The Microfoundations of Diversionary Conflict

ABSTRACT Diversionary conflict theorists assert that leaders can become more popular at home by pursuing conflict abroad. At first glance this claim appears counterintuitive in light of the hardship conflict often imposes on ordinary citizens. Relying on social identity theory (SIT), I deduce two hypotheses to help explain why conflict can increase popular support for leaders. First, conflict with an outgroup can make people identify more strongly with their ingroup. Second, stronger ingroup identification can lead to increased support for leaders inside the group. The second part of the article applies these two hypotheses to Russia's seizure of Crimea in early 2014. Attitude surveys show that the Crimea conflict increased national pride among Russians while support for President Vladimir Putin rose dramatically, and they suggest that the two processes were causally linked. These findings support the article's two hypotheses.

applies these hypotheses to Russia's seizure of Crimea in early 2014. Attitude surveys show that the Crimea conflict boosted national pride among Russians while support for President Putin rose dramatically, and they imply that the two processes were causally linked. These findings support the article's two hypotheses.

Conflict and Support for Leaders
Conscious of the questions it raises, diversionary conflict theorists have sought to corroborate their claim that conflict can strengthen popular support for leaders in three main (potentially overlapping) ways. According to the first explanation, conflict can benefit struggling leaders by distracting their publics from domestic problems and/or by allowing them to demonstrate their supposed competence and decisiveness. Even where this effect is relatively short-lived, it might allow a leader to weather a temporary crisis at home. 6 A second explanation builds on the observation that diversionary leaders often pursue conflicts they can portray to their constituents as delivering tangible benefits, for instance by claiming that a preemptive attack is necessary to stop the target state becoming an unmanageable threat in the future. 7 Where citizens accept such justifications, and because of this support a belligerent leader, such support derives from utilitarian means-ends calculations focusing on the expected outcome of the conflict rather than from the experience of conflict as in the first explanation. The third explanation draws on the classic conflict-cohesion hypothesis developed by sociologists, including Emile Durkheim, Georg Simmel, and Lewis A. Coser. 8 It purports that conflict with an outgroup increases people's commitment to their ingroup and that this translates into support for group leaders. 9 This explanation thus focuses once again on people's experience of conflict (rather than on how they anticipate its outcome) but asserts that this experience stimulates more deeply rooted processes of social identification beyond mere "distraction" effects. The group identities strengthened through conflict may endure long after the conflict itself is over.
None of these three accounts fully corroborates the diversionary conflict theorists' claim that conflict can make people rally behind their leaders. The problem with the first two explanations is not that they are necessarily wrong but rather that they already presuppose what diversionary conflict theorists need to explain.
As for the first explanation, it assumes that citizens are attracted to conflict without justifying why this is supposedly so. If citizens were not already receptive to being mobilized for external belligerence, leaders would lose rather than gain support by using such belligerence as a "distraction" device. The second explanation faces a similar problem. Even if citizens believe their diversionary leader's (disingenuous) claim that a conflict he or she has started is in the national interest, it is not clear why they should wish to prioritize such collective aims over their individual welfare and security that the conflict endangers. This explanation, too, therefore leaves the central puzzle unsolved and brings us back to the original question.
The conflict-cohesion-based explanation, by contrast, appears to offer a path to understanding people's underlying motivations. Not only does the conflict-cohesion claim enjoy relatively solid empirical backing, 10 but since international conflict is quintessentially an intergroup phenomenon it seems plausible to assume that it affects people's political attitudes somehow "through" their groups. Nonetheless, reflecting the diversity of their analytical backgrounds and empirical concerns, sociological conflict-cohesion theorists continue to debate the same questions that diversionary conflict theorists also need to address. Critically, these include the questions of why and when groups matter to people in the first place and, by extension, how different scope conditions (pertaining to different kinds of conflicts, political systems, ingroups and outgroups, etc.) might mediate the effect of conflict on people's relationship with their groups. 11 Second, the conflict-cohesion claim could conceivably support the diversionary conflict hypothesis only if it were shown in a second theoretical step that increased group cohesion engenders greater support for group leaders. 12 Several analysts suspect such a causal relationship in different contexts and from various perspectives, 13 but this, too, might easily seem counterintuitive. If conflict-cohesion theorists are correct in arguing that conflict makes people value their group more strongly but at the same time puts that group at greater risk, should one not expect group members to reign in and punish leaders whose behavior exposes the group to these risks rather than reward them with more support?
Adopting the conflict-cohesion approach as a theoretical starting point, the following sections seek to address: (1) how and when conflict can increase people's commitment to their groups; and (2) how and when this can strengthen support for group leaders. To account for both processes I turn to social identity theory, a social-psychological approach conceived around the question of how relations between groups and developments inside them affect each other.

Social Identity Theory
Social identity theory asserts that people have an intrinsic need to form groups and to identify with them, and that this influences how they perceive and act inside groups and across group boundaries. Central to SIT is the phenomenon of categorization, which stems from our quest for cognitive parsimony in a world that is inherently fluid and ambiguous. In order to reduce this complexity we divide our social and physical environment into categories, overestimating both the similarity of items placed in the same category and their difference from items placed in other categories. By categorizing others and ourselves we generate human categories; by internalizing these categories and our membership in them we turn these into groups, defined as bounded human collectivities that attract the identifications of their members and influence their cognitions, emotions, and actions. 14 Inside groups, their members to various degrees conform to what SIT terms the "group prototype," that is, the norms and practices through which group members define their group and distinguish it from other groups-shared conceptions of "what it means to be" Russian, Irish, or a Rotarian. 15 For their part, relations between groups are often influenced by social comparison, as group members form judgments about their ingroups' qualities and relative status by comparing them against relevant outgroups. Being self-esteem seekers overall, we tend to compare our ingroups to outgroups along dimensions that reflect favorably upon our ingroups (and thus upon ourselves) and, by implication, less favorably upon outgroups.

Groups and Emotions
Implicit in this brief sketch is that SIT treats groups as partially constitutive of our cognitions (how we apprehend and categorize the world) and of our emotions (how we feel toward the world). Since SIT was first developed, the emotional aspects of group identification have received increasing attention and helped spawn several theoretical offshoots, notably intergroup emotions theory. 16  "people appraise ongoing events based on [their] relevance to the extended self (the social group) … and respond to them with corresponding emotions." 17 Accordingly, we may feel proud, angry, embarrassed, sad, or insulted because of something that affects a group with which we identify, such as the football team we support winning, our nation being insulted, or our gender suffering discrimination. We may experience these group-based emotions even if we are not directly affected by the events that give rise to them. 18 In line with SIT's basic premises, the strength of our group-based emotions depends in part on the strength of the relevant group identity. 19 The more deeply we have internalized a group, the more we feel "as" the group and appraise objects and events in terms of their significance for the group. The next section builds on the role of groups as objects of emotion. It models how conflict with outgroups can cause people to identify more strongly with their ingroups and to experience group-based emotions more intensely-positive emotions related to the ingroup as well as negative emotions related to the outgroup. The subsequent section charts how group leaders can act as "entrepreneurs of emotion" by channeling positive ingroup emotions toward themselves. 20

Conflict and Group Identification
According to SIT, groups' ongoing efforts to derive status by comparing themselves against other groups often leads to competition but not necessarily to aggression or even violence. 21 Pacified Western Europe, among other examples, illustrates that social comparison and status competition between states and societies can be entirely peaceful. Yet SIT suggests that where full-scale conflict does break out, it can rally people to the defense of their groups and thereby strengthen their identification with these groups and their experience of group-based emotions. For social identity theorists, an important determining factor of how strongly we identify with a group is the group's salience to us, which in part reflects the frequency and intensity with which we experience the group and become conscious of its defining norms, its boundaries, and our membership in it. Conflict is a powerful driver of group salience. By placing a group's status or even its survival at risk, conflict (and especially violent conflict) can commit group members to an intense communal and community-affirming effort that defines their self-and other-understandings in terms of their contribution to the group. As they experience one another working, sacrificing, and fearing for their group, its members affirm their communality at the inside and shared difference from the outside, each subjectively magnified by its contrast with the other. Their experience of group-related emotions intensifies, while other social categories to which they simultaneously belong (temporarily) recede into the background. Increasingly, group members perceive in-and outgroup members alike in group-stereotypical terms, which both stems from and further reinforces the group's importance to them. 22 "The more intense is an intergroup conflict, the more likely it is that the individuals who are members of the opposite groups will behave toward each other as a function of their respective group memberships, rather than in terms of their individual characteristics or interindividual relationships." 23 If conflict can increase our attachment to a group while our willingness to defend the group depends in part on the strength of this attachment, the relationship between intergroup conflict and ingroup identification can become reciprocal and drive escalating conflictual spirals that characterize many real-world conflicts. 24 In the present context, however, the critical causal link is from conflict to group identification. SIT's conceptualization of groups as internalized social categories and objects of emotions helps explain why people are willing to sacrifice for their groups and how intergroup conflict can make them identify more strongly with their groups. This link between conflict and group identification is subject to various scope conditions that I further discuss below. Before doing so, I turn to the second question posed at the outset: if conflict causes people to identify more strongly with their group, how can this rally support for group leaders? SIT suggests two mutually compatible processes.

Group Identification and Support for Group Leaders
The first process is a straightforward extension of the conflict-cohesion logic just outlined. If conflict with outgroups makes people more committed to their ingroups, it will also strengthen support for group leaders provided their constituents consider them effective defenders of the group. 25 Yet belligerent leaders who rely on this support-generating logic alone remain vulnerable to competitor elites who seek to turn it to their own advantage by accusing the incumbent of weakness or betrayal and demanding even more ferocious action against the outgroup. The ethnic conflict literature describes this as "ethnic outbidding," 26 but interstate conflicts, too, can unleash such internal contests. They can leave belligerent leaders torn between trying to appease increasingly radicalized domestic competitors while seeking to limit the damage this does to their standing with more moderate internal constituents and the wider outside world. Former Serb leader Slobodan Milo sevi c's predicament during Serbia's wars against its neighbors in the 1990s is a good example of a diversionary leader squeezed between more moderate forces and radicalized competitors calling for an even more uncompromising stance against outgroups. 27 Second, group leaders may benefit from increasing group identification among their constituents by virtue of one of the most fundamental mechanisms postulated by SIT as previously discussed: when identification with a group intensifies, attachment to that group's defining norms and symbols also grows. Strengthening group identification therefore can boost support for leaders and hierarchies inside the group provided these have appropriated group-defining norms and symbols, approximating what Johann P. Arnason describes as a "symbolic fusion" of political power with social identifications. 28 In such a sociopolitical setting, the norms and symbols that represent a group also represent the group's leader with the two dimensions drawing on strongly overlapping symbolic vocabularies. In SIT terminology, to the extent that a leader embodies the group prototype, attachment to the group equates with attachment to the leader. As the group's sway over its members strengthens, the leader's position ipso facto consolidates. Support that emerges through this second process is harder to usurp by internal competitor elites. For it arises not because group members consider their leader simply a means to defend the group (as in the first scenario) but rather see him or her as the group's embodiment and personification and thus value the leader in the same way they value the group itself: intrinsically rather than just instrumentally. Where "the leader constructs him-or herself as the embodiment of the ingroup … anything the leader says or does by definition encapsulates the group identity and anyone who opposes the leader by definition becomes an opponent of the group. The extent to which leaders become "fused" to their groups in turn depends in part on the political system in which they operate. It is a hallmark of many authoritarian systems that they seek to infuse norms, symbols, and rituals signifying the group with those representing the leader. At the extreme end of the authoritarian spectrum from this perspective are systems in which power and its symbolic manifestations are highly personified and the leader's attempts to appropriate symbols of national identity and communal belonging are correspondingly intense-take the omnipresent cult of the leader and its interwovenness with nationalist group imagery that itself is fashioned in the leader's image in the likes of Turkmenistan and North Korea. 30 How a regime's subjects respond to such efforts needs to be established in any given instance. Overall, however, findings by some researchers that highly personalized "strongman" regimes are more likely than other authoritarian governments to react to domestic turmoil with external belligerence square with such an account. 31 That same logic would also seem to explain why a slide towards more personalized leadership styles in (semi)authoritarian and semidemocratic systems often goes hand-in-hand with more confrontational policies towards the outside world and/or internal minorities-Hugo Ch avez's Venezuela and Robert Mugabe's Zimbabwe are relevant examples, as is Putin's Russia, to which I will return below.
Democracies by contrast differ from autocracies, among other things, in that they typically separate more strongly between the day-to-day exercise of political power and symbols of national communal belonging. This separation helps insulate national identifications from frequent changes of political personnel. However, as with authoritarian systems, here, too, different subtypes have different characteristics. The separation between political decision makers and communal symbols is typically strongest in parliamentary systems where the nation's embodiment and personification falls to constitutional monarchs or presidents with little political power. By contrast, in (semi)presidential democracies that (partially) merge the roles of political chief executive and symbolic head of state, the exercise of political power may be deeply embedded in symbols and rituals that for their part also serve as communal identity markers. Compare the prosaic daily functioning of British and Dutch prime ministers with the symbolically highly charged and ritualized presidential role in the United States and some similar systems.
Diversionary conflict theorists have not systematically separated between different democratic systems from that perspective. Yet the argument in this section accords with findings that struggling presidents in presidential democracies overall are more likely to use military force abroad than floundering leaders of coalition governments in parliamentary systems. 32 This discrepancy exists even though leaders in both settings tend to face powerful veto players that make it hard for them to remedy the crisis at hand through domestic reforms while they enjoy a comparatively freer hand abroad. Following the argument in this section, one reason for this difference in outcome might be that, compared to leaders of coalition governments, presidents in presidential democracies have a greater incentive to choose external conflict given that they stand to reap particularly strong popularity gains as a result.

Scope Conditions
The argument so far offers a basic rationale for the conflict-cohesion and the diversionary conflict propositions respectively. It shows (1) that conflict can make people identify more strongly with their groups; and (2) that stronger group identification can increase attachment to group leaders. Yet the strength of both processes depends on several factors beyond the type of political system in which leaders operate. The first process depends in part on how ingroup members perceive the outgroup that is being targeted by the belligerent leader, and how the leader and his or her domestic allies frame such perceptions. The second process partially depends on how leaders position themselves domestically in relation to the conflict. The following two subsections discuss these in turn.

Perceptions and Framing of the Outgroup
For a conflict to increase people's attachments to their ingroup, they must have become mobilized in support of that conflict. In turn, for mobilization to occur they must hold negative feelings toward the relevant outgroup. Such emotions can include hatred, disdain, distrust, a desire for revenge, feelings of humiliation, victimization and of being threatened and, most of all, anger. 33 As conceptualized by SIT and intergroup emotions theory discussed earlier, such emotions are groupbased and mostly derive from stereotypical representations of the outgroup. Consequently, leaders are often able to manipulate such emotions through propaganda campaigns and other means. Yet even where leaders manage to fuel negative emotions toward an outgroup, this does not inevitably mobilize their constituents in support of conflict against that group. Instead, various factors mediate between group-based emotions and action tendencies. For example, Eran Halperin found that people tend to become more supportive of aggression against an outgroup (1) if they harbor entrenched negative sentiments toward that group revolving around hatred; and (2) once they learn of an (actual or fictional) event that induces acute anger directed at the same group. 34 Entrenched hatred and acute anger combined mobilize people much more powerfully than either emotion in isolation. In a similar vein, where ingroup members have long perceived the outgroup as posing a threat, they "will be more attuned to threatening cues and will be more likely to interpret individual events as threatening and as calling for a violent response." 35 Against this backdrop, leaders wanting to mobilize their constituents for conflict often seek to influence their feelings toward the outgroup on both dimensions. On the first dimension, this typically involves attempts to "construe the nature of self and other in order to create appraisals of threat and hence legitimate hostility against the outgroup. The Nazi portrait of Jews as vermin, the Hutu extremist portrait of Tutsis as cockroaches [and] the Hindu nationalist portrait of Muslims as cow killers" exemplify such entrenched negative group-based stereotypes. 36 On the second dimension, attempts to promote acute anger against an outgroup often center on blaming the whole group for the anger-inducing actions of individual members. Some nationalist movements in Europe routinely linking terrorist attacks by Muslim extremists to some supposed character trait of the general Muslim population exemplify such a generalization of blame. In other cases, behavior attributed to the outgroup may be entirely fictional, such as in Nazi propaganda films depicting "Aryan" Germans being victimized by conspiring Jews. Whichever strategy elites choose, it is often at the point where the long-term buildup of negative stereotypes intersects with a dramatic anger-inducing event that belligerent elites manage to turn latent hostility toward an outgroup into full-scale violence. The genocide of Tutsis and moderate Hutus in Rwanda is a poignant example. The mass killing was sparked by the shooting down of Rwandan president Juv enal Habyarimana's plane in 1994 but occurred on the back of a long campaign of Tutsi dehumanization through mass media and other means. 37 Yet even where hatred and anger combine, their effect may be tempered by other factors. Hatred, anger, and feelings of humiliation tend to trigger inwardly directed resignation rather than outwardly projected aggression if ingroup members perceive the outgroup superior to the point of rendering the ingroup impotent. 38 Fear may have a similar constraining effect. In line with SIT, people are more likely to rally against outgroups if they believe that these groups pose an acute threat to the ingroup. For example, in the United States, public support for US military involvement abroad is typically highest if citizens judge it necessary to avert a direct threat to the United States. 39 Such threat perceptions involve fear. Yet studies found that fear by itself may dissuade group members from wanting to challenge the fear-instilling outgroup unless they expect their group to win, in which case their fear has in a sense been eclipsed by confidence. 40 Where fear outweighs confidence it tends to paralyze rather than mobilize, which also would seem to explain why scaremongering alone rarely serves as an effective mobilization strategy. 41 When seeking to mobilize citizens for conflict with an outgroup, diversionary leaders therefore must perform delicate balancing acts. While they need to convince their constituents that the outgroup is strong enough to pose a realistic threat to the ingroup, they at the same time tend to portray the outgroup as inferior and facing likely defeat. Here again, former president Milo sevi c is a good example. From the late 1980s onwards, his attempts to mobilize Serbs in support of conflict with their neighbors revolved around cultivating a sense of Serb victimization and humiliation at the hands of various ougroups (Muslims, Croats, the EU, etc.). Simultaneously, the Serb leader overlaid these victimization and humiliation themes with a chauvinistic nationalism that hailed Serbia's supposed superiority over its neighbors. 42 As argued below, President Putin's depiction of Russia as the victim of a sneaky and aggressive yet degenerate West reflects a similar mixing of the themes of danger, victimhood, strength, and superiority.
Yet, even where leaders manage to rally their constituents in support of conflict and through this strengthen their attachment to the group, this does not inevitably rally the public behind their rule. As was suggested, whether increased group identification transfers to leaders depends in part on how strongly they are "fused" to their group and thus also on the political system in which they operate. Moreover, as the next section shows, it also depends on how leaders position themselves domestically in relation to the conflict.

Perceptions and Positioning of the Leader
Diversionary leaders seeking to use a conflict to rally public support for themselves may do so in two main ways. First, conflict typically allows leaders to present themselves in various group-protecting and group-defending roles that strengthen their perceived "fusion" with the group-meeting wounded soldiers, awarding bravery medals, visiting civilians in air raid shelters, appealing for national unity in televised addresses to the nation, and so forth. 43 The potential of conflict to tie leaders to their groups in this way is especially strong since during national emergencies domestic opponents often desist from criticizing the leader and from questioning her or his motives. Furthermore, group members tend to experience (especially violent) conflict both intensely and over protracted periods. This creates a combination of impact and length of exposure, both of which promote the internalization of emotional associations and the development of durable affective dispositions. 44 Second and parallel to this, leaders may instrumentalize a conflict to further amplify their constituents' experience of positive group-based emotions as well as the extent to which these emotions become associated with their rule. To that end, they may frame the conflict and their role in it in terms of familiar historical templates that for their part are already positively charged. 45 For example, overwhelming public support in Britain for the Falklands/Malvinas War in 1982 and rising support for (the hitherto unpopular) Prime Minister Margaret Thatcher were in part driven by several mass-circulation popular newspapers persistently covering the war through the lens of Winston Churchill fighting Nazi aggression and of Thatcher as the archetypical embodiment of British grit and resilience. 46 Such highly personalized associations and comparisons transformed the conflict into a perceived rerun of a positively charged experience with Thatcher at the helm. Whether leaders can cultivate such associations obviously depends, among other things, on the availability of suitable historical templates, on their personality, and on the support they enjoy among opinion multipliers.
In a similar vein, leaders may seek to frame conflict-related events in a way that converts negative emotions (such as fear and doubt) into positive emotions (such as confidence, solidarity, and resolve) while placing themselves at the center of this emotional "conversion" process. For example, President George W. Bush's popularity rose by around forty percentage points after the September 11 attacks. Yet as James N. Schubert and his collaborators' detailed tracing of public sentiments from the immediate aftermath of the attacks shows, the surge in Bush's approval ratings was not caused by the attacks themselves, whose primary effect was to instill fear and confusion. 47 What instead ignited Bush's rise in popularity was his rhetorical response to the atrocities, encapsulated by his carefully calibrated television address on the evening of September 11. Though Bush arguably had no diversionary intentions, his appearance "triggered a rally effect [around the president] by presenting Bush to the nation in an appropriate leadership role, reassuring the nation of the government's stability and capacity to function, while committing the nation to a forceful, retaliatory response to the attacks." 48 Long before his administration had decided on its political and military response to the atrocities, Bush's rhetoric helped to convert public fear and confusion into a (however fragile) sense of continuity and reassurance.
As Lloyd Cox and Steve Wood argue, the killing of Osama bin Laden by US forces in 2011 benefitted President Barack Obama in analogous ways, though he, too, had no obvious diversionary motive. Obama's somber address to the nation after the killing, in which he recalled bin Laden's atrocities and claimed that justice had been served, was complemented by images depicting Obama with his aides in the White House Situation Room supervising the raid on bin Laden's hideout in real time. Beyond just emphasizing leadership, such depictions reinforced and channeled the emotions bin Laden's killing stimulated among the American public. They helped Obama to convert anger, fear, and humiliation invoked by the memory of 9/11 into a shared sense of justice, revenge, resilience, and closure to trauma. 49 As argued below, President Putin's framing of the Crimea annexation as symbolizing Russia's transition from post-Cold War humiliation, victimization, and betrayal to resurgent strength and self-assertion represented a similar effort to transform negative into positive group-based emotions and to personify that shift. In these and other examples, conflict can increase popular support for leaders if their words and actions have the "emotional effects on public opinion of quelling fears while instilling confidence and efficacy in the political response…. Simultaneously, the response should stimulate positive feelings of reassurance, hope, and enthusiasm." 50 Finally, group-based perceptions and emotions "can create their own dynamics or spirals of action and reaction" 51 across group boundaries, making the conditions discussed in this and the previous subsection subject to various interaction effects. Where hostility between two groups is entrenched, leaders on one or both sides find it relatively easy to frame the other group in such a way as to induce support for aggression and thus aggravate long-term mutual hostility still further. The Israeli-Palestinian conflict is a frequently invoked example of entrenched negative perceptions facilitating regular flare-ups of violence while being further reinforced by them. 52 Conversely, where long-term perceptions between two groups are largely positive, their leaders tend to refrain from diversionary belligerence, which then further solidifies benign mutual attitudes and makes it even harder and less likely for leaders to mobilize their publics in support of conflict. The diversionary conflict literature has not systematically explored such interaction effects and doing so would go beyond the scope of this article. However, an interaction perspective could help address puzzles such as the far-reaching absence of diversionary violence (accompanied by the absence of other forms of violence) in democratic dyads, 53 which coincides with generally trusting mutual perceptions among their mass publics. 54 Diversionary conflicts and their enabling conditions may constitute as well as de-constitute each other over time.

Rallying Around Russia and Putin: Public Opinion and the Crimea Conflict
This second part of the article applies the two SIT-deduced processes modeled in the first part to Russia's annexation of Crimea in 2014, assessing how the annexation affected national identifications and public support for President Putin. To do so, it treats the two proposed processes as the basis for two respective hypotheses: Hypothesis 1 (H1). This hypothesis states that conflict with an outgroup can cause people to identify more strongly with the ingroup. Applied to the Crimea conflict, if H1 holds we would expect the conflict to have been accompanied by strengthening national identifications and growing ingroup pride among Russians. In parallel, we would expect the Crimea conflict to have worsened attitudes toward Ukraine and the West since the conflict was primarily in opposition to these two outgroups, and since a SIT-based framework suggests that during conflict positive ingroup and negative outgroup perceptions become mutually reinforcing. If the matching null hypothesis holds, we would find no significant change in national identifications resulting from the conflict.
Hypothesis 2 (H2). This second hypothesis states that where group attachments strengthen, these can become channeled toward the leader and boost his or her popularity. Applied to the Crimea conflict, evidence for H2 would be in the form of increased popular support for Putin. Because such support would derive from intensifying national identifications and benefit Putin by virtue of him being perceived as embodying the nation, we would expect it to outlast the immediate context of the Crimea annexation and to extend to Putin's policies beyond the Crimea conflict itself. The matching null hypothesis states that group identification does not affect support for the leader. According to the null hypothesis, even if the Crimea conflict increased national identifications (as per H1), support for Putin would have remained unaffected by this. To the extent that the annexation nonetheless boosted support for Putin, this would have stemmed exclusively from factors other than a projection of national attachments onto the Russian president, for example, from the annexation diverting public attention from domestic problems and/or from the public admiring Putin's handling of the annexation. Such support, however, would likely be volatile (tied to the specific event of the Crimea annexation) and could be eroded relatively easily by countervailing factors such as the experience of economic hardship resulting from the conflict. If the argument in this article is correct, the Crimea conflict had several dimensions that made it prone to strengthen both national sentiments and support for President Putin. As is shown below, over many years prior to the annexation Putin had sought to "fuse" his personae with nationalist themes and symbols. Moreover, the Russian mass media's framing of the annexation consistently emphasized national themes while placing Putin's personae center stage. If, in light of these presumably favorable scope conditions, the Crimea conflict turned out not to have affected group identification and attitudes toward Putin as predicted by the two hypotheses, the social-psychological logic outlined in this article would likely be flawed or at least in need of further specification. Moreover, that same logic would be unlikely to apply to conflicts whose scope conditions appear less conducive to it than did those in Russia at the time of the Crimea annexation. 55 If, by contrast, it became evident that the Crimea conflict did enhance national identifications, and through this support for Putin, this would back the article's theoretical argument while illustrating how the social-psychological processes outlined can manifest themselves in a real-world scenario. In that sense, the Crimea conflict serves to empirically test and, potentially, to illustrate the article's theoretical claims.
The section begins by briefly discussing Putin's efforts to "nationalize" his leadership prior to the annexation. It then ascertains how the Crimea seizure, Putin's rhetorical framing of it, and Russia's ensuing standoff with Ukraine and the West affected national sentiments and popular attitudes towards Putin. Both developed in ways consistent with H1 and H2 respectively.
Putin's Positioning: From the "Nationalist Turn" to the Crimea Annexation Russia's annexation of Crimea, which began in late February 2014, followed sustained efforts by Putin to embrace policies, rhetoric, and symbolic gestures with a broadly national and nationalist slant. These had origins in the early years of Putin's reign but gained added momentum after the anti-Putin street protests of 2011-12 and his government's relatively poor showing in the parliamentary elections in December 2011. Surveys suggested that much of this opposition to Putin emanated from a broadly defined nationalist political spectrum, 56 which many of Putin's post-election policies seemed designed to appease.
Enthusiastically backed by a largely Putin-friendly mass media and various Kremlin-sponsored youth movements and civic organizations, Putin's nationalist reorientation encompassed an eclectic mix of rhetoric, symbolic gestures, and policy initiatives at home and abroad. Many of these were predicated on claims that the West had betrayed post-Soviet Russia and that this justified Russia adopting a more confrontational stance. As encapsulated in Putin's landmark Munich speech of 2007, such grievances had been fueled notably by NATO's enlargement into Central Europe and by alleged US attempts to forge a unipolar world order through its policies in the Middle East and elsewhere. 57 In addition, the Russian government became increasingly outspoken against several former Soviet republics, accusing them of maltreating Russian-speaking minorities. Concerns for what it had started to refer to as "compatriots abroad" 58 (some of whom were granted Russian citizenship) evolved into a central theme for the Russian government, which used them notably to justify its military intervention in South Ossetia in August 2008 and the war with Georgia it ignited. Referring to the conflict, President Dmitry Medvedev (Putin was prime minister at the time 59 ) proclaimed it his government's "duty to protect the lives and dignity of Russian citizens wherever they may be." 60 These themes of threat and confrontation accompanied a growing official emphasis on Russia's social and cultural distinctiveness and an increasingly virulent rejection of Western-style liberal democracy as a worthy model for Russia to emulate. Underscoring its growing self-demarcation from the liberal West, Putin's government fashioned closer links with the socially conservative Russian Orthodox Church while stepping up persecution of critical artists, NGOs, and sexual minorities, often accusing them of subversion and of acting as Western agents. Rhetorically, meanwhile, Kremlin officials, pro-Putin public intellectuals and the state- controlled media depicted Western Europe as a repellent social model in decline, in the thralls of excessive individualism, runaway materialism, moral relativism, and homosexuality, while lacking the strength and character to free itself from US domination. 61 As one commentator observed, what these various policies and portrayals aimed for was to construct Russia not as a society apart from Europe so much as an embodiment of "the True Europe, while Western Europe is a corrupted version occupied, influenced, and suborned by the United States." 62 Throughout Putin's reign, the Russian mass media became subjected to steadily tightening formal and informal state control. 63 Its coverage of the government became more homogenously supportive and more strongly focused on Putin himself, an emphasis that became especially strong during the war with Georgia in 2008. 64 Representations of Putin as defender and protector of the nation were visually reinforced by a continuous stream of media images of Putin in martial, "manly," and "patriotic" poses-brandishing rifles on hunting expeditions, piloting fighter jets, inspecting military maneuvers in combat uniform, flooring opponents during martial arts sessions, and bare-chested on horseback riding through the Russian steppes. While such poses were widely derided abroad, the message conveyed to Putin's domestic audience was relentless and unambiguous: Russia reasserting itself guided by a powerful and respected leader, and a "communal outpouring of love and appreciation for Putin, with a focus on his strength, patriotism, and protection of Russia's national interests." 65 None of this meant that Putin could jump on a preexisting nationalist frenzy, anti-Western or otherwise. On the contrary, research consistently showed that post-Soviet Russian identity was relatively weak, fragmented, and demoralized. 66 What Putin rather appeared to aim for was the partial (re)construction of post-Soviet Russian national identity in his own image: it was as much about "nationalizing" Putin as it was about "Putinizing" national sentiments. These efforts appeared to resonate with a broader Russian public from relatively early on. As far back as the middle of the last decade, surveys indicated that a growing number of Russians were defining their leader in terms of the very characteristics his policies, pronouncements, and publicity campaigns sought to convey: as a tough and manly defender of the nation against its enemies at home and abroad. 67 The Crimea annexation fit into this pattern of national assertiveness tailored around President Putin. Taken from Russia and "gifted" to Ukraine by Khrushchev, the fate of the peninsula and its mainly Russian-speaking population had long occupied Russian nationalists of various stripes and had become a set-piece feature in the Kremlin's "compatriots abroad" rhetoric over the preceding years. 68 Russian resentments were further aggravated by widespread popular protests in Ukraine throughout the winter of 2013-14 demanding democratic reforms and closer ties with the European Union, and by the overthrow of Ukraine's pro-Russian president in February 2014. The Kremlin blamed both on Western efforts to undermine Russia's regional influence and to turn Ukraine into a client state of the West.
Crimea's return to Russia was swift. Taking advantage of Ukraine's domestic turmoil, Putin employed a "hybrid strategy" that began with an intensive pro-Russian media campaign targeting the Crimean population. This was followed on the last day of February 2014 by Russian soldiers without insignia occupying strategic locations across Crimea and subsequently by regular Russian troops consolidating Russia's hold over the territory. 69 Two weeks later, a large majority of Crimeans voted to join the Russian Federation in a Kremlin-backed referendum. At the earliest stages of the intervention Russia denied direct military involvement, but it was clear that the Russian government was guiding events. After the referendum had sealed Crimea's return to Russia, pro-Kremlin media were quick to celebrate it as one of Putin's foremost patriotic achievements. 70 At the same time, the annexation threw Russia into a prolonged standoff with Ukraine and the West. It became defined by tightening economic sanctions and counter-sanctions, and by ongoing tensions in eastern Ukraine where the Ukrainian and Western governments accused Putin of arming pro-Russian separatists in the Donbass region with a view to dismembering Ukraine still further.
Putin's portrayal of the seizure slotted in seamlessly with the general nationalism-and patriotism-centered themes that had marked his rule from the middle of the preceding decade. They revolved around anger and a sense of betrayal at the West's alleged disrespectful meddling in Ukraine and elsewhere, mapped onto long-lasting grievances over Russia's post-Cold War loss of status and the fate of "compatriots abroad." These themes of humiliation and anger were mixed with expressions of pride over the annexation, which was held up as a symbol of Russia's national reawakening inexorably linked to Putin's leadership. Putin developed these themes extensively in a televised address on 18 March 2014 that marked Crimea's formal reincorporation into Russia. 71 Adopting a tone that "mixes politics, national resentments and nationalism, all overlaid with a religio-mystical tone that sounds, at times, almost messianic," 72 Putin proclaimed that everything in Crimea was "symbolising Russian military glory and outstanding valor." Yet when the Soviet Union collapsed, Crimea's Russian-speaking majority joined many other Russian speakers in various Soviet republics who "went to bed in one country and awoke in different ones, overnight becoming ethnic minorities … while the Russian nation became one of the biggest, if not the biggest ethnic group in the world to be divided by borders." At that point, Russia realized "that it was not simply robbed, it was plundered." Russia's humiliation, according to Putin, was compounded by what he described as a Western-inspired coup against Ukraine's pro-Russian government, which he claimed had left the country in the hands of "nationalists, neo-Nazis, Russophobes and anti-Semites" bent on threatening Russian-speakers in Crimea and elsewhere in Ukraine. Putin assured his audience that Ukraine's new rulers had already decided to pay back their masters in the West by joining NATO and allowing NATO to move into Crimea. Had this been allowed to happen, "NATO's navy would be right there in this city of Russia's military glory, and this would create … a perfectly real threat to the whole of southern Russia." "I simply cannot imagine that we would travel to Sevastopol to visit NATO sailors." 73 In the same speech, Putin conceded that the Crimea annexation would make Russia's relations with the West more confrontational, equating potential Western countermeasures with domestic opposition to his rule and thus, implicitly, his rule with the Russian nation's independence and resistance to foreign interference. "Some Western politicians are already threatening us with not just sanctions but also the prospect of increasingly serious problems on the domestic front. I would like to know what it is they have in mind exactly: action by a fifth column, this disparate bunch of 'national traitors,' or are they hoping to put us in a worsening social and economic situation so as to provoke public discontent?" Yet Putin assured his audience that with "the will of millions of our people, our national unity and the support of our country's main political and public forces," Russia's enemies would find that the nation could not be divided. 74 In the months following the annexation and against the backdrop of mounting international criticism and hardening economic sanctions, Putin's rhetoric became ever more dramatic. Addressing a meeting of Russian ambassadors, the president claimed that "everything Russia has fought for since Peter the Great was threatened" by the overthrow of Ukraine's pro-Russian president. 75  and pride. In a documentary titled "Crimea: Way Back Home," aired on the popular pro-Kremlin Rossiya 1 television network on the first anniversary of the annexation, Putin offered an extensive account of his role in planning and supervising the annexation and claimed that he would have placed nuclear weapons on alert to defend Russia's hold over Crimea had this become necessary. 76 The documentary and many other portrayals in the largely Putin-friendly Russian media rehearsed the same set of basic themes discussed earlier: Russia and Russians had suffered humiliation and disrespect and faced acute dangers (a Western-engineered coup in Ukraine, threats to "compatriots abroad," NATO eyeing Crimea, etc.); Russia had successfully fought off these dangers and asserted itself against a sneaky and treacherous West; and Crimea's return to Russia corrected a historical wrong while representing "a triumph of security planning and execution, with Mr. Putin at its heart" 77 -a leader who, according to his spokesman Dmitry Peskov, was now "probably the main guarantor of the safety of the Russian world." 78 In all these depictions relentlessly reiterated to the Russian public, the return of Crimea symbolized the end of Russia's post-Cold War humiliation and a resurgence of national pride with Putin cast as presiding over and embodying this transformation, animated by the same virtues of strength, pride, and patriotism that his rhetoric conferred upon the Russian nation at large.

Effect on National Identifications (Hypothesis 1)
Survey data indicate that the Crimea annexation was extremely popular among the Russian public. A spring 2014 Pew Research Center survey showed that 89 percent of Russians welcomed Crimea's return to Russia. 79 Moreover, the seizure was followed by significant increases in national pride and in general ingroup positivity among Russians and by an even sharper worsening of perceptions of Ukraine and the West. The share of respondents who professed a "very favorable" view of Russia soared from 29 percent in the spring of 2013 to 51 percent immediately following the annexation one year later. It rose further to 63 percent the year after as Russia's standoff with Ukraine and the West hardened, Western sanctions tightened, and the Kremlin's defense of the annexation became ever more dramatic. 80 Several related indicators compiled by Russia's respected Levada Center point in the same direction. The share of respondents satisfied with Russia's level of "national well-being" increased from 41 percent in November 2013 to 59 percent in May 2014 following the annexation. 81 The proportion of Russians who believed that Russia was "moving in the right direction" likewise rose from 43 percent in January 2014 to 60 percent immediately after the annexation in March 2014 and remained at similarly high levels throughout 2015. 82 For many Russians, the boost in national self-esteem was so strong that they even projected it onto how they thought foreigners perceived their country. Immediately after the annexation, 43 percent believed that it had improved Russia's standing in the world, with only 27 percent sensing the opposite effect. 83 Meanwhile, Russian perceptions of the West deteriorated sharply during the same period. By early 2015 only 15 percent still had a positive image of the United States (down from 51 percent two years earlier), with perceptions of Germany, NATO and the European Union experiencing similarly steep drops. 84 Judged by these figures, the effect of the Crimea conflict on Russian national sentiments broadly corresponds to the conflict-cohesion claim and its elaboration by social identity theory and thus accords with H1. Extremely popular in its own right, the Crimea annexation boosted ingroup pride among Russians and a sense of collective optimism and well-being. Ingroup positivity intensified still further as Russia became entangled in a sustained conflictual relationship with Ukraine and the West and as Putin's tone hardened. The simultaneous and sharp deterioration in outgroup perceptions is in line with a SIT-style competitive social comparison process as part of which outgroup negativity and ingroup positivity reinforce each other. It was not only the annexation itself that appears to have boosted positive ingroup emotions but also a perception among many Russians that their country was holding its own in a standoff with parts of the outside world they viewed with growing suspicion. Their increasingly hostile attitudes towards the West, soaring levels of national pride, and their stated belief that Russia had regained respect in the world suggests that the Russian public had largely absorbed its president's framing of the annexation. As did Putin's rhetoric, public attitudes revealed a sense of shared victimhood paired with feelings of regained national efficacy, self-assertion, and pride.

Effect on Support for President Putin (Hypothesis 2)
Throughout the first eleven months of 2013, Putin's approval ratings had remained relatively stable. They started to rise sharply as the popular uprising against Ukraine's pro-Russian president gained momentum and Russia's stance towards its neighbor hardened, from 61 percent in November 2013 to 69 percent in February 2014. They rose even more dramatically-from 69 to 80 percent-between February and March 2014, coinciding with Russia's military intervention in Crimea and the territory's return to Russia during that period. Support for Putin continued to climb to 88 percent in late 2014 as Russia's suspected involvement in eastern Ukraine exacerbated tensions with Ukraine and the West still further. 85 Parallel to this, satisfaction with the Russian government's performance in many areas with no direct bearing on Ukraine or Crimea also rose sharply. This included Putin's perceived record in fighting corruption, in respecting personal freedoms and in managing the economy. 86 Growing approval of Putin's economic performance is especially remarkable, since over the same period Russians judged their actual economic circumstances to have gone from bad to worse. By early 2015, 73 percent deemed the economy in a bad state (up 9 percent from 2013), reflecting a shrinking GDP, rising inflation, and a collapse of the value of the Ruble on currency markets. Russia's economic woes were clearly exacerbated by tightening Western economic sanctions in response to the Crimea seizure and Russia's suspected support for separatists in eastern Ukraine. Despite this, only 25 percent of Russians thought that Putin's policies bore any responsibility for the economic downturn. 87 In sum, Russian public opinion in response to the Crimea annexation developed in ways consistent with the article's two SIT-derived hypotheses. The annexation appeared to stimulate national pride and a sense of collective well-being among Russians (H1) alongside a dramatic and sustained increase in President Putin's popularity (H2). Both intensified still further as Russia became entangled in a sustained standoff with Ukraine and the West and faced mounting economic and political pressure from abroad and as the Putin-allied mass media relentlessly hailed the annexation as a symbol of Russia's national revival inexorably linked to Putin's leadership. Having "fused" his personae to national themes and symbols over the years prior, it appears that Putin successfully channeled the surge in ingroup pride sparked by the annexation toward himself and thereby created a diversionary effect.

Alternative Interpretations
The interpretation outlined over the previous pages might attract two objections. First, the data presented allows for the possibility that the relationship between soaring ingroup esteem and growing support for Putin could have operated in the opposite direction: the highly popular annexation could have lifted support for Putin, which then boosted group pride, given that Putin had sought to "fuse" himself with national symbols in the ways discussed. However, such a scenario does not contradict my argument. If group and leader are partially "fused" one would expect the relationship to work in both directions, with pride in one also augmenting pride in the other. Modeling a link from positive ingroup emotions to support for the leadership as I have tried to do does not rule out that the relationship may be bidirectional.
A second and potentially more serious objection is that the data presented are logically consistent with an alternative scenario in which national pride and support for Putin increased independently of each other-the former driven by ingroup/outgroup dynamics a la SIT (and in line with H1) but the latter by nothing more than popular admiration for Putin's handling of the highly popular annexation. Yet several factors would seem to contradict such an interpretation. For a start, support for Putin continued to climb long after the annexation had been wrapped up successfully and thus probably beyond any direct boost Putin might have received from his handling of it, but in parallel with rising levels of national pride and worsening perceptions of Ukraine and the West. Second, Putin's actual policy performance at any rate does not appear to have been the sole criterion by which Russians judged their leader. As was shown, after the Crimea seizure popular support for Putin's policies soared even in areas that had no direct link to Crimea or Ukraine and where objective conditions manifestly worsened, such as the economy and civil liberties. What this implies is a kind of "reverse causality" dynamic, which is in line with a social identity-centered interpretation captured by H2. Having "fused" his leadership to national themes and symbols in the years prior, the surge in national pride unleashed by the Crimea conflict also benefited Putin and rubbed off on how Russians judged their leader's performance across the board. This surge in popular approval of Putin appeared to fuel growing support for his policies, not vice versa.
In its essence, such an interpretation is not new. It closely relates to what Russian sociologists and political scientists describe as a process of "negative mobilization"-driven by a presumed interconnectedness of conflict, hostile perceptions of the outside, national pride, and support for Putin's leadership. 88 Such a link was already evident during previous conflicts in Chechnya and Georgia, both of which boosted popular support for Putin significantly. The Crimea annexation has produced its strongest and most sustained manifestation to date. In the words of one Russian political scientist: "By annexing Crimea and backing pro-Russian separatists in eastern Ukraine, the Kremlin was able to justify its military-patriotic mobilization of society and its transformation of Russia into a 'besieged fortress.' … 88 For a detailed discussion, see Gudkov, "Putin's Relapse." Public mobilization around the leader and the motherland rose to a new pitch … Individuals are invited to compensate for their helplessness by looking for meaning in collective national 'successes' that promise to bring them together and restore their pride. The annexation of Crimea has become such a 'success,' giving ordinary Russians a chance to forget their woes and feel a surge of vicarious optimism." 89 None of this indicates whether Putin intervened in Crimea in order to enhance his popularity, even though his framing of the conflict to his domestic audience might hint at such an underlying motive. What the Crimea experience suggests with greater certainty is that for as long as the Russian president remains "fused" to national themes and symbols in a way that makes external conflict liable to enhance his domestic position he will face a latent diversionary incentive. As the Crimea annexation has demonstrated, even conflicts that reduce the quality of life for most ordinary citizens can boost national attachments and support for the leader and thereby produce a diversionary effect.

Countervailing Pressures and the Limits of Diversion
Diversionary conflict theorists assert that conflict between groups and states can increase popular support for their leaders. This claim is counterintuitive, as it implies that people choose to compromise their physical security and material welfare by supporting to the very leaders who put these at risk. Relying on social identity theory, this article deduced an explanation for why and when conflict can mobilize popular support for leaders. SIT offers a parsimonious yet powerful social-psychological rationale for how conflict between groups can cause people to identify more strongly with these groups and to experience positive ingroup emotions and how this in turn can boost support for group leaders. SIT also highlights the scope conditions under which these processes are most likely to gain momentum. By dramatically boosting national pride and support for President Putin, Russia's annexation of Crimea helps corroborate the article's central argument.
Several questions flow from this account. First, the Crimea conflict, among many others, shows that conflict can make people more attached to their group and leader even when it exacts material sacrifice; indeed, for SIT the shared experience of hardship at the hands of outgroups can itself be a source of ingroup solidarity. Nonetheless, such an effect may not last indefinitely. As a conflict intensifies and the suffering it imposes on citizens mounts, support for diversionary leaders may gradually diminish and with it the conflict's attractiveness as a power-preservation strategy. Similarly, while for SIT the shared experience of physical danger can make groups more cohesive (and thus potentially benefit leaders), this effect, too, is bound to wane once individual self-preservation instincts start to override loyalties to the group and leader. High desertion rates experienced by some armies during war are one manifestation of this. While a conflict's ferocity would appear to be the key determinant of how quickly warring groups reach these tipping points, various social, political, and ideological background conditions might also play a role and merit closer investigation by diversionary conflict researchers.
Finally, the effect of political and economic sanctions against diversionary governments requires further investigation. On the one hand, by making a conflict even more costly to citizens, sanctions might bring it closer to the threshold beyond which social identity mechanisms fail to override their individual selfinterest. On the other hand, sanctions might play into the hands of diversionary leaders by further hardening their and their constituents' conflictual posture towards the outside world. Complicating matters even more, different types of sanctions might have different effects, and this, too, might differ between different kinds of conflicts, types of diversionary regimes, and stages in the conflict cycle.
The deductive logic this article has developed cannot answer such questions, but it provides a basis for further empirical inquiry. By deepening our understanding of how and why conflict can strengthen leaders and the conditions under which this is most likely to happen, it has laid theoretical groundwork for further analyzing diversionary conflicts as conflicts with distinct causes and inhibitors and a trajectory of their own.