THE POWER OF VALUES IN DETERMINING INTERSTATE THREAT PERCEPTION

(1)

THE POWER OF VALUES IN DETERMINING INTERSTATE THREAT PERCEPTION

KATRĪNA MARIJA SITNIECE Master's Thesis

Spring 2021

Department of Peace and Conflict Research, Uppsala University Supervisor: Magnus Öberg

Word count: 21391

(2)

ACKNOWLEDGEMENTS

This is dedicated to my support team - my family and friends who have become my home away from home. I want to personally thank my advisor Magnus for his words of wisdom and unwavering belief in my abilities and Louis for his patience and support in helping me shape this study.

(3)

ABSTRACT

Interstate threat perception plays a vital role in peace and conflict, having the potential to lead to pre-emptive war. Despite the significance of the theme, little is known about the elements and mechanisms at play guiding threat perception at the state level. This study contributes to said gap by focusing on values as the driver of threat perception. The argument explores the regional constraints of threat perception, focusing on regional military interventions and their effects on the threat perception of the states within said region. Thus, the paper addresses the question of what role value congruence plays in interstate threat perception following a regional military intervention. The paper hypothesises that a higher perceived value congruence between the perceiving state and the intervening state lowers the degree of threat perception. Noting data limitations and isolation challenges, the study finds support for this hypothesis by exploring the case of 2014 Russian intervention in Ukraine and its effects on Belarus, Poland, and Sweden.

(4)

INTRODUCTION

The perception of threat plays a crucial role in the lives of states and individuals. By providing the link between “action and reaction” (Cohen 1978, 93), threat perception constitutes a vital component of self-defence mechanics, contributing to the survival of both humanity and political entities. Despite its pragmatic role, the perception as well as misperception of threat can also lead to conflict, potentially resulting in pre-emptive war. Although receiving attention in the last decades, the dynamics of interstate threat perception are extremely limited in terms of systematic research. Even though violent conflict has recently seen a trend of increased intrastate and non- state conflict ratio vis-à-vis interstate conflict, interstate conflict research continues to hold a significant role in peace and conflict research. States maintain vast resources and mobilisation capacity (Themnér and Wallensteen 2012, 67), thus interstate conflict can lead to more fatalities than other types of conflict. These motifs highlight the magnitude of exploring the dynamics of interstate threat perception within the context of peace and conflict studies.

A military intervention in one’s state is generally perceived as a clear threat to said state’s national security. But what about an intervention by a neighbouring state within one’s regional security complex?¹ This study examines this question, contributing to the general puzzle of how states establish whether external military interventions are perceived as security threats, narrowing the scope to a regional lens. The regional focus highlights the importance of states’ geographically immediate security landscape.

Previous literature on interstate threat perception has largely come from the field of international relations. With different schools of thought contributing various perspective on security threats, the common foundation of these views rests on the materialist notion of the balance of power. This material focus is countered by the ideational approach, highlighting cultural factors as the driving forces behind threat perception. Although previously theorised, the concept of threat perception still remains under-researched in a systematic and empirically driven matter, specifically in relation to outlining the causal chain and assessing the role of specific ideational factors in shaping state threat perception. The objective of this thesis is to offer a contribution to this research gap in both

1 A play on Raymond Cohen’s quote: “Troop concentrations on one's border will usually be perceived as threatening.

But what about troop concentrations on somebody else's border?” (Cohen 1978, 95).

(7)

7 the theoretical and empirical components of the field by proposing a potential causal link between values and threat perception at the state level of analysis.

The paper addresses the question of what role value congruence (used interchangeably with value similarity) plays in interstate threat perception following a regional military intervention. The main theoretical argument asserts that states perceiving a high level of value similarity to the intervening state should have a lower perception of threat on the basis of categorising similar states as belonging to the in-group. Value congruence and the subsequential in-group categorisation creates a sense of similarity and identification, functioning as a mediating element. The theoretical scope of the paper is framed by the regional security complex notion, pinpointing the centrality of regional focus in interstate security research. Thus, the theory limits the assessment of threat perception to states within the constraints of the immediate region encompassing the intervening and target states.

The thesis hones in on the regional impact of the 2014 Russian Federation’s (Russia) military intervention in Ukraine. Through structured, focused comparison, the research analyses the post- intervention perception of threat in Belarus, Poland, and Sweden and attempts to assess whether value congruence between Russia and these three cases bears any effect on their threat perception of the 2014 Russian intervention. The case selection is established, firstly, by the theoretical scope, thus belonging to the Baltic Sea – Eastern European regional security complex, and, secondly, by selecting three cases with different values for the independent variable according to the World Values Survey data. Despite being constrained by several empirical limitations and isolation challenges, the analysis demonstrates support for the hypothesised relationship. Although all cases demonstrate signs of threat perception in response to the 2014 Russian intervention in Ukraine, the results find variation in the magnitude of threat perception. The case variation in threat perception is consistent with the theoretical argument, with Belarus demonstrating the lowest level of threat perception, followed by Poland, and Sweden.

The thesis consists of five chapters. The first chapter offers an overview on threat perception in international relations scholarship. The second chapter outlines the theoretical argument of the paper. The third chapter lays out the thesis’ research design. The fourth chapter delves into case analyses. Finally, the fifth chapter discusses and evaluates the empirical and theoretical findings of the paper.

(8)

8

1. THREAT PERCEPTION IN EXISTING SCHOLARSHIP

Interstate conflict is a result of calculated decision-making. Actors strategize and make decisions based on the available information they obtain and perceive. Understanding threat perception is vital to understanding international conflict and crisis escalation. By providing a connection between external actions and state reactions (Cohen 1978, 93), threat perception sheds light on the interpretation of external foreign policy, mediating between recognition and action formation. An external action triggering another state’s threat perception might result in conflict onset or escalation when the perceiving state identifies a viable endangerment to its national security. This state then has the potential to respond to the identified threat with political or military means, potentially triggering or further escalating the crisis. The study of threat perception and its influence on interstate relations, thus, contributes to the objective of peace by highlighting the early warning signs and potential pathways to violence. Exploring the field could provide insights to the conditions and issues to be addressed in order to prevent the breakdown of the security dilemma and conflict escalation in general.

Threat perception has long been a vital facet for scholars of politics and conflict either implicitly or explicitly, with different schools of thought placing aspects at the forefront of theoretical considerations. At the core of the academic discussion is the debate comparing the effects of material versus ideational drivers of threat perception. While material factors are defined by objective ontology, ideational factors are defined by subjective ontology. Thus, an example of a material driver of threat is the balance of military power, whereas normative divergence is a potential ideational component of threat perception. The ideational approach, supposedly fuzzier and less straightforward than its material counterpart, includes a broad spectrum of elements relating, but not limited, to culture, ideas, and intentions (Darwich 2016, 10).

In international relations literature, historically the most explanatory power has been associated with the realist tradition and the realist notion of balance of power. This strand of thought has been countered by the constructivist and liberal camps that highlight the significance of ideas and domestic factors. This section outlines the main international relations notions in the threat perception domain, creating the basis for the subsequent theoretical argument of the paper.

(9)

9 1.1. Understanding threat perception

As stated by Cohen, “[t]hreat perception is the decisive intervening variable between action and reaction in international crisis. When threat is not perceived, even in the face of objective evidence, there can be no mobilization of defensive resources.” (Cohen 1978, 93). This link is a key aspect in understanding the foreign policy choices of states in or preceding international crisis situations.

Cohen interprets threat as an awareness of potential looming endangerment (ibid, 95). His understanding of threat perception engages two psychological dimensions of perception:

observation and appraisal. While observation refers to the detection of signals, appraisal refers to the assessment and categorisation (threatening/non-threatening) of the said signals. This process of perception is driven by certain predispositions (ibid., 95-97). There is no absolute consensus on these predispositions that determine the assessment of received signals, hence providing a relevant gap for the theoretical discussion of this paper.

In a similar vein, Janice Gross Stein highlights a simple yet crucial notion that “[t]hreats do not unambiguously speak for themselves” (Stein 2013, 365). By this, she stresses the definitive nature of a subject’s perception, in this case, the perception of a state’s decision-makers. Stein also considers perception through a psychological dimension; through cognition and the interpretation of catalyst-induced sensory processes (ibid.). This understanding is similar to the one discussed by Cohen, who also refers to cognitive deliberative processes of perception. Stein, however, identifies a distinction between individual and collective perceptions, noting that collective perception mechanisms are more complex to trace. She features the role of idea and emotion dissemination as well as the socially constructed nature of threats (ibid.). Stein highlights that perception motivates action, which, in basic terms, explains the general relevance of the topic to the study of international relations and peace and conflict.

Both Cohen and Stein present rather similar explanations for threat perception. From their arguments, it can be derived that signals (foreign policy and military actions) are assessed and interpreted through the lens of the perceiving actor. Mediated by predispositions, signals are received and categorised as either threatening or non-threatening, subsequently motivating the corresponding reaction. As it pertains to this paper, the working definition of threat perception is the identification and interpretation of external foreign policy and military actions that could potentially inflict harm on the perceiving state’s national security.

(10)

10 1.2. Threat perception in international relations academia

Although threat perception functionally plays a key role in international relations, the theoretical concept itself is largely understood implicitly in the international relations literature. This aspect can especially be attributed to the realist strand of scholarship, which regards power as the key determinant of interstate affairs (Cashman 2014, 375). Realism is a rather materialist explanation of peace and conflict, highly focusing on the latter. Realism regards material variables as the true measure of status and interstate relations (Mearsheimer 2001, 386). Realist focus on the material world determines ideational and subjective elements as minor at best. The explanatory weight’s fixation on material indicators deems the cognitive dimension largely void. This constrains the conceptual exploration of threat perception as perception itself entails a subjective and cognitive perspective.

The realist scholarship esteems power to be the focal point of all interstate relations. Following Robert A. Dahl’s conceptualisation (Dahl 1957, 202-203), in international relations, power is generally understood as the ability to manipulate the behaviour of other actors (Garcia-Retamero et al. 2012, 180). Neorealist Kenneth Waltz understands power as “the combined capability of a state” (Waltz 1990, 36). The material nature of the theory assumes that actors behave rationally and similarly if placed in the same situation regardless of ideational aspects (Waltz 1979, 80;

Jackson and Sørensen 2013, 66, 79). Even though different scholars present different versions of power indicators, including the size of territory, economic and political influence, the amount of natural resources, etc., realism largely regards military capabilities the main indicator of power (Waltz 1979, 107; Mearsheimer 2001, 30). Derived from this, threat can be defined in terms of military inferiority vis-à-vis other states.

The state-centred core assumption of the world as an anarchical system stands true with the realist view of threat. As power is the foundation of realism, threats are linked to the balance of power. Respectively, power asymmetry gives rise to threat. This, in turn, prompts states to balance against powerful actors that can potentially cause military harm to other states in the system (Cashman 2014, 376-377). In this vein, an actor surrounded by stronger actors should perceive threat (Garcia-Retamero et al. 2012, 180). This goes in hand with the security dilemma concept, which outlines a spiral power maximisation battle. The increase of an actor’s security ironically actually ends up decreasing it as other actors will try to change the balance of power (Herz 1950,

(11)

11 157). In this case, threat is perceived when other actors increase their security capabilities, thus triggering a change in the balance of power.

Stein states that traditional realist texts largely equate material power measures to threat assessment, with modern scholarship only starting to analytically examine intentions, perceptions, and misperceptions of threat more recently (Stein 2013, 364). Thus, the traditional realist approach largely assumes that threat comes from power asymmetry, not focusing on the dynamics of perception itself or exploring ideational aspects as potential influences on threat perception. The objective ontology and epistemology based core theoretical assumptions of realism hinder the investigation of non-material aspects, failing to consider alternative approaches or conditioning factors. The theory pays little attention to the non-material differences between actors that motivate different perceptions and foreign policy decisions based on those perceptions. The presupposed power asymmetry equivalence to threat determination offers a highly simplified version of international relations and politics as such. Although providing merit to the debate, the realist “one size fits all” approach fails to explain the differences in foreign policies of states in similar power relations and the process of threat assessment and perception outside of the immediate power balance. This is not to say that power does not play a role in interstate threat perception. Power is a vital element for all security concerns and interstate matters. The balance of power in the international system outlines the basic groundwork of threat perception, but cannot solely explain the variation in interstate threat perception itself.

Building on the theoretical notions of the realist school of thought, Stephen M. Walt proposes an altered concept of the balance of power. Alternatively, he presents the balance of threat theory that puts the perception of threat at the core of state alignment behaviour (cooperation and allyship) as opposed to purely objective power asymmetry (Walt 1985, 8-9). Moreover, Walt critiques the simplified view of interstate system and power relations, declaring other factors such as offensive intentions as valuable theoretical considerations. Although not completely siding with the ideational approach to threat perception, he alludes to the potential of ideologically motivated state alignment (ibid., 18-19). Walt also notes that ideological similarity can encourage allyship while ideological difference may cause actors to categorise divergent actors as potential enemies (ibid., 25-26). Therefore, from this argumentation, it can be derived that the similarity of ideological elements may hypothetically lower threat perception whereas divergence may heighten it. This

(12)

12 conclusion is significant in the context of both the addition of an ideological dimension to the realist tradition as well the development of the theoretical argumentation of this paper. With Walt’s considerations of perception and intentions, the following theoretical discussion delves deeper into the ideological and psychological scholarship within the theme of threat perception in international relations.

Perception is the central tenet that underpins many facets of rationalist theory - information availability, credibility, and signalling - just to name a few. The theory asserts that misrepresented information, incorrectly perceived signals, and subsequently chosen strategy may lead to conflict (Fearon 1995, 381). Following rationalist logic, uncertainty and inconsistent information about the capabilities and motives of other actors in foreign policy decision-making triggers threat perception. This is closely connected to the anarchical structure of the international system. The absence of an “international police” enforcing actors to limit their use of force itself stimulates threat perception resulting from uncertainty and suspicion about the others’ intentions regarding potential use of force (ibid., 401).

Combining the study of international relations, security, and psychology, the perception of state intentions and overall image plays the central role in Robert Jervis’ work. Jervis declares that states’

domestic beliefs and internal and international history experiences impact their perceptions (Jervis 1968, 466-470). Thus, Jervis suggests that ideational factors have a conditioning effect on states’

perception. Importantly, he also alludes to the shaping power of the “evoked set” - the prevailing concerns of the perceiving actor at the given moment they receive and evaluate signals (ibid., 472- 473). Therefore, perception can face influences from features of both past and present.

The liberal strand of international relations further widens the scope of explanatory variables by exploring domestic, ideational, and institutional subjects of international relations. Highly relevant to the concept of threat perception is the democratic peace theory. At the core of the theory lies an expectation that states of democratic nature abstain from engaging in military conflict with other states classified as democratic (Owen 1994, 87). The explanation of this phenomenon is positioned between structural and normative realms. Common justifications include shared ideology, economic interdependence, membership in common organisations, incumbents’ political accountability as a policy restrictor, democracies’ credibility of signalling their intentions, or a combination of some of the said aspects (Fearon 1994, 577; Owen 1994, 89, 99, 102; Farnham

(13)

13 2003, 397-398). The implication of this expected relationship is that democracies should have lower threat perception of one another. By default, this suggests that democratic states vis-à-vis states with divergent political regimes should have higher threat perceptions of one another.

A proponent of the ideological approach, John M. Owen, for instance, asserts that liberal democracies abstain from engaging in armed conflict with one another as a result of adherence to mutual liberal ideas. These notions also drive liberal democracies to conflict with illiberal states (Owen 1994, 88). Notably, Owen places state perceptions at the focal point of the discussion. He argues that for the democratic peace hypothesis to succeed, democratic states must perceive the other (potential opponent) as a liberal democracy (ibid., 96). Essentially, the perception of the other determines state action. This conclusion is consistent with the previously discussed concept of threat perception, particularly the notion that perception motivates action. Thus, the central motif alludes to the perception of regime type as barometer for threat perception, with the democracy- democracy dyadic relationship demonstrating a promise for lower threat perception. Depending on the scholar, this relationship may be specified to liberal democracies.

The relationship is further explored by Barbara Farnham, a contributor to the democratic peace theory scholarship, who argues that the failure to adhere to internationally and commonly understood liberal-democratic norms increases threat perception. Farnham notes that actors having

“no regard for the rules of the game”, meaning mutual respect and the commitment to peaceful conflict resolution, are perceived as more threatening (Farnham 2003, 403-404). Contrary to the popular assumption that democratic states automatically categorise non-democratic states as animus, Farnham’s argument extends outside the constraints of this dyadic relationship. She states that democratic actors might expect autocratic regimes to adhere to “the rules of the game” and rely on more than regime type for threat assessment (ibid., 406). Farnham’s theoretical contribution is linked to Cohen’s discussion on the relation between international law and norm violation and increased threat perception. Building on Thomas Schelling’s work (Schelling 1963), Cohen states that international rules (formal and informal) function as a state behaviour coordination and action boundary demarcation mechanism, hence helping avert conflict (Cohen 1978, 106).

The liberal tradition brings the discussion to the constructivist understanding of interstate threat perception. Constructivism maintains that ideational factors such as culture, beliefs, and identities shape foreign policies and state actions in the international system (Wendt 1999; Hopf 2002).

(14)

14 Constructivism places norms and ideas in the focal point of interstate behaviour and attitudes toward other states. Perhaps the most notable scholar in the realm of constructivism is Alexander Wendt. He argues that contrary to the realist and other international relations mainstream theories’

focus on material ontology, the world is actually a social phenomenon (Wendt 1999, 20, 370-371).

Thus, the tradition maintains that politics are subjective. As stated by Wendt, ideas shape how power is perceived by different actors (ibid., 24-25). This suggests that the similarity or difference of ideational factors might impact the way a certain state and its actions are perceived by other states. Wendt talks about how the perceived image of other actors influences their policy assessment and behaviour. Specifically, he notes that the intentions of an actor viewed as an enemy are considered negatively rather than optimistically (focus on the worst-case scenario).

Furthermore, this image has a conditioning effect on the perception of a different actor’s capabilities (ibid., 262). Thus, this suggests that ideas and perceptions are an intervening variable in the realist relationship between the balance of power and threat perception.

Wendt focuses on collective identity formation, stimulating the classification and division of the

“self” and “other” (similar or different), where shared (collective) identity places actors in the same in-group (ibid., 338). He offers several potential causes for collective identity formation, one of which is of particular interest in reference to this paper. Wendt states that the similarity of actors is a driver of collective identity formation. This alikeness can be evaluated via different dimensions, such as political regime, economic ideology, and values. The author proceeds by referring to Samuel Huntington’s notion of the “clash of civilizations” (Huntington 1996) as one of the principal theoretical foundations of heterogeneity-based conflict. Wendt suggests that when controlling for other elements, the lack of heterogeneity increases states’ interest harmonisation, thus advancing collective identity development. Based on this, he even goes on to offer actor heterogeneity a causal effect on threat perception (Wendt 1999, 354-356). This line of argumentation is crucial for the further discussion of the paper as it contributes to the hypothesis that ideational elements have an influential effect on interstate perception.

1.3. Identifying the research gap

The threat perception literature in the field of international relations is vague by nature, attempting to simplify inherently complex interstate politics. While outlining numerous notions

(15)

15 and having relevance for the subject as a whole, the simplification of threat perception construction leaves many avenues unexplored. Although the existing scholarship offers different perspectives on the subject of threat perception, there is a lack of systematically explored mechanisms and empirically testable models driving the dynamics of threat perception in interstate context.

Moreover, much of the international relations scholarship neglects to examine the dynamics and processes underlying the conception of perception itself, necessitating the reader to “read between the lines” and draw their own conclusions about the subject-related implications. This ultimately convolutes the translation of theoretical explanations into testable notions.

As previously stated, academics allude to both material and ideational elements contributing to threat perception. While the material approach considers power to be the strongest explanatory variable driving threat perception, the notions in the ideational approach are rather abstract, offering fuzzy concepts such as “identity”, “culture”, and “ideas” to manifest variation in threat perception. As such, it fuses notions about ideational characteristics without distinguishing specific components and examining their theoretical weight. In failing to provide specifications, the approach cultivates a lack of credibility. Furthermore, the absence of distinguished ideational elements hinders the provision of empirical inference, considering the conceptual complexity.

Building on this, this paper aims to contribute to both the theoretical and empirical research gap of the international relations threat perception literature by focusing on a specific component of the ideological approach, namely, values. In the context of this thesis, values are defined as “desirable, transsituational goals, varying in importance, that serve as guiding principles in people’s lives” as defined by Shalom H. Schwartz and Anat Bardi (Schwartz and Bardi 2001, 268).

Although the link between value congruence and threat perception has been previously brought up in the realm of social psychology, the previous studies explore the individual level of analysis (Garcia-Retamero, Rousseau et al., 2007, 2012). This paper, however, changes the focal point to evaluate the interstate dynamics of value similarity and threat.

(16)

16

2. THEORETICAL ARGUMENT

Following the previously established research gap in the area of ideational approaches to threat perception, this chapter proposes a theoretical foundation to answer the research question of what role value congruence plays in interstate threat perception following a regional military intervention. The section begins with outlining the general argument, then specifies the phases of the causal mechanism, finally adapting the argument to the state level. The model below (Figure 1) visually demonstrates a simplified version of the proposed causal mechanism between the independent variable (values) and the dependent variable (threat perception) mediated by categorisation.

VALUES CATEGORISATION THREAT PERCEPTION

Figure 1 As previously mentioned, social psychology researchers have theorised about the link between the similarity of values and the perception of threat, contributing to the understanding of intergroup relations. Rocio Garcia-Retamero and David L. Rousseau et al. (2007, 2012) have conducted experiments demonstrating that shared identity and values influence threat perception at the individual level. In their work, the authors hypothesise that individual perception of a high level value similarity with a foreign nation has a diminishing effect on individual threat perception of the foreign nation. In this vein, perceived value divergence inflames feelings of discomfort in association with the said state. Conditioned by power asymmetry, this can contribute to a higher level of threat perception (Garcia-Retamero, Rousseau et al. 2012, 182). In another paper, they explore a similar theme by focusing on the notion of shared identity and its conditioning effect on threat perception, which, although encompasses, does not explicitly outline values as the focal point in the theoretical argument (Garcia-Retamero, Rousseau et al., 2007, 744). Similarly, research by Geoffrey Wetherell, Or'Shaundra Benson et al. proposes that feelings of value similarity influence individual attitudes towards states and support for foreign policies concerning said states (Wetherell, Benson et al. 2015, 3).

Based on the theoretical notions and findings of the studies, this paper introduces a more elaborate theoretical argument adapted to state level dynamics, linking values and threat perception

(17)

17 through categorisation. The outline of the proposed mechanism is seen in Figure 2. It is expected that if actor A perceives actor B to have similar values, it will categorise actor B as belonging to the in-group. This, in turn, will decrease threat perception. In the same vein, if actor A perceives actor B to adhere to different values, actor B will be categorised as belonging to the out-group, increasing the perception of threat.

Figure 2

This mechanism is loosely based on the theoretical foundation developed by Shalom H.

Schwartz, who connects values to action. As cited by Steven Hitlin and Jane Allyn Piliavin, Schwartz’s mechanism is connected through (1) value activation, (2) “privileging of certain actions over others”, (3) the interpretation and perception of situations, and (4) action planning (Hitlin and Piliavin 2004, 365). Steps two and three bear most relevance to the theoretical argument of the thesis. While action prioritisation can be linked to the concept of categorisation, privileging actions and characteristics of the in-group over those of the out-group, the situational perception clearly manifests itself in threat perception.

This theoretical argumentation employs the notion of categorisation as the mediating step between values and threat perception. Although categorisation has been mentioned in previous studies in one way or another (such as Garcia-Retamero, Rousseau et al. 2007, 2012), the concept itself is not explicitly defined in the context of the causal mechanism. In their 2007 paper, Garcia- Retamero, Rousseau et al. note the significance of in-groups and out-groups in establishing threat perception. This paper does not thoroughly explicate the definitions of in-group and out-group, often referring to similar ideas with different terms, but contextually the groups relate to a shared

(18)

18 or divergent sense of identity. Authors rely on the social identity approach, anchoring categorisation as the starting point and placing intragroup value adoption after the categorisation process (Garcia-Retamero, Rousseau et al., 2007, 747). The social identity approach in social psychology stipulates that, to understand and simplify the social world, individuals sort the available information into social categories based on observed and perceived patterns purposeful to the individual (Tajfel 1974, 69). The approach argues that the categorisation of in-groups and out-groups promotes bias. Respectively, belonging to a group (in-group) creates favourable feelings toward its members as opposed to outsiders (out-group), thus sustaining the perceiver’s positive image of oneself and, by extension, the social category or group to which the perceiver belongs (ibid., 311). Although categorisation fosters discrimination, it does not create automatic animosity towards the out-group.

Furthermore, the notion of categorisation is explored building on the constructivist ideas regarding the division of “self” and the “other” noted by Wendt. Wendt’s categorisation rests on the assumption that groups or categories are divided along the borders of identity, which can be assessed on the basis of a multitude of aspects, including values (Wendt 1999, 338) In his work, Wendt distinguishes between the dichotomous classifications of “self” and “other” as well as “us”

and “them”. Whereas “self” and “us” carry a similarity connotation, “other” and “them” display divergence. The author elaborates on these classifications by stating that the concepts encompassing congruence indicate a shared identity among separate actors. When actors identify with one another, common interests are developed, producing altruistic behaviour (ibid., 305-306).

Both perspectives offer interesting ideas for understanding categorisation. While the constructivist approach is vague in its theoretical operationalisation of categorisation and the relevant mechanism, the social identity approach assumes categorisation as the baseline for further social interaction. Considering the interstate level of interaction, the social psychology concept has to be adjusted to the structure and dynamics of the international system. Thus, the fusion of both outlined understandings provides the optimal conceptualisation of categorisation for the study of interstate threat perception.

Building on this, categorisation has to be understood as the process of systematically sorting subjects into groups on the basis of their similarity or difference. Combining this with values as the hypothesised indicator of similarity or difference, categorisation helps actors interpret the actions

(19)

19 of others by placing them into in-groups and out-groups on the principle of values, which is the first step of the causal mechanism. Contrary to the social identity approach, this mechanism switches the temporal order of the argument, positioning values as the defining criteria for categorisation. As previously stated, this argument uses Schwartz and Bardi’s definition of values, understanding them as “desirable, transsituational goals, varying in importance, that serve as guiding principles in people’s lives” (Schwartz and Bardi 2001, 268). The complexities of the concept and the multitude of value dimensions necessitate studies employing this theory to narrow the focus, specifying the exact values used for the assessment of the hypothesis. The unspecified value dimensions are a current limitation of the theory that can be rectified with future research on how different value dimensions affect threat perception. For the purposes of this paper, the specification is provided by the empirical operationalisation in the research design.

The second step of the causal mechanism connects categorisation to threat perception.

Following the previous argumentation, identification with another actor, thus, its placement within the in-group should generate favourable feelings towards the said actor. Building on the promotion of bias in favour of the in-group, this implies subsequential intergroup prejudice. Although prejudice does not automatically create threat, identifying another actor as similar fosters positive attitude and altruistic behaviour regarding the said actor. Furthermore, perceived similarity and identification stimulates the feeling of security because of the belief that actors think and act in a similar manner, thus, are more predictable in their policies. It follows that that the value-based categorisation of actors into in-groups and out-groups conditions the perception of threat. In that respect, the actions of actors in the in-group are perceived as less threatening than those conducted by actors in the out-group.

To further understand the application of the causal mechanism, it is necessary to set the scope conditions limiting the theoretical argument. As previously specified, this argument examines threat perception following a regional military intervention. The dynamics of threat perception are assessed on the state level. The argument extends to the states in the regional security complex, excluding the target state of the intervention (the target state). Respectively, the theory evaluates how states perceive a military intervention by and targeting a state within their region. To reiterate the leading theme from the introduction: A military intervention in one’s state is generally

(20)

20 perceived as a clear threat to the target state’s national security. But what about an intervention by a neighbouring state within one’s regional security complex?

The theoretical argument is framed by the notion of regional security complexes. Respectively, the argument views cases of threat perception as a response to military interventions within regional security complexes. Popularised by Barry Buzan, the regional security complex theory investigates security issues through a regional perspective. According to Buzan, such complexes are to be understood as “a set of states whose major security perceptions and concerns are so interlinked that their national security problems cannot reasonably be analyzed or resolved apart from one another”

(Buzan 2003, 141). Although the theoretical definition of regional security complexes remains vague and open to the researcher’s interpretation, and, thus, is further operationally defined in the research design section, the main criteria for states to be categorised as belonging to the same regional security complex has two underlying assumptions. First, the states must be a part of the same geographical region, thus be in near geographic proximity to one another. And, second, be interdependent in terms of security challenges. Condensed by Matteo Legrenzi and Fred H.

Lawson, such regions are grouped by “countries whose security interests are deeply affected by the externalities of policies that their respective governments adopt in response to severe internal and external challenges” (Legrenzi and Lawson 2018).

It has to be noted that, for the purposes of this argument and the framework of the paper, security and military challenges refer to conventional hard security threats, excluding the realms of information, cyber, and other non-physical forms of security and military. Furthermore, the perceived threat relates to the perceiver’s fear about potentially experiencing a similar scenario, thus, becoming a target state vis-à-vis the intervening state. Therefore, the argument does not evaluate threats resulting from the intervention such as cross-border migration, drug and arms trafficking, domestic unrest, and other similar effects of conflict.

Based on this framework, the theoretical argument has several scope conditions, limiting the population of cases. Firstly, the theory specifically addresses threat perception following a military intervention. Secondly, both the target of the intervention and the actor conducting the military intervention (intervening state or intervener) have to belong the same regional security complex.

The third and fourth scope conditions refer to the actor(s) whose threat perception is measured (perceiving state or perceiver). The third condition specifies that the perceiver also belongs to the

(21)

21 same regional security complex as the intervener and target state. The fourth condition establishes that the intervening state is militarily superior to the perceiver. It is important to state that the theoretical argument only examines the perspective of the perceiving state. The argument does not zoom in on the intentions of the state undertaking the military intervention.

Based on this theoretical argument, one can derive an empirically testable hypothesis.

Hypothesis: The greater the perceived similarity of values between the intervener and the perceiver, the lower the degree of threat perception.

Null hypothesis: The perceived similarity of values between the intervener and the perceiver does not influence the degree of threat perception.

(22)

22

3. RESEARCH DESIGN

This section outlines the research design of the thesis, discussing the methodological choices of the paper. The methodological framework builds on the research question and the theoretical argument presented earlier. This chapter firstly addresses the chosen method, followed by the operationalisation of the key variables, and the justification of case selection.

3.1. Method

To test the hypothesis, this study employs the method of structured, focused comparison. This decision stems from the nature of the research question and data availability. The method of structured, focused comparison allows the researcher to systematically assess different case studies in a qualitative way by focusing on the same set of questions or variables in each of the cases (George and Bennett 2005, 69-71). This study aims to evaluate threat perception in three cases.

The structured, focused comparison provides a solid framework for the assessment of operationalised indicators across all cases.

The methodological choice to frame the study through a structured, focused comparison is pragmatic in nature. Considering the fuzzy notion of the dependent variable (threat perception), there currently is no existing quantifiable indicator to measure this variable, thus necessitating an in-depth case study approach. Furthermore, the highly contextual and cognitive essence of the variable calls for a qualitative approach as material indicators alone do not provide enough evidence to make an inference about the hypothesised relationship. Threat perception includes many dimensions that have to be assessed in combination with one another within cases to provide an aggregated assessment of the variable that can be appraised in a cross-case comparison.

Therefore, the method of structured, focused comparison provides the optimal foundation for this study.

3.2. Operationalisation of the key variables

Building on the theoretical argument, the key variables in this paper are values (independent variable) and threat perception (dependent variable). While the independent variable is measured by an existing indicator, the study offers a novel operationalisation of the dependent variable.

(23)

23 3.2.1. Independent variable

As stated by the theory, the independent variable is values. In this paper, values are measured via the World Values Survey. Specifically, the operationalisation builds on value scores of the World Values Survey’s Inglehart–Welzel cultural map.²

As demonstrated by Figure 3, the map includes two dimensions representing cross-cultural variation of values: traditional versus secular-rational values and survival versus self-expression values (World Values Survey, n.d.b). Whereas the traditional versus secular-rational values dimension includes values related to authority and conformity, the survival versus self-expression values dimension includes values related to achievement, stimulation, and security. As stated by the World Values Survey, these dimensions have been inductively created by running factor analysis on the World Values Survey data, uncovering that the value orientations of different societies are aggregated within specific cultural traditions (World Values Survey, n.d.b). Thus, the scores reflect values of different societies categorised by dimension. Building on this, the Inglehart–Welzel cultural dimensions set the operational values framework for the paper.

The independent variable is measured by the 2014 Inglehart–Welzel cultural map scores on both dimensions. The map is built on the World Values Survey Wave 6 dataset for 2010 – 2014 (Inglehart, Haerpfer et.al. 2014). To assess the congruence of the values, the dimensions are aggregated and compared. Respectively, the aggregated value scores of the perceivers are compared to the aggregated value score of the intervener to calculate the value distance between the actors. A smaller distance indicates higher value congruence, whereas a greater distance indicates a lower value congruence. The aggregated scores of each state are visualised with a data point on the map’s plane, with the dimension scores representing coordinates (Figure 3).

The independent variable is, thus, operationalised as value distance in the 2014 Inglehart–

Welzel cultural map. Value distance is calculated as the length of the vector between the respective data points (Table 1). As the dimension axes range from -2.5 to 2.5 and -2.0 to 2.5, the minimum value the independent variable can take is 0 (complete value congruence with data points overlapping), but the maximum value is 6.36 (complete value divergence with data points at the opposite corners of the plane) (Figure 3).

2 The numerical scores of both dimensions were provided via e-mail communication with the World Values Survey Secretariat. They represent the coordinates of country locations in the cultural map. Included in the annex.

(24)

24 The value distance (vector length) is calculated via the geometric formula for vector length:

|vector|=√(x²+y²)

Independent variable value component operationalisation

Mathematical term Operationalised term

Vector Value distance

X X axis coordinate point (perceiver) – X axis

coordinate point (intervener)

Y Y axis coordinate point (perceiver) – Y axis

coordinate point (intervener)

Table 1 Inglehart–Welzel cultural map – World Values Survey wave 6 (2010-2014)³

Figure 3 (World Values Survey, n.d.b).

3 Within the context of this paper, the map only serves as a measurement tool for the independent variable. The cultural traditions/groupings are not taken into account.

(25)

25 The indicator offers an acceptable level of validity as the concept of values is rather narrow. The scores reflect ten aggregated survey indicators⁴, assessing only specific values (World Values Survey, n.d.b). The map covers ten World Values Survey dataset indicators chosen on the basis of data availability as the questions representing the indicators have been systematically measured in both spatial and temporal dimensions (ibid). The limited coverage of values, however, constrains the theoretical validity of the measurement as other values are not included in the assessment.

Furthermore, as the selection of dimension indicators for the map is of pragmatic nature, it is plausible that the survey indicators do not provide the most accurate measurement of the value dimensions.

The theory employs the concept of perceived values. However, based on the existing indicators for value measurement, the paper uses a proxy measure – tangible values instead of perceived values. This methodological choice reflects the availability of data, as there is no available measure of perceived values. Although not a perfect match, the tangible value measurement is indicatory of value perception. Reflecting back on Schwartz’ theoretical mechanism for value – action nexus (p.

17), it has to be considered that values motivate attitudes and actions. This implies that actions are perceived as indicatory of values. In this vein, it can be argued that perceived values of another actor should be rather similar to the actual values of said actor assuming the actor acts on its values.

Although score discrepancies are plausible, the World Values Survey indicator provides the best available option for the measurement of the independent variable considering its focus on value evaluation and systematic measurement of the data. Notably, the World Values Survey data is well- established in the social sciences field as it offers high quality data covering temporal and spatial variation (World Values Survey n.d.a.).

The indicator provides high reliability as the data is consistent and publically available. This, thus, allows for the replication of the indicator measurement. The World Values Survey dataset at large contains national level scores for a multitude of value indicators. The data collection method is in-person interviews conducted at the interviewees’ homes (World Values Survey n.d.c.). The sampling strategy is not standardised across countries, but is approved at the institution’s governing level. The data collection attempts to counter bias by offering the survey in all languages used as the primary communication language by at least 15 percent of the society. Furthermore, the

4 For more information on the indicators, see World Values Survey (n.d.b.)

(26)

26 guidelines specify a distance between a concrete number of respondents in a given area. The data covers the adult population of the states (18 to 85) and specifies a sample size of at least 1200 respondents per country (ibid.). Although providing a general outline of the data collection and sampling methodology, the dataset does not specify whether and how the samples take into account the different demographics of the countries, specifically relating to the gender dimension. The stated information notes that “[s]amples must be representative of all people in the age 18 and older residing within private households in each country, regardless of their nationality, citizenship or language.” (ibid.). Without the specification of the sampling strategies, one cannot be convinced about the absence of sampling bias.

3.2.2. Dependent variable

The dependent variable of the study is threat perception. Based on the theoretical argument pinpointing states as the perceiving actors, the nature of state decision-making, and the availability of data, the perceiver is operationalised as the governing body of the perceiving state. This operationalisation selection reflects the decision-making structure of states. As the variable of interest is threat perception at the state level, it is necessary to specify the actor(s) representing the state and responding to security threats. Therefore, the chosen unit is the governing body. Although potentially differing in composition for each country, the governing body leads the state. As such, it both interprets the actions of other states and reacts.

Thus, the questions addressing cases via structured, focused comparison refer to the actions and statements made at the governing level. As this paper assumes that the governing body controls the official political discourse of the state, which is both published in the official state websites and stated by members of the governing body in media, the analysed information is within the realm of official state discourse. It has to be noted, however, that the governing body is composed of different smaller scale actors, thus the official state position and discourse is the result of inner political discussions and aggregated perceptions and their interpretations. This paper makes the assumption that the statements of the governing body reflect the overall unitary position of the state. As it pertains to this paper, the governing body has to be understood as the decision-making aggregate body of the state. Considering the differences in state structures, the body is case-

(27)

27 specific. Generally, it refers to the executive branch of the state, thus, the government or the cabinet of ministers. Depending on the country, this may include the president.

Having outlined the operationalisation of the perceiver, it is necessary to operationalise the dependent variable itself by setting the questions and subsequent indicators guiding the structured, focused comparison to assess the dependent variable.

The measurement of threat perception on a state level necessitates several dimensions. To advance the validity of the measurement, this paper builds on the operational definition of threat perception outlined by Cohen. He specifies such indicators as (1) decision-maker statements about the threatening event, (2) observations by “contemporary spectators” on the thoughts and expressions of the decision-makers, (3) “evidence of exploration by decision makers of alternative responses to the threat (such as [..] search for external support)”, (4) political and military actions in response to the threatening event (Cohen 1978, 95). Based on data access and time constraint, this paper reflects on three of the provided indicators: statements, politico-military actions, and alternative responses to threat (specifically zooming in on reaching out for external support). The indicators are assessed via structured, focused comparison.

In this vein, the questions this research asks each of the case studies in reference to the intervention (threatening event) are as such:

1. Have there been any changes to the actor’s military capabilities?

2. Has there been a change to the actor’s politico-military alignment or search for external support?

3. Have there been changes in actor’s military activities?

4. Has the actor made any changes to its military policy?

5. What has the actor’s discourse been like in reference to the intervention?

To establish inference, it is necessary to derive specific indicators addressing the questions. The indicators are empirically relevant as they aim to measure the different components of threat perception mentioned in the previous framework outline provided by Cohen. The indicators referring to each of the questions are provided by Table 2 below. The indicators also serve as the structure guiding the case studies.

(28)

28 Operationalisation of the dependent variable

Questions Indicators

Have there been any changes to the actor’s military capabilities?

Defence expenditure Arms imports Has there been a change to the actor’s politico-

military alignment or search for external support?

Participation in alliances

Cooperation with external actors (incl. arms trade partnerships)

Have there been changes in actor’s military activities?

Military activities (troop movements, military exercises, etc.)

Has the actor made any changes to its military policy?

Conscription status

Strategic changes to military policy What has the actor’s discourse been like in

reference to the intervention?

Statements referencing the intervention

Table 2 Defence expenditure and arms imports provide valid indicators to measure changes in military capabilities as both constitute military resources. Material resources are necessary to maintain security of the state. Their increase indicates a decision to advance the level of security. This, set in the context of a potentially threatening action, can indicate a response to threat. The specific measurement of these indicators is produced by the Stockholm International Peace Research Institute (SIPRI), an academically valued source for data pertaining to security and defence.

Indicators covering changes in military activities, policy, alignment, cooperation, and conscription status all allude to actions taken by a state to maximise its security. Variation and enhancement in military activities can indicate a potential mobilisation of resources, indicating an increased level of military readiness. Similarly, the introduction of conscription suggests the necessity to increase the size and capabilities of the forces. In the same vein, alignment and politico- military partnerships with other states demonstrate the actors’ position vis-à-vis the intervener.

Whereas cooperation with the intervener post-intervention indicates a favourable relationship, decision to side with other actors or alliances indicates an attempt to maximise the actor’s security via balancing.

All of the mentioned indicators, although relevant for the measurement of the dependent variable, require a contextual analysis to isolate the motivation for their variation. To do this, the study includes discursive indicators such as statements and discourse in policy papers. By placing the factual indicators contextually in reference to the intervention, an assessment can be made about

(29)

29 the level of threat perception. The consideration of discourse provides an explanation for the actions, linking the other variables to the source of threat.

Apart from the mentioned SIPRI indicators, sources used in the case study analysis are country- specific. Most sources in the case analyses are published by state authorities, reflecting the official state position. This is supplemented by information provided by international organisations, research papers, and media reports. The media sources are both international and local. Most of the cited sources are in the national languages, with a part of the data in English. The official state data is both valid and reliable, demonstrating the official discourse. The previous assumptions about nature of the official state discourse predicate their accuracy. There is potential bias and accuracy issues in reference to the use of secondary sources. In attempt to eliminate this, the paper strives to make use of data triangulation. Furthermore, for the sake of data validity and accuracy, majority of the sources used are produced by official state authorities and posted on their websites.

The concept of threat perception is complex in the operational sense as it is not possible to directly measure an essentially cognitive process. Despite this, the use of a theoretically driven operational definition creates a solid foundation for the measurement of the variable. The specified indicators reflect the concept of threat perception well as they cover both material and discursive components of the concept, thus all in all establishing an adequate level of validity. The measurement of the dependent variable is comprised of several elements, causing some reliability concerns. The material/factual indicators have a high degree of reliability. The discursive elements themselves do not predicate high reliability as they are contextual and open to the researcher’s interpretation. In attempt to strengthen the reliability, the chosen method for the research is structured, focused comparison. By posing clear and specific questions to the cases, the precision of the indicators increases, thus making it more accurately applicable and replicable.

Constrained by data access, the thesis employs separate units of analysis for the independent variable and the dependent variable. This discrepancy constitutes a measurement problem. The operationalisation of the independent variable is measured by a national indicator, while the operationalisation of the dependent variable is measured through a qualitative analysis of actions and statements produced at the governing body level. Thus, whereas the independent variable reflects the societal unit, the dependent variable reflects the governing body unit. The reason for this is the availability of data for the independent variable. As there is no existing high-quality

(30)

30 public measure of governing body values, the thesis uses the World Values Survey data on national values as a proxy measure.

Despite not being absolutely representative of the governing level, the values of the society at large are indicatory of values at the governing level. According to Schwartz, the development of values is influenced by both biological and social factors. He mentions factors such as genetics, psychological dispositions, and socio-economic characteristics (Schwartz 2016, 78-80). From a biological standpoint, there should not be any unit-specific differences. From a social standpoint, one could argue that the socio-economic standing of the governing body differs from that of the general population. This argument could be made in reference to the divergent level of education, social class, and the profession-specific socialisation. While the argument has merit, it does have to be noted that the governing body is generally not isolated from the general population, being subjected to socialisation with the public. Although not disputing potential value variation, the governing body does constitute a part of the general society. Thus, one can assume that the main pillars of the nation’s social environment overlap for both units.

Furthermore, values play an important role in selecting representatives for the governing body.

Voter values influence their political choices, voting for candidates that they perceive represent similar values and views to their own (Tatarko 2017, 4). This argument is conditioned by regime type, thus might be more applicable to Poland and Sweden. However, even in autocratic regimes like Belarus, the governing body has to be sensitive to values and views of the society. As stated by Machiavelli, the prevention of popular discontent is necessary to stay in power (Machiavelli 1988, 63-64). This can also be attributed to values. The governing body has to be perceptive of the society’s values and beliefs to maintain the status quo. Not to say that all autocratic leaders are perceptive and act upon the society’s beliefs, but some concessions in favour of the public should lessen the general discontent with the regime. Therefore, it can be argued that the values of the general population, although not a perfect match, are sufficient for making an inference about the values of the governing body.

3.3. Case selection

Case selection for this thesis is twofold. While the first selection relates to the chosen military intervention, the second acknowledges the perceivers.

(31)

31 The military intervention in focus of this paper is the 2014 Russian intervention in Ukraine. In this context, military intervention relates to a military action taken by a state to locate their forces into external state’s territory or territorial waters with a political aim “to influence either domestic disputes or political conditions in a target state” (Pearson and Baumann 1987, 173). The Russian intervention provides an interesting study as the last recent case of an interstate conflict that resulted in de facto territory loss in the continent of Europe. The case constitutes an outlier as the continent has been largely peaceful and void of active interstate conflict with territory loss in the 21^st century. In the context of the paper, it is necessary to note that the start of the intervention is dated between the last days of February and March 1^st, 2014 (USASOC5 2016, 51-52). Another important aspect to remember in light of the case studies and changes in policy and activities is the use of hybrid tactics: covert operations, propaganda, proxies, and unmarked soldiers (“green men”) (ibid., 53-56). The intervention itself was comprised of different stages and locations. In the context of the thesis, the focus is placed on the initial phase of intervention, starting from the Russian invasion of Crimea to the annexation of the territory. However, these changes have to be considered in the context of the larger Russian involvement in Ukraine. The period of analysis is, thus, 2014 – 2019, allowing for the implementation of policy changes. The statement analysis, however, reflects the initial (immediate) period following the intervention, thus referring to the already mentioned initial phase of the conflict.

Considering the intervention, it is necessary to operationalise the relevant regional security complex. As it pertains to this paper, the regional security complex is constrained by the location of both the intervener and the target state, thus, Russia and Ukraine. Both countries are located in Eastern Europe, surrounded by Eastern, Central, and Northern European states. The theory establishes regional security interdependence. Considering Russia is the main military player in the region, the security complex is highly connected to Russian actions. Based on these considerations, this paper identifies the Baltic Sea – Eastern European regional security complex. To specify which countries are considered within the complex, the paper employs an extra operational condition – boundary with Russia. Thus, the population of cases are states in the Baltic Sea – Eastern European regional security complex having a land or maritime boundary with Russia, which are militarily inferior to the intervener (Russia).

(32)

32 This limits the population of cases to Belarus, Estonia, Finland, Poland, Sweden, Latvia, Lithuania, Moldova, and Norway. Considering the Russian military occupation of a Moldovan territory, Moldova has been eliminated from the case list. Picking on the different values of the independent variable (value distance to Russia), this paper assesses threat perception in three cases – Belarus, Poland, and Sweden. The independent variable values of the potential cases can be seen in Table 3. To establish temporal order, the independent values are measured before the intervention, thus using the World Values Survey Wave 6 (2010 – 2014).

Independent variable values for all potential cases

State Value distance

Belarus 0.513945789

Latvia 0.53775106

Lithuania 0.62731778

Estonia 0.920611881

Poland 1.832525588

Finland 2.592094944

Norway 3.351088941

Sweden 3.62583384

Table 3: based on Table 1 and Annex 1, derived from Inglehart, Haerpfer et.al. (2014).

Thus, case selection is justified by the different value distances. The lower distance indicates a higher level of value congruence between the actors. Thus, it can be derived that Belarus has high level of value congruence with Russia, Poland has a medium level of value congruence with Russia, and Sweden has a low level of value congruence with Russia.

Based on case selection, it is necessary to view the cases in light of potential confounding variables derived from the previous theoretical discussion of threat perception, giving most confounding power to the realist power arguments (Table 4). The first confounder is pre-existing military alignment. Military alignment can potentially influence values through socialisation and threat perception based on both ideological and balance of power dimensions. In this vein, the second confounder is the state’s military power, linking back to the realist argument about threat perception.

THE POWER OF VALUES IN DETERMINING INTERSTATE THREAT PERCEPTION