Adult Attachment Styles in the Workplace

Prior research has demonstrated that attachment styles are important antecedents of interpersonal relationship quality and psychological well-being. Despite this, the theory of attachment styles has been largely ignored by researchers interested in workplace phenomena. The present paper aims to explain the theory of attachment styles, why researchers have overlooked attachment styles as an antecedent of organizational behavior, and a possible means of reconciling attachment theory with current models of personality. Moreover, I will review what existing research has actually demonstrated in terms of linking attachment styles to leadership, trust, satisfaction, performance and other outcomes. Finally, I will explore what possible future directions may be taken by researchers in the future in order to broaden and deepen our understanding of the role of attachment styles in the workplace. Although it is generally considered one of the “grand theories” in personality research, attachment theory has received scant attention from researchers investigating the role of individual differences in the workplace. Possible reasons for the failure to address the role of attachment styles in the workplace range from overcoming conceptual boundaries and potential assessment issues to the predominance of trait models and the general disdain for psychodynamic models in the applied literature. Nonetheless, over the last two decades, a small number of studies have attempted to examine the role of attachment styles and a variety of behaviors, attitudes, and experiences in the workplace setting. These studies have focused primarily on issues of leader– follower dynamics and perceptions, job attitudes and stress, and performance outcomes. While these studies represent some progress in the field for integrating attachment theory into standard organizational behavior models, there remains a great deal of research to be done in order to integrate attachment theory into current models of leadership, performance, and job satisfaction. Attachment Theory Attachment theory, based on the work of John Bowlby (1982), postulates that all individuals are born with an innate desire to seek proximity to others in times of need or distress in order to enhance their survival prospects. To the extent to which these efforts to gain proximity are successful, individuals develop a sense of security. This sense of security (or lack thereof) then becomes the basis of their own individual attachment style which then remains relatively fixed over the lifespan of the individual. Bowlby’s theory of attachment was originally inspired by his observation that socially maladjusted and delinquent boys were disproportionately likely to have experienced some sort of severe disruption in their early home life (Bowlby, 1944). To explain these findings, Bowlby integrated research from psychodynamic theory, comparative psychology, cognitive developmental psychology, and the principles of control systems (Fraley & Shaver, 2008). In particular, Bowlby focused his attention on the attachment behaviors (e.g. crying, grabbing and clinging, and frantic searching) he observed in young infants who were separated from their caregivers. Bowlby postulated that because mammalian infants are largely unable to feed or protect themselves, that their survival is dependent on their ability to maintain proximity with older, wiser, and more capable adults. Consequently, their actions, which may seem extreme, function as an adaptive response to separation from a primary attachment figure. That is, they engage in these behaviors in order to attract the attention and care of someone with a history of providing support, protection, and affection to the child. Bowlby argued that over time, evolutionary processes would select for individuals who were more successful at attracting and maintaining proximity to attachment figures. Over time, humans (and other species) developed an “attachment behavioral system” that is triggered whenever an infant is separated from its primary caregiver. According to this system, if an infant is in proximity to their caregiver, they will experience security, love, and confidence and will tend to be more sociable and engage in exploratory behavior. However, if the infant is separated from their primary attachment figure, they will display attachment behaviors ranging from visually monitoring their attachment figure to vocal signaling, clinging, and actively searching for their attachment figure. These behaviors persist until either the desired level of proximity and attention is reached or the child becomes exhausted. Failures to reestablish proximity were believed to shape a child’s expectations of their relationship with their caregiver as well as influencing their own conceptions of self-worth. While Bowlby’s model describes the basic processes by which the attachment system operates, it was not until later that researchers established the basic attachment patterns that emerged in response to histories of successful and unsuccessful attachment-seeking efforts. The primary attachment styles used in research today were based on research by Mary Ainsworth (Ainsworth, Behar, Waters, & Wall, 1978) using young infants assessed using a technique called the “Strange Situation.” This procedure involved separating infants from their parent for a short period of time and observing their reactions. The majority of children behaved in a manner that corresponded to Bowlby’s attachment theory. When their parent left, they engaged in attachment behaviors and/or became upset, but when their parent returned they were easily soothed. These infants were referred to as “secure” in their attachment orientation. Other infants (approximately 20%) also displayed attachment-seeking behaviors upon separation, but when their parents returned were not easily soothed and 286 Harms in Human Resource Management Review (2011) 21 continued in displays of distress. Researchers interpreted this response as still reflecting a desire for proximity to the attachment figure, but also a desire to punish their parent for leaving them in the first place. Infants with this style of response were labeled as “anxious” in their attachment orientation. The final group of infants (approximately 20%) failed to show much distress when separated from their parents. Moreover, when their parent returned, they appeared to be actively avoiding contact with their parent. Infants displaying this pattern of behavior were labeled as “avoidant” in their attachment orientation. Both of the latter styles were considered “insecure” attachment styles. Ainsworth’s research not only provided the first basic taxonomy of attachment styles, but also demonstrated that the individual differences in attachment responses witnessed in the strange situation were related to prior histories in the parent-child relationship. That is, secure infants typically had parents who were responsive to their needs while insecure infants often had parents who were either insensitive to their needs or inconsistent in their responses to the attachment-seeking behaviors of their children. Interestingly, additional research established that although there is correspondence between the attachment styles displayed towards fathers and mothers, there is also a large degree of relationship-specificity (Fox, Kimmerly, & Schafer, 1991). Consequently, it is believed that attachment styles reflect more than temperamental differences in infants (Fraley & Shaver, 2008). Bowlby hypothesized that the experiences that infants had with their parents would result in scripts or working models of attachment that would continue to influence interpersonal experiences throughout the lifespan of the individual. Recent research on adult attachment styles has largely supported this belief (Fraley & Brumbaugh, 2004; Fraley & Shaver, 2008). Research on adult attachment has largely focused on romantic relationships as an alternative context for the attachment behavioral system to operate in. In these relationships, we see functional similarities between the infant–parent and romantic partner relationships (Shaver, Hazan, & Bradshaw, 1988). For example, in both cases individuals feel more at ease when their attachment figure is present and insecure when separated. When the attachment figure is present, individuals tend to engage in close physical contact and pay special attention to their attachment figure. While the majority of prior research has been conducted on romantic relationships, it is believed that the same patterns of attachment would be found in other relationships that may activate attachment scripts such as leader–follower relationships (Kahn & Kram, 1994; Keller, 2003; Troth & Miller, 2000). Assessment of Adult Attachment Tests for adult attachment are of three primary types: interview, self-report typologies, and self-report dimensional questionnaires. The Adult Attachment Interview (Main, Kaplan, & Cassidy, 1985) focuses primarily on an individual’s attachment orientation with regard to their family of origin. Individuals are quizzed regarding the amount of contact they had with other relatives, experiences of loss (i.e. death) or separation, quality of relationship with attachment figures, feelings of rejection, beliefs concerning the motivations of attachment figures, and the presence of alternative attachment figures. Early self-report tests of adult attachment were based on Ainsworth’s taxonomy and involved giving individuals descriptions of the three primary attachment patterns and have them rate themselves according to which description best characterized them (Hazan & Shaver, 1987). Research using this tool found strikingly similar results to the strange situation technique used on infants with regard to the distributions of attachment styles in the population. Approximately 60% of individuals described themselves as generally have securely attached relationship with 20% of those surveyed describing themselves as being more similar to each of the insecure types of attachment. Later measures of attachment (e.g. Brennan, Clark, & Shaver, 1998; Fraley, Waller, & Brennan, 2000) have tended to be dimensionally-based with individuals responding to a large number of attachmentrelated statements (e.g. I worry a lot about my relationships). The dimensional models of attachment generally have two primary dimensions: attachment-related anxiety and attachment-related avoidance. Individuals high on attachment-related anxiety report greater anxiety with regard to whether their partners are available and responsive to them. Individuals high on attachment-related avoidance report disliking it when others open up to them emotionally and being less prone to relying on the support of others. Secure individuals would be those who are low on both of these individuals and report not only being more secure in terms of their expectations of others, but also more willing to be intimate with others and offer support when needed. While each of these techniques can be used to assess an individual’s attachment style, the differences in targets (e.g. parents/partners), methods (interview coding vs. self-report), and content do produce unique (Shaver, Belsky, & Brennan, 2000) and sometimes substantially different (Roisman et al., 2007) results. Although there is an emerging consensus that measures derived from dimensional models of attachment are the most precise (Fraley & Waller, 1998), many researchers continue to use the typological approach when describing attachment styles in their writings. Attachment Styles and Contemporary Personality Theory There can be little argument that although attachment theory has been quite influential among basic researchers, the theory of attachment has been largely ignored by researchers investigating the role of individual differences in applied workplace settings. The author of this paper conducted an informal survey of 19 introductory textbooks in Organizational Behavior and Human Resource Management and found that although each of them had sections dealing with personality, not a single one made any mention of attachment theory. Instead, there was an overwhelming emphasis on the Five Factor Model (FFM) of phenotypic personality traits (Goldberg, 1993; McCrae & Costa, 1995) across the textbooks surveyed. Even academic books aimed specifically at examining the role of individual differences in the workplace (e.g. Barrick & Ryan, 2003; Hogan, 2007; Roberts & Hogan, 2001) or dealing with highly relevant topics such as leadership (e.g. Bass, 1990) either fail to mention Bowlby’s attachment theory or make only cursory mention of it. As with the introductory texts, the academic books generally frame their discussions in FFM terms. The predominance on the FFM can be seen as both a blessing and curse. The FFM has enabled researchers to .nd a common language for talking about individual differences (John & Srivastava, 1999). The FFM theory emerged in the context of the existential crisis faced by the field of personality research that resulted from claims made in Walter Mischel’s (1968) book. With numerous personality programs being shut down or reduced in size (Swann & Seyle, 2005), personality researchers were eager to find an idea to latch onto that Adult Attachment Styles in the Workplace 287 would be “scientific” enough to silence their critics. They found such scientific cover in the “lexical hypothesis.” The lexical hypothesis postulated that the most important differences in human transactions would be encoded in most, if not all, of the world’s languages (Goldberg, 1993). By cataloging the adjectives of different languages and utilizing factor-analytic techniques to reduce them to manageable numbers, personality researchers arrived at the FFM or Big Five traits of Extraversion, Neuroticism, Agreeableness, Conscientiousness, and Intellect. With widespread replication of this model, researchers were able to create a common language that enabled them to compare findings across different measures and evaluate the distinctiveness of new constructs (John & Srivastava, 1999). The predominance of the FFM in the .eld of personality has led to a number of studies exploring the relationships between attachment styles and the Big Five. The first such study by Shaver and Brennan (1992) found that secure attachment was most associated with extraversion, agreeableness, and low neuroticism. Anxious attachment was primarily associated with high neuroticism and avoidant attachment was most closely linked with disagreeableness, introversion, and neuroticism. Roisman et al. (2007) showed moderate relationships between anxious attachment and neuroticism and disagreeableness. Avoidance was associated with reduced extraversion. Noftle and Shaver (2006) showed substantial relationships between anxious attachment and neuroticism, but only slight negative relationship between avoidant attachment and the Big Five traits of extraversion and conscientiousness. Gillath, Shaver, Baek, and Chun (2008) found no significant relationships between avoidant attachment and the Big Five, but did find a relationship between neuroticism and anxious attachment. Similarly, using an Italian sample, Picardi, Caroppo, Toni, Bitetti, and Di Maria (2005) found a strong relationship between anxious attachment and neuroticism, but no significant relationships between the Big Five traits and avoidant attachment. Overall, studies have generally found fairly strong relationships between anxious attachment and neuroticism while avoidant attachment is weakly associated with introversion and disagreeableness (Fraley & Shaver, 2008). Despite several studies finding links between attachment and the Big Five, it has been noted that attachment styles typically show significant predictive power above and beyond the Big Five traits (e.g. Noftle & Shaver, 2006; Roisman et al., 2007), particularly when relationship outcomes are the criteria of interest. Moreover, the relationships between the Big Five traits and attachment styles are generally small or insignificant. Indeed, Bowlby’s writings make it clear that attachment styles were never meant to be considered a composite of general personality traits and efforts to define attachment in those terms are misguided (Fraley & Shaver, 2008). For example, attachment theory postulates that when individuals feel they have secure base of attachment that they can trust and rely on, they are more willing to engage in exploratory behaviors. In FFM terms, this would be akin to making the argument that individuals low in neuroticism become high in intellect/openness to experience. Not only does this violate the orthogonal nature of FFM phenotypic traits, but it implies that there is a causal relationship between the traits themselves. To address this issue, more recent models of personality such as the Neo-Socioanalytic model (Roberts, Harms, Smith, Wood, & Webb, 2006) make it clear that traits do not subsume other domains of personality (e.g. goals, interests, and motives), but that each of these domains has unique relationships with important life outcomes and should be considered in the context of other personality factors rather than independently. Consequently, from a Neo-Socioanalytic perspective, researchers would be advised to take into account both traits and attachment styles simultaneously in order to gain a fuller understanding of personality effects. This would be particularly true when interpersonal relationships are the criteria of interest. Although it appears that attachment styles cannot be subsumed under Big Five traits, one recent model of personality presents an opportunity for integration at a deeper level by taking into account the motives, abilities, and perceptions (MAP) that underlie both phenotypic traits and attachment styles. The MAP model of personality Wood and Hensler (2010) was inspired by other similar frameworks for explaining general behavior across multiple disciplines such as social psychology (Ajzen, 1991; Reis, 2008; Snyder & Cantor, 1998), personality psychology (Fleeson & Jolley, 2006; Mischel & Shoda, 1995), industrial organizational psychology (McClelland, 1985, 1987; Porter & Lawler, 1968; Vroom, 1964), and economics (Manski, 2000). Across these models, common ideas are shared as to the basic determinants of behavior. First, many of these models note that in order to understand an individual’s behavioral tendencies, it is important to be aware of what goals the person is trying to achieve or the outcomes that they find desirable. Second, it is importance to understand personal capacities or situational attributes that influence the difficulty of performing a behavior. Finally, these models note that there exist differences in a person’s perceptions of the situation, their role, and the expectations of setting which function to make the behavior (or desired outcome) seem more available, appropriate, or functional. Under the MAP framework, Big Five traits and attachment styles and the covariation between them are best explained as a composite of several MAPs (Wood & Hensler, 2010). Put in the context of MAP units, we get a clearer picture of the psychological mechanisms behind attachment behaviors. For example, it has been noted that secure individuals possess a greater capacity to regulate emotions than both avoidantly and anxiously attached individuals (Cooper, Shaver, & Collins, 1998). Specifically, anxious individuals tend to react more strongly to negative emotions while avoidant individuals have trouble disengaging from negative emotions (Gillath, Bunge, Shaver, Wendelken, & Mikulincer, 2005). In terms of goals, avoidant individuals seem to actively avoid closeness in relationships while anxious individuals are more driven by the fear of being distant from others (Locke, 2008). Similarly, while both anxious and secure individuals may desire intimacy, avoidant individuals do not (Tidwell, Reis, & Shaver, 1996). Another defining difference between secure attachment and the two insecure types is that secure individuals perceive others as being trustworthy or good and that they can depend on others (Fraley & Shaver, 2008). Attachment styles as MAPs can also be seen in the nature of the items used to assess attachment styles. For example, in Griffin and Bartholomew’s (1994) 30-item Relationship Scales Questionnaire 9 of the items reflect motives (e.g. “I want emotionally close relationships”), 7 reflect abilities (e.g. “I find it relatively easy to get close to others”), and 14 reflect perceptions (e.g. “I find that others are reluctant to get as close as I would like”). This ratio of components, with an overweighting of perception items, reflects the general tendency of attachment researchers to refer to attachment as a working model or schema (Fraley & Shaver, 2008). On the whole, however, it is important to remember that attachment styles (and phenotypic traits) should be thought of as an integrated system of MAP units which interact with one another to produce patterns of behavior. It is these patterns that dictate the individual’s interpersonal experiences. 288 Harms in Human Resource Management Review (2011) 21 Attachment Theory and Research in the Workplace Personality has received fairly extensive attention with regard to its links with organizational outcomes such as job performance (Barrick & Mount, 1991; Hogan & Holland, 2003; Organ & Ryan, 1995), leadership (Bono & Judge, 2004; Judge, Bono, Ilies, & Gerhardt, 2002; Zaccaro, 2007), and job attitudes (Judge, Heller, & Mount, 2002). Despite this, none of the major reviews has ever included attachment styles as a domain of personality to be evaluated as a possible antecedent of important workplace outcomes. This oversight is no doubt partially due to the overwhelming tendency to categorize personality variables according to the Big Five which, as discussed above, is a framework ill-suited for attachment-related constructs. But it is also likely that attachment theory has gone overlooked by most applied researchers for historical and theoretical reasons. One reason is that while attachment styles have been studied extensively in the developmental psychology literature, they have only come into widespread use in mainstream personality psychology very recently. For example, the widely read second edition of the Handbook of Personality (Pervin & John, 1999) included no chapter on attachment theory. However, one decade later the field of attachment had received enough attention to warrant its own chapter in the third edition of that same publication (John, Robins, & Pervin, 2008). Another related reason may be linked to the popularity of attachment theory in the developmental psychology literature itself. By being closely linked with research on relationships between parents and children, many researchers may have dismissed attachment styles as not being relevant to workplace settings. More recently, when attachment researchers began exploring the nature of adult attachment styles, they primarily focused on romantic relationships. Once again, it is likely that this research was either ignored by applied researchers or dismissed as being irrelevant to workplace settings beyond issues related to workplace romances. Despite the obstacles faced in having attachment theory accepted by organizational researchers, a number of studies linking attachment styles to workplace outcomes have been published on a variety of topics. Moreover, these studies are being conducted with ever increasing frequency which seems to reflect the increased interest in attachment theory among personality researchers in general. Most of these studies have been conducted on topics such as leader–follower dynamics, where the parallels to parent– child relationships is fairly obvious, or on other issues closely linked with the experiences of attachment and loss such as job satisfaction and trust. Leadership Emergence and Effectiveness Ever since Freud (1939, 1961) first spoke about leaders as proxy father-figures, researchers have speculated as to the relationship between parent–child relationships and those between leaders and their followers. Since then, several theoretical reviews have attempted to link attachment styles and childhood experiences with leader perceptions (Kahn & Kram, 1994; Keller, 2003) and leader performance outcomes (Avolio, 1994; Breshanan & Mitroff, 2007; Keller, 2003). The logic of the attachment system as an antecedent of leadership outcomes is based on the idea that attachment relationships are formed with individuals that one is close to, who can provide a safe haven in times of stress, and who can be relied on to encourage and support exploration and new experiences (Fraley & Shaver, 2000). Beyond one’s own parents, other individuals can assume these roles (e.g. close friends, coaches, and romantic partners; Ainsworth, 1991) and that the functions of attachment figures can even be split between several individuals (Trinke & Bartholomew, 1997). The transference of attachment relationships to non-parents is even more likely to occur when one is no longer able to rely on their original attachment figures or have de-idealized them (Mayseless & Popper, 2007), particularly in situations where stress is felt acutely (Kahn & Kram, 1994; Mayseless, 2005). Because of the implicit conceptions that many individuals have of leaders (e.g. they ought to provide support and encourage autonomy), it can be anticipated that followers will have a tendency towards establishing attachment relationships with their leaders (Keller, 2003). Moreover, that having the attachment system triggered by leaders will be pronounced in stressful organizational settings (Kahn & Kram, 1994). Beyond the tendency to see leaders as potential attachment figures, it is now well-accepted in the field of leadership that relationships are the foundation of leadership (Uhl-Bien, 2006). As a primary causal agent in the perceptions of close relationships and their success over time well beyond other personality constructs (Noftle & Shaver, 2006) attachment styles are an obvious candidate for explaining the interpersonal dynamics of leaders and followers in the workplace. The first empirical research linking parent attachment with leadership was actually conducted by a contemporary of Bowlby’s who argued that a pattern of insecure attachment relationships was associated with the failure to develop the independence necessary to be a good leader. Using a sample of 50 supervisors and employees, Tarnopol (1958) found that individuals with distant attachments to their fathers and overly close relationships with their mothers were less likely to be nominated by peers as being “natural leaders.” Tarnopol’s argument that the lack of self-reliance and independence in insecurely attached individuals is likely to cause career derailment has been echoed by others in more recent reviews (Quick, Nelson, & Quick, 1987). While Tarnopol failed to report the actual statistical relationships between attachment and leader emergence, Berson, Dan, and Yammarino (2006) recently published findings showing that securely attached team members were more likely to emerge as leaders in experimental groups. These results were partially replicated in sample of Israeli military recruits where individuals with secure or avoidant attachment were more likely to be nominated as a leader by their peers than individuals with anxious/ambivalent orientations (Mikulincer & Florian, 1995). With regard to leadership styles and behaviors, attachment has been shown to be linked to a variety of constructs. For example, secure attachment has been associated with a relational (as opposed to task) leadership style using Fiedler’s Least Preferred Coworker scale (Doverspike, Hollis, Justice, & Polomsky, 1997). In that same study, avoidant attachment was associated with a tendency towards task-oriented leadership. In line with the theory that secure attachment relationships foster exploration and support, Johnston (2000) demonstrated that securely attached leaders were more likely to delegate while avoidant leaders reported the least amount of delegation. In a study of Israeli officers (Davidovitz, Mikulincer, Shaver, Izsak, & Popper, 2007), leaders with an anxious orientation were described by their followers as having lower task efficacy while of.cers with avoidant orientations were described as having lower emotional efficacy. Moreover, the units of of.cers with avoidant attachment styles were reported as being less cohesive. Interestingly, both insecure styles of attachment in leaders were associated with followers reporting that their own performance was poorer. Over time, the followers of leadAdult Attachment Styles in the Workplace 289 ers with higher levels of avoidant attachment tended to show decreases in mental health as well. In a related study, Nelson and Quick (1991) found that the presence of a supervisor as a source of social support for newcomers in organizations was a significant determinant of psychological distress symptoms. More recently, Ronen and Mikulincer (2010) have demonstrated that both leader and follower attachment insecurity contribute to follower burnout and job satisfaction. Surprisingly, although there has been a good deal written about the presumed relationship between charismatic/transformational leadership (House & Shamir, 1993) and attachment styles (e.g. Popper & Mayseless, 2003; Troth & Miller, 2000), there has been very little research. There are a number of reasons to suspect that attachment styles should be linked with effective leadership styles such as transformational leadership. It has been argued that self-confidence and empathic ability, both of which as features of secure attachment orientation, are essential to visionary leadership (Goleman, Boyatzis, & McKee, 2002). It has also been argued that individuals who have greater capacity for emotion regulation will be more likely to promote positive emotions in their work group, be more encouraging of the creative efforts of their followers, and be more likely to put the needs of others before their own (Sosik & Megarian, 1999). Others have argued that the features of transformational leadership emphasize mutual reliance, shared responsibility, transparent communication, and trust (Troth & Miller, 2000). Specifically, the Individualized Consideration dimensions of transformational leadership may be closely linked with leaders who have a secure attachment. It is also likely that dysfunctional, passive leadership styles such as Management-by-Exception Passive and Laissez-Faire leadership may be consequences of leaders having an avoidant attachment style. In a series of three studies using Israeli military and police samples, Popper and colleagues showed that while secure attachment was associated with higher scores across transformational leadership dimensions (Popper, Mayseless, & Castlenovo, 2000). There was also a general tendency for avoidant/ dismissing attachment to be negatively associated transformational leadership. In a related study on empowering leadership, Towler (2005) found that individuals whose relationships with parents could generally be described as secure were more likely to describe their leadership style as charismatic. Another study of attachment styles and transformational leadership in an Australia sample found that the followers of securely attached leaders described their leaders as being more effective than the followers of insecurely attached leaders. Those followers also reported higher levels of job satisfaction (de Sanctis & Karantzas, 2008). Trust Trust in the workplace, in both leaders and coworkers, is almost by de.nition an outcome of attachment styles. Indeed, it could be argued that the driving force of attachment orientation is the perception that others are worthy of trust and the ability/willingness to make oneself vulnerable in interpersonal relationship. Curiously, no major review of antecedents of trust has made mention of attachment styles as a relevant antecedent (Colquitt, Scott, & LePine, 2007; Dirks & Ferrin, 2002). However, in terms of individual differences that have been mentioned as being closely linked with trust, attachment orientations would be expected to be closely linked with propensity to trust. Further, in the same way that a general propensity to trust is eventually replaced with person-specific trust relationships built of a history of experiences (Burke, Sims, Lazzara, & Salas, 2007; Mayer, Davis, & Schoorman, 1995; Whitener, Brodt, Korsgaard, & Werner, 1998), attachment relationships are thought to begin with a generalized attachment orientation based on a history of prior attachment relationships, but then change to reflect the experiences and expectations specific to new relationships. Interestingly, the parallels between trust and attachment extend even to their expected outcomes. Individuals with secure attachment relationships are thought to be more willing to engage in exploratory and risk-taking behavior because they perceive their partner as being a secure base (Fraley & Shaver, 2008). Similarly, individuals in high trust relationships have been shown to be more likely to take risks due to their positive expectations and willingness to be vulnerable (Colquitt et al., 2007). In terms of empirical research, both avoidance and anxious attachment have been linked with lower levels of trust and subsequent caregiving behaviors (Feeney & Collins, 2001). Mikulincer and Nachshon (1991) have demonstrated that secure individuals were more willing to open themselves up and disclose information to others. Likewise, Adams (2004) found significant positive relationships between secure attachment and trust in supervisors, peers, and upper management. More recently, Cranshaw and Game (2010) found that having an avoidant attachment relationship with one’s supervisor was associated with lower levels of trust and, in turn, career satisfaction. Interestingly, in a study of the reasons for developing trust with others (Mikulincer, 1998), securely attached individuals tended to report their goal tended to report that their goal was to gain intimacy. By contrast, insecure individuals reported that gaining a sense of security was of more importance to them. Further, in response to trust violations, secure individuals reported attempting to communicate with partners to resolve the problem, avoidant individuals reported distancing themselves from those relationships, and anxious individuals reported increases in rumination and worry. These reactions to trust violations seem to suggest that attachment styles may represent an important antecedent of con.ict resolution patterns in the workplace. Job Attitudes, Stress, Health, Coping, and Work-Family Balance In the same way that secure attachment enables individuals to regulate emotions in romantic relationships, it is anticipated that attachment styles will influence emotional reactions to other people and to stressful workplace situations (Hazan & Shaver, 1990). The strong link between neuroticism and anxious attachment (Fraley & Shaver, 2008) should provide particular insight into emotional reactions in the workplace. It is anticipated that while secure individuals will generally form secure, supportive, and happy relationships with coworkers, anxiously attached individuals will be more prone to worrying about their relationships in the workplace and will generally report less job satisfaction along with higher stress and burnout. Avoidantly attached individuals may be less prone than anxious individuals to report job dissatisfaction, but it is anticipated that their inability to disengage from negative emotions (Gillath et al., 2005) may have consequences for burnout in the workplace and work–family balance issues. In a broad survey of the workplace, Hazan and Shaver (1990) found that securely attached individuals reported significantly higher satisfaction with most aspects of their workplace (e.g. coworkers, job security, recognition, etc.). Secure individuals were also less likely 290 Harms in Human Resource Management Review (2011) 21 to report hostile outbursts in the workplace, were less prone to psychosomatic illnesses, and less prone to experiencing actual physical illnesses (Hazan & Shaver, 1990). In a second study by the same authors, insecurely attached workers reported greater anxiety over rejection by others if their work was of poor quality. Anxious individuals in particular felt unappreciated and misunderstood in the workplace (Hazan & Shaver, 1990). Similarly, in a sample consisting mostly of computer software workers, securely attached individuals reported higher levels of work satisfaction and various aspects of their jobs (Krauz, Bizman, & Braslavsky, 2001). Likewise, in a large sample of university employees, securely attached individuals reported higher levels of job satisfaction while anxiously attached individuals reported significantly lower levels of job satisfaction (Sumer & Knight, 2001). In an experimental study of reactions to a supervisor being dismissive or distant, insecurely attached individuals reported that they would be more prone to experiencing anger or distress emotions (Game, 2008). Hardy and Barkman (1994) found that individuals higher on anxious and avoidant attachment were less likely to report being satisfied with various aspects of their jobs including relationships with coworkers, levels of autonomy, and whether or not they felt they were getting the attention they deserved. In a study of nurses, secure attachment was significantly related to both hope and higher levels of self-reported health (Simmons, Nelson, & Quick, 2003). Nurses with insecure attachment styles reported less hope and those with avoidant attachment reported being less healthy. Joplin, Nelson, and Quick (1999) found similar results in a sample of students who worked full time. Individuals with higher levels of avoidant attachment reported experiencing psychological problems in addition to insomnia and social dysfunction. Individuals higher on anxious attachment reported poorer physical health along with somatic symptoms, insomnia, and social dysfunction. Securely attached individuals were less likely to report social dysfunctions, but did not report significantly less psychological and physical problems. In terms of burnout in the workplace, Ronen and Mikulincer (2009) found strong relationships with insecure attachment in a large sample of working adults in Israel. Interestingly, the effects of attachment on feelings of burnout were largely mediated by team cohesion and perceived organizational fairness. There is also some evidence that attachment styles are linked with coping styles such as seeking help from others, particularly when dif.culties are more pronounced (Lopez, Melendez, Sauer, Berger, & Wyssmann, 1998). For example, in a study of military recruits, individuals with avoidant attachment reported significantly less support-seeking behaviors and more attempts at distancing (Mikulincer & Florian, 1995). Moreover, anxious/ambivalent individuals were less prone to use using emotion-based coping strategies when faced with highly stressful situations. Interestingly, securely attached individuals were less likely to report stressful situations as a threat to themselves and were significantly more likely to describe them as being opportunities for growth and challenge (Mikulincer & Florian, 1995). Similarly, Richards and Schat (2007) found that secure individuals were more likely to engage in support-seeking behaviors while those with avoidant attachment were significantly less likely to seek support when facing a problem. Hazan and Shaver (1990) also found that securely attached individuals were less likely to report that work was interfering with their home life. Moreover, insecurely attached individuals were more likely to report that their work (as opposed to their home life) was more important to them in terms of their overall happiness. Similarly, Sumer and Knight (2001) reported that securely attached individuals tended to report positive spillover effects between work and home while individuals with insecure attachment orientations were significantly more likely to report negative spillover between work and home life. Avoidant individuals were also significantly more likely to report attempts at segmentation of the two domains. Job Performance Although job performance may not immediately seem like an outcome closely associated with attachment styles, it is one of the most studied aspects of organizational behavior in the attachment literature. Of particular interest to most researchers are the aspects of job performance that are more interpersonal in nature, namely discretionary work behaviors such organizational citizenship behaviors (OCBs) and counter-productive work behaviors (CWBs) (Dalal, 2005). This is not only because of the interpersonal nature of attachment orientation itself, but also because discretionary work behaviors are more likely when individuals have formed (or failed to form) a close bond with their leaders, coworkers, and organizations. In terms of discretionary helping behaviors, Geller and Bamberger (2009) found that both avoidant and anxious attachment were associated with less instrumental helping behaviors in the workplace. Erez found that individuals higher on avoidant attachment are particularly unlikely to report engaging in volunteer activities (Erez, Mikulincer, van Ijzendoorn, & Kroonenberg, 2008). In terms of OCBs, significant positive relationships have been found with secure attachment in a number of studies (e.g. Frazier et al., 2009; Little, Nelson, Wallace, & Johnson, 2010). Similarly, high scores on anxious and avoidant attachment have also been associated with reduced OCBs in the workplace (Desivilya, Sabag, & Ashton, 2006; Little et al., 2010). Likewise, Rom and Mikulincer (2003) reported that individuals with higher levels of anxious and avoidant attachment report putting less effort into team tasks. In addition, individuals with higher levels of avoidant attachment are less likely to help other group members or facilitate team cohesion. In terms of CWBs, Richards and Schat (2007) found that individuals with avoidant attachment were more likely to report engaging in CWBs, but anxious and secure individuals were not. These results were largely replicated by Little et al. (2010) who found a positive relationship between avoidant attachment and CWBs, but who also found a negative relationship with secure attachment. Other studies have failed to find relationships between attachment and performance. For instance, in a study of employees at an assisted living center, no relationship was found between secure attachment overall performance (Simmons, Gooty, Nelson, & Little, 2009). Likewise, using student grades as proxy for performance, Joplin et al. (1999) failed to find significant effects for any of the three major attachment styles. In a more recent study, Ronen and Zuroff (2010) failed to find significant relationships between either avoidant or anxious attachment and supervisor-rated performance. Using a group-level performance outcome, Daus and Joplin (1999) failed to find support for either leader or follower attachment having an impact on performance. However, it should be noted that in each of these cases, the performance outcome was not interpersonal in nature and would not have been hypothesized to be closely related to attachment orientation. Adult Attachment Styles in the Workplace 291