CAHRS / Cornell University 187 Ives Hall Ithaca, NY 14853-3901 USA Tel. 607 255-9358 www.ilr.cornell.edu/CAHRS/ W O R K I N G P A P E R S E R I E S Organizational Change and the Identity Cycle: Understanding the Effect of Change Theresa M. Welbourne Daniel M. Cable Working Paper 9 4 - 0 1 Advancing the World of Work Organizational Change and the Identity Cycle WP 94-01 Organizational Change and the Identity Cycle: Understanding the Effect of Change on Individual Attitudes and Behaviors Through a Combined Social Identity Theory/ Identity Theory Perspective Daniel M. Cable and Theresa M. Welbourne Center for Advanced Human Resource Studies School of Industrial and Labor Relations Cornell University Ithaca, NY 14853-3901 http://www.ilr.cornell.edu/CAHRS Working Paper #94-01 This paper has not undergone formal review or approval of the faculty of the ILR School. It is intended to make results of Center research, conferences, and projects available to others interested in human resource management in preliminary form to encourage discussion and suggestions. Page 1 Organizational Change and the Identity Cycle WP 94-01 Abstract The study of roles and role behavior is particularly relevant today as individuals acquire more roles in the complexity of the 1990s. One environment that has been significantly prone to change is the workplace, where multiple committees, teams, and departments have transformed the nature of work and are altering the way that jobs are defined. In addition to the fact that workers are now taking on multiple roles within organizations, the roles themselves are changing at an accelerated pace. Reengineering, downsizing, mergers, acquisitions, and total quality initiatives are just a few of the interventions that businesses are implementing in order to become leaner, flatter, and more responsive to their environments. This paper draws from social identity theory and identity theory to understand how employees respond to organizational change. Given the fairly low success rates of major change efforts (estimates are between 10% and 50%), it is suggested that a more thorough understanding of the effect of these programs on an individual's role within the organization is necessary. This paper begins by first defining, comparing, and contrasting social identity theory and identity theory. This has, to date, not been done, and it is particularly important because a number of authors appear to be using the two theories interchangeably. Next, social identity theory and identity theory are used to build a broader framework for understanding human behavior, and this model, called the identity cycle, is used to develop a set of propositions regarding the effects of organizational change on employee attitudes and behavior. Page 2 Organizational Change and the Identity Cycle WP 94-01 Examples of successful reengineering or large-scale organization change efforts are rare, with estimates of success ranging between 10% to 50% of attempted projects; at the same time, surveys show that approximately 70% of large, U.S. organizations claim to be reengineering their corporations (Cafasso, 1993; Hall, et. al., 1993). Given the high failure rates, the persistent need for change, and the fact that there is no consensus as to why so many change efforts fail, it seems that a theoretical model focusing on the effect of organizational change on individual behavior can make a contribution in understanding this complex process. Our premise, shared by a number of authors (e.g. Drucker, 1992; Danter, 1983; Mitroff, 1987) is that organizational change, combined with massive societal change, has resulted in not only new roles but also an increased number of roles that each person must internalize and act upon (Drucker, 1992). For example, the working mother often takes on the roles of parent, spouse, and employee. However, the employee role now is more complex, possibly involving roles such as accountant, quality team member, selection committee member, accounting department member, accounts receivable department member, new product project team, and the role associated with identification with the overall organization. In addition, the parent role is often just as complex, possibly encompassing the roles of mother, step-mother, primary care giver for aging parents, religious educator, and school volunteer (to name only a few). Then interjected into this scenario, many organizations are reengineering or transforming themselves, which means that all of those organizational roles (or at least a significant number of them) are prone to change. This change will more than likely also alter the various roles established outside of the organization. However, little attention has been paid to the effect of these organizational changes on the multiple roles that compose an individual employee's identity. This paper proposes that a deeper understanding of the effect of organizational change efforts on identity structure (defined by the number and importance of these various roles) can aid in predicting the effect of these changes on employee attitudes and behavior. Two theories that consider the relationship between roles and behavior are social identity theory and identity theory. Social identity theory concentrates on individual roles within a group context, and it has been applied to a variety of research contexts, including community development (Wells, 1990), power and status (Sachdev & Bourhis, 1991), and ethnicity (Garza, Lipton, & Isonio, 1989). Identity theory seeks to explain attitudes, behaviors, and reactions to events as a function of individual identities. Identity theory focuses on the individual as the unit of analysis and has been used in research on stress and psychological well-being (Burke, 1991; Simon, 1992; Thoits, 1992; Thoits, 1991; Wiley, 1991). Page 3 Organizational Change and the Identity Cycle WP 94-01 In several ways, social identity theory and identity theory are conceptually related. In fact, the theories have enough overlap that they have inadvertently been combined by researchers. In an effort to understand the unique contributions of each theory, this paper will first define and contrast social identity theory and identity theory. After reviewing the two theories and selected research, a model that merges social identity theory and identity theory, called the identity cycle, will be introduced. A set of propositions derived from this model that focus on the way in which employees respond to organizational change will then be introduced. Identity Theory Defined Identity theory proposes that individual identities can be used to predict performance aril that identities are structured by differential commitment to roles (Stryker, 1987). The theory assumes that the self is composed of internalized roles arranged in a hierarchy of salience; these roles are "ordered" by the probability that a particular identity or role will come into play within and across situations (Stryker, 1987). Thus, the more likely that one will refer to a role in a situation, the higher or more dominant that role will be on the identity hierarchy. For example, an employee's reaction to a child's sickness might be predicted through identity theory. If the employee has a hierarchy with a dominant parenting role, the theory suggests that this employee might stay home with the child. On the other hand, if the hierarchy reflects a dominant managerial role, the theory predicts the employee should seek alternative arrangements for the child so the employee can work. Identity theory, therefore, can be used to make specific predictions about individual reactions to events (Stryker, 1968). The structural interactionist perspective, which is at the root of identity theory, suggests that social interaction is critical for self development and that the self is a product of these interactions (Stryker, 1987; Stryker and Statham, 1985; Thoits, 1992). Identity theory proposes that individuals come to understand themselves and their environment through knowledge of the roles that they and others assume within society. Identity theory is graphically depicted in Figure 1. Essentially, the model indicates that individual behaviors are based on identity hierarchies, that a given event can contain different meanings as a result of these identities, and that events can be expected to induce a range of reactions from different individuals because each person has a unique hierarchy. Page 4 Organizational Change and the Identity Cycle WP 94-01 As Figure 1 indicates, individuals have a set of roles that defines the self, but they are not equally committed to each role. Based on the importance of and the commitment to each role, a hierarchical identity structure results. The identity structure acts as a filter that enables individuals to interpret their environment and respond to the events in it (Burke, 1991). Both emotional and behavioral reactions to events are a function of these identity structures. Thus, greater commitment to a specific role suggests that actions and attitudes consistent with that role will be invoked in a given situation. More role salience leads to more meaning, purpose, and behavioral guidance from that role (Thoits, 1991). Page 5 Organizational Change and the Identity Cycle WP 94-01 Social Identity Theory Defined Social identity is defined as that part of an individual's self concept that derives from knowledge of membership in a social group together with the value and emotional significance attached to that membership (Ashforth & Mael, 1989; Tajfel, 1978). Although social identity theory may have roots in group identification concepts (e.g., Tolman, 1943, Ashforth & Mael, 1989), the theory evolved primarily from social comparison theory (Festinger, 1954; Turner, 1975) and from empirical research on intergroup discrimination studies (e.g., Tajfel, 1970). Social comparison theory proposes that humans are driven to evaluate their attributes and abilities and that they also have a need for positive self-evaluation (Turner, 1975). Because no objective standard of comparison exists, individuals evaluate themselves relative to others, often based on comparative dimensions in which they excel (Turner, 1981). Social identity theory provides a theoretical explanation for inter-group discrimination and the repeated research findings that individuals favor their in-group (Brewer, 1979; Tajfel, 1970). The model is a response to a number of experiments conducted by Tajfel and his colleagues demonstrating that individuals attempt to maximize the differences between their group and other groups, even at the cost of in-group rewards (Tajfel, 1970; Tajfel & Billig, 1974; Tajfel, Billig, Bundy, & Flament, 1971). Figure 2 graphically depicts social identity theory. According to the tenets of social identity theory, individuals have a need to order their social environment (Abrams et al., 1990; Tajfel, 1970; Turner, 1975). As a result, individuals place themselves and others into social categories that define group membership. However, a group only can exist if it is defined through some distinctive characteristic relative to other groups and can only maintain an identity if it maintains meaningful differences from other groups. This need for inter-group differentiation ultimately leads to in-group and out-group biases (Turner, 1981). Furthermore, groups tend to highlight those differences that reflect themselves in a positive light (Turner, 1975). Page 6 Organizational Change and the Identity Cycle WP 94-01 While positive in-group biases affect subsequent inter-group perceptions and interactions, these relationships appear to be moderated by several factors. For instance, researchers have investigated the role of group identification (Brown & Williams, 1984), amount of contact between the groups (Oaker & Brown, 1986) and similarity of the groups (Turner, 1978). In summary, social identity theory proposes that groups have a need to build positive identities by heightening positive distinctiveness through inter-group comparison, and this process leads to in-group biases and inter-group differentiation. Social Identity Theory versus Identity Theory In addition to sharing the word "identity" in their tittles, social identity theory and identity theory have several key theoretical tenets in common. In fact, the theories are similar enough that some researchers inadvertently have combined social identity theory and identity theory in Page 7 Organizational Change and the Identity Cycle WP 94-01 their work. However, the similarities and differences between the two theories have yet to be succinctly defined. Therefore, this section of the paper compares and contrasts social identity theory and identity theory, highlighting the similarities and differences that exist in addition to providing examples of representative research. Social identity theory and identity theory share at least two similarities. First, both theories claim that an understanding of roles is critical for predicting human behavior. Second, both social identity theory and identity theory are based on socially-defined self-concepts and realities. In this imagery, society and social roles do not "exist" but are continuously created and recreated. Accordingly, identity theorists posit that we come to know who we are through interactions with others and society and that our behavior is based on these socially defined roles (Stryker, 1968; Stryker and Serpe, 1982; Stryker & Statham, 1985). Social identity theory is based largely on Tajfel's (1970) contention that individuals construct "webs of social affiliation" by applying principles of order and simplification in an effort to reduce the complexities of society (p. 98). For example, Tajfel suggests that one of the most important subjective interpretations of our social environment is the classification of groups as "we" and "they" -- as in-groups and out-groups. Tajfel, Billig, and Bundy (1971) have pointed out that there can be no inter-group behavior before the social environment has been categorized and individuals have been divided into defined groups. Furthermore, "no group lives alone"; groups can be defined only in relation to other groups. Due perhaps to the similarities inherent between social identity theory and identity theory, recent interpretations and applications of social identity theory and identity theory suggest that the theories are being inadvertently integrated. Because this integration is occurring without clearly defining the differences between the theories and their associated research streams, there may be several unfortunate results. For instance, researchers focusing on one theory may not be aware of research conducted with the other, and work may be unknowingly replicated. Also, without first clearly communicating the fundamental differences between the theories, social identity theory and identity theory may be misspecified. One research literature in which social identity theory and identity theory are being integrated is organizational behavior. Perhaps because of its relevance to employee socialization, role conflict, and inter-group relations concepts of self-identity have recently received considerable attention in organizational research (e.g. Abrams, Jackson, & St. Claire, 1990; Ashforth & Humphrey, 1993; Ashforth & Mael, 1989; Lobel, 1991; Lobel & St. Clair, 1992). Although each of the above authors claim to apply "social identity theory" in their research, a gradual merging of social identity theory and identity theory is apparent in their work. Page 8 Organizational Change and the Identity Cycle WP 94-01 Lobel (1991), for instance, applied a "social identity theory" perspective to time invested in work versus family roles. However, social identity theory was integrated with identity theory in several ways. Although social identity theory traditionally has focused on explaining inter-group behavior, Lobel's discussion and propositions concerned the actions and emotions in an individual, rather than a group, context. Furthermore, the research propositions "developed from social identity theory" were almost identical to hypotheses developed by Stryker (1969) and tested by Stryker and Serpe (1982) utilizing identity theory. Specifically, Lobel (1991, p. 513) proposed that "an individual's investment in a role will increase as his or her identification with the role increases" and "the relative salience of career and family identities will determine an individual's relative investment in career and family." However, Stryker and Serpe (1982, p. 212), have already supported the hypotheses that "the higher the commitment to a role, the higher the time spent in role" and "the higher the identity salience, the higher the time spent in that role." Furthermore, although Lobel cites identity theory researchers (e.g., Stryker, 1968; Stryker & Serpe, 1982) no attempt to made to isolate the two theories, or even define identity theory as a separate model of behavior. This is unfortunate, because identity theory is more relevant than social identity theory to several of the propositions suggested. Ashforth and Humphrey (1993) also cited the research of identity theorists (e.g., Burke, 1991; Thoits, 1991) and developed research questions consistent with identity theory (e.g., modeling the psychological well-being of individuals). However, the authors clearly stated that they utilized social identity theory. Like Lobel (1991) the authors appear to have merged social identity theory and identity theory. Lobel and St. Clair (1992, p. 1058) provide a third example of the gradual convergence of the two theories. The research proposed that "social identity theory offers an approach to explaining differences in work effort and performance outcomes." However, the authors then describe identity theory: "According to that theory, people classify themselves into multiple hierarchically organized social categories." Furthermore, the authors claim that "identity salience is a key concept of social identity theory" when identity salience represents the core of identity theory and is not a major component of social identity theory. Finally, Lobel and St. Clair also make propositions that model individual behaviors (e.g., work effort) for which identity theory is more appropriate. Thus, recently ft appears that some researchers are recognizing the inherent similarities between social identity theory and identity theory. Although identity theory and social identity theory share some common elements, they can also be contrasted on several dimensions. Table 1 represents several attributes on which the two models can be compared. Page 9 Organizational Change and the Identity Cycle WP 94-01 Table 1 Comparing Social Identity Theory and Identity Theory Identity Theory Social Identity Theory Attribute Disciplines Social Psychology Sociology Roots Derives from Symbolic Interactionism: Derives from Social Comparison Theories: Humans must evaluate Social interaction defines self and behavior themselves and in the absence of an absolute standard, we seek others to define ourselves. We also have a need for positive self-evaluation, especially of important roles. Founders Sheldon Stryker Henri Tajfel, Michael Billig, and John Turner Key Idea Commitment affects identity hierarchies which are used Groups and individuals build positive self identity through to predict emotions, role performance, and behaviors positive group distinctiveness and intergroup comparisons Goals The goal is to explain and predict individual behavior The goal is to explain group relationships and conflict Level of Individual-oriented Group-oriented Analysis (eg, deals with how a person will behave) (eg, deals with how a groups interact) Conception of Encompasses both established societal roles and types Deals with established roles that are publicly recognized Roles of people it is possible to be (e.g., helping role) social units (e.g., assigned group membership, employee, wife) Processes More 'information processing oriented' More emphasis on group characteristics and perceptions (eg, how information is interpreted depends and less on an 'event'. (eg, ingroup biases are a function of on identity structures) the positive distinctiveness of the group) Dependent Primarily individual-based: Time spent in role, Primarily group-based: Group differentiation, commitment to group Variables psychological distress, substance abuse intergroup attitudes, estimated contributions of groups Independent Individual-based: Personal identities, identity rankings, Group-based: Group identification, exposure to other groups, Variables number of roles held, identity salience ingroup/outgroup status, descriptions of groups Page 10 Organizational Change and the Identity Cycle WP 94-01 Roles and self-concerts As shown in Figure 3, self-concepts are thought to be composed of two elements: a personal identity encompassing idiosyncratic characteristics (e.g., interests, traits) and a social identity encompassing salient group classifications (Ashforth & Humphrey, 1993; Ashforth & Mael, 1989; Tajfel & Turner, 1985). Most identity theorists conceptualize roles as free from (but inclusive of) highly organized social units (Stryker & Stratham, 1985). According to identity theory, the positions or roles that compose a person's identity do not necessarily require group affiliation, and it is possible for an individual's identity to be based on things such as a bad marriage (e.g., divorcee role), a desire to learn more about the world (e.g., intellectual role), or a desire to solve others' problems (e.g., helping role). As Stryker and Statham (1985, p. 354) suggested, identity theory conceptualizes roles relatively broadly "as the kinds of people it is possible to be in society (thus freeing the term from highly organized, publicly recognized social units)". Page 11 Organizational Change and the Identity Cycle WP 94-01 Social identity theory, on the other hand, is based on the assumption that inter-group perceptions and conflict arise because groups attempt to maintain distinctiveness from other groups. A prerequisite component of social identity theory, then, is a socially defined and recognized group that members can seek to protect. Social identity theory does not focus on personal, non-group identities (e.g., an "intellectual" role), but concentrates on those roles that are derived from socially established groups such as "departmental committee member" or "republican." Goals and Levels of Analysis Perhaps one of the largest differences between social identity theory and identity theory is the type of behavior they seek to model and the measures they use to predict behavior. Table 2 depicts several important characteristics of published articles that serve to illuminate differences between social identity theory and identity theory and demonstrates how researchers actually have applied identity theory and social identity theory to date. Page 12 Organizational Change and the Identity Cycle WP 94-01 Table 2 - Examples of Research Employing Social Identity Theory and Identity Theory Theory Author Date Title Book / Journal DV's IV's Type of Study SIT Tajfel, H. (1970) Experiments in intergroup Scientific American In-group favoritism, Group Lab discrimination out-group categorization, differentiation meaning of categorizations SIT Tajfel, H., Billig,(1971) Social Categorization and European Journal of Choices of reward, Neutral or Lab M.G., Bundy, R.P. intergroup Behavior Social Psychology allocation to value-laden & Flament, C. in-group and group divisions out-group SIT Tajfel, H., Billig,(1974) Familiarity and categoriza- European Journal of Choices of reward, Familiarity with lab Lab M.G. tion in intergroup behavior Social Psychology allocation to procedure, neutral or in-group and value- laden group out-group divisions SIT Turner, J. (1978) Social comparison, similarity Differentiation Between Merit ratings, Importance of Lab and ingroup favoritism Social Groups performance ratings, comparison criteria, Lab In H. Tajfel (Ed.). stability of status differences, London: membership preferences similarity of outgroup Academic Press Identity Santee, R.T. & (1979) Commitment to self- Human Relations Commitment / goals, Identity rankings Field Theory Jackson, S.E. identification: A socio- expectations, psychological approach to intensity personality Identity Burke, P.J. & (1981) The link between identity Social Psychology Educational plans, Measured meanings Field Theory Reitzes, D.C. and role performance Quarterly participation in of different roles student social activities Identity Stryker, S. & (1982) Commitment, identity Personality, Roles, Time spent in Religious Field Theory Serpe, R.T. salience and role behavior: and Social Behavior religious role, commitment, Theory and research salience of religious salience of religious example role, religious identity, religious satisfaction satisfaction Page 13 Organizational Change and the Identity Cycle WP 94-01 Theory Author Date Title Book / Journal DV's IV's Type of Study Identity Hoelter, J.W. (1983) The effects of role Social Psychology Role salience role evaluation, role Field Theory recreational commitment Quarterly commitment on identity salience SIT Brown, R. & (1984) Group Identification: The Human Relations Intergroup Group Identification, Field Williams, J. same thing to all people differentiation, perceived conflict, estimation of group perceived stability, contribution in-group status discrepancy SIT Oaker, G & (1986) Intergroup relations in a Human Relations Group differentiation Amount of contact, Field Brown, R hospital setting: A further computed from in-group / out-group test of social identity theory general question- status, responses to naire intergroup identification scale, attitudes interview responses SIT Brown, R., (1986) Explaining intergroup Journal of Occupa- Estimation of the Group identification, Field Condor, S, differentiation in an tional Psychology contribution of other exposure to other Mathews, A., industrial organization groups groups Wade, G. & Williams, J. Identity Leary, M.R., (1986) Aspects of identity and Social Psychology Desire for personally Personal identity, Field Theory Wheeler, D.S. & behavioral preference: Quarterly or socially rewarding social identity, Jenkins, T.B. Studies of occupational and job recreational choice Identity White, C.L. & (1987) Ethnic role identity among Sociological Ethnic identity, self Ethnic identity Field Theory Burke, P.J. black and white college evaluation salience, self-esteem, students role commitment, self evaluation, SES Identity Serpe, R.T. (1987) Stability and change in Social Psychology Stability of Commitment Field Theory self: A structural symbolic Quarterly commitment and (affective and interactionist perspective salience interactional), identity salience Page 14 Organizational Change and the Identity Cycle WP 94-01 Theory Author Date Title Book / Journal DV's IV's Type of Study Identity Kuchera, M.E. (1988) The effects of perceptions Sociology of Education Time spent in Identity salience Field Theory & Miller, S.I. of the academic job market identity role, role commitment, job on adjunct faculty: An satisfaction market perceptions identity-theory analysis. Identity Charno, H. W., (1988) Role identity and reasoned Social Psychology Intentions and Role identity, social Field Theory Pilliavin, J.A., & identity-in the prediction of Quarterly behaviors of blood relations connected Callero, P.L. repeated behavior donation to donation, habit, behavioral intention, attitude, subjective norms SIT Abrams, D., (1990) Social identity and the Human Relations Labeling of Social distance to Field Jackson, D. & St. handicapping functions of handicapped persons handicapped persons, Claire, L. stereotypes: Childrens descriptions of others understanding of mental and with different physical handicap handicaps Identity Burke, P.J. & (1991) An identity theory approach Social Psychology Student identity, time Commitment to role Lab Theory Reitzes to commitment Quarterly in role, GPA, adjustment to role Identity Thoits, P.A. (1992) Identity structures and Social Psychology Psychological Roles held, identity Field Theory psychological well-being: Quarterly distress, substance claims, Gender and Marital status use identity salience, comparisons stress in role identity Identity Simon, R.W. (1992) Parental role strains, Journal of Health and Psychological Parental role Field Theory salience of parental identity Social Behavior distress strains, parental and gender differences in identity salience, psychological distress Page 15 Organizational Change and the Identity Cycle WP 94-01 As verified by the dependent variables that researchers have studied, identity theory is explicitly concerned with the choices made by individuals in situations where alternative courses of action exist (Stryker & Serpe, 1982). By knowing an individual's salience hierarchy, identity theorists have sought to explain and predict role-related attitudes and behaviors. Examining Table 2, it is apparent that both the predictor and outcome variables studied by identity theorists have been individual-based. For instance, Stryker and Serpe (1982) explained the time invested in religious duties by priests with measures of religious commitment and the salience of the religious identity. The researchers also predicted religious satisfaction with measures of commitment. Thoits (1992) assessed the roles in individuals' lives, the salience of those roles, and the stress experienced in those roles to predict psychological distress and substance abuse. Simon (1992) also studied psychological distress using measures of role salience and strain as predictor variables. In summary, identity theorists have attempted to explain emotions, attitudes, and behaviors based on the identity hierarchy. Social identity theory, conversely, has offered a theoretical means to model such outcomes as group interaction and group conflict in a number of situations. For instance, Brown and Williams (1984) used social identity theory to help model inter-group differentiation and employee estimations of other groups' contributions. The researchers measured group identification, perceived conflict, perceived stability of group status, and the discrepancy between current and desired group status. Turner (1978) examined estimates of group performance and group membership preferences based on measures of the importance of the criteria used to compare the groups, the stability of the status differences between the groups, and the similarity of the groups. Abrams, Jackson, and St. Claire (1990) examined one group's (e.g., non-handicapped) social distance from handicapped persons to predict subsequent labeling of handicapped persons. Oaker and Brown (1986) utilized social identity theory to predict that group identification would be positively related to inter-group differentiation, moderated by the amount of group contact, and group status. In summary, social identity theory has offered a means to explain group differentiation and behavior, based in part on characteristics of the groups. Integrating Social Identity Theory and Identity Theory Social identity theory and identity theory share several basic assumptions about how individuals structure and perceive their environment, which in turn affect how they respond to that environment. As indicated in the discussion above and in Table 1, however, each theory makes a unique contribution to our understanding of behavior. Although researchers recently Page 16 Organizational Change and the Identity Cycle WP 94-01 have begun to blend the theories, a more valuable approach might be one that creates a more general model of identity, incorporating aspects of social identity theory and identity theory. It is argued that social identity theory and identity theory can be merged into a single model that encompasses self-identity development, role commitment, the identity hierarchy, and related attitudes and behaviors. A framework that integrates the theories, called the identity cycle, is depicted in Figure 4. Essentially, this framework is an attempt to create a more general model of how individuals (either alone or as part of a group) can be expected to respond to environmental stimuli of any type. The model is iterative because identity development and event interpretation are continuous processes. Page 17 Organizational Change and the Identity Cycle WP 94-01 Page 18 Organizational Change and the Identity Cycle WP 94-01 Social identity theory and identity theory both assume that society is not a static entity but is continually defined and redefined through social interaction. This assumption is an explicit tenet in social identity theory (e.g., Turner, 1981) while it is inherent in identity theory through is evolution from symbolic interactionism (e.g., Stryker, 1987). Thus, consistent with both social identity theory and identity theory, the identity cycle posits that individuals have a need to define and order their social environment before they can respond to it. By ordering their social environment, individuals place themselves and others into roles and groups. According to social identity theory, ft is this membership in certain roles and groups that defines self concepts. (Turner, 1981; Tajfel & Turner, 1979). Accordingly, Figure 4 depicts individuals utilizing the roles they take on in society to locate and define themselves within that society. Consistent with the above discussion that the types of people ft is possible to be in society encompasses not only formally recognized roles (e.g., university department member) but also personal roles (e.g., risk-taker), the identity cycle utilizes the more comprehensive definition of roles inherent in identity theory. According to the identity cycle, then, individuals define their selves based on the roles they assume within a social environment. Although an individual may hold multiple roles and be involved in many groups, each making some contribution to his or her self-identity, the roles are not equally important. As Stryker (1987) suggested, the distribution of identities in an individual's salience hierarchy reflects the varying levels of commitment to the roles underlying the identities. One is committed to a specific role to the degree that extensive and intensive social relationships are based on that role (Stryker & Statham, 1985). Thus, individuals are expected to be more committed to those identities upon which relationships to important others depend (Hoelter, 1983; Stryker, 1968). In addition to offering continued involvement in important relationships, certain roles also offer other positive feedback. For instance, roles that are associated with financial rewards, social support, or that are enacted competently are likely to be more salient in an individual's identity hierarchy than less rewarding, nonprestigious, or incompetently enacted roles (Thoits, 1991). This argument is also consistent with social identity theory, which assumes that individuals desire positive self-esteem and therefore seek positive distinctiveness for their in-groups in comparison to out-groups (Tajfel & Turner, 1979; Turner, 1981). Thus, each individual has many roles (both personal and social) arranged in a hierarchy of salience, which is based upon one's commitment to a number of various roles or identities. Although social identity theory and identity theory traditionally have been applied to different levels of analysis (e.g., individual versus group), both theories seek to explain attitudes and behavioral choice when alternatives are present. Accordingly, the goal of the recommended Page 19 Organizational Change and the Identity Cycle WP 94-01 framework is to help explain and predict attitude responses and behavioral choices. Furthermore, both social identity theory and identity theory recently have been developed to model responses to certain event characteristics. Thus, while identity theory may be used to understand attitudes and behaviors based on identity hierarchies (e.g., holding the event constant; see Stryker, 1968; Stryker & Serpe, 1982), more recently the theory has been developed to model interpretations of and reactions to dynamic events and environments (e.g., the identity structure as a filter; see Burke, 1991; Simon, 1992; Thoits, 1991; Welbourne & Cable, 1993). Similarly, although social identity theory was utilized to predict general group responses to other groups (see Tajfel, 1970; Turner, 1975), recent refinements of the theory consider both the characteristics of the groups and the environment in which the groups exist (e.g., Brown & Williams, 1984; Oaker & Brown, 1986). In an effort to encompass both the original and more recent theoretical developments of social identity theory and identity theory, the identity cycle assumes behavioral choices are a function of both the individual (e.g., salience hierarchy) and the characteristics of an event. The identity cycle allows researchers to predict responses to both daily events and significant life-changing events (such as job loss, death, etc.). It is assumed that some events, by the nature of event itself and is significance to the individual, will result in very strong emotional and behavioral responses, which in turn can cause a reevaluation of one's social categorization schema, commitment to roles, and finally an alteration in the identity hierarchy. It is the nature of the event itself and the interaction of the event characteristics with the individual's identity hierarchy that determine whether the subsequent response set will eventually alter an individual's hierarchy. To date, identity theorists have discussed the choice of behavior and the interpretation of events through the role hierarchy, assuming a static hierarchy. It appears that a more general model could also explain how new roles may be developed and how role commitment (and subsequently role hierarchy and future reactions) can change. Consider, for instance, the case of a someone whose career role is dominant and who experiences a layoff. According to the identity cycle, the initial reaction should be intense because the career role was very salient to the individual. This significant emotional or behavioral response can be expected to affect social categorization (e.g., from job holder to unemployed) and subsequent commitment to the career role (e.g., reevaluating the importance of the career role as opposed to the parent role). Future events (perhaps the same event) may be interpreted differently by the individual as a function of the significant, identity-altering event. Because the major theoretical contributions of both theories were incorporated into the model, both social identity theory and identity theory can be viewed as specific applications of Page 20 Organizational Change and the Identity Cycle WP 94-01 the more general identity cycle. However, this framework also appears to make important contributions to both social identity theory and identity theory. Concerning social identity theory, the identity cycle permits a more specific conceptualization of the relationship between groups and individuals. For example, while social identity theory proposes that individuals are motivated to seek positive comparisons between their group and out-groups, the identity cycle suggests that groups that are more amenable to positive comparison and subsequent self-esteem are expected to be more salient to their members' identities (e.g., be more central to defining their behaviors). The identity cycle also models personal roles when predicting behavior in inter-group comparison events. Thus, although researchers employing social identity theory would propose that two similar, interacting groups will tend to stereotype and differentiate, the identity cycle would make this prediction more specific by suggesting that group members' identity hierarchies might moderate the tendency to stereotype. For example, group members whose "intellectual" or "pacifist" roles are very salient might be less likely to view other groups negatively, as out-groups. Although certainly these propositions demand further research and development, they represent theoretical contributions of the identity cycle. Finally, as discussed, the identity cycle extends social identity theory to encompass not only socially defined roles (e.g., job holder) but also the more personal roles (e.g., risk-taker). Concerning identity theory, the identity cycle makes the source of roles, and individual's commitment to various roles, more explicit. While identity theory assumes that both roles and role commitment are already in place (e.g., "commitment determines identity determines role performance"; Stryker, 1987), the identity cycle encompasses the origin of individual roles (e.g., categorizing self and others in society) and considers the sources of role commitment. Although role development and commitment are implicit in the symbolic interactionist framework, identity theory does not model the means by which individuals acquire roles and become differentially committed to those roles. This contribution is related to the identity cycle's ability to account for changes that events may have on social categorizations, roles, and identity hierarchies. Finally, the identity cycle specifically acknowledges that the "event" need not be held constant. Thus, although identity theory has been applied to help explain reactions to stressful events (Burke, 1991; Hammen et al, 1985; Simon, 1990), the research to date has not specifically considered varying characteristics of the event itself, such as varying degrees of stress. To date, the event has been defined as simply existing or not existing. Page 21 Organizational Change and the Identity Cycle WP 94-01 The Identity Cycle and Organizational Change The identity cycle may be applied to behavior in any situation, however, the framework will be utilized here to understand how employees respond to organizational change efforts. Particularly given the number of changes being implemented today within organizations, the effect these changes are having on work and family roles, and the rate at which change is taking place, the identity cycle seems to provide some useful contributions for researching the change process. Although succinctly defining change is difficult (Tichy, 1983), change has been categorized as being either first order or second order change, and some researchers refer to second order change processes as transformations or reengineering efforts (Hammer & Champy, 1993; Tichy & Devanna, 1986). The difference between these types of change seems to stem around the magnitude of and pace of change. First order change has been defined by Levy and Merry (1986: 5) as "minor improvements and adjustments that do not change the system's core and occur as the system naturally grows and develops". Second order change, on the other hand, has been defined by the same authors as "a multidimensional, multi-level, qualitative, discontinuous, radical organizational change involving a paradigmatic shift". Second order change is viewed as discontinuous, "deep" structural and cultural change, while first order change is considered part of a continuous process (Levy, 1986). Traditional organizational development interventions are used to implement both first and second order change, however, they are primarily designed to implement first order or gradual change (Quinn, 1980). This type of change proceeds in small steps and attempts to secure each employee's "conversion" to the new system or process (Dunphy & Stace, 1988). Total quality management programs are often implemented in a way that could be characterized as first-order change. In fact, many successful quality initiatives have actually been initiated as grass roots efforts, where supervisors or managers at manufacturing plants decide to make the change long before the executive management team considers the idea (Beer, et. W., 1990). First order changes are generally more successful than second order changes. Resistance to change is often seen as the culprit in understanding why rapid, transformational-type changes often do not succeed (Goldstein, 1988). The identity cycle can be used to develop a set of propositions regarding the relative effect of first-order and second-order change on individual employees. First order change can be viewed as an event that would not generally result in a emotional or behavioral response strong enough to cause an individual to reevaluate an existing social categorization. If, for example, a total quality initiative is introduced, without layoffs or any other significant, Page 22 Organizational Change and the Identity Cycle WP 94-01 non-gradual changes, it would be unlikely to cause members to reevaluate their views of the social environment because there is no observable change in the work environment. This, then, would more than likely not alter role commitment or subsequently change an identity hierarchy. Put simply, small, incremental changes would not be expected to cause an emotional or behavioral response significant enough to alter an individual's social categorization schema. In general, resistance to the change effort should be minimized when first-order changes are initiated. Proposition 1: First-order change will lead to minor emotional and/or behavioral responses, which will not cause subsequent changes in role commitment or identity hierarchies. Proposition 2: First-order changes will result in minor resistance to change. However, second order change, such as a total quality initiatives combined with a 40% reduction in the workforce, a restructuring of the pay system, and changes to self-managed work teams, all implemented concurrently, might result in significant emotional and/or behavioral responses that could then trigger immediate social recategorization. Now an employee is a "survivor" rather than simply an employee. The worker's job has changed, and the nature of the management/worker relationship has probably been altered. This social recategorization should then result in changes in role commitment and eventually in the identity hierarchy. Perhaps the employee will reevaluate the importance of the career role, the role within their department, and the family role, most likely resulting in work-related roles being less dominant than before the change. Proposition 3: Second order changes will result in significant emotional and/or behavioral responses from individual employees. Proposition 4: Second order changes will result in changes in social categorization, role commitment, and the identity hierarchy. Of course, the degree to which these organizational changes actually cause emotional or behavioral responses and subsequent changes in role commitment and identity hierarchies is the result of an interaction between an individual's hierarchy before the change occurs and the characteristics of the event. Therefore, although overall, the above-noted patterns would be expected to occur, the theory indicates that differences in individual hierarchies will be important in determining the actual response of an individual worker. First order changes can result in social recategorization if the nature of the event is such that it affects an important role within an employee's hierarchy. There are circumstances where Page 23 Organizational Change and the Identity Cycle WP 94-01 even a first order, fairly insignificant change might have a dramatic effect on an employee. Within the work place a number of roles exist. For example, if an individual perceives the role of "information gatekeeper" as highly important, and the total quality management initiative results in all employees receiving instantaneous information via a computer system, then one might expect this event to result in a more severe emotional and behavioral reaction from the "gatekeeper" because his/her role is significantly affected by the change. This might be particularly significant if the employee places the work role as the most dominant in the identity hierarchy, and it can be even more significant if there is a considerable amount of "distance" between the work-related roles and other roles. Distance refers to the degree to which one role is more important than another; it is suggested that distance, as well as location in the hierarchy, should be significant in predicting emotional or behavioral responses. This same phenomenon would be predicted in the case of an individual whose job or work-related roles were very insignificant in the identity hierarchy (particularly with a large "distance" between the work and non-work roles) in the event of a large scale, second-order change. Although, generally, second order change should elicit strong emotional or behavioral reactions, the individual whose hierarchy reflects work-related roles as very low and distanced from other non-work roles, should not experience significant emotional or behavioral responses. Proposition 5: The effect of both first-order and second-order changes on a particular individual can be predicted by evaluating the individual's identity hierarchy. If the event is associated with a dominant role, then the event will result in a significant emotional or behavioral reaction, regardless of whether the change is considered first- or second-order change. Proposition 6: It is not only the location of the particular identity on one's hierarchy that can be used to predict reactions to events, it is also the "distance" between roles in a hierarchy that will be significant in determining the nature of a response to an organizational change effort. Given the fact that the characteristics of the event, not simply the event, interact with identities to determine how employees will respond to change efforts, one can speculate about the ways in which employees will respond to second order change. Basically, second-order changes efforts are enacted because immediate behavioral change is needed within the organization. Resistance to change curtails the organization's ability to adapt. Therefore, given the relationships posited in the identity cycle, organizations have two vehicles for encouraging behavioral changes supportive of change. First, they can tailor the event to somehow tap each employee's most dominant identity, or the organization can enact a change effort that has the most likely chance of altering employees' hierarchies so that the organization (not career, job, Page 24 Organizational Change and the Identity Cycle WP 94-01 department, team, or committee roles within the organization) role is dominant. If the organization role is dominant, then the firm might have the greatest chance of encouraging employees to go along with change efforts that are needed for organizational survival. Designing change efforts to maximize the needs associated with the most dominant role in each individual employee's hierarchy would be fairly difficult, if not impossible to do. However, it might be probable to group employees and at least tailor training or rewards programs to minimize resistance within certain work groups. For example, a particular division within an organization might have strong affiliation with the division, not the organization, thus resulting in a majority of employees viewing their "division-member role" as more salient than any other work-related roles. The organization might then tailor the change effort so that it appeals to their divisional role, thus inciting quick attitudinal and behavioral responses. For example, the business might make the desired behavior somehow the target of a competition between divisions. Thus, behaving in a way that is needed by the organization means changing behavior in a way that supports the division member role. Of course, the problem with this course of action is that organizations undergoing radical change do not have time to create tailor-made programs for targeted employees. Proposition 7: Implementation of second order change can be enhanced by implementation methods that are targeted at an employee group's most dominant role within their hierarchy (assuming the group can be defined by sharing a common dominant role). Given the low probability of identifying groups with common dominant roles, the more feasible way to pursue second-order change might be to implement a process that has the most likely chance of encouraging emotional and behavioral reactions desired by the organization. Given the same division, with low organizational commitments (thus the organizational member role is low on the hierarchy), the organization's strategy for changing behavior would require methods that would actually have a significant chance of changing commitment away from the division to the organization. Thus, these employees will have a resulting hierarchy that emphasizes the organizational role. If the business shows that change is necessary for organizational survival, then the probability that behaviors will be consistent with the needs of the organization increases. This is important because organizational survival might necessitate deletion of a job, team, department or division. Therefore, organizational identity is important in supporting radical change efforts. In order to enact this strategy the organization might make a change in top management in the division, alter the rewards system to support organizational rather than division objectives, Page 25 Organizational Change and the Identity Cycle WP 94-01 or make future promotional opportunities available outside the division. These "events" have a chance of eliciting a strong emotional and behavioral reaction that might result in changes in role commitment, changes in the hierarchy, and subsequent behaviors that encourage a dominant organizational role. Proposition 8: Implementation of second order change that encourages a change in social categorization, commitment, and hierarchy to emphasize the organizational member role will enhance an organization's ability to implement the change. Organizational transformations and reengineering efforts have not been very successful. Can this be explained with the identity perspective? Consider the fact that most second order changes (transformations, reengineering) have been associated with intentional organizational efforts to disassociate the employee from the organization. Organizations are telling their employees that they no longer will enjoy long-term careers with them. They should start to see themselves as "careerists", not organizational members. Employees are responding to this, and employee loyalty at many organizations is at record lows (Fisher, 1991). Thus, organizations, rather than creating change efforts that encourage a reordering of identity hierarchies so that the organization is more important, have created situations where the organizational role is probably the least important. Should we be surprised that many organizational change efforts have failed? If employees do indeed react to events through their identity filter, and the filter has changed so that the roles associated with their affiliation with the organization are not only low but probably significantly distanced from all other roles, why should we expect employees to behave in ways that are supportive of organizational survival when it threatens more valued roles? Conclusion The identity cycle provides an interesting vehicle for analysis of a phenomenon that is gripping corporations in the 1990s. This paper has shown how some of the ideas from the identity cycle can be used to understand the ways in which employees respond to organizational change. The model takes a "bit picture" topic that is often difficult to define and research (second-order change, transformation, reengineering) and provides a useful framework for studying these phenomenon. Both the social identity theory and identity theory literature are replete with examples of successful research designs, survey instruments, and construct validation studies. These tools have, to date, been limited in their domain, but the research is ready to be transferred to other Page 26 Organizational Change and the Identity Cycle WP 94-01 fields. This paper suggests that the identity cycle, which not only incorporates social identity theory and identity theory, but also extends these two theories, can be used to understand complex organizational change. According to Robertson et. al. (1993: 619), "general theoretical formulation of the dynamics of planned change processes - formulations not tied to specific types of interventions - remain particularly undeveloped". It is hoped that the identity cycle will provide a framework for extending the research on organizational change efforts by providing a reasonable and workable model for guiding future research. In addition, the propositions developed in this paper could also be extended to other domains, such as the family, school, or home where change of various magnitudes also occurs. Page 27 Organizational Change and the Identity Cycle WP 94-01 References Abrams, D., Jackson, D. & St. Claire, L. 1990. Social identity and the handicapping functions of stereotypes: Children's understanding of mental and physical handicap. Human Relations, 43:1085-1098. Adler, P. & Adler, P.A. 1987. Role conflict and identity salience: College athletics and the academic role. The Social Science Journal, 24, 443-455. Ashforth, B.E. & Mael, F. 1989. Social identity theory and the organization. Academy of Management Review, 14:20-39. Ashforth, B.E. & Humphrey, R.H. 1993. Emotional labor in service roles: The influence of identity. Academy of Management Review, 18:88-115. Beer, M., Eisanstadt, R. A., & Spector, B. (1990). Why change programs don't produce change. Harvard Business Review, 68, 158-166. Blumer, H. 1969. Symbolic interactionism: Perspective and Method. Englewood Cliffs, N.J.: Prentice Hall. Brewer, M.B. 1979. In-group bias in the minimal intergroup situation: A cognitive motivational analysis. Psychological Bulletin, 86:307-324. Brown, R. & Williams, J. 1984. Group Identification: The same thing to all people. Human Relations, 37: 547-564. Brown, R., Condor, S, Mathews, A., Wade, G., & Williams, J. 1986. Explaining intergroup differentiation in an industrial organization. Journal of Occupational Psychology, 59:273-286. Burke, P.J. 1991. Identity processes and social stress. American Sociological Review, 56: 836- 849. Burke, P.J. & Reitzes. 1991. An identity theory approach to commitment. Social Psychology Quarterly, 54:239-251. Calfasso, R. (1993). Rethinking reengineering. Computerworld Special Report, 102. Charno, H.W., Pilliavin, J.A., & Callero, P.L. 1988. Role identity and reasoned action in the prediction of repeated behavior. Social Psychology Quarterly, 51:303-317. Dunphy, D. C., & Stace, D. A. (1988). Transformational and coercive change: Beyond the O.D. Model. Organizational studies, 9, 317-334. Drucker, P. F. (1992). The new society of organizations. Harvard Business Review, 70, 95-104. Festinger, L. 1954. A theory of social comparison processes. Human Relations, 7:117-140. Fisher, A. B. (1991). The morale crisis. Fortune, 124, 70-80. Garza, R.T., Lipton, J.P., & Isonio, S.A. 1989. Group ethnic composition, leader ethnicity, and task performance: An application of social identity theory. Genetic, Social, and General Psychology, 115: 297-314. Page 28 Organizational Change and the Identity Cycle WP 94-01 Goldstein, J. (1988). A far-from-equilibrium approach to resistance to change. Organizational Dynamics, 17, 16-26. Hall, G., Rosenthal, J., & Wage, J. (1993). How to make reengineering really work. Harvard Business Review, 71, 119-131. Hammer, M. & Champy, J. (1993). Reengineering the corporation. New York, NY: Harper Business. Hoelter, J.W. 1983. The effects of role evaluation and commitment on identity salience. Social Psychology Journal, 46:140-147. James, W. 1890. The principles of psychology. New York: Holt. Kanter, R. M. (1983). The change masters. New York: Simon & Schuster. Kuchera, M.E. & Miller, S.I. 1988. The effects of perceptions of the academic job market on adjunct faculty: An identity-theory analysis. Sociology of Education, 61:240-254. Leary, M.R., Wheeler, D.S. & Jenkins, T.B. 1986. Aspects of identity and behavioral preference: Studies of occupational and recreational choice. Social Psychology Quarterly, 49:11-18. Levy, A. & Merry, U. (1986). Organizational transformations: Approaches, strategies, theories. New York: Praeger. Levy, A. (1986). Second-order planned change: Definition and conceptualization. Organizational Dynamics, Summer, 5-20. Lobel, S.A. 1991. Allocation of investment in work and family roles: Alternative theories and implications for research. Academy of Management Review, 16:507-521. Lobel, S.A. & St. Clair, L. 1992. Effects of family responsibilities, gender, and career identity salience on performance outcomes. Academy of Management Journal, 35:1057-1069. Mitroff, I. I. (1987). Business not as usual. San Francisco, CA: Josey-Bass. Oaker, G & Brown, R. 1986. Intergroup relations in a hospital setting: A further test of social identity theory. Human Relations, 48:767-778. Quinn, J. B. (1980). Strategies for change: Logical incrementalism. Homewood, IL: Irwin. Sachdev, I. & Bourhis, R.Y. 1991. Power and status differentials in minority and majority group relations. European Journal of Social Psychology, 21:1-24. Santee, R.T. & Jackson, S.E. 1979. Commitment to self-identification: A sociopsychological approach to personality. Human Relations, 32:141-158. Serpe, R.T. 1987. Stability and change in self: A structural symbolic interactionist perspective. Social Psychology Quarterly, 50:44-55. Page 29 Organizational Change and the Identity Cycle WP 94-01 Simon, R.W. 1992. Parental role strains, salience of parental identity and gender differences in psychological distress. Journal of Health and Social Behavior, 33:25-35. Stryker, S. 1968. Identity salience and role performance: The relevance of symbolic interaction theory for family research. Journal of Marriage and the Family, 30:558-564. Stryker, S. & Serpe, R.T. 1982. "Commitment, identity salience and role behavior: Theory and research example." Pp. 199-218 in Personality, Roles, and Social Behavior, edited by W. Ickes & E.S. Knowles. New York, NY: Springer Verlag. Stryker, S. & Statham, A. 1985. "Symbolic interaction and role theory." Pp. 311-378 in Handbook of Social Psychology, edited by C. Lindzey & E. Aronson. New York, NY: Random House. Stryker, S. 1987. "Identity theory: Developments and extensions." Pp. 89-103 in Self and Identity: Psychosocial Perspectives, edited by K. Yardley & T. Honess. John Wiley & Sons. Tajfel, H. 1970. Experiments in intergroup discrimination. Scientific American, 223:96-102. Tajfel, H., Billig, M.G. 1974. Familiarity and categorization in intergroup behavior. European Journal of Social Psychology, 10:159-170. Tajfel, H., Billig, M.G., Bundy, R.P., & Flament, C. 1971. Social Categorization and intergroup Behavior. European Journal of Social Psychology, 1:149-178. Tajfel, H., & Turner., J. C. 1985. "The social identity theory of intergroup behavior." Pp. 7-24 in Psychology of intergroup relations, edited by S. Worchel & W. G. Austin. Chicago: Nelson-Hall. Tajfel, H., & Turner., J. C. 1979. "An integrative theory of intergroup conflict." Pp. 33-47 in The social psychology of intergroup relations, edited by W. G. Austin & S. Worchel. Monterey, CA: Brooks/Cole. Thoits, P.A. 1991. On merging identity theory and stress research. Social Psychology Quarterly, 54:101-112. Thoits, P.A. 1992. Identity structures and psychological well-being: Gender and Marital status comparisons. Social Psychology Quarterly, 55:236-256. Tichy, N. M. (1983). Managing strategic change. NY: J. Wiley & Sons. Tichy, N. M. & Devanna, M. A. (1986). The transformational leader. NY: J. Wiley & Sons. Tolman, E.C. 1943. Identification and the post-war world. Journal of Abnormal and Social Psychology, 48:141-148. Turner, J.C. 1975. Social comparison and social identity: Some prospects for Intergroup behavior. European Journal of Social Psychology, 5:5-34. Page 30 Organizational Change and the Identity Cycle WP 94-01 Turner, J.C. 1981. "The experimental social psychology of intergroup behavior." Pp. 66-101 in Intergroup Behavior, edited by J.C. Turner & H. Giles. Chicago, IL: University of Chicago Press. Turner, J. 1978. "Social comparison, similarity and ingroup favoritism." Pp. 235-250 in Differentiation Between Social Groups, edited by H. Tajfel. London: Academic Press. Welbourne, T., and Cable, D. (in press). Group incentives and pay satisfaction: Understanding the relationship through an identity theory perspective. Human Relations. Wells, B. 1990. Social identity as a tool to build multi-community clusters. Sociological Practice, 8:157-166. White, C.L. & Burke, P.J. 1987. Ethnic role identity among black and white college students. Sociological Perspectives, 30:310-331. Wiley, M.G. 1991. Gender, work, and stress: The potential impact of role-identity salience and commitment. The Sociological Quarterly, 32:495-510. Page 31