
Relationship
-decreased
coercion
+ increased
coercion
Together all of these findings further confirm that when these countries are
challenged domestically (through terrorist acts, protest, or defection) civil society
51


linkages generally decrease the use of repression, while social and communication
linkages increase the use of coercion. Conversely, the models that test socio-economic
challenges after 1996, find that social and communication linkages decrease the use of
violent and subtle forms of repression, while civil society linkages increase the use of
repression in these Central Asian nations. All of the linkage models together produce
three important trends: in most cases linkages without challenges increase low intensity
coercion, geopolitical linkages have a positive conditional relationship to all of the
different types of challenges, and that the effect of linkages depends on whether the
regime is experiencing socio-economic problems or challenges from society or the elite.
52


CHAPTER V
DISCUSSION
The previous chapter presented the results from the six models in this study.
There are several important findings produced by these models that require further
discussion because, for the most part, many of the results were contrary to expectations.
The following chapter will propose explanations for the surprising findings from the first
model and the models that tested Western linkages. This discussion will focus on the
unexpected findings produced in the first model and the three important patterns that
emerge from the five models that test the influence of Western linkages. After a brief
review of the first model, the three major trends in the linkage models will be reviewed.
The linkage models show: that geopolitical linkages interact with the various challenges
in a way that tends to increase the use of coercion, that when there are Western linkages
without challenges these regimes tend to heighten the restrictions on the rights and
liberties of the citizens, and finally that some types of Western linkages can increase or
decrease the use of repression depending on whether a nation is being challenged socio-
economically or domestically.
The First Model
The first model, which did not test Western linkages, but rather tested how the
regimes in these Central Asian nations react to vertical, horizontal, and socio-economic
challenges exposed some unexpected findings that require some explanation. This model
found that when these regimes are being challenged by terrorist acts or through the
instability created by ethnic violence they will react violently to these problems. These
findings provide support for the proposed hypothesis that social challenges will increase
53


the use of coercion, though the third component to this concept, protests, did not increase
the use of force in these countries. This shows that these regimes find terrorist acts and
ethnic violence worthy enough challenges to react with more violent and forceful types of
repression in order to maintain stability. This reaction from these regimes would be
predicted, surprisingly though, the other concepts tested in this model did not conform to
the hypothesized expectations.
Table 12: Model1 Expected and Directional Results
Model1:
High Intensity Low Intensity
Coercion Coercion
Protest
Terrorism
Ethnic Violence
Economic Decline
Inflation
Arrests of Elite or
Opposition
Violence Against
Elite or Opposition
Public Criticism
from Elite or
Opposition
New Opposition or
Coalitions
Defection of Elite
+ increased
coercion
-decreased
Unexpected Results coercion
No Relationship
Expected Results
54


One of the most important findings in this study, while being one of the most
unexpected, was that it appears that when these regimes face economic problems they
actually increase freedoms and reduce their more brutal policies. It was theorized that
when a nation is facing economic hardships or increased inflation, because of fears of
social unrest, these regimes would increase their use of repression in order to maintain
control over their dissatisfied population. The results from this study show, however, that
tms is not the case in these nations; in fact it appears that these types of challenges
actually induce these regimes to loosen their grip on this type of control.
Because these findings were so contrary to expectations, it was proposed that
there was a possibility that the first few years for these nations may have skewed the data.
When the Soviet Union collapsed and these nations in Central Asia became independent
they were reeling economically, while being pressured by the West to institute
democratic systems. Dropping these first years show that inflation does increase the use
of violent forms of repression. Tms is important because the same expected findings do
not exist when the years 1993, 1994 andl995 are dropped when examining the influence
on economic decline. This same process only confirmed that economic declines decrease
the use of low intensity repression. This shows that during periods of increased inflation
these regimes do respond with the expected violent crackdowns, the same cannot be said
in the case of periods of economic decline. Surprisingly, these findings show that when
the economy is experiencing a downturn these regimes actually expand the rights and
freedoms of the people.
55


Table 13: Socio-Economic Hypotheses 1993-2010 vs. 1996-2010
Socio-Economic Hypotheses 1993-2010 vs. 1996-2010
1993-2010
1996-2010
High Intensity Low Intensity High Intensity Low Intensity
Coercion Coercion Coercion Coercion
Economic
Decline
Inflation
Expected
Results
Unexpected
Results
No Relationship
+ increased
coercion
-decreased
coercion
It would be expected that these regimes would react with violence when
challenged by the instability of high inflation. It makes sense that in order to maintain
control that this form of repression would be used to combat the possible dissatisfaction
created by this threat to the economic wellbeing of the people. It is a bit more difficult to
explain why a regime would broaden freedoms when the economy is bottoming out. It is
possible that these regimes may be willing to loosen their control over some aspects of
life and broaden the rights, liberties and freedoms in these nations in order to possibly
56


attract Western investment. These regimes may believe that appearing to uphold more
Western ideals of may help attract businesses, which when the economy is faltering
would be exactly what is needed. It would certainly help to soothe the civilian
dissatisfaction with economic hardships if there were new jobs and opportunities arising
in these nations. These findings indicate that these regimes may view inflation as a
greater danger to stability than economic decline. Though this distinction requires further
research, it is very interesting that they type of socio-economic challenge changes how
these regimes respond. Although this is one of the most interesting findings produced by
this model, the fourth hypothesis tested and its results must also be reviewed.
The elite divisions hypothesis also produced unexpected findings, although most
of the variables did not significantly affect coercion. This model found that when there is
an increase in arrests of the opposition or the elite that these regimes will actually
decrease the use of low intensity repression. This model also found that when the elite
begin defecting from the leaders party the regime will reduce violently repressive
measures. These are interesting findings because it was hypothesized that arrests would
indicate deepening elite divisions, which could challenge the regime in power, and in
order to combat this possible growing problem these regimes would increase in the use of
coercion. It is surprising that these finding indicate that these regimes actually loosen
their repressive grip even when it appears that the elite might be fracturing. Since these
two significant results are so unexpected, and many of the components in this hypothesis
did not produce any important results, possible explanations for these findings must be
discussed.
57


First, the data collection process through document analysis may not have
exposed all of the elite divisions that were present. As was discussed in the third chapter,
divisions emerging within the elite are not necessarily public knowledge and may occur
behind the scenes. That being saidthe instances that were found in the Keesings World
News reports may not have fully captured the instances of these types of divisions. The
other possibility is that there may be ways in which these regimes respond to horizontal
threats that is not through repression of the entire society, either through high intensity or
low intensity forms of coercion. These regimes could influence the elite through
promotions or demotions to maintain loyalty when challenged or punish or reward
specific groups, tribes, or families, which would not translate into a restriction of
freedoms or violent forms of repression on the ground. It is possible that there exists a
much more dynamic use of power within the elite power structure that was not captured
by the dependent variables in this study.
The first model indicates how dynamic the responses of authoritarian regimes
may actually be. These findings show that these regimes react with violence when
stability is challenged by inflation, but actually increase rights and freedoms when the
economy is declining. Though surprising, these results show that there are important
distinctions between the type of socio-economic challenge that drastically alters the
reactions of these regimes. The following discussion will also highlight how important
understanding the type of challenge is when predicting the reactions of these regimes,
which will be shown in each of the Western linkage models. As mentioned before, there
are three major trends that can be found across all of linkage models, which include: the
interaction of geopolitical linkages heighten the use of repression, linkages without
58


challenges increase the use of low intensity coercion, and that socio-economic and
domestic challenges change the influence of Western linkages.
Geopolitical Linkages
The first interesting trend in the interaction models shows that geopolitical
linkages, when a regime is challenged by inflation, economic decline, terrorism,
defection or protests, tend to increase the use of repression. In other words, when these
regimes are challenged the presence of geopolitical linkages actually induces these
regimes to increase their use of repression. This is most likely attributable to the fact that
the measure did not accurately capture the concept. The level of geopolitical linkages was
determined by counting the number of UN treaties that were signed in each year by the
Kyrgyzstan, Kazakhstan, Tajikistan and Uzbekistan. Though this was not the most
optimal way to measure this type of linkage, the limited information from the region
constrained the possibilities for other types of measure. This, however, may help to
explain the positive results for most of the interaction terms that included geopolitical
linkages.
It is possible that these nations signed these treaties in an attempt to appear to be
working with the international community, but that the intentions behind these gestures
(such as signing the Convention on the Political Rights of Women 44 years after it was
first ratified) was actually to distract from the high intensity and low intensity coercive
techniques that these regimes were using on the ground (Kyrgyzstan UNTS). In other
words, the regimes were using repression to control their people, and choose to sign on to
these treaties to signal to the international community that they were making an attempt
to work within the expected framework of norms, even though their use of coercion
59


during these periods was not what the West may have approved of. Unfortunately, this
measure may have failed because the intention of these regimes to enter into these treaties
was not assumed to be connected to their use of repression, which these finding indicate
that it might be.
Without Challenges, Low Intensity Coercion Increases
Table 14: Linkages Increase Low Intensity Repression when Challenges are Absent
Low Intensity Coercion when Challenges are Absent
Economic
Linkages
Social Linkages
Communication
Model 2 Model 3 Model 4 Model 5 Model 6
Linkages
Civil Society
Linkages
Expected
Results
Unexpected
Results
No
Relationship
-decreased
coercion
+ increased
coercion
The second interesting trend that emerges from the five Western linkage models
finds that when challenges are absent these regimes actually begin restricting the rights of
the people through low intensity coercive techniques. The purpose of this research was to
explore how Western linkages interact with challenges in these regimes, but these models
60


actually show how these linkages interact with the levels of repression when the specific
challenges are not present..
Table 14 shows that the more subtle forms of coercion actually increase with
Western linkages. These results indicate that Western influence does not necessarily
guarantee the removal of repression, as these linkages do not show a decrease of high
intensity coercion, but that it may have influenced these regimes toward using low
intensity forms of coercion. Though these are not decisive findings, as they do not show
that there may be a trend away from high intensity repression, they do show that there
seems to be an increase in less visible forms of repression with these linkages. One of the
ways to explain this would be to view these increases as a movement away from violent
repression because these visible actions may draw negative attention from the West.
These findings are not completely inconsistent with the previously discussed theories of
Levitsky and Way (2006) though. Linkages are thought to increase the risk of using
violent repression, because as linkages increase, the possibility that these bmtal
techniques will be noticed and condemned will be much higher, while low intensity
forms of coercion might be thought of as preferable because they are much more difficult
for Western nations to notice. This helps to explain why these regimes are opting for low
intensity repression while these Western linkages are being formed. This is an important
finding that also requires further study into the relationship between linkages and low
intensity coercion in authoritarian regimes.
Socio-Economic vs. Domestic Challenges
The final important trend this research exposes is that the type of problem that
challenges these regimes changes how Western linkages influence their behavior. There
61


appear to be two distinct types of challenges that alter the effect of some linkages. The
findings from the final five models show that if there is a socio-economic challenge that
communication and social linkages decrease the use of coercion, while when the
challenge is from the people or the elite, economic and civil society linkages can decrease
the use of repression. These models also produce findings that show that, depending on
the type of challenge, there are some linkages that increase the use of coercion. When
there are challenges from society and the elite, communication and social linkages
interact with these types of challenges, and these regimes respond with higher levels of
repression. Additionally, when the first three years are removed from the inflation and
economic decline models, civil society and economic linkages sometimes increase the
use of repression.
Challenges and Effect of
Linkages
Socio-Economic
Challenges
Domestic
Challenges
Y Communication
Y Social
Economic (after 1995)
A Civil Society (after 1995)
Y Economic
Y Civil Society
A Communication
A Social
Figure 1:Socio-Economic and Domestic Challenges and Linkages
62


Communication and social linkages seem to be particularly important in
understanding the responses from these regimes, both of which influence the regimes
differently depending on the type of problem that arises. Though it is proposed, within
the previously discussed theories of Levitsky and Way (2006), that increases in linkages
will help further democratization, these findings indicate that the type of linkage may be
considered by these regimes to be liabilities in specific situations. Because the interaction
between domestic challenges, communication and social linkages in these models
actually show an increased use of repression, it is possible that these regimes may be
responding to these types of linkages as if they are also threats as well.
Communication and social linkages seem to be liabilities to these regimes when
terrorist acts, defection, and protests arise, but when challenged by inflation and
economic decline these linkages decrease the use of coercion used by these regimes.
These two forms of linkages capture the level of information and connection between the
citizens of these nations and the West. These two types of linkages are thought to
influence regimes because they open up opportunities for people within these countries to
access information through communication, and promote the transfer of ideas across
borders, which may have a democratizing influence on the populous in these nations,
while also making these regimes vulnerable to international criticism. It is possible that
access to information and ideas can at times be considered a liability to these regimes, but
it depends on the type of information that could challenge the status quo. This possible
consideration may help to explain why these regimes have such varied reactions
depending on the type of challenge.
63


The models that look at challenges that arise domestically, whether from the
people or from the elite, show that in most cases the presence of communication and
social linkages in the face of these problems actually increase the use of violent
repression and some times low intensity forms as well. From these findings it is clear that
not only do these types of linkages not deter these regimes, when faced with domestic
challenges, but they actually increase the use of repressive techniques. This access to
information and migrants in Western countries may be viewed by these nations as
possible challenges to the authority of the regime. This may induce reactions from these
governments that utilize both forms of repression, because of fears of losing power. If a
regime were losing legitimacy, or felt that it might be, then it could consider access to
other ideas through communication and social ties to the West a liability to their
continued control. The particular types of ideas that might be of concern to these regimes
might challenge the level of democratization, freedoms, and human rights concerns, just
to name a few. Increased interaction with the West could allow the people in these
Central Asian nations to question the ideology of the regime and the leadership. Access
to these types of ideas, while these regimes are facing challenges from popular protests,
elite defection, or terrorist acts, could be seen as a source of vulnerability to their power,
thus exposure to different ideas that may challenge the regimes behavior or ideology
could be considered a liability and might induce these regimes to repress their people.
Conversely, these types of linkages would not be nearly as concerning to these
regimes when faced with inflation or economic decline, because the ideas that the people
in these nations may gain from access to the outside world would not challenge the
current economic systems in these countries. In other words, because these nations have
64


embraced free market capitalism, access to the West would not challenge this system,
since these nations have embraced the same economic forms. Though there are other
possible explanations to these findings there does seem to be different effects of
communication and social linkages depending on the type of challenge these regimes in
Central Asia face.
Civil society linkages to the West, as measured through the directed aid in these
nations that promote democratic policies, also has adverse effects depending on whether
the challenge is socio-economic or through a loss of support from the people or the elite.
The models that look at terrorist acts and elite defection show that civil society linkages
interact with the challenges and decrease the use of violent forms of coercion. This
indicates that targeted monies from the West will actually deter these regimes from using
violence against their people when they face issues within the ailing elite or because of
terrorist acts.
Using the years from 1996 to 2010 the socio-economic models show that civil
society linkages conditionally increase the use of both forms of coercion. Civil society
linkages do decrease the use of violent forms of repression in the case of protest or
defection, but the positive relationship that emerges after 1995 is difficult to comprehend.
It is possible that these regimes are pressured by democratization programs during
episodes of domestic challenge, as would be predicted, but that when the economy is
faltering these funds may disruptor be seen to disrupt societythus leading to increased
use of both forms of coercion. As the previous findings show, these regimes do take
domestic challenges seriously, if they did not they would not resort to repression as often,
so it is possible that these civil society linkages are actually deterring these regimes away
65


from repression as a way to control society. Because they react with repression in the face
of socio-economic challenges when there are civil society linkages, it seems as though in
these situations this form of aid may also be viewed by these regimes as a problem, or a
financial source of disruption in society.
The impact of economic linkages seems much less influential, compared to the
other forms of linkage, but these findings are also worth mentioning. The models show
that economic linkages, when they interact with the challenge of protest, actually reduced
these governments restriction of the rights and freedoms of the people. Which means
that when protests occur the presence of economic linkages decrease the use of low
intensity coercion. It is possible that tms is because as there are increased economic ties
between these nations and the West there will be a pressure for these regimes to expand
some of the freedoms that are normally restricted when a regime is attempting to control
through low intensity forms of coercion. This is similar to the explanation proposed
earlier for the decreased use of this form of repression when regimes were directly faced
with economic decline in the first model. Possibly these regimes, in an attempt to further
their attractiveness to Western investors, at times choose to loosen their repressive hold
on their people in order to further their economic prospects. This finding, because it does
not exist in any of the other models must be further explored before any broader
conclusions about the connection between these pressures be espoused.
What is interesting about this type of linkage though, is that when the first three
years after independence are removed and these regimes experience an economic decline,
economic linkages actually increase the use of violent forms of coercion. This shows that
economic ties, when there are economic challenges, increase the more negative aspects of
66


authoritarianism. It could be proposed that during periods of economic decline that
because of the possibility for social disruption it is better for these regimes to repress
through high intensity coercion in order to show Western investors that they have control,
rather then appear less like dictators and more like Presidents. In other words, it is more
important to appear to be a stable regime that can protect investments, rather then a weak
government that might fall because of the social unrest produced from an economic
decline. Again this is another finding that indicates that these regimes act very differently
depending on whether the challenge comes from the people or if it is socio-economic.
These findings do not confirm the proposed influence of Western linkages, but
they also do not entirely disprove the theories proposed by Levitsky and Way (2006).
Rather these findings indicate that the type of challenge changes the influence of the type
of linkage and that some linkages have more influence than others. The findings in this
study indicate that there may be a much more complicated relationship between the use
of coercion and specific types of challenges, that linkages to the West seem to increase
the use of low intensity coercion, and that some linkages can have positive and negative
effects depending on whether a regime is facing a socio-economic challenge or a
domestic problem. From the findings in this study it is clear that there are many
opportunities for further research, which will be discussed in the final chapter.
67


CHAPTER VI
CONCLUSION
The findings presented in this study, though they contribute to the underwhelming
catalog of Central Asian studies, only indicate that further work needs to be conducted.
Because many of the results produced by the models explored in this work were far from
the hypothesized relationships in many cases, it is clear that there are several areas of
research that can benefit from further exploration The major findings in this study,
considering they are not at all what were expected, must be further researched and
confirmed before conclusions can be drawn about their meaning and relation to the
broader understanding of authoritarian regimes. There are three major findings produced
by this research, each of which clearly indicates the dynamic nature of these authoritarian
regimes and the complicated relationship between the use of repression and Western
influence.
Though the first model confirmed the positive relationship between the use of
repression when a regime was faced with terrorist acts and ethnic violence, this model
contained one of the more perplexing results: that economic decline and inflation
decrease the use of repression in Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan.
Though the removal of the first three years after independence produced the expected
positive relationship between violent repression and inflation, this was not the case with
the economic decline variable. The robust and negative relationship produced requires
further research in Central Asia, without which these findings will merely remain
confusing. These results, although somewhat problematic to understand, underscore how
difficult it is to completely predict the behavior of these regimes.
68


The second important finding produced by this research was found in almost all
of the models that examined the conditional relationships between the use of low
intensity coercion and Western linkage. All of the linkage models, to varying degrees,
show that increased Western linkages in these nations tends to increase the use of low
intensity coercion when challenges are absent. This finding is contrary to what was
expected and is also justification for further research into the relationship between low
intensity coercion and different forms of Western linkage broadly. The previous chapter
proposed the possible explanation that these increases in low intensity forms of
repression were actually because these regimes were moving away from high intensity
tactics, but the models did not confirm the latter part of this assertion. The study did not
show a decrease in high intensity coercion that was consistent, this of course, further
proving that more research with a broader selection of cases is required to explore the
meaning of these findings.
The final important finding is that there appears to be specific types of challenges,
whether they are socio-economic or domestic, that change the impact of the different
types of linkages proposed by Levitsky and Way (2006). In the cases selected from
Central Asia, communication and social linkages seem to conditionally decrease the use
of repression when the challenge is socio-economic. Conversely, these types of linkages
appear to conditionally increase the use of coercion when these regimes are confronted
by domestic challenges such as protest, terrorist acts, or defection of the elite. Civil
society linkages also produce different findings depending on the type of challenges.
These results indicate that the type of challenge that these regimes in Central Asia face
changes the directional impact of the linkages that are present. These findings show that
69


it is centrally important to understand the type of problem a regime is facing because it
can fundamentally change the influence of Western linkages in these countries. Because
these findings indicate that some forms of linkage may actually contribute to the use of
repression, further research that includes a broader selection of cases is necessary to
determine if these conditional relationships between linkages and coercion are unique to
Central Asia or not. Additionally, further research into these nations may require
exploration of the relationships these nations have with the other important political
actors in the region, such as Russia or China. Because China and Russia do not tend to
promote democratization policies, such as the West, the interaction between these
strategic relationships may assist in creating a more complete understanding of these
nations and their behavior. Broadening this understanding of the foreign pressures within
these nations may also help to explain some of the more confusing findings produced by
this research.
Though there were some important findings that indicate a need for further
research, the almost completely insignificant findings for the elite division components
also merits further inquiry. As the first model shows there were only two significant
findings for a relationship between the elite division variable and those results were in the
opposite of the hypothesized relationship. As proposed in the last chapter, it is possible
that these variables failed to produce the expected result because these regimes are
actually combating certain forms of horizontal challenges in a different way that does not
directly impact the whole of society. It is possible that there are more dynamic reward
and punishment mechanisms that exist in these nations that help to combat these types of
challenges that this research missed. Although there could have also been issues with the
70


matter in which the information for these variables was collected, there are still clearly
questions about how these regimes are able to work against elite divisions. Further
research into these relationships is required to fully understand how control is maintained
within the regimes of Central Asia.
Together many of the findings, or lack there of, show that the regimes in
Uzbekistan, Kyrgyzstan, Kazakhstan and Tajikistan do not easily fit into the theoretical
models or meet the expectations that were proposed. Continued research into these
nations and the theoretical models proposed is required to determine whether these
findings indicate the uniqueness of these cases, or if there are broader patterns that this
study may have unearthed.
71


REFERENCES
Aleman, Jose and David D. Yang. 2011.A Duration Analysis of Democratic Transitions
and Authoritarian Backslides. Comparative Political Studies 44:9 1123-1151.
BBC. 2005. How the Andijan Killings Unfolded. BBC. May 17, 2005.
http://news.bbc.co.Uk/2/hi/4550845.stm> [accessed September 24, 2013].
BeilinEva. 2012. Reconsidering the Robustness of Authoritarianism in the Middle
East.^ Comparative Politics. 44:2, 127-149.
BramborThomasWilliam R. Clarkand Matt Golder. 2006. Understanding Interaction
Models: Improving Empirical Political Analysis. 14:63 63-82.
Brooker, Paul. 2009. Non-Democratic Regimes. New York: Palgrave Macmillian.
Cingranelli, David L. and David L. Richards. 2010. The Cingranelli-Richards (CIRI)
Human Rights Dataset. Availaole from http://www.humanrightsdata.org
[accessed 1 March 2013].
Department of Homeland Security. 1993-2011. Yearbook of Immigration Statistics.
Available from http://www.dhs.gov/yearbook-immigration-statistics
[accessed 3 March 2013].
Francisco, Ronald A. 2005. After the Massacre: Mobilization in the Wake of Harsh
Repression. Mobilization 9:107-126.
Gerschewski, Johannes. 2013. The Three Pillars of Stability: Legitimation, Repression,
and Co-optahon^ Democratization. 20:1, 13-j8.
Gandhi, Jennifer and Adam Przeworski. 2007. uAuthoritarian Institutions and the
Survival of Autocrats.Comparative Politics. 40,1279-1301.
Gasiorowski, Mark J.1995. Economic Crisis and Political Regime Change: An Event
History Analysis. The American Political Science Review 89:4 882-897.
Gilbert, Leah and Payam Mohseni. 2011.Beyond Authoritarianism: The
Conceptualization of Hybrid Regimes. Studies in Comparative International
Development 46: 270-297.
Hale, Henry E. 2011.Formal Constitutions in Informal Politics: Institutions and
Democratization in Post-Soviet Eurasia. World Politics 63(4): 581-617.
Jaccard, James, Robert Tumsi and Choi K. Wan. 1990. Interaction Effects in Multiple
Regression. Newbury Park: Sage Publications.
72


Juraev, Shairbek. 2008. Kyrgyz Democracy? The Tulip Revolution and Beyond. Central
Asian Survey 27(3-4): 253-264.
Johnston, Hank. 2011.States and Social Movements. Cambridge: Polity Press.
Karagaiannis, Emmanuel. 2006. Political Islam in Uzbekistan: Hizb Ut-Tahrir Al-Islami.
Europe-Asia Studies. 58(2): 261-280.
Keesings World News Archive. 1993-2011.Available from http://www.keesings.com
[accessed March 2013]
KuranTimur. 1991. Now Out of Never: The Element of Surprise in the East European
Revolution of 1989.World Politics. 44:1, 7-48.
Levitsky, Steven and Lucan Way. 2002. The Rise of Competitive Authoritarianism.
Journal of Democracy. 13:2 51-65.
Levitsky, Steven and Lucan Way. 2006. Linkage versus Leverage. Rethinking the
International Dimension of Regime Change. Comparative Politics. 38:4 379-400.
Levitsky, Steven and Lucan Way. 2010. Competitive Authoritarianism. New York:
Cambridge University Press.
Liu, Morgan Y. 2011.Central Asia in the Post-Cold War World. Annual Review of
Anthropology 11-131.
MegoranNick. 2004. Revisiting the pivot the Influence of Halford Mackinder on
Analysis of Uzbekistan5 s International Relations. The Geographical Journal
170(4): 347-358.
MurphyJonathan. 2006. Illusory Transition? Elite Reconstitution in Kazakhstan 1989-
2002 Europe-Asia Studies. 58:4, 523-554.
NeundorfAnja. 2010. Democracy in Transition: A Micro Perspective on System Change
in Post-Socialist Societies. The Journal of Politics 72:4 1096-1108.
Petrie, Boris-Mathieu. 2005. Post-Soviet Kyrgyzstan or the Birth of a Globalized
Protectorate. Central Asian Survey 24(3): 319-332.
Proksch, Marc, [no date] Regional and bilateral trade
agreements in Asia-Pacific and Central Asia: An Evaluation. UNESCAP. <
http://www.unescap.org/tid/projects/globalize_wgmarc.pdf> [accessed September
12,2013].
Radio Free Europe/Radio Liberty. 2010. Five Years After Andijon Events, Key
Questions Remain Unanswered. Radio Free Europe/Radio Liberty. May 11,2010.
<
73


http ://www.rferl.org ontent/FiveYearsAfterAndij onEventsKeyQuestions
_______RemainUnanswered/2039096+html> [accessed September 24, 2013).
Schedler, Andreas. 2002. The Menu of Manipulation. Journal of Democracy.13:2 36-
50.
Spechler, Martin C. 2007. Authoritarian Politics and Economic Reform in Uzbekistan:
Past, Present, and Prospects. Central Asian Survey. 126:2 185-202.
Sindelar, Daisy. With Pakistan Supply Routes in Doubt, Washington Looks to
Uzbekistan. Radio Free Europe/Radio Liberty. September 29, 2011.<
http://www.rferl.org/articleprintview/24344138.html> [accessed December 1
2011].
United Nations Economic and Social Commission for Asia and the Pacific. 2013.
UNESCAP Dataset. Available from http://www.unescap.org/stat/data/ [accessed 3
March 2013].
United States Census. 2013. Trade in Goods Kazakhstan. Available from
http://www.census.gov/foreign-trade/balance/c4o34.html [accessed 3 March
2013].
United States Census. 2013. Trade in Goods Kyrgyzstan. Available from
http://www.census.gov/foreign_trade/balance403).htm [accessed 3 March
2013].
United States Census. 2013. Trade in Goods Tajikistan. Available from
http://www.census.gov/foreign-trade/balance/c4642.html [accessed 3 March
2013].
United States Census. 2013. Trade in Goods Uzbekistan. Availaole from
http://www.census.gov/foreign-trade/balance/c4644.html [accessed 3 March
2013].
United Nations. 2013. UNTS Treaty Collection. Available from
http://treaties.un.org/pages/UNTSOnline.aspx?id=J [accessed 30 March 2013].
World Bank. 2013. World Bank Dataset. Available from http://data.worldbank.org/
[accessed 4 March 2013].
74


APPENDIX
A. Document Analysis Coding
Coding 0 1 2 3 4
HI: Protest no events 1-2 minor protests, relatively small in size (not against the regime, mainly basic complaints) 1 anti-gov. 2+minor Or widespread mass protest, minors can be elevated if violently repressed 2 or more antigov., 2+ big protracted demonstratio ns, 4+minor 6 + widespread, antigov.
Terrorism no events 2+minor,1-2 2-3 (3 multiple fatalities) 3-5 6+ (4+ if they are high profile or high fatalities)
Ethnic Violence no events minor medium major major with high fatalities
H4: Arrests no events 1 arrest minor 1 major arrest (PM or VP) 2-3 arrests 2 major 3-5 minor 3 major 6+ minor
Violence no events 1 minor (Beatings) 2-3 minor 1 major (murders or disappearanc es) 3-5 minor 2 major 6+ minor 3 major
Public Criticism no events 1 2-3 3-5 6+
Emergence of New Parties no events 1 new party 1 bloc (2-9 parties) 2-3 new parties 3-5 new parties 2 blocor 1 bloc (10+parties) 6+ new parties, 2 large bloc
Defection no events 1 minor player 1 major or 2- 3 minor 2+ major 3+minor 3+ major 4+minor
75


B. Variable Ranges
Ranges
Dhigh 0 2
Dlow 0 2
Protest 0 4
Terrorism 0 4
Ethnic Violence
Inflation 2 1546.7
Econdown
(econdownneg) -0.4979757 0.3308958
Arrests 0 4
Violence 0 4
Public Criticism 0 4
New Opp 0 4
Defection 0 4
Centered Ranges
Protest -0.4605263 3.539474
Terrorism -0.5921053 3.407895
Inflation
Econdown -0.4341506 0.3947209
Defection -0.4078947 3.592105
Lecon -343.2803 2245.62
Lgeopol 2 29
Lsoc -1322.461 4941.539
Lcomm -14.54531 68.05656
Lcivsoc -1.92E+08 3.23E+08
76


Full Text

PAGE 1

CENTRAL ASIAN REPRESSION AND WESTERN INFLUENCE by SARA L. GOUDGE B.A., University of Colorado at Denver, 2010 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements f or the degree of Master of Arts Political Science 2013

PAGE 2

! 2013 SARA GOUDGE ALL RIGHTS RESERVED

PAGE 3

ii This thesis for the Master of Arts degree by Sara L. Goudge has been approved for the Political Science Program by Christoph H. St efes, Chair Michael Berry Thorsten Spehn October 8 th 2013

PAGE 4

iii Goudge, Sara L. (M.A., Political Science) Central Asian Repression and Western Influence Thesis directed by Associate Professor Christoph H. Stefes ABSTRACT Repression, used by dictators as a tool to control the populous of a nation, is a defining feature of authoritarian regimes. Often the most concerning response, the use of coercion is centrally important to a greater understanding of how dictators maintain stability. Emerging theories ha ve proposed that autocracies may be influenced the West toward democratic transition, thus away from repressive autocratic behavior. For both intellectual and ethical reasons, it is important to know when a regime will use repression and how the West may b e able to modify these responses. This study explores when the regimes of Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan have used violent forms of coercion or restricted the rights and freedoms of their people in order to maintain stability, and if ti es to Western nations influenced this behavior. Looking at problems that arise vertically (from the people), horizontally (from elite divisions), or socio economically (from economic decline or rising inflation), this study explores how these authoritarian regimes respond to a variety of challenges, and how Western linkages alter these responses. Using interaction terms to test the conditional relationship between challenges to regime stability and Western linkages, this study explores how ties with the Wes t may be able to influence coercive responses in these authoritarian regimes. The following analysis finds that these regimes view rising inflation as a greater threat to their stability, resulting in an increased use of violence against their people, but interestingly, when challenged during periods of economic decline these dictators expand freedoms and

PAGE 5

iv rights. Additionally, when these regimes increase their ties to the West they utilize more difficult to distinguish forms of repression and are more apt t o restrict the rights of their people. Finally, the following analysis shows that whether a challenge is socio economic or domestic alters the influence of Western linkages, and that in some cases the growth of these ties may be viewed by these regimes as liabilities to their rule. The form and content of this abstract are approved. I recommend its publication. Approved: Christoph H. Stefes

PAGE 6

v ACKNOWLEDGEMENTS Though there have been many individuals that contributed to this thesis, I would like to first thank all of the wonderful professors in the Political Science Department at the University of Colorado at Denver. Specifically, I must thank Christoph Stefes and Michael Berry for all of their personal assistance with this project. Thank yo u both for all of your insight and direction, because this would have been impossible without both of you. I would also like to thank James Myers for his astute suggestions and his extremely detailed editing technique, and though we may disagree on stylis tic issues, I still occasionally respect your opinion. And finally, I must thank my parents, Linda and Chuck Goudge, for all of their support throughout the years, because without them none of this would have been possible.

PAGE 7

vi TABLE OF CONTENTS C HAPTER I. INTRODUCTION...... 1 II. REVIEW OF THE LITURATURE 6 III. HYPOTHESES, MEASUREMENT, AND MODELS 16 Hypotheses 16 Measurement. 18 Models... 27 IV. RESULTS..... 28 Model 1: Coercion and Horizontal, Vertical and Socio Economic Challenges. 28 Model 2: Inflation, Western Linkages, and Repression 35 Model 3: Economic Decline, Western Linkage s, and Repression 37 Model 4: Terrorism, Western Linkages, and Repression......................... 42 Model 5: Elite Defection, Western Linkages, and Repression..... 46 Model 6: Protests, Western Linkages, and Repression..... 49 V. DISCUSSIO N........... 53 The First Model. 53 Geopolitical Linkages... 59 Without Challenges, Low Intensity Coercion Increases... 60 Socio Economic vs. Domestic Challenges... 62 V I. CONCLUSION.68 REFERENCES. 72

PAGE 8

vii APPENDIX A. Document Analysis Coding.... 75 B. Variable Ranges.. 76

PAGE 9

viii LIST OF TABLES Tables 1. Model 1 Coercion and Challeng es.. 30 2. Inflation and Coercion.... 33 3. Economic Decline and Coercion 34 4. Model 2 Inflation, Western Linkage and Repression. 35 5. Model 3 Economic Decline, Western Linkages, and Repression... 38 6. Inflation, Western Linkages, and Repression 1996 2010... 40 7. Economic Decline, Western Linkages, Repression 1996 2010.. 41 8. Terrorism, Western Linkages, and Repression... 43 9. Elite Defection, Western Linkages, and Repression... 47 10. Protests, Western Linkages, and Repression.. 49 11. Models 2 6 Linkage Interaction Terms ... 51 12. Model 1 Expected and Directional Results. 54 13. Socio Economic Hypotheses 1993 2010 vs. 1996 2010 56 14. Linkages Increas e Low Intensity Repression when Challenges are Absent... 60

PAGE 10

ix LIST OF FIGURES Figures 1. Socio Economic and Domestic Challenges with Linkages 62

PAGE 11

1 CHAPTER I INTRODUCTION "We never expected they would shoot at women and children. I've got three children, I was trying to cover them up. I was crying: 'Take my life, don't shoot my children.'" (BBC 17 May, 2005). After several days of peaceful antigovernment protest, on the evening of May 13, 2005 the military began openly firing on the crowds that had gathered in Andijon, Uzbekistan ( Radio Free Europe/Radio Liberty 2010) Most accounts conclude that these demonstrators went to the streets to protest the injustice, harsh repression and poverty they were experiencing under t he rule of Islam Karimov (Radio Free Europe/ Radio Liberty 2010). On that day it is estimated that somewhere between several hundred and nearly 1,000 people lost their lives (Radio Free Europe/ Radio Liberty 2010). Islam Karimov's regime has not only argu ed that the casualties were much less, but that this harsh response was necessary because these were not protesters, but religious extremists (Radio Free Europe/ Radio Liberty 2010). What actually happened during the Andijon Massacre is still somewhat uncl ear, but what is known is that this action was a striking example of an authoritarian regime reacting with brutal violence when challenged by popular protests. Although it may seem that the choice to fire on peaceful protesters was an overreaction, this behavior is not only common in authoritarian regimes, but can at times be considered acceptable. Knowing when these regimes resort to violence is not only of importance for a greater understanding of political science, but also because this knowledge could be the basis for informed and responsible foreign policy around the

PAGE 12

2 globe. The following study hopes to provide insight into when these regimes react with repression and how ties with Western nations may be able to influence these leaders away from using coercion to control their people. The study that follows utilizes a theoretical framework that recognizes the dynamic nature of authoritarian regimes, rather than focusing on the possibility for democratization. Using the cases of Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan, this research hopes to avoid the misconception that these regimes are merely failed or semi democracies. Much of the research conducted on this region has focused primarily on the state, the culture, or the possibilities of demo cratic transition (Liu 2011, Spechler 2007, Hale 2011). Though work in this region has been preformed in many different academic disciplines, each with their own important contributions, political science studies have mainly focused on the process of democ ratization (Petric 2005, Juraev 2008, Spechler 2007). Although this research is important, it makes the assumption that these nations are working toward greater liberalization or expanded human rights (as if this were the natural and expected progression) By utilizing theories that explore the dynamic nature and durability of authoritarian regimes, the following research will examine these nations through the lens of understanding authoritarian forms of governance, rather than imperfect democracies. The following study is interested in determining when theses Central Asian nations react not only with violence against their people, but also by restricting the rights of their citizens more generally. The research that follows will explore the challenges th at have arisen in the selected cases from 1993 2010. By looking at challenges that may arise in these nations, such as socio economic problems, societal instability, or elite

PAGE 13

3 divisions, this work hopes to contribute to a nuanced understanding what kinds o f problems elicit repressive responses from these regimes and how the West might assuage these responses. The two central research questions that have guided this research are: under what circumstances do the authoritarian regimes in Central Asia use repr ession or coercion to control their people, and what types of Western linkages deter these regimes from using these techniques? The measurement for the level of coercion in these nations is based off the ratings provided by the Cingranelli Richards (CIRI) Human Rights Dataset. Although it can be difficult to find information from this region of the world, this study will use data collected from the World Bank, the US Department of Homeland Security, US Census, and the United Nations Economic and Social Comm ission for Asia and the Pacific (UNESCAP). For the concepts that are more difficult to measure, or where reliable information was unavailable, the Keesing's World News Reports are used and a process of document analysis is performed. This research uses a t ime series panel design that will test not only the use of coercion when these regimes face challenges, but also how ties to Western nations interact with these responses. The following study focuses on the nations of Kazakhstan, Kyrgyzstan, Tajikistan a nd Uzbekistan, but omits Afghanistan and Turkmenistan Although it would have been optimal to use all of the nations in Central Asia in order to make broad assertions about this region, these four were selected because they are each interesting cases indiv idually. This research hopes to explore Western influence in the region, thus Afghanistan has been omitted because the Western led war being conducted in this nation changes the dynamic that hopes to be explored in this research. Because this nation was in vaded by

PAGE 14

4 the US and NATO, this use of force does not allow for this to be a possible case in which Western influence is examined, since for the last 12 years it has been subject to only Western intervention. Turkmenistan was omitted because the data from t his nation are even scarcer than the data that are available for the other countries in the region. Kazakhstan was selected because it is the largest nation in the region, which also possesses the greatest oil wealth. To examine the range of authoritarian practices, Kyrgyzstan was selected because it is arguably the most democratic in the region, and conversely Uzbekistan was selected because it falls on the more repressive end of the authoritarian continuum. And finally, Tajikistan was chosen because the n ation experienced a civil war after independence. The limited scope of this study did not allow for all of the cases in Central Asia to be examined, but those that were selected each in their own way expose the complicated and diverse nature of the nations in this region. Using theories of authoritarianism, guided by an understanding of the dynamic nature of these types of regimes, this study produces several major findings. First of all, this research shows that when these regimes are experiencing econom ic downturns, they surprisingly, respond by broadening the rights and freedoms of their people. Theoretically, it is predicted that authoritarian regimes, when faced with economic hardships will increase repression in order to maintain stability, rather th an expanding the rights of the citizen. This finding shows that these regimes react contrary to anticipated expectations. Additionally, during periods of increased inflation from 1996 2010 these regimes tend to increase their use of violent repression. Thi s study also finds that these regimes are more likely to restrict the rights of their people when there are greater ties to the West. And finally the following research shows that the type of problem that arises in

PAGE 15

5 these nations, whether socio economic or domestic, can dramatically change the influence of ties to Western nations. Though these findings are ultimately important contributions to the research on these types of regimes, as well as Central Asia, they highlight how dynamic, yet responsive dictator ships can be. The study that follows will show that understanding the behavior of authoritarian regimes is complex and that the power and influence of the West in changing the coercive behavior of these regimes is dependent on the type of problems that th e leaderships face in these nations.

PAGE 16

6 CHAPTER II REVIEW OF THE LITURATURE When the Soviet Union collapsed the initial reaction was one of hope for the future of democracy around the globe. Many political scientists claimed that the democrat ic model had succeeded in conquering all other political systems, but as the dust finally settled it became clear that although many of the newly independent nations had pursued democratic institutions, these arrangements were merely facades for the increa singly authoritarian characteristics of these regimes (Schedler 2002, Levitsky and Way 2010). Because these governments were unlike the traditional democracies, or outright autocracies, theorists were forced to address the possibility that many of these ne w nations were not just semi democracies, but had become a new form of authoritarianism (Gilbert and Mohseni 2011). In facing the emergence of systems that appear to have the democratic prerequisites, but are very clearly not democracies theorists have pro ceeded to reconsider previous understandings of these types of regimes. This has led to a broader understanding of the complexity of authoritarian governance rather than just viewing them as imperfect forms of democracy (Brooker 2009, Levitsky and Way 2002 ). Recognizing that authoritarian regimes are their own distinct form of governance, and that these systems do not necessarily lead to democracy, studies have begun examining the durability and responsiveness of these regimes and how they respond to chall enges. Theorists have proposed that challenges to a regime can emerge from within the governing elite, from the people, or the international community (Gerschewski 2013, Levitsky and Way 2010). Although there are many ways that a regime can respond to thes e challenges, repression is not only one of the most visible, but

PAGE 17

7 one of the most definitive authoritarian reactions. By looking at the work of multiple theorists that explore the plethora of challenges that face authoritarian regimes and the importance of understanding when a regime uses repressive or coercive techniques, it will become clear that the dynamic nature of these regimes must be further explored. Regimes can face challenges that come from vertical, horizontal, and international sources. More specifically, a regime can fall victim to cleavages within the ruling elite, which can be considered a horizontal problem. Problems may also arise vertically originating from the populous. And finally, the pressure of international actors to democratize h as also been theorized to influence the behavior of regimes or destabilize them and possibly dislodge a dictator. The central concern of a regime is to maintain its power. To do this leadership must be able address specific problems. One of the most impo rtant challenges can come from within the regime's elite. Gerschewski (2013) proposes a theory of three pillars within an authoritarian regime that can predict the stability of the government in power. These pillars include: repression, cooptation, and leg itimation (Gerschewski 2013). He defines cooptation "as the capacity to tie strategically relevant actors (or group of actors) to the regime elite" (Gerschewski 2013, 22). It is centrally important for the leadership to continue to maintain the support of other powerful people within the ruling elite, because these actors have their own power base, whether it may be financial, tribal, or regional (just to name a few) and can easily defect from the ruling party and join the opposition. Gandhi and Przewors ki (2007) explain that there are multiple levels to the elite power structure, which they call trenches. The first trench is made up of the central members of a leader's administration, such as those that make up "[c]onsultive councils,

PAGE 18

8 juntas, and politic al bureaus" (Gandhi and Przeworski 2007, 1289). The second trench is comprised of "a legislature that encapsulates some opposition, a party that mobilizes popular support for the dictator, or even multiple parties" (Gandhi and Przeworski 2007, 1293). The c ooperation from these political elite is thought to enhance the durability of the authoritarian leader. These theories suggest that some of the greatest challenges to a regime can be horizontal. A dictator must be concerned with the leadership that is clos est to them, and the opposition within the governing bodies. The complexity of this situation requires a leader to maintain many relationships and ward off challenges from these groups and individuals. Vertical challenges can also come in many forms, and can also require some astute maneuvering for the regime to combat these problems. Gerschewski's (2013) pillar of legitimation can be used to explain challenges that arise from the society's discontent with the leadership, as well as the socio economic fac tors that may destabilize a regime. He explains, "Legitimation seeks to guarantee active consent, compliance with the rules, passive obedience, or mere toleration within the population" (Gerschewski 2013, 18). He breaks up this concept into the two distin ct categories of "diffuse" and "specific support" (Gerschewski 2013, 20). Specific support is gained when a regime is able to fulfill demands, such as guaranteeing security and improved economic conditions (Gerschewski 2013, 20). Diffuse support refers to more general ideas about a regime, and includes political ideologies promoted, national identity and leadership (Gerschewski 2013, 20). Problems that arise from below generally challenge the legitimacy of a regime. One of the most important vertical cha llenges to authority can come in the form of mass protests, which are generally a sign that a regime has lost legitimacy. These types of

PAGE 19

9 challenges can appear instantaneous and remove a leader very quickly (Kuran 1991). Although protests are the most stud ied of the various vertical challenges, terrorism can also be considered a vertical challenge, though generally from a smaller faction in the society. Using Gerschewski's (2013) concept of legitimation, both of these challenges are symptoms of a regime los ing support from the people (20). Although terrorist acts are generally committed by a small minority of people in society, they are by definition intended to destabilize a nation and challenge the regime in power. The final type of vertical challenge, tha t is particularly important in Central Asia, are instances of ethnic violence. Though this factor does not specifically challenge the central government's policies, it does challenge a regime's ability to ensure the security of the people in society. It has also been theorized that a regime can face challenges when there are dramatic economic fluctuations. These economic shocks are not in origin vertical, but they can manifest themselves into problems that challenge the regime from below ( Alem‡n and Yang 2011) In other words, though the economic problems can directly disrupt the government, the major concern is that it will promot e social unrest. This idea fits nicely into Gerschewski's (2013) definition of specific support that translates into a stable pillar of legitimation. If a nation is facing economic hardships, and the regime is unable to remedy these problems, the pain will be felt by those at the bottom, which can translate into a loss of specific support, and thus legitimation (Gerschewski 2013). Economic challenges can arise from decreases in economic performance as well as increases in inflation. Studies have found tha t there is a connection between economic growth and development, which can decrease the human rights abuses and promote

PAGE 20

10 democratization in Central Asia and other post Soviet cases (Spechler 2007, Neundorf 2010). In the same strand of logic that economic gr owth would promote democratization, the inverse, that economic decline could induce a government to become more repressive and authoritarian, might also be true. Gasiorowski (1995) finds that when regimes are faced with economic crises there is an incre ased likelihood for regime breakdown. He finds that there are differences in the effect of specific types of crises depending on the time period in which they occur (Gasiorowski 1995). The author explains: We can conclude from these findings that econom ic crises do not simply undermine the legitimacy of whatever type of regime currently exists in a country, thus triggering regime change in either direction, as argued by several authors. However, we must also conclude that while inflationary crises inh ibited democratization in both contexts in a complementary manner in the 1950's and 1960's [. .] they may actually have facilitated democratic transitions in the late 1980's; and recessionary crises did not have this kind of complementary effect thro ughout 1950 1989. Gasiorowski 1995, 892 This study shows that the impact of inflation has different effects depending on when the challenge to the regime occurs. It is clear from these findings that inflation and economic decline are important challenges that may lead to regime breakdown and should be centrally important in the study of authoritarian regimes. Though a regime can encounter problems from multiple levels within society, it may also be challenged by the international community Challenges th at come from international organizations, such as the UN, generally, come in the form of pressure to democratize. Within the theory of competitive authoritarianism proposed by Levitsky and Way (2010), the importance of Western influence on democratic polic ies and the

PAGE 21

11 possibility of democratic regime change is explored. These theorists explain that international pressure may influence regimes though both linkages and leverage exercised by Western powers, particularly the United States (Levitsky and Way 2010) Leverage is defined as the level of vulnerability of a regime to pressures from the international community to democratize, which can come in the form of "positive conditionality (for example, EU membership), punitive sanctions (aid withdrawal, trade sa nctions), diplomatic persuasion, and military force" (Levitsky and Way 2006, 382). The influence of these forms of leverage may be diminished by the size of the country that the West is attempting to influence, for instance, a large nation with many resour ces would be less vulnerable to these types of pressures in comparison to a smaller nation (Levitsky and Way 2006, 383). Additionally, the authors explain that there are two other limitations that may negate the influence of Western leverage, which are: an alternative regional power that can offer support to the regime that does not push democratization (such as Russia or China in the case of Central Asia), or there may be other Western policy objectives that outweigh the importance of democratization of th e nation (Levitsky and Way 2006, 383). The authors explain that leverage can increase the costs to the regime to commit authoritarian abuses, but that there are limitations to using leverage to change the behavior of authoritarian regimes (Levitsky and Way 2006, 383). The authors explain that linkages "can be defined as the density of ties and cross border flows between a particular country and the U.S, the EU and western dominated multilateral institutions" (Levitsky and Way 2006, 383). These ties can be broken up into five dimensions that include: economic, geopolitical, social, communication, and through transnational civil society (Levitsky and Way 2006). These types of linkages apply

PAGE 22

12 pressure on authoritarian regimes because when a government commits a n abuse there is a heightened chance of international salience and thus an increased chance that Western governments may act to combat these abuses (Levitsky and Way 2006, 384). The authors explain that linkages help to make leverage more effective because they increase the "boomerang effects" that can draw attention to even minor abuses, which increases the costs to the regime for acting outside of the expected norms promoted by the West (Levitsky and Way 2006, 386). These types of bonds change the expecta tions from society of the government and also make the use of coercion or repression less likely. This theory indicates that countries that have higher linkages and leverage will be pressured to democratize. The authors emphasize that without linkages the influence of leverage is lower, if not entirely ineffective. Because linkages are so important to the successful use of leverage, these types of bonds are of central importance in understanding how to change the repressive or coercive behavior of authorit arian regimes. As the previous discussion has shown there are many types of challenges that authoritarian regimes may face and although there are many ways that a government may respond to these problems one of the most visible and significant is through their use of repression to maintain power. One of the defining characteristics of authoritarian regimes is their use of repression or coercion (Gerschewski 2013, 21). Though these types of governments have other tools that can be used to maintain thei r power, a regime's use of coercive techniques helps to differentiate this type of governance from more democratic systems. One of the most important ways an authoritarian regime can respond to challenges and

PAGE 23

13 maintain stability is through the use of repres sion and the ability to use the coercive capability of the security apparatuses, which has been found to be centrally important to the stability and success of a regime (Gerschewski 2013, Bellin 2012). The use of coercion is not a constant factor though; r ather it increases and decreases in response to challenges that face the regime in power. Levitsky and Way (2010) explain that conceptually, repression or coercion is not only limited to the level of visible repression, but also includes more subtle tech niques. These authors propose that there are two types of coercion: high intensity and low intensity (Levitsky and Way 2010, 57 58). High intensity coercion, they explain, is more visible because the regime may target large groups, important people, or ins titutions (Levitsky and Way 2010, 57). Specifically, high intensity coercion could "include campaigns of violence against opposition parties (e.g. Cambodia and Zimbabwe), imprisonment (e.g. Malaysia and Russia), attempted assassination of major opposition leaders (e.g. Belarus and Ukraine), and high profile assaults on democratic institutions such as parliament (e.g. Russia in 1993)" (Levitsky and Way 2010, 57 58). Low intensity coercion is much less visible and can take multiple forms, such as surveillance harassment, denial of public services, or employment (Levitsky and Way 2010, 58). This type of coercion can take many forms, but the use of this type or repression is always to accomplish the same goal: to diminish the power of the possible opposition to such a degree that they no longer are able to pose a serious challenge (Levitsky and Way 2010, 58). By creating these conceptual categories Levitsky and Way (2010) bring attention to the fact that these regimes have many tools that they can use in respon se to possible challenges that may threaten their regime.

PAGE 24

14 There are also limitations to the ability of a regime to use some forms of repression. Johnston (2011) explains dictators can respond to protests with repression, but that increased brutality can, in some cases, embolden the protesters. It is thought that low levels of repression or high levels will both result in an increase in protest, and that a government needs to find a level of repression that is neither too tough or too lenient to quell soci al movements (Johnston 2011: 108 109). Attempting to find this line of proportionality is referred to as the "dictator's dilemma" (Johnston 2011, Francisco 2005). Repression, particularly high intensity or more visible types of coercion, also carry risks in the eyes of the West. Regimes may also be influenced by international actors to decrease the use of repression. Because there are many concerns that surround the use of coercion, whether it is in the form of an increased reaction from the people, or fro m the West, it is clear that there are many concerns that must be weighed by a leader or regime when faced with the many types of problems that may arise. Though repression is not the only technique that a regime can choose to use, how and when they cho ose to use coercion to elicit desired responses is arguably the most valuable to understand for political scientists and for the discussion international foreign policy. This can be said because knowing when a regime may respond with high intensity or low intensity coercion could help to guide policies that may influence democratization and at the very least could limit morally negligent policies. A regime can use repression or coercive techniques whenever it chooses, but some have recognized that there are different challenges from different sources that may destabilize a regime. By recognizing multilevel challenges and particular forms of repression that a regime may use to combat them, the following research will examine how these pressures interact and w hether

PAGE 25

15 Western linkages are able to mediate the coercive responses from these regimes in Central Asia.

PAGE 26

16 CHAPTER III HYPOTHESES, MEASUREMENT, AND MODELS Hypotheses There are many ways that an authoritarian regime can be challenged, b ut how the government responds to these problems speaks to the dynamic nature of these regimes and the necessity for scholars to understand their behavior. One of the most important reactions that a regime may have to particular issues is to repress its pe ople, through either high intensity or low intensity forms of coercion. To explore the relationship between the level and type of coercion implemented by a regime, this study will test various types of challenges and the specific actions taken by the gover nments in Central Asia. By applying the previously discussed theories to the cases of Uzbekistan, Tajikistan, Kyrgyzstan and Kazakhstan, the relationship between coercion or repression and vertical, horizontal, and international challenges will be explore d. The discussion presented in the literature review lends itself to four different hypotheses. HYPOTHESIS 1: Societal challenges to a regime will lead to an increase in the level of coercion HYPOTHESIS 2: Pronounced economic downturns will cause an increase in the level of coercion. HYPOTHESIS 3: Rises in inflation will result in an increased level of coercion. HYPOTHESIS 4: Increasing divisions within the ruling elite will challenge the stability of the current regime and will result in increase d level of coercion. HYPOTHESIS 5: The effect of rising inflation on the use of coercion is expected

PAGE 27

17 to be negatively conditioned by the presence of Western linkages. HYPOTHESIS 6: Western linkages will negatively condition the use of coercion when a r egime is faced with economic decline. HYPOTHESIS 7: When a regime is faced with the challenge of terrorist acts linkages to the West will negatively condition the use of coercion. HYPOTHEISIS 8: The use of coercion when a regime is challenged with elite defection will be negatively conditioned by the presence of Western linkages. HYPOTHESIS 9: Western linkages, when a regime is challenged by protests, will negatively condition the use of coercion. For this study the dependent variable will be the leve l of coercion or repression that a regime implements, which will be broken up into high intensity and low intensity forms of coercion. The first hypothesis will test the types of vertical challenges discussed earlier. These types of problems will come from the people and can challenge the stability of the regime. The second and third hypotheses test the economic challenges that can manifest themselves into vertical challenges. And the fourth hypothesis will examine the horizontal challenges that a regime ma y face that come from the ruling elite within the regime or the opposition parties or leaders. Regarding the international dimension, as the previous chapter briefly discussed, these types of challenges to a regime are not direct, instead they have subtle effects that alter the expectations and, theoretically, would mediate a regime's use of repression or coercion. Because this is not a direct relationship, these factors will be tested for their conditional effects in the responses of these regimes These c onditional relationships will be tested using the fifth thru ninth hypotheses.

PAGE 28

18 Measurement This study will cover Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan during the period from 1993 2010. These cases and years produce a sample size of 72. Thou gh all of these countries became independent from the Soviet Union in the final days of 1991, 1992 was omitted because the turmoil during this year led to a lack of reliable data. The use of repression or coercion will be the dependent variable in thi s study. And in light of the discussion presented by Levitsky and Way (2010), this concept must be divided into two different types of coercion, high intensity and low intensity. This keeps these two forms of coercion conceptually separate, which is useful because in some instances a regime may use one form and not the other. This distinction is important for a broader understanding of the circumstances in which these regimes use particular types of repression. For the purpose of this study high intensity c oercion is considered the more violent and visible form of coercion, whereas low intensity involves more subtle techniques that remove the liberties and rights of the people in order to minimize their possible oppositional power. The dependent variables ar e based off of the ratings produced in the CIRI Human Rights Data set' (Cingranelli and Richards 2010). This dataset looks at nations using various factors that contribute to the level of human rights in the nation. These ratings are scaled from 0 to 2 fo r each component, with 0 indicating higher levels of the repressive technique being used and 2 being little to no instances. These ratings give a higher value to the lower levels of repression. Because the direction of these ratings is negative, for the pu rpose of this study the ratings that make up the two dependent variables have been reversed by multiplying them by 1, thus making a higher rating indicate a higher level of repression or coercion.

PAGE 29

19 The measure for high intensity coercion is constructed f rom the ratings for extrajudicial killings and torture as presented in the dataset. The measure for low intensity coercion is comprised of the ratings for freedom of assembly and association, foreign movement, domestic movement, speech, electoral self dete rmination, religion, worker's rights, and the independence of the judiciary. Because there are a different number of components included in each of the dependent variables when they are combined (resulting in a high intensity range of 0 4 and low intensity range of 0 16) the coefficients produced should not be directly compared, but rather understood with regard to their specific range. The first and the fourth hypotheses selected for this study create independent variables that lend themselves to a pan el design because they have many components that should be included. The first hypothesis predicts a relationship between coercion and problems that arise in society, which is a fairly broad statement that requires a multiple component understanding of the se types of challenges. For this study, problems that may arise in society that can challenge the stability and legitimacy of a regime include protests, terrorist acts and ethnic violence. The previous chapter discussed the important role that protests can have in removing dictatorships and disrupting a regime, so it is of paramount importance that this type of challenge be tested in relationship to the possible types of repressive responses that may come from the regime in power. Though terrorist actions h ave not been studied to the same extent as the influence of protests, the same rationale seems to apply: discontent within society can spread to the point that the regime no longer can claim it is able to represent the people or protect them. And finally, instances of ethnic violence can be seen as problems where a regime, as Gerschewski

PAGE 30

20 (2013) might put it, has lost legitimacy because they are unable to keep up their promises to ensure the security of the people. The fourth hypothesis proposes that there is a relationship between elite divisions and the level of repression used by the government. As the previous discussion showed, some regimes may face challenges that can come from people within the government that are close to the leader, and other horiz ontal challenges that may come from the opposition that maintains some level of power ( Gandhi and Przeworski 2007) It is particularly difficult to measure the level of division that can exist at a given time within a regime, mainly because these types of problems are not nearly as visible as societal or vertical challenges. That being said, there are some events that could signal these types of divisions. For the purpose of this study there are five observable events that may occur which might be signs tha t there are broadening elite divisions occurring. These include: arresting or charging individuals within the government or the opposition, violent acts perpetrated against members of the elite or the opposition (these can be with or without ties to the go vernment), members of the elite publicly criticizing the regime, high ranking members of the government defecting, or the creation of new parties or opposition blocs. The first four indicate that there are problems within these groups that may be challengi ng the regime's control. The final component is slightly different because it measures the growth of opposition voices within the nation, which could, given enough power, become enticing places to defect to for members of the elite if the regime appears to be weakening. Measuring concepts such as the levels of protest, terrorism, or elite divisions is primarily difficult because of the lack of available research and data from these Central

PAGE 31

21 Asian nations. That being said a process of document analysis was selected for the components for the first and the fourth hypotheses. Using Keesing's World News archive, each of the countries names were searched and all relevant returns were read and coded according to their content regarding the component of the variab les that they included (Keesing's Record of World Events 2010). The totals were then collected for each year and weighted on a 0 4 scale, with 0 indicating that there were no instances of the event, whether it be a protest or an instance of elite criticism occurring and 4 meaning that they were comparatively frequent. This technique of document analysis was used for each of the components of the first and the fourth hypothesis. Although document analysis is not always the most reliable way to generate meas urements for concepts, it was really the only way to examine these types of events in these countries. There are two reasons that using this form of document analysis in this case was a better option than all other available choices. One of the major iss ues with researching these nations is that there is a very limited pool of data. Many of the indicators used in political science are just not available consistently for these countries. The second, and most important reason for choosing document analysis is that it allows for concepts that are difficult to measure to be examined, such as the components that might indicate elite divisions. It would be difficult and possibly fraught with mistakes to measure the divisions within the elite in another way than through the possible signs that might indicate that there are fractures emerging within the ruling class in these nations. It is particularly difficult to measure these divisions without tracing the paths of all of the elite in each of these cases. Other s tudies have tried to follow the elite movement, but this is a particularly difficult process in nations, such as those selected for this research, because

PAGE 32

22 even though a member of the ruling elite might leave the government this does not necessarily mean th at they have lost favor with the regime, often it is quite the contrary. Jonathan Murphy (2006), in his study of the elite in Kazakhstan from 1989 2002, finds that although many of the elite left the government for positions within the private sector, they eventually returned, indicating that their departure was more of a reshuffle because they had earned favor with the regime, rather then an indication that there were elite divisions. The findings presented in this piece point out how difficult it is to tr ace the behavior of the elite and that even though a member of the ruling class might leave the government this does not necessarily mean that this is a sign of elite division. It would be impossible to follow all of the members of the elite for the two de cades at issue in this study, not only because it would be an enormous task, but because the available information about these individuals is very limited and could lead to many errors in the conclusions. The second and third hypotheses, the influence of economic decline and inflation in relation to the use of coercion, are much easier concepts to measure. To measure the economic decline in the second hypothesis the gross national income ($US per capita) was used from the database created by the United Na tions Economic and Social Commission for Asia and the Pacific (UNESCAP). The thought behind this hypothesis is that the greater the impact on the people as their economic wellbeing diminishes, the greater the challenge will be to the legitimacy of the gove rnment, to which the regime must respond. Thus the magnitude of the economic decline is important. To capture this the percent change from the year previous was produced and these figures were lagged one year to allow for the impact of the economic downtur n to reach society. In order to

PAGE 33

23 measure the level of decline and to make the coefficients more easily interpretable these data was reversed by multiplying all of the figures by 1. The third hypothesis, regarding inflation, is based on the same logic discu ssed in the literature review. To measure this concept, the indicator from the World Bank for inflation as GDP deflator (annual %) was used (World Bank). These data were also lagged one year, as to be able to measure the societal pressure, rather than the immediate reaction. As the previous chapter explained, challenges to an authoritarian regime can also come from the international community through the emergence of Western Linkages. As Levitsky and Way (2010) explain the level of linkages and leverage e xperienced by a nation has an impact on the nation's behavior and the pressure to democratize. Leverage and linkage work together, but linkages are centrally importan t. Though both of these types of influence could be examined, this study will concentrate on the influence of linkages. This was selected because the theories presented by Levitsky and Way (2010) indicate that leverage without linkage is ineffective. That being said even if there were years within these nations in which leverage was high, witho ut linkage the leverage would theoretically be futile. Because authoritarian regimes are not constantly repressive, rather they are repressive because they are challenged, this study will explore whether international linkages may play a meditating role in the behavior of an authoritarian regime. Although the relationship between coercion and linkages could be tested directly, this work recognizes that these regimes responsive rather than maintaining a static level of repression, thus the influence of We stern linkages is thought to condition the responses to specific types of challenges. In other words a regime may have low linkages and low

PAGE 34

24 repression, but also no challenges to the regime, which does not prove that this theory presented by Levitsky and Wa y (2006) is incorrect, rather it shows that there must be challenges to a regime to actually test the influence of the Western linkages. For that reason this study will apply the theory that linkages have a conditioning effect on the behavior of authoritar ian regimes. Using this concept as an interaction term, rather than an independent variable allows for the conditional effects of increased linkages with the West to be explored. The use of interaction terms have been argued by many, that when used correct ly, these types of models can explain more complicated relationships (Jaccard, Turrisi, and Wan 1990, Brambor, Clark, and Golder 2006). In order to make the coefficients more meaningful in relation to each other, each of variables and interaction terms use d in the all of the models, excluding the first, have been centered (Jaccard, Turrisi, and Wan 1990, Brambor, Clark, and Golder 2006). Hypotheses five thru nine show the five challenges that have been selected in this study to test the influence of Wester n linkages, which include: inflation, economic decline, terrorism, elite defection, and protests. Unfortunately, all of the other types of challenge components could not be tested in this matter because the study size limited the ability for such a lengthy process. Using the five different types of linkages explained by Levitsky and Way (2006) this study will also explore what types of linkages produce conditioned responses in the use of coercion from these regimes. Though these measures could be combined to make a composite score for linkages, keeping them separate helps to keep the results distinct and find a possible hierarchy in types of linkages. The first type of linkage explored is economic. To measure this concept, this study uses the exports and i mports between each nation and the United States, which is obtained through the US census website (United

PAGE 35

25 States Census). The exports and imports were combined to show the amount of commerce, without focusing on one of the particular aspects of it. The the ory proposed by Levitsky and Way (2006) does not focus on something like economic dependency, but rather just the amount of interaction, so it is coherent to combine the imports and exports within the theoretical framework. The second linkage, geopolitic al ties, is a much more difficult concept to measure. These types of ties could be measured using treaties or agreements that the nations of focus have decided to enter into with Western powers. Unfortunately, the data regarding these types of interactions is difficult to find. What are available though are the dates that these nations became signatories to United Nations treaties (UNTS Treaty Collection). Though the UN is not run entirely by the West, it does promote more democratizing ideals. By going thr ough the dates on which the nations in this study became signatories to different treaties within the UN, though not the perfect measure for geopolitical ties, does capture the years in which these nations were attempting to appear more open to participati on in the international community. Becoming signatories to these treaties does not necessarily prove that these nations are trying to democratize, but it does indicate that they were interested in appearing to be cooperating with the international communit y. The annual figures created to measure geopolitical linkage were obtained by doing a participant search with the all matching' specification selected (UNTS Treaty Collection). The returns generated were reviewed and each treaty and the year in which the nation became a signatory were collected. While this may not be the best measure for geopolitical linkages, with the limitations to the available data, this is the closest and best measure available for these nations.

PAGE 36

26 The third type of linkage discussed by Levitsky and Way (2010) is social, that is "the flow of people across borders, [which] includes migration, tourism, refugees, and diaspora communities, as well as elite education in the West (Levitsky and Way 2006, 383). To capture this type of linkage this study uses the figures for immigrants admitted to the United States by region and country of birth from the United States Department of Homeland Security's yearbooks of statistics (DHS). The fourth linkage, communication, fortunately, is a much eas ier concept to measure. It would be optimal for the data regarding telephone and Internet access to be used, but the data for Internet users is incomplete, so it has been omitted (World Bank). There were consistent figures available though for cell phones and telephones per 100 people from the World Bank (World Bank). To create a single annual figure the data was combined resulting in a value for cell phones and telephone lines per 100 people in each of these nations. The final type of linkage, transnati onal civil society, is also one of the more difficult concepts to measure. Levitsky and Way (2006) explain that this concept includes "local ties to western based nongovernmental organizations, religious groups, and party organizations" (384). To measure t his concept this study uses the World Bank data on net official development assistance and official aid received (current $US) (World Bank). These forms of aid that are received from outside governments are generally for the purposes of promoting democrati zation programs on the ground in these nations. Though this is not the perfect measure it is similar to the concept presented by Levitsky and Way (2006), while not being a direct measure for aid, which they would consider a type of leverage, as discussed i n the literature review (Levitsky and Way 2006).

PAGE 37

27 Though some of these measures may not fit the concepts perfectly, they are the best options because the data from this region are so limited. The following discussion will explain the specific models that were used in this research. Models This study, because each of the concepts covered in the hypotheses have multiple components, will utilize a panel design. To explore the relationship between the levels of coercion and the different types of challenge s that a regime may face, the first model will compare the two dependent variables, high and low intensity coercion, and each of the components that for the first four hypotheses. The first model will test the direct relationship between the dependent vari ables and the levels of terrorism, ethnic violence, protest, economic downturn, inflation, arrests, violence, public criticism, new opposition, and defection. The next five models will test the fifth thru ninth hypotheses using interaction terms for the le vel of Western linkage during periods of inflation, economic downturn, terrorism, defection and protest. Each of the models that include the interaction terms all of the variables will be centered and will test both of the dependent variables.

PAGE 38

28 CHAPTER IV RESULTS This chapter will review the models used in this research that directly test the relationship between the dependent variables of coercion and the nine hypotheses and also will explore the conditioning relationships between Western linka ges and various challenges. These results will expose some of the possible patterns between the models that will then be discussed in the following chapter. Although many of the findings in these models do not appear to conform to the expectations, they do expose some interesting trends that will become apparent in the following review. From this discussion it will become clear that the type of challenge that a regime faces changes how Western linkages influence the behavior of these regimes in Central Asia Dependent on whether the challenge is socio economic or from the people or the elite can change how particular types of Western linkage influence the behavior of these regimes. Additionally, although this research is primarily interested in exploring how Western linkages condition the coercive responses of these regimes when faced with challenges to legitimacy, these models will show that in many cases Western linkages alone help to increase the use of low intensity coercion. Model 1: Coercion and Horizo ntal, Vertical and Socio Economic Challenges The first model shows that specific types of challenges can result in different types of behavior from the regimes of Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan. The first hypothesis, that challenges t hat arise from society will result in an increase in the use of coercion from these regimes, produces some interesting results. The findings show that when there are instances of terrorism and ethnic violence the

PAGE 39

29 government will increase violent forms of c oercion. These results estimate that for every three terrorist acts or instances of large scale ethnic violence in these nations there is an expected one point increase in the level of high intensity coercion, which on a 4 0 scale is dramatic. The relatio nship between the instances of protest and high intensity coercion, is very close to being statistically significant (.12), but has a negative impact on the use of this more violent form of repression from these regimes. These findings show that when these regimes face terrorist acts or there is an outbreak of ethnic violence that they will respond with an increase in the use of high intensity coercion, but when there are protests this form of repression decreases. Though this is not a significant relations hip, it is interesting to note that the type of vertical challenge produces different reactions from these regimes. Regarding low intensity coercion, the model results provide no support for the first hypothesis. This indicates that there does not appear to be a direct relationship among protests, terrorism, and ethnic violence and low intensity coercion. Though it might be assumed that there would be a restriction of the freedoms included in low intensity repression, such as freedom of the press or speec h, in the face of vertical challenges this does not appear to be the case in these Central Asian nations. No evidence is found in support of the second hypothesis, that economic downturn will increase the use of coercion, as it pertains to the use of hi gh intensity coercion. However, a strong relationship with low intensity coercion is produced, though it is not in the predicted direction. The negative coefficient indicates that as the economy declines the use of low intensity coercion actually decreases It was hypothesized that when a regime is challenged with the instability created by economic decline this

PAGE 40

30 problem could translate into dissatisfaction from the people and in response to this type of challenge a regime will use repression to maintain con trol over the population. The results from this model indicate that these regimes react oppositely and actually decrease the level of low intensity coercion in the face of economic decline. Table 1: Model 1 Coercion and Challenges High Intensity Coercion Low Intensity Coercion H1: Protest .325 (.21) .010 (.40) H1: Terrorism .345*** (.05) .124 (.18) H1: Ethnic .370** (.18) .069 (.59) H2: Economic Decline .064 (.62) 2.876*** (.06) H3: Inflation .001*** (.00) .002* (.00) H4: Arrest s .045 (.14) .569** (.28) H4: Violence .257 (.22) .323 (.22) H4: Public Criticism .143 (.13) .178 (.28) H4: New Opposition .1178 (.15) .109 (.17) H4: Defection .154* (.08) .388 (.34) N 72 72 R squared 0.367 0.301 *p<.10 **p<.05 ***p<.01 The third hypothesis, which proposes that the challenge of inflation, because it may destabilize society, will increase the use of repression is also tested. The results indicate that there is a relation ship between inflation and levels of coercion, but one that

PAGE 41

31 is negative. Both dependent variables are statistically significant, but indicate that when inflation increases these regimes actually decrease the use of violent repression and low intensity coer cion. Though it was theorized that as inflation increases there will be a corresponding increase in the possible vertical challenges in a nation, and thus the regime will choose to implement coercive techniques to maintain stability, this does not appear t o be the case. The fourth hypothesis, which states that elite divisions will lead to increased levels of repression or coercion, also did not produce the expected relationships. Most of the components used to measure the level of elite division did not p roduce significant findings, with two exceptions. Surprisingly, this model shows that when defection of the elite increases there is actually a decrease in the use of violent forms of repression. For instance in 2004 Kyrgyzstan experienced a very high leve l of defection, but there was little violent repression. Even though important members of the ruling party, such as Kurmanbek Bakiev, left the governing party and established new voting blocs to oppose the current regime of Askar Akayev, this did not tran slate into an increase in the use of more visible forms of repression in Kyrgyzstan. These defections in 2004 show the elite movement away from Akayev's regime, who was ultimately overthrown the next year and replaced by a government headed by the defector Bakiev. Though it would be expected that when a regime is losing legitimacy to such a degree that important elite leaders are defecting, the regime would use both forms of coercion to maintain power, this model and example find the opposite relationship. Additionally, this model also finds that low intensity coercion also decreases with arrests of elite and opposition. This shows that as arrests of leadership and opposition

PAGE 42

32 increase there will be a decrease in the use of the non violent forms of coerci on, such as restricting freedoms of speech, press, or the judiciary as a few examples. These findings indicate that there is not the expected reaction from these regimes in their use of coercion when faced with these types of elite divisions. In sum, thi s model shows that there are specific types of challenges like terrorist acts and ethnic violence that do result in the hypothesized reactions from these regimes, but the other variables show that challenges such as economic declines, rising inflation and elite divisions, do not lead to increased levels of coercion. One of the most surprising findings in this model was with the socio economic variables of inflation and economic decline which show that when challenged by these types of problems, these regim es will actually move away from violent repression and broaden liberties and freedoms, in most cases. However, it is possible that these contrary findings are being skewed by the first few years of independence in these nations. After the collapse of the Soviet Union the newly independent nations in Central Asia experienced economic instability and inflation at staggering levels. For instance within the first few years of independence the Kyrgyz people experienced a drop in their per capita GDP of 20 perc ent in 1994 compared to the previous year, while the Tajik people saw almost a 30 percent drop in 1992. During this time there was an intense push from the West for these nations to transition from communism to democracy. This pressure in the face of stagg ering socio economic challenges may explain the negative findings for the first model. As the work by Gasiorowski (1995) discussed, timing matters when understanding the effects of economic challenges, and this period in the history of these nations may be unique, which may help to explain these negative

PAGE 43

33 findings. To explore this possible explanation the first three years were dropped from the model to test if these years were skewing the data in the negative direction. Table 2: Inflation and Coercion 19 93 2010 1996 2010 High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion Inflation .001 (.00) .003*** (.00) .003***(.00) .004(.00) N 72 72 56 56 R Squared .051 .132 .072 .001 *=p< .1 **= p< .05 ***=p< .01 ****=p< .001 By removing the first three years the significant relationship between inflation and for low intensity coercion does not hold but the expected relationship between rising inflation and an increase in high i ntensity coercion emerges. This shows that from 1996 2010 when these nations are experiencing periods of high inflation that the regimes do tend to resort to violent forms of repression in order to maintain control. This indicates that the first three year s were, to some degree, skewing the data in the case of inflation. Al though these years being removed exposed the expected relationship for inflation, the removal of these years actually indicates the robustness of these findings for economic decline. As the third table shows, even with the removal of the first three years the negative and significant findings that show that economic decline actually decreases the use of low intensity coercion, though lower is still significant. Unfortunately these findin gs together do not help to explain the results completely; rather they only further

PAGE 44

34 indicate the complicated nature of the relationship between these types of challenges and the regimes' use of coercion. Although, the inflation findings, with the removal o f the first years of independence, do produce the predicted relationship, the economic decline results are still difficult to explain. To make broad assertions about the possible meaning of these findings would require further research in Central Asia and the passage of more time to test whether these findings persist. Table 3: Economic Decline and Coercion 1993 2010 1996 2010 High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion Economic Decline .345 (1.10) 3.410*** (1.29) .662 (.66) 1.805** (.99) N 72 72 56 56 R Squared .002 .083 .008 .012 *=p< .1 **=p< .05 ***=p< .01 ****=p< .001 To provide another perspective, the following models incorporate interaction terms to t est how linkages to the West can influence the behaviors of these regimes. As discussed in the literature review, it is thought that when nations increase their linkages with the West, there will be a normative pressure to reduce the use of repression. The se models will explore whether these nations in Central Asia conform to these theoretical expectations. The second model tests both of the dependent variables against increased levels of inflation with the five linkage components that include the levels of economic, geopolitical, social, communication, and civil society ties to the West.

PAGE 45

35 Model 2: Inflation, Western Linkages, and Repression The second model, which tests the fifth hypothesis, indicates that there is a conditional effect of Western linkage s to a regime's use of coercion when faced with the challenge of rising inflation. All of the p values for the interaction terms, excluding economic linkages for both dependent variables and communication linkages for low intensity coercion, are significan t. What is interesting about these findings though, is that they each have a negative conditional effect on the dependent variable, except for geopolitical linkages. Table 4: Model 2 Inflation, Western Linkage and Repression Impact of Inflation on Coercio n with Linkages (Model 2) High Intensity Coercion Low Intensity Coercion H3: Inflation .015*** (.00) .008*** (.00) Link: Economic .001** (.00) .000 (.00) Link: Geopolitical .028*** (.01) .043 (.04) Link: Social .000 (.00) .001*** (.00) Link : Communications 0.179*** (.01) .058 (.06) Link: Civil Society .000 (.00) .000 (.00) Inflation*Economic .000 (.00) .000 (.00) Inflation*Geopolitical .000*** (.00) .000*** (.00) Inflation*Social .000*** (.00) .000*** (.00) Inflation*Commu nication .001*** (.00) .000 (.00) Inflation*Civil Society .000*** (.00) .000* (.00) N 72 72 R squared 0.356 0.538 *p<.10 **p<.05 ***p<.01

PAGE 46

36 Interpreting the interaction terms, there are three e xpected and one unexpected finding for the conditional influence of Western linkages and the use of coercion during periods of inflation. This model shows that the interaction between inflation, social and civil society linkages helps to decrease both form s of coercion in these nations. Additionally, the results show that communication linkages help to decrease the use of violent repression while inflation increases. Each of these interaction terms shows a negative relationship to the dependent variables, w hich is what was expected. In other words, when there are communication, social, and civil society linkages their interaction with inflation puts pressure on these regimes, to which they will respond with less violent repression and a loosing of restrictio ns on personal freedoms Looking at the Kazakhstan in 1993 confirms these findings. In this year, this nation was experiencing extraordinary levels of inflation, though accompanied a high amount of civil society linkages, which together produced no insta nces of high intensity coercion. This example indicates the conditioning effect of the civil society linkages, that the above table shows, that helped to deter the regime of Nursultan Nazarbayev away from more violent forms of repression. The unexpected finding in this model is the positive impact of geopolitical linkages on the use of coercion. This finding indicates that when a regime has increased levels of geopolitical linkages and is facing rising inflation the leadership will increase the use of bot h subtle and violent forms of repression This model also identifies a significant relationship between the inflation variable and both types of coercion in the absence of linkages. These findings show that when

PAGE 47

37 there is only inflation, and Western linkage s are missing, these regimes will move away from using both types of coercion. Though this study is interested in the interaction between Western linkages when these regimes are faced with challenges, these models also automatically measure the influence of these ties when the specific challenges are absent. Even though it seems contrary to the intent of this study to include a discussion of these findings, because they are so often significant within these models and they also produce interesting trends, they merit discussion. The coefficient for geopolitical and communication linkages in the case of high intensity coercion also shows that these types of linkages when inflation is zero will decrease violent repression in these countries. Of course this is the expected relationship, because as discussed before it is thought that as linkages increase, violent forms of repression will become too costly for these regimes to use. Conversely, this model finds that when there is no inflation economic linkages inc rease the use of violent forms of repression. Additionally, this model finds that when inflation is zero, social linkages increase the use of low intensity forms of coercion. These findings show that there are some types of linkages that reduce the use of the different types of coercion, but others that do not. The following models will continue to expose the complicated nature of these relationships. Model 3: Economic Decline, Western Linkages, and Repression The sixth hypothesis proposes that when a regime is faced with economic decline that Western linkages will decrease the use of repression in these regimes. The third model tests the influence of these linkages on the use of both forms of coercion when a regime is faced with economic decline. The coefficients for the variables show an

PAGE 48

38 unexpected relationship for most of the linkages when a regime is not experiencing a period of economic hardship. This model shows that when these regimes are not experiencing periods of decreased economic performance economic, social and civil society linkages increase the use of low intensity coercion. Interestingly, this model seems to indicate that the less visible forms of coercion may be preferable when Western linkages increase, even when the regime is not expe riencing any problems. Table 5: Model 3 Economic Decline, Western Linkages, and Repression Impact of Economic Decline on Coercion with Linkages (Model 3) High Intensity Coercion Low Intensity Coercion H2: Economic Decline 1.125 (.82) 1.003 (1.06) Link: Economic .000 (.00) .002*** (.00) Link: Geopolitical .030*** (.01) .077* (.04) Link: Social .000 (.00) .001*** (.00) Link: Communications .002 (.02) .026 (.03) Link: Civil Society .000 (.00) .000* (.00) Econdown*Economic .002 (.00 ) .004 (.00) Econdown*Geopolitical .261** (.11) .150 (.33) Econdown*Social .001 (.00) .000 (.00) Econdown*Communication .055* (.03) .156 (.11) Econdown*Civil Society .000*** (.00) .000 (.00) N 72 72 R squared 0.217 0.488 *p<.10 **p<.05 ***p<.01

PAGE 49

39 Additionally, this model finds that when the economy is not declining, geopolitical linkages increase the use of high intensity coercion. The only expected finding in the linkage variables shows tha t when there is no economic decline geopolitical linkages will decrease the use of the more subtle forms of repression in these countries. Like the model that tested inflation, this model shows that when the economies in these nations are not experiencing economic hardships, economic, social, and civil society linkages actually appear to increase restrictions on civil liberties and freedoms, which as the following discussion will show is a pattern between each of the models that explore Western linkages. I n this model only one interaction term has a significant coefficient estimate for the dependent variable of high intensity coercion. Communication linkages in periods of economic decline condition the reaction of these regimes to decrease their use of v iolent repression. Surprisingly, there are significant, but positive findings for the conditional effect of geopolitical and civil society linkages on the use of high intensity coercion though. These results show that economic decline interacts with geopol itical and civil society linkages in a way that actually increases the use of the more brutal forms of repression. Interestingly, these findings show that in some cases certain forms of linkage may actually increase the chances that these regimes will use coercion. Altogether these findings show that when these regimes are experiencing economic hardships communication linkages will decrease the use of high intensity coercion, but that civil society and geopolitical linkages can increase violent repression. Because the second and third models examined the socio economic variables that produce such contrary results in the first model, the same process of dropping the first

PAGE 50

40 three years was conducted to see how the omission of these turbulent years changes th e results. As the following discussion will show, dropping the first years of independence in these nations helps to expose some more interesting trends in the data. Table 6: Inflation, Western Linkages, and Repression 1996 2010 1993 2010 1996 2010 H igh Intensity Coercion Low Intensity Coercion High Intensity Coercion Low Intensity Coercion Inflation .015*** (.00) .008*** (.00) .002(.01) .008 (.01) Link: Economic .001** (.00) .0003 (.00) .008 (.01) .020 (.01) Link: Geopolitical .0 28*** (.01) .043 (.04) .242 (.18) .875*** (.26) Link: Social .000 (.00) .000*** (.00) 0.000 (.00) .004 (.00) Link: Communications .179*** (.01) .058 (.06) .396** (.19) .684 (.50) Link: Civil Society .000 (.00) .000 (.00) .000 ** (.00) .000** (.00) Inflation*Economic .000 (.00) .000 (.00) .000 (.00) .000 (.00) Inflation*Geopolitical .000*** (.00) .000*** (.00) .001 (.00) .005** (.00) Inflation*Social .000*** (.00) .000*** (.00) .000 (.00) .000* (.00) Inflation*Communication .001*** (.00) .000 (.00) .002** (.00) .004 (.00) Inflation*Civil Society .000*** (.00) .000* (.00) .000** (.00) .000** (.00) N 72 72 56 56 R Squared 0.356 0.538 .246 .592 *=p< .1 **=p< .05 ***=p< .01 ****=p< .001

PAGE 51

41 Table 7: Economic Decline, Western Linkages, and Repression 1996 2010 1993 2010 1996 2010 High Intensity Coercion Low Intensity Coercion High Intensity Coercion Low I ntensity Coercion Economic Decline 1.043 (.83) .848 (.99) 1.452* (.88) .885 (2.44) Link: Economic .000 (.00) .002*** (.00) .003 (.00) .001 (.00) Link: Geopolitical .030*** (.00) .077* (.04) .027* (.02) .032 (.08) Link: Social .000 ( .00) .001*** (.00) .000 (.00) .001*** (.00) Link: Communications .002 (.00) .026 (.03) .025 (.03) .008 (.04) Link: Civil Society .000 (.00) .000* (.00) .000 (.00) .000 (.00) Econdown*Economic .002 (.00) .004 (.00) .003* (.00) .000 ( .00) Econdown*Geopolitical .261** (.11) .150 (.33) .127 (.12) .529 (.61) Econdown*Social .001 (.00) .001 (.00) .002 (.00) .002 (.00) Econdown*Communicati on .055* (.03) .156 (.11) .096** (.04) .076 (.15) Econdown*Civil Society .000 *** (.00) .000 (.00) .000** (.00) .000 (.00) N 72 72 56 56 R Squared 0.217 0.488 .012 .518 *=p< .1 **=p< .05 ***=p< .01 ****=p< .001 When the first three years are dropped the negative relationship between civil society linkages and inflation on high and low intensity coercion reverses, but remains positive in the case of economic decline. In other words, the 1996 2010 model shows that civil society linkages interact with the c hallenge of inflation and actually increase the use of repression in these nations. The removal of these years also confirms that the presence of communication linkages negatively conditions the violent responses of these regimes.

PAGE 52

42 This shows that when the re are socio economic challenges, communication linkages help to decrease the use of coercion. The models for 1996 2010 also show a positive conditional relationship between economic decline and economic linkages with high intensity coercion. This means th at when there are economic linkages and these economies are declining, this model predicts that these regimes will increase their use of more brutal forms of repression. From all of this, it appears that socio economic challenges increase the use of coerci on when there are civil society linkages, but decrease the use of high intensity coercion with communication and social linkages. These findings seem to indicate that the type of challenge changes the effect of Western linkages, considering these two mode ls share several important similarities in their results. The following discussion will show that the models that test the socio economic factors that can destabilize a regime, inflation and economic decline, produce completely opposite effects to those th at test domestic challenges, such as terrorist acts, defection and protest. There seem to be distinct differences between how a regime responds while in periods socio economic challenge that changes the effect of specific types of Western linkages. The fin dings from the following models show that the type of challenge, whether socio economic or domestic, changes which types of linkage increase or decrease the use of repression in these cases. Model 4: Terrorism, Western Linkages, and Repression Based on the seventh hypothesis, the fourth model tests the conditional effects of Western linkages on the use of high intensity and low intensity coercion when the regimes in Central Asia have been challenged by acts of terrorism. This model shows that when face d with terrorist acts, without linkages, these regimes respond with the use

PAGE 53

43 of violent forms of repression. This finding is to be expected, but many of the others produced by this model are not. Table 8: Terrorism, Western Linkages, and Repression Impact of Terrorist Acts on Coercion with Linkages (Model 4) High Intensity Coercion Low Intensity Coercion H1: Terrorism .575*** (.11) .189 (.45) Link: Economic .000 (.00) .001** (.00) Link: Geopolitical .021* (.01) .032 (.07) Link: Social .000 (.00) .001*** (.00) Link: Communications .016 (.02) .018 (.04) Link: Civil Society .000 (.00) .000 (.00) Terrorism*Economic .001 (.00) .000 (.00) Terrorism*Geopolitical .010 (.00) .091* (.06) Terrorism*Social .000** (.00) .000** (.00) Terroris m*Communication .032*** (.01) .005 (.05) Terrorism*Civil Society .000** (.00) .000 (.00) N 72 72 R squared 0.383 0.514 *p<.10 **p<.05 ***p<.01 The coefficient estimate s for economic and social linkages when tested against low int ensity coercion are both positive and significant, which indicate that if there are no terrorist acts these linkages actually have a positive impact on the use of subtle forms of repression. This is similar to positive relationship found for low intensity coercion in the

PAGE 54

44 previous models, with the exception of geopolitical linkages in the case of economic decline. This model also shows that geopolitical linkages, when there are no instances of terrorism, decrease the use of violent repression. The interact ion terms in this model produce more surprising findings. Out of the five significant interactions using both of the dependent variables, three of these conditional effects are in the positive direction, which is not what the theories would predict. In the case of high intensity coercion a positive conditional relationship exists between social linkages and terrorism, as well as between communication and terrorism. This means that in these Central Asian nations when these regimes are challenged through ter rorist acts, communication and social linkages actually increases the use of the more violent a visible forms of coercion. The relationship here is positive, unlike the previous models that looked at inflation and economic decline, which found that social and communication linkages decrease the use of repression. This discrepancy can be explained though. In the context of the US led "War on Terror" it may be expectable for these regimes to respond by any means necessary when they are able to convincingly a rgue that the challenge is from a terrorist group. Of course in many cases these regimes quickly defend their more brutal reactions on fighting terrorism, which may be a reason why some Western linkages actually positively affect the use of coercion. In th ese cases it may be seen as an acceptable excuse to the West. Additionally, in the case of low intensity coercion there exists a positive conditional relationship between terrorism and geopolitical ties. This indicates that this form of linkage actually increases the likelihood of more subtle forms of coercion, such

PAGE 55

45 as the restriction of freedoms, when a regime is faced with increased terrorism. Possible explanations for all of these unexpected findings will be discussed in the following chapter. There are two interaction terms though, that produce the expected negative conditional effects. This model finds that civil society linkages decrease the use of high intensity forms of repression when these regimes are experiencing increased acts of terrorism. This shows that when there is an increase in civil society linkages, such as funding for democratization programs or NGOs on the ground, that even when a regime faces terrorist actions these linkages help to decrease the use of violent repression. What is important to understand about these findings is that with low levels of civil society linkages these regimes still tend to continue to resort to high intensity forms of coercion. For instance, in Tajikistan in 1997 there was great social unrest that was ex hibited through the high level of terrorist actions that year, including the kidnapping of 15 Red Cross workers (Keesing's, 41Feb41500). This year also experienced an increased level of instances of torture and extrajudicial killings. This relationship is to be expected, particularly with the knowledge that this year was also accompanied by very low, comparatively, levels of civil society linkages. The final expected finding for this model is that it shows, that when faced with terrorist acts, social lin kages can decrease the use of low intensity coercion. This indicates that increases in social linkages, such as heightened levels of migration to the US, will decrease the use of less intense forms of repression when a regime is challenged by acts of terro rism.

PAGE 56

46 This model, compared to the inflation and economic decline linkage models, produced similar findings in that they each seem to show that linkages, when challenges are absent, seem to increase the use of low intensity coercion. The findings containe d in this model though, stand apart from the previous models in that it shows that there is an increase in the use of violent forms of repression when communication linkages and terrorism are present, while the socio economic models found that these linkag es decreased the use of coercion. This indicates that these types of linkages have the opposite effect when these regimes are challenged by terrorist acts. Model 5: Elite Defection, Western Linkages, and Repression The eighth hypothesis predicts that r epression will decrease in the face of elite defection when there are increased linkages to Western nations. The fifth model tests the conditional effect of these ties on the use of both types of coercion when the regime is challenged horizontally by elite s defecting from the leader's party. This model identifies multiple factors that can influence these regimes' use of coercion, which further indicate the complicated relationship that exists between different forms of linkages and how these relationships t ranslate into a regime's use of coercion. One of the most surprising findings is that defection has a positive effect on the use of high intensity coercion, but a negative impact on the use of low intensity forms of coercion when there are no linkages. Thi s shows that when these regimes are challenged in this way, their response is to use more violent forms of repression rather than more subtle forms that restrict the freedoms of the people. In fact when these regimes are experiencing an increase in elite d efection they increase their use of violent and visible repression while decreasing the use of low intensity. Additionally, this model finds that in years when there is no elite defection,

PAGE 57

47 economic, social, and civil society linkages actually increase the use of low intensity coercion, which is similar to the previous linkage models. Just as in the second and fourth models, the fifth model finds that geopolitical ties decrease the use of high intensity coercion when defection is zero. Table 9: Elite Defec tion, Western Linkages, and Repression Impact of Defection on Coercion with Linkages (Model 5) High Intensity Coercion Low Intensity Coercion H4: Defection .241*** (.02) .536*** (.14) Link: Economic .000 (.00) .001** (.00) Link: Geopolitical .05 6*** (.02) .039 (.03) Link: Social .000 (.00) .001*** (.00) Link: Communications .006 (.02) .0317 (.02) Link: Civil Society .000 (.00) .000*** (.00) Defect*Economic .000 (.00) .000 (.00) Defect*Geopolitical .043*** (.01) .125*** (.02) Defe ct*Social .000 (.00) .000 (.00) Defect*Communication .030*** (.01) .004 (.00) Defect*Civil Society .000*** (.00) .000 (.00) N 72 72 R squared 0.174 0.518 *p<.10 **p<.05 ***p<.01

PAGE 58

48 The interac tion terms show several unexpected positive conditional relationships. Similar to the other interaction models discussed, these findings show that positive conditional effect of geopolitical linkages and defection on the use of both forms of repression. Th e previous models show, that geopolitical linkages actual increase the use both types of coercion, in most cases, when confronted by the various challenges. Additionally, this model identifies a positive conditional relationship with communication and defe ction and the use of high intensity coercion. This means when experiencing increasing levels of defection, higher communication ties will actually increase the use of violent repressive techniques. Just like the model that looked at how regimes react to te rrorism, this model shows Western communication linkages, when faced with defection, actually increases the use of violent repression. These findings indicate that, contrary to what was expected, communication linkages may be viewed by these regimes as li abilities to which, when challenged domestically, these Central Asian nations react violently. In this model there is one interaction term with a significant and negative effect. This model finds that the interaction between defection and civil society linkages decreases the use of high intensity coercion. Similar to the model that tested terrorism, though it is the opposite when compared to the socio economic models after 1996. These findings suggest an emerging trend between the models: one that indic ates that socio economic challenges and domestic challenges, such as terrorism or in this case defection, have dramatically different effects on the influence of Western linkages. The other important trend in all of theses linkage models discussed thus far show that linkages when the specific challenges are absent increase the use of lower intensity forms of

PAGE 59

49 repression. Although this study intended to explore how Western linkages condition the repressive responses of these regimes when they are challenged, because these models also inadvertently tests the lack of challenge as well, these findings must also be discussed particularly because they expose some interesting trends. Model 6: Protests, Western Linkages, and Repression The final hypothesis in this study states that Western linkages when regimes are faced with protests will decrease the use of repression. The sixth and final model tests this idea, but finds that only economic linkages produce the predicted decrease in the use of coercion. Table 10: Protests, Western Linkages, and Repression Impact of Protest on Coercion with Linkages (Model 6) High Intensity Coercion Low Intensity Coercion H1: Protest .263 (.21) .060 (.20) Link: Economic .000 (.00) .000 (.00) Link: Geopolitical .022 (.02) 0356 (.02) Link: Social .000 (.00) .001*** (.00) Link: Communications .006 (.02) .005 (.02) Link: Civil Society .000 (.00) .000** (.00) Protest*Economic .001 (.00) .001*** (.00) Protest*Geopolitical .016 (.07) .114*** (.03) Protest*Social .0 00*** (.00) .001*** (.00) Protest*Communication .003 (.01) .004 (.01) Protest*Civil Society .000 (.00) .000 (.00) N 72 72 R squared 0.284 0.566 *p<.10 **p<.05 ***p<.01

PAGE 60

50 This model finds two significant linkages when tested against t he dependent variable of low intensity coercion. According to the model's estimates, when there are no protests civil society and social linkages increase the use of less visible forms of coercion. This means that when there are no vertical challenges, tha t come in the form of popular protests, these types of linkages tend to result in a restriction of freedoms and liberties of the people. This model, just like all of the other linkage models, for the most part, shows that increased linkages in these nation s increase the use of low intensity coercion. This model also produces four significant interaction terms, three of which impact the use of low intensity repression and one that influences the use of high intensity coercion. There is a positive conditiona l effect on the use of low intensity coercion with the interaction between protest and geopolitical ties. Which means that when there are periods of increased protest, higher levels of geopolitical ties in these nations will increase the use of less violen t forms of repression. Similarly, social linkages together with protests also increase the use of both violent and subtle forms of repression in these nations. This model, interestingly, finds a negative conditional effect of economic linkages and protest on the use of low intensity coercion. Unlike any of the previous models, this shows that when faced with protests, economic linkages actually deter these regimes from restricting the rights of the people through low intensity repression. The increased us e of coercion when social linkages and protest work together also mimics the findings in the models that tested terrorism and defection.

PAGE 61

51 Table 11: Models 2 6 Linkage Interaction Terms Together all of these findings further confir m that when these countries are challenged domestically (through terrorist acts, protest, or defection) civil society Linkage Interaction Terms High Intensity Coercion Model 2 (1996 2010) Model 3 (1996 2010) Model 4 Model 5 Model 6 Challenge* Economic + Cha llenge* Social + + Challenge* Communication + + Challenge* Civil Society + + Low Intensity Coercion Model 2 (1996 2010) Model 3 (1996 2010) Model 4 Model 5 Model 6 Challenge* Economic Challenge* Social + Challenge* Communication Challenge* Civil Society + Expected Results decreased coercion Unexpected Results + increased coercion No Relationship

PAGE 62

52 linkages generally decrease the use of repression, while social and communication linkages increase the use of coercion. Conversely, the m odels that test socio economic challenges after 1996, find that social and communication linkages decrease the use of violent and subtle forms of repression, while civil society linkages increase the use of repression in these Central Asian nations. All of the linkage models together produce three important trends: in most cases linkages without challenges increase low intensity coercion, geopolitical linkages have a positive conditional relationship to all of the different types of challenges, and that the effect of linkages depends on whether the regime is experiencing socio economic problems or challenges from society or the elite.

PAGE 63

53 CHAPTER V DISCUSSION The previous chapter presented the results from the six models in this study. There are several important findings produced by these models that require further discussion because, for the most part, many of the results were contrary to expectations. The following chapter will propose explanations for the surprising findings from the first m odel and the models that tested Western linkages. This discussion will focus on the unexpected findings produced in the first model and the three important patterns that emerge from the five models that test the influence of Western linkages. After a brief review of the first model, the three major trends in the linkage models will be reviewed. The linkage models show: that geopolitical linkages interact with the various challenges in a way that tends to increase the use of coercion, that when there are Wes tern linkages without challenges these regimes tend to heighten the restrictions on the rights and liberties of the citizens, and finally that some types of Western linkages can increase or decrease the use of repression depending on whether a nation is be ing challenged socio economically or domestically. The First Model The first model, which did not test Western linkages, but rather tested how the regimes in these Central Asian nations react to vertical, horizontal, and socio economic challenges exposed some unexpected findings that require some explanation. This model found that when these regimes are being challenged by terrorist acts or through the instability created by ethnic violence they will react violently to these problems. These findings provi de support for the proposed hypothesis that social challenges will increase

PAGE 64

54 the use of coercion, though the third component to this concept, protests, did not increase the use of force in these countries. This shows that these regimes find terrorist acts a nd ethnic violence worthy enough challenges to react with more violent and forceful types of repression in order to maintain stability. This reaction from these regimes would be predicted, surprisingly though, the other concepts tested in this model did no t conform to the hypothesized expectations. Table 12: Model 1 Expected and Directional Results Model 1: High Intensity Coercion Low Intensity Coercion Protest Terrorism + Ethnic Violence + Economic Decline Inflation Arrests of Elite or Opposition Violence Against Elite or Opposition Public Criticism from Elite or Opposition New Opposition or Coalitions Defection o f Elite Expected Results + increased coercion Unexpected Results decreased coercion No Relationship

PAGE 65

55 One of the most important findings in this study, while being one of the most unexpected, was that it appears that when these regimes face econ omic problems they actually increase freedoms and reduce their more brutal policies. It was theorized that when a nation is facing economic hardships or increased inflation, because of fears of social unrest, these regimes would increase their use of repre ssion in order to maintain control over their dissatisfied population. The results from this study show, however, that this is not the case in these nations; in fact it appears that these types of challenges actually induce these regimes to loosen their gr ip on this type of control. Because these findings were so contrary to expectations, it was proposed that there was a possibility that the first few years for these nations may have skewed the data. When the Soviet Union collapsed and these nations in Ce ntral Asia became independent they were reeling economically, while being pressured by the West to institute democratic systems. Dropping these first years show that inflation does increase the use of violent forms of repression. This is important because the same expected findings do not exist when the years 1993, 1994 and1995 are dropped when examining the influence on economic decline. This same process only confirmed that economic declines decrease the use of low intensity repression. This shows that d uring periods of increased inflation these regimes do respond with the expected violent crackdowns, the same cannot be said in the case of periods of economic decline. Surprisingly, these findings show that when the economy is experiencing a downturn these regimes actually expand the rights and freedoms of the people.

PAGE 66

56 Table 13: Socio Economic Hypotheses 1993 2010 vs. 1996 2010 It would be expected that these regimes would react with violence when challenged by the instability of high inflation. It makes sense that in order to maintain control that this form of repression would be used to combat the possible dissatisfaction created by this threat to the economic wellbeing of the people. It is a bit more difficult to explain why a regime would broaden freedoms when the economy is bottoming out. It is possible that these regimes may be willing to loosen their control over some aspects of life and broaden the rights, liberties and freedoms in these nations in order to possibly 1993 2010 1996 2010 High Intensity Coercion Low Intensity Coercion High Intensit y Coercion Low Intensity Coercion Economic Decline Inflation + Expected Results + increased coercion Unexpected Results decreased coercion No Relationship Socio Economic Hypotheses 1993 2010 vs. 1996 2010

PAGE 67

57 attract Weste rn investment. These regimes may believe that appearing to uphold more Western ideals of may help attract businesses, which when the economy is faltering would be exactly what is needed. It would certainly help to soothe the civilian dissatisfaction with e conomic hardships if there were new jobs and opportunities arising in these nations. These findings indicate that these regimes may view inflation as a greater danger to stability than economic decline. Though this distinction requires further research, it is very interesting that they type of socio economic challenge changes how these regimes respond. Although this is one of the most interesting findings produced by this model, the fourth hypothesis tested and its results must also be reviewed. The elit e divisions hypothesis also produced unexpected findings, although most of the variables did not significantly affect coercion. This model found that when there is an increase in arrests of the opposition or the elite that these regimes will actually decre ase the use of low intensity repression. This model also found that when the elite begin defecting from the leader's party the regime will reduce violently repressive measures. These are interesting findings because it was hypothesized that arrests would i ndicate deepening elite divisions, which could challenge the regime in power, and in order to combat this possible growing problem these regimes would increase in the use of coercion. It is surprising that these finding indicate that these regimes actuall y loosen their repressive grip even when it appears that the elite might be fracturing. Since these two significant results are so unexpected, and many of the components in this hypothesis did not produce any important results, possible explanations for th ese findings must be discussed.

PAGE 68

58 First, the data collection process through document analysis may not have exposed all of the elite divisions that were present. As was discussed in the third chapter, divisions emerging within the elite are not necessarily public knowledge and may occur behind the scenes. That being said, the instances that were found in the Keesing's World News reports may not have fully captured the instances of these types of divisions. The other possibility is that there may be ways in which these regimes respond to horizontal threats that is not through repression of the entire society, either through high intensity or low intensity forms of coercion. These regimes could influence the elite through promotions or demotions to maintain lo yalty when challenged or punish or reward specific groups, tribes, or families, which would not translate into a restriction of freedoms or violent forms of repression on the ground. It is possible that there exists a much more dynamic use of power within the elite power structure that was not captured by the dependent variables in this study. The first model indicates how dynamic the responses of authoritarian regimes may actually be. These findings show that these regimes react with violence when stabi lity is challenged by inflation, but actually increase rights and freedoms when the economy is declining. Though surprising, these results show that there are important distinctions between the type of socio economic challenge that drastically alters the r eactions of these regimes. The following discussion will also highlight how important understanding the type of challenge is when predicting the reactions of these regimes, which will be shown in each of the Western linkage models. As mentioned before, the re are three major trends that can be found across all of linkage models, which include: the interaction of geopolitical linkages heighten the use of repression, linkages without

PAGE 69

59 challenges increase the use of low intensity coercion, and that socio economi c and domestic challenges change the influence of Western linkages. Geopolitical Linkages The first interesting trend in the interaction models shows that geopolitical linkages, when a regime is challenged by inflation, economic decline, terrorism, defe ction or protests, tend to increase the use of repression. In other words, when these regimes are challenged the presence of geopolitical linkages actually induces these regimes to increase their use of repression. This is most likely attributable to the f act that the measure did not accurately capture the concept. The level of geopolitical linkages was determined by counting the number of UN treaties that were signed in each year by the Kyrgyzstan, Kazakhstan, Tajikistan and Uzbekistan. Though this was not the most optimal way to measure this type of linkage, the limited information from the region constrained the possibilities for other types of measure. This, however, may help to explain the positive results for most of the interaction terms that included geopolitical linkages. It is possible that these nations signed these treaties in an attempt to appear to be working with the international community, but that the intentions behind these gestures (such as signing the Convention on the Political Rights of Women 44 years after it was first ratified) was actually to distract from the high intensity and low intensity coercive techniques that these regimes were using on the ground (Kyrgyzstan UNTS). In other words, the regimes were using repression to contro l their people, and choose to sign on to these treaties to signal to the international community that they were making an attempt to work within the expected framework of norms, even though their use of coercion

PAGE 70

60 during these periods was not what the West m ay have approved of. Unfortunately, this measure may have failed because the intention of these regimes to enter into these treaties was not assumed to be connected to their use of repression, which these finding indicate that it might be. Without Challen ges, Low Intensity Coercion Increases Table 14: Linkages Increase Low Intensity Repression when Challenges are Absent The second interesting trend that emerges from the five Western linkage models finds that when challenges are absent these r egimes actually begin restricting the rights of the people through low intensity coercive techniques. The purpose of this research was to explore how Western linkages interact with challenges in these regimes, but these models Low Intensity Coercion when Challenges are Absent Model 2 Model 3 Model 4 Model 5 Model 6 Economic Linkages + + + Social Linkages + + + + + Communication Linkages Civil Society Linkages + + Expected Results decreased coercion Unexpected Results + increased coercion No Relationship

PAGE 71

61 actually show how these linka ges interact with the levels of repression when the specific challenges are not present.. Table 14 shows that the more subtle forms of coercion actually increase with Western linkages. These results indicate that Western influence does not necessarily gu arantee the removal of repression, as these linkages do not show a decrease of high intensity coercion, but that it may have influenced these regimes toward using low intensity forms of coercion. Though these are not decisive findings, as they do not show that there may be a trend away from high intensity repression, they do show that there seems to be an increase in less visible forms of repression with these linkages. One of the ways to explain this would be to view these increases as a movement away from violent repression because these visible actions may draw negative attention from the West. These findings are not completely inconsistent with the previously discussed theories of Levitsky and Way (2006) though. Linkages are thought to increase the risk of using violent repression, because as linkages increase, the possibility that these brutal techniques will be noticed and condemned will be much higher, while low intensity forms of coercion might be thought of as preferable because they are much more di fficult for Western nations to notice. This helps to explain why these regimes are opting for low intensity repression while these Western linkages are being formed. This is an important finding that also requires further study into the relationship betwe en linkages and low intensity coercion in authoritarian regimes. Socio Economic vs. Domestic Challenges The final important trend this research exposes is that the type of problem that challenges these regimes changes how Western linkages influence their behavior. There

PAGE 72

62 appear to be two distinct types of challenges that alter the effect of some linkages. The findings from the final five models show that if there is a socio economic challenge that communication and social linkages decrease the use of coerc ion, while when the challenge is from the people or the elite, economic and civil society linkages can decrease the use of repression. These models also produce findings that show that, depending on the type of challenge, there are some linkages that incre ase the use of coercion. When there are challenges from society and the elite, communication and social linkages interact with these types of challenges, and these regimes respond with higher levels of repression. Additionally, when the first three years a re removed from the inflation and economic decline models, civil society and economic linkages sometimes increase the use of repression. Figure 1: Socio Economic and Domestic Challenges and Linkages Challenges and Effect of Linkages Socio Economic Challenges Domestic Challenges Communication Social Economic (after 1995) Civil Society (after 1995) Economic Civil Society Communication Social

PAGE 73

63 Communication and social linkages seem to be particularly important in understanding the responses from these regimes, both of which influence the regimes differently depending on the type of problem that arises. Though it is proposed, within the previously discussed theories of Levitsky and W ay (2006), that increases in linkages will help further democratization, these findings indicate that the type of linkage may be considered by these regimes to be liabilities in specific situations. Because the interaction between domestic challenges, comm unication and social linkages in these models actually show an increased use of repression, it is possible that these regimes may be responding to these types of linkages as if they are also threats as well. Communication and social linkages seem to be liabilities to these regimes when terrorist acts, defection, and protests arise, but when challenged by inflation and economic decline these linkages decrease the use of coercion used by these regimes. These two forms of linkages capture the level of infor mation and connection between the citizens of these nations and the West. These two types of linkages are thought to influence regimes because they open up opportunities for people within these countries to access information through communication, and pro mote the transfer of ideas across borders, which may have a democratizing influence on the populous in these nations, while also making these regimes vulnerable to international criticism. It is possible that access to information and ideas can at times be considered a liability to these regimes, but it depends on the type of information that could challenge the status quo. This possible consideration may help to explain why these regimes have such varied reactions depending on the type of challenge.

PAGE 74

64 The models that look at challenges that arise domestically, whether from the people or from the elite, show that in most cases the presence of communication and social linkages in the face of these problems actually increase the use of violent repression and s ome times low intensity forms as well. From these findings it is clear that not only do these types of linkages not deter these regimes, when faced with domestic challenges, but they actually increase the use of repressive techniques. This access to inform ation and migrants in Western countries may be viewed by these nations as possible challenges to the authority of the regime. This may induce reactions from these governments that utilize both forms of repression, because of fears of losing power. If a reg ime were losing legitimacy, or felt that it might be, then it could consider access to other ideas through communication and social ties to the West a liability to their continued control. The particular types of ideas that might be of concern to these reg imes might challenge the level of democratization, freedoms, and human rights concerns, just to name a few. Increased interaction with the West could allow the people in these Central Asian nations to question the ideology of the regime and the leadership. Access to these types of ideas, while these regimes are facing challenges from popular protests, elite defection, or terrorist acts, could be seen as a source of vulnerability to their power, thus exposure to different ideas that may challenge the regime' s behavior or ideology could be considered a liability and might induce these regimes to repress their people. Conversely, these types of linkages would not be nearly as concerning to these regimes when faced with inflation or economic decline, because t he ideas that the people in these nations may gain from access to the outside world would not challenge the current economic systems in these countries. In other words, because these nations have

PAGE 75

65 embraced free market capitalism, access to the West would no t challenge this system, since these nations have embraced the same economic forms. Though there are other possible explanations to these findings there does seem to be different effects of communication and social linkages depending on the type of challen ge these regimes in Central Asia face. Civil society linkages to the West, as measured through the directed aid in these nations that promote democratic policies, also has adverse effects depending on whether the challenge is socio economic or through a loss of support from the people or the elite. The models that look at terrorist acts and elite defection show that civil society linkages interact with the challenges and decrease the use of violent forms of coercion. This indicates that targeted monies fr om the West will actually deter these regimes from using violence against their people when they face issues within the ruling elite or because of terrorist acts. Using the years from 1996 to 2010 the socio economic models show that civil society linkages conditionally increase the use of both forms of coercion. Civil society linkages do decrease the use of violent forms of repression in the case of protest or defection, but the positive relationship that emerges after 1995 is difficult to comprehend. It i s possible that these regimes are pressured by democratization programs during episodes of domestic challenge, as would be predicted, but that when the economy is faltering these funds may disrupt, or be seen to disrupt society, thus leading to increased u se of both forms of coercion. As the previous findings show, these regimes do take domestic challenges seriously, if they did not they would not resort to repression as often, so it is possible that these civil society linkages are actually deterring these regimes away

PAGE 76

66 from repression as a way to control society. Because they react with repression in the face of socio economic challenges when there are civil society linkages, it seems as though in these situations this form of aid may also be viewed by thes e regimes as a problem, or a financial source of disruption in society. The impact of economic linkages seems much less influential, compared to the other forms of linkage, but these findings are also worth mentioning. The models show that economic linka ges, when they interact with the challenge of protest, actually reduced these government's restriction of the rights and freedoms of the people. Which means that when protests occur the presence of economic linkages decrease the use of low intensity coerci on. It is possible that this is because as there are increased economic ties between these nations and the West there will be a pressure for these regimes to expand some of the freedoms that are normally restricted when a regime is attempting to control th rough low intensity forms of coercion. This is similar to the explanation proposed earlier for the decreased use of this form of repression when regimes were directly faced with economic decline in the first model. Possibly these regimes, in an attempt to further their attractiveness to Western investors, at times choose to loosen their repressive hold on their people in order to further their economic prospects. This finding, because it does not exist in any of the other models must be further explored bef ore any broader conclusions about the connection between these pressures be espoused. What is interesting about this type of linkage though, is that when the first three years after independence are removed and these regimes experience an economic declin e, economic linkages actually increase the use of violent forms of coercion. This shows that economic ties, when there are economic challenges, increase the more negative aspects of

PAGE 77

67 authoritarianism. It could be proposed that during periods of economic dec line that because of the possibility for social disruption it is better for these regimes to repress through high intensity coercion in order to show Western investors that they have control, rather then appear less like dictators and more like Presidents. In other words, it is more important to appear to be a stable regime that can protect investments, rather then a weak government that might fall because of the social unrest produced from an economic decline. Again this is another finding that indicates that these regimes act very differently depending on whether the challenge comes from the people or if it is socio economic. These findings do not confirm the proposed influence of Western linkages, but they also do not entirely disprove the theories pr oposed by Levitsky and Way (2006). Rather these findings indicate that the type of challenge changes the influence of the type of linkage and that some linkages have more influence than others. The findings in this study indicate that there may be a much m ore complicated relationship between the use of coercion and specific types of challenges, that linkages to the West seem to increase the use of low intensity coercion, and that some linkages can have positive and negative effects depending on whether a re gime is facing a socio economic challenge or a domestic problem. From the findings in this study it is clear that there are many opportunities for further research, which will be discussed in the final chapter.

PAGE 78

68 CHAPTER VI CONCLUSION The findings pre sented in this study, though they contribute to the underwhelming catalog of Central Asian studies, only indicate that further work needs to be conducted. Because many of the results produced by the models explored in this work were far from the hypothesiz ed relationships in many cases, it is clear that there are several areas of research that can benefit from further exploration The major findings in this study, considering they are not at all what were expected, must be further researched and confirmed be fore conclusions can be drawn about their meaning and relation to the broader understanding of authoritarian regimes. There are three major findings produced by this research, each of which clearly indicates the dynamic nature of these authoritarian regime s and the complicated relationship between the use of repression and Western influence. Though the first model confirmed the positive relationship between the use of repression when a regime was faced with terrorist acts and ethnic violence, this model c ontained one of the more perplexing results: that economic decline and inflation decrease the use of repression in Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan. Though the removal of the first three years after independence produced the expected posit ive relationship between violent repression and inflation, this was not the case with the economic decline variable. The robust and negative relationship produced requires further research in Central Asia, without which these findings will merely remain co nfusing. These results, although somewhat problematic to understand, underscore how difficult it is to completely predict the behavior of these regimes.

PAGE 79

69 The second important finding produced by this research was found in almost all of the models that exa mined the conditional relationships between the use of low intensity coercion and Western linkage. All of the linkage models, to varying degrees, show that increased Western linkages in these nations tends to increase the use of low intensity coercion when challenges are absent. This finding is contrary to what was expected and is also justification for further research into the relationship between low intensity coercion and different forms of Western linkage broadly. The previous chapter proposed the poss ible explanation that these increases in low intensity forms of repression were actually because these regimes were moving away from high intensity tactics, but the models did not confirm the latter part of this assertion. The study did not show a decrease in high intensity coercion that was consistent, this of course, further proving that more research with a broader selection of cases is required to explore the meaning of these findings. The final important finding is that there appears to be specific t ypes of challenges, whether they are socio economic or domestic, that change the impact of the different types of linkages proposed by Levitsky and Way (2006). In the cases selected from Central Asia, communication and social linkages seem to conditionally decrease the use of repression when the challenge is socio economic. Conversely, these types of linkages appear to conditionally increase the use of coercion when these regimes are confronted by domestic challenges such as protest, terrorist acts, or defe ction of the elite. Civil society linkages also produce different findings depending on the type of challenges. These results indicate that the type of challenge that these regimes in Central Asia face changes the directional impact of the linkages that ar e present. These findings show that

PAGE 80

70 it is centrally important to understand the type of problem a regime is facing because it can fundamentally change the influence of Western linkages in these countries. Because these findings indicate that some forms of linkage may actually contribute to the use of repression, further research that includes a broader selection of cases is necessary to determine if these conditional relationships between linkages and coercion are unique to Central Asia or not. Additionally further research into these nations may require exploration of the relationships these nations have with the other important political actors in the region, such as Russia or China. Because China and Russia do not tend to promote democratization policies such as the West, the interaction between these strategic relationships may assist in creating a more complete understanding of these nations and their behavior. Broadening this understanding of the foreign pressures within these nations may also help to explain some of the more confusing findings produced by this research. Though there were some important findings that indicate a need for further research, the almost completely insignificant findings for the elite division components also merits furth er inquiry. As the first model shows there were only two significant findings for a relationship between the elite division variable and those results were in the opposite of the hypothesized relationship. As proposed in the last chapter, it is possible th at these variables failed to produce the expected result because these regimes are actually combating certain forms of horizontal challenges in a different way that does not directly impact the whole of society. It is possible that there are more dynamic r eward and punishment mechanisms that exist in these nations that help to combat these types of challenges that this research missed. Although there could have also been issues with the

PAGE 81

71 matter in which the information for these variables was collected, ther e are still clearly questions about how these regimes are able to work against elite divisions. Further research into these relationships is required to fully understand how control is maintained within the regimes of Central Asia. Together many of the f indings, or lack there of, show that the regimes in Uzbekistan, Kyrgyzstan, Kazakhstan and Tajikistan do not easily fit into the theoretical models or meet the expectations that were proposed. Continued research into these nations and the theoretical model s proposed is required to determine whether these findings indicate the uniqueness of these cases, or if there are broader patterns that this study may have unearthed.

PAGE 82

72 REFERENCES Alem‡n JosŽ and David D. Yang. 2011. A Duration Analysis of Demo cratic Transitions and Authoritarian Backslides. Comparative Political Studies 44:9 1123 1151. BBC. 2005. How the Andijan Killings Unfolded BBC. May 17, 2005. < http://news.bbc.co.uk/2/hi/4550845.s tm > [accessed September 24, 2013]. Bellin, Eva. 2012. Reconsidering the Robustness of Authoritarianism in the Middle East." Comparative Politics. 44:2, 127 149. Brambor, Thomas, William R. Clark, and Matt Golder. 2006. Understanding Interaction Mode ls: Improving Empirical Analyses. Political Analysis. 14:63 63 82. Brooker, Paul. 2009. Non Democratic Regimes New York: Palgrave Macmillian. Cingranelli, David L. and David L. Richards. 2010. The Cingranelli Richards (CIRI) Human Rights Dataset. Avai lable from http://www.humanrightsdata.org [accessed 1 March 2013]. Department of Homeland Security. 1993 2011. Yearbook of Immigration Statistics. Available from http://www.dhs.gov/yearbook immigration statistics [accessed 3 March 2013]. Francisco, Ronald A. 2005. After the Massacre: Mobilization in the Wake of Harsh Repression. Mobilization 9: 107 126. Gerschewski, Johannes. 2013. The Three Pillars of Stability: Legitimation, Repression, and Co optation." Democratization. 20:1, 13 38. Gandhi, Jennifer and Adam Przeworski. 2007. "Authoritarian Institutions and the Survival of Autocrats." Comparative Politics. 40, 1279 1301. Gasioro wski, Mark J. 1995. Economic Crisis and Political Regime Change: An Event History Analysis. The American Political Science Review 89:4 882 897. Gilbert, Leah and Payam Mohseni. 2011. Beyond Authoritarianism: The Conceptualization of Hybrid Regimes. Stud ies in Comparative International Development 46: 270 297. Hale, Henry E. 2011. Formal Constitutions in Informal Politics: Institutions and Democratization in Post Soviet Eurasia. World Politics 63(4): 581 617. Jaccard, James, Robert Turrisi and Choi K Wan. 1990. Interaction Effects in Multiple Regression. Newbury Park: Sage Publications.

PAGE 83

73 Juraev, Shairbek. 2008. Kyrgyz Democracy? The Tulip Revolution and Beyond. Central Asian Survey 27(3 4): 253 264. Johnston, Hank. 2011. States and Social Moveme nts. Cambridge: Polity Press. Karagaiannis, Emmanuel. 2006. Political Islam in Uzbekistan: Hizb Ut Tahrir Al Islami. Europe Asia Studies. 58(2): 261 280. Keesing's World News Archive. 1993 2011. Available from h ttp://www.keesings.com [accessed March 2013] Kuran, Timur. 1991. "Now Out of Never: The Element of Surprise in the East European Revolution of 1989." World Politics. 44:1, 7 48. Levitsky, Steven and Lucan Way. 2002. The Rise of Competitive Authorita rianism. Journal of Democracy. 13:2 51 65. Levitsky, Steven and Lucan Way. 2006. Linkage versus Leverage. Rethinking the International Dimension of Regime Change. Comparative Politics. 38:4 379 400. Levitsky, Steven and Lucan Way. 2010. Competitive Authoritarianism. New York: Cambridge University Press. Liu, Morgan Y. 2011. Central Asia in the Post Cold War World. Annual Review of Anthropology 40: 11 131. Megoran, Nick. 2004. Revisiting the pivot': the Influence of Halford Mackinder on Analysi s of Uzbekistan's International Relations. The Geographical Journal 170(4): 347 358. Murphy, Jonathan. 2006. "Illusory Transition? Elite Reconstitution in Kazakhstan 1989 2002." Europe Asia Studies. 58:4, 523 554. Neundorf, Anja. 2010. Democracy in T ransition: A Micro Perspective on System Change in Post Socialist Societies. The Journal of Politics 72:4 1096 1108. Petric, Boris Mathieu. 2005. Post Soviet Kyrgyzstan or the Birth of a Globalized Protectorate. Central Asian Survey 24(3): 319 332. Pr oksch, Marc. [no date] Regional and bilateral trade agreements in Asia Pacific and Central Asia: An Evaluation. UNESCAP. < http://www.unescap.org/tid/projects/globalize_wgmarc.pdf > [accessed September 12, 2013]. Radio Free Europe/Radio Liberty. 2010. Five Years After Andijon Events, Key Questions Remain Unanswered Radio Free Europe/Radio Liberty. May 11, 2010. <

PAGE 84

74 http://www.rferl.org/content/Five_Years_After_Andijon_Events_Key_Questions_ Remain_Unanswered_/2039096.html > [accessed September 24, 2013). Schedler, Andreas. 2002. The Menu of Manipulation. Jou rnal of Democracy. 13:2 36 50. Spechler, Martin C. 2007. Authoritarian Politics and Economic Reform in Uzbekistan: Past, Present, and Prospects. Central Asian Survey. I26:2 185 202. Sindelar, Daisy. With Pakistan Supply Routes in Doubt, Washington Lo oks to Uzbekistan. Radio Free Europe/Radio Liberty. September 29, 2011. < http://www.rferl.org/articleprintview/24344138.html > [accessed December 1, 2011]. United Nations Economic an d Social Commission for Asia and the Pacific. 2013. UNESCAP Dataset. Available from http://www.unescap.org/stat/data/ [accessed 3 March 2013]. United States Census. 2013. Trade in Goods Kazakhstan. Ava ilable from http://www.census.gov/foreign trade/balance/c4634.html [accessed 3 March 2013]. United States Census. 2013. Trade in Goods Kyrgyzstan. Available from http://www.census.gov/foreign trade/balance/c4635.htm [accessed 3 March 2013]. United States Census. 2013. Trade in Goods Tajikistan. Available from http://www.census.gov/foreign trade/balance/c4642.html [accessed 3 March 2013]. United States Census. 2013. Trade in Goods Uzbekistan. Available from http://www .census.gov/foreign trade/balance/c4644.html [accessed 3 March 2013]. United Nations. 2013. UNTS Treaty Collection. Available from http://treaties.un.org/pages/UNTSOnline.aspx?id=3 [ac cessed 30 March 2013]. World Bank. 2013. World Bank Dataset. Available from http://data.worldbank.org/ [accessed 4 March 2013].

PAGE 85

75 APPENDIX A. Document Analysis Coding Coding 0 1 2 3 4 H1: Protest no events 1 2 minor protests, relatively small in size (not against the regime, mainly basic complaints) 1 anti gov. 2+minor Or widespread mass protest, minors can be elevated if violently repressed 2 or more antigov., 2+ big protracted demons tratio ns, 4+minor 6 + widespread, antigov. Terrorism no events 2+minor, 1 2 2 3 (3 multiple fatalities) 3 5 6+ (4+ if they are high profile or high fatalities) Ethnic Violence no events minor medium major major with high fatalities H4: Arrests no events 1 arrest minor 1 major arrest (PM or VP) 2 3 arrests 2 major 3 5 minor 3 major 6+ minor Violence no events 1 minor (Beatings) 2 3 minor 1 major (murders or disappearanc es) 3 5 minor 2 major 6+ minor 3 major Public Criticism no events 1 2 3 3 5 6+ Emergence of New Parties no events 1 new party 1 bloc (2 9 parties) 2 3 new parties 3 5 new parties 2 bloc, or 1 bloc (10+parties) 6+ new parties, 2 large bloc Defection no events 1 minor player 1 major or 2 3 minor 2+ major 3+mino r 3+ major 4+minor

PAGE 86

76 B. Variable Ranges Ranges Dhigh 0 2 Dlow 0 2 Protest 0 4 Terrorism 0 4 Ethnic Violence Inflation 2 1546.7 Econdown (econdownneg) 0.4979757 0.3308958 Arrests 0 4 Violence 0 4 Public Criticism 0 4 N ew Opp 0 4 Defection 0 4 Centered Ranges Protest 0.4605263 3.539474 Terrorism 0.5921053 3.407895 Inflation Econdown 0.4341506 0.3947209 Defection 0.4078947 3.592105 Lecon 343.2803 2245.62 Lgeopol 2 29 Lsoc 1322.461 4 941.539 Lcomm 14.54531 68.05656 Lcivsoc 1.92E+08 3.23E+08