Almost 60 years ago Robinson (1949) defined supervision in the context of social work as “an educational process in which a person with a certain equipment of knowledge and skill takes responsibility for training a person with less equipment” (p. 53). This emphasis on the educational aspect of supervision has over the years been combined with a second emphasis on administration that includes efforts to control and coordinate social workers to get the job done.

In another definition of supervision, Kadushin (1976) added to these two sets of tasks the “expressive–supportive leadership function” (p. 20) that focuses on the problem of sustaining social workers by offering emotional support and making efforts to assist them when they have “job-related discouragements and discontents.” Kadushin (1976), combining these three major functions, provided a definition of supervision that serves the purposes of this entry well:

A social work supervisor is an agency administrative staff member to whom authority is delegated to direct, coordinate, enhance, and evaluate on-the-job performance of the supervisees for whose work he [or she] is held accountable. In implementing this responsibility the supervisor performs administrative, educational, and supportive functions in interaction with the supervisee in the context of a positive relationship. The supervisor's ultimate objective is to deliver to agency clients the best possible service, both quantitative[ly] and qualitatively, in accordance with agency policies and procedures. (p. 21)

A crucial aspect of this definition is the emphasis on carrying out these tasks in interaction with the supervisee “in the context of a positive relationship.” This relationship between supervisor and supervisee, which parallels the relationship between social worker and client, has been described as consisting of three elements: (1) rapport (general ability to get along); (2) trust (the ability of the social worker to be open with the supervisor and to share mistakes and failures as well as successes); and (3) caring (the communication by the supervisor of concern for the social worker as well as for the client) (Shulman, 1993).
Kaiser (1992), addressing the issue of supervisory relationship in family therapy supervision, highlighted the “phenomenon of isomorphism; what happens in supervision is reflected in the therapy” (p. 284). This is referred to later in this entry as the “parallel process,” in which the interaction between supervisor and supervisee directly affects the relationship between the social worker and the client. In reviewing the literature from the fields of social work, psychology, and marriage and family therapy, Kaiser identified four consistently cited issues: “the process of accountability (maintaining objectivity), the need for promoting the supervisee's personal awareness, the importance of establishing trust, and the need to attend to power and authority issues” (p. 284).

Dynamic System
One descriptive model of supervision provides a dynamic system conceptualization in which staff members constantly interact with a number of systems that are directly related to their work (Shulman, 1993). A social worker in a child welfare agency, for example, must deal with clients, foster parents, the agency administrators, the supervisor, professional colleagues, clerical staff, the court system, and other agencies or institutions, such as the schools. At any moment in their work day, social workers could be called on to negotiate one or more of these systems. The relationship with each system places unique demands on the social worker and requires specific knowledge and skills if he or she is to negotiate it effectively. This model describes the supervisor's role as in the middle between the social worker and these important systems, helping the social worker to negotiate them more effectively.

This model generates four major agendas for the supervisor and the social worker. First, there is the question of job management. A social worker must be able to work within the structure of the agency in terms of time (for example, being on time for work and meetings, meeting deadlines on reports, and timely recording and developing the skills needed for effective management of caseloads). Second, the social worker must relate effectively to agency policy and procedures. Social workers must implement policies and follow established procedures while they simultaneously develop the skills necessary to influence them. Third, for effective practice, social workers must develop skills to deal with professional colleagues, support staff, and supervisors. As social workers attempt to deliver a service, their efforts must be coordinated with those of other staff members. Harmonious work relationships are required for staff members to effectively provide help to clients. When a breakdown occurs in team relationships, the outcome is an almost inevitable deterioration of client service. Fourth, social workers must also deal with supervisors—symbols of authority—and must learn how to use this relationship to their advantage.

Supervisor–Supervisee Relationship
Other models of supervision focus on the nature of the supervisor–supervisee relationship. These have been described in the literature as ranging from more traditional, authoritarian models in which the supervisor's authority emerges from agency sanction on one end, to more collaborative models in which the authority emerges essentially from the supervisor's competence on the other end.

Munson (1981, 1983) surveyed 65 supervisees and 64 supervisors. He focused on models of supervision in three areas: (1) structure (traditional–individual, group, and independent); (2)authority (sanction versus competence); and (3)teaching (Socratic, growth, and integrative). He examined the impact of the use of different models on social worker satisfaction with supervision and integration. Munson (1981) found that “The structural models did not produce significantly different outcomes regarding interaction and satisfaction, but the authority models did. The competence model of authority was the most productive in all respects” (p. 71). This was the model in which the supervisor's authority was derived from competence and skill rather than from agency sanction.

Munson's (1981) findings supported those of Kadushin (1974). Traditional supervision, in which authority for the supervisor flows from agency sanction, was questioned. Munson pointed to “the need to encourage greater independence and autonomy” (p. 296). In a study of supervision, Kadushin (1974) conducted a national survey of 750 supervisors and an equal number of supervisees. The purpose was to identify the sources of satisfaction for both supervisors and supervisees. Supervisors in the study took great satisfaction in helping supervisees grow and develop professionally; their greatest source of dissatisfaction related to dealing with administrative red tape. Supervisees identified being able to share responsibility with supervisors and being able to obtain support for difficult cases as their greatest source of satisfaction. A majority of both the supervisors and  supervisees in Kadushin's study believed that as the supervisee gained experience, the relationship became one of consultant–consultee, a form of supervision preferred by many social workers. Strong dissatisfaction with supervision was reported by many social workers, who believed the authority of the supervisor was exercised in a negative manner.
Consider that most of the research findings discussed in this entry focus on the supervisor's or the supervisee's perceptions of the process rather than on the impact of supervision on client services and the outcomes of those services. Harkness and Portner (1989) reviewed a number of these studies and examined the underlying conceptualization of social work supervision guiding research efforts in the field. They pointed out that a view of supervision as a training process shifted the research focus from the impact of supervision on client services to the impact on social workers. In 1991 Harkness and Hensley, citing Shulman et al.'s (1981) proposal for a change in the paradigm, argued for a shift back to evaluating supervision in terms of client outcomes. Their own study (1991), for example, suggested a link between a client-focused supervision and an improvement in client ratings of helpfulness and goal attainment.

EDUCATIONAL FUNCTION

Kadushin (1976) described educational supervision as a more specific kind of staff development in which “training is directed to the needs of a particular worker carrying a particular caseload, encountering particular problems and needing some individualized program of education” (p. 126). Four major areas of the curriculum are (1) professional practice, (2) professional impact, (3) job management, and (4) continued learning.

In one supervision study (Shulman, Robinson, & Luckyj, 1981), social workers were asked to identify what they would like to have discussed in their supervision sessions, compared with the actual content of those sessions. Their first preference was that supervisors should devote more time to teaching practice skills, followed by more time on discussing research information and providing feedback on performance. Interestingly, such supervision–consulting roles also were the favored tasks of supervisors queried in the study.

Other research has supported the idea that both supervisors and supervisees regard the educational function of supervision as important and a source of satisfaction. In Kadushin's (1974) study, two of the three strongest sources of supervisor satisfaction were found to be related to helping the supervisee grow and develop professionally and to sharing social work knowledge and skill. In the same study, social workers indicated that two of the three main sources of their satisfaction with supervision were receiving help in dealing with clients and in developing as professionals. In another study (Scott, 1969), professionally oriented social workers expressed a preference for supervisors who knew their theoretical fundamentals, were skilled in teaching, and were capable of offering professional assistance.

The notion of the parallel process is central to the educational function of supervision. There are assumed parallels between the dynamics of supervision and any other helping relationship. Therefore, the skills that are important in direct practice with clients or patients also are important to the supervisory relationship. A number of authors have identified these similarities (Arlow, 1963; Doehrman, 1972; Schwartz, 1968). Much of what is known about effective communication and relationship skills can be useful in implementing diverse aspects of the supervisory function, such as coordination, education, and evaluation.
In addition, the way the supervisor demonstrates the helping relationship with social workers will influence the manner in which staff members relate to clients. For example, when supervisors attempt to help staff members develop a greater capacity for empathy with difficult clients, they ought also to simultaneously demonstrate their own empathy for the staff. Supervisees learn what a supervisor really feels about helping by observing the supervisor in action. More is “caught” by staff than is “taught” by the supervisor.

Frankel and Piercy (1990) described the potential positive benefits of this modeling in a study of the impact of supervisory phone-ins on family therapists and the resultant therapist responses to clients. She found that effective “support” and “teach” behaviors of the supervisor were replicated by the therapists and resulted in more positive client outcomes. Jacobs (1991) discussed this process in terms of student–supervisor relationships, stressing the problem of power abuses and boundary violations in clinical supervision. She suggested that clients can become “victims of a dysfunctional supervisory system if the student replicates the harmful interaction with current or future clients” (p. 130).

The administrative role of the supervisor contains a number of elements, all of which are designed to aid in the implementation of the mission. These elements include coordination of activities between staff members and between one's unit or department and other parts of the organization, as well as between staff and the community (for example, other agencies). This rule also involves working with staff and administration to design and implement policies and procedures for supporting the work of the setting.

Conflict between Staff and Administration
An area of stress reported by supervisors is the feeling of being caught in the middle between staff and administration on a point of conflict. Conflicts between staff members and the administration are often the rule rather than the exception, and they make up a large part of the interaction in the formal and informal systems. Administrators may sometimes set unworkable policies, because they are too far removed from the realities of practice to understand their effect on services. New programs or procedures may be developed by study groups or outside consultants who have little understanding of the actual nature of the practice. Cost-containment and funding cutbacks can lead to organizational stress and force decisions that can affect programs and positions.

One study (Erera, 1991), which involved 62 public welfare supervisors, used qualitative and quantitative methods to examine the impact of incompatible policies as one of four examples of role conflict experienced by supervisors. Incompatible policies emerging from state, county, or public agencies were found to be associated with the supervisory role for all supervisors except for two who were involved in specialized programs guided by few policies. Erera concluded that the study supported the notion “that the middle managers are located at that critical point where the organization's structure impinges on the individual worker” (pp. 46–47).

At the same time, staff members may resist changes that are threatening to the status quo or require sacrifices such as weekend or evening work. They may be unwilling to consider the requirements of the entire institutional setting and may stubbornly resist pleas for flexibility or consideration.

Mediating Role
The “third force,” or mediating role, for the supervisor is that of providing a framework in which the supervisor does not choose between identifying with either the staff or the administration. In most cases, rather than taking sides with one over the other, the supervisor can take a stand with respect to the process. Being caught in the middle can be a most effective position for stimulating change. Bunker and Wijnberg (1985) described supervisors as mediators of organizational climate who serve as a buffer between frontline staff and administration.

This role for supervisors, however, is not one in which they never take a position or are neutral on every issue. It does not mean that supervisors will shy away from conflict in attempts to smooth over real differences of interest between social workers and the administration. Just the opposite is true. Effective implementation of this role requires that conflicts smoldering beneath the surface be brought to light. There will be times when advocacy of a staff position and confrontation of the administration are essential tools for supervisors, although careful thought must be given to how these tools are used. Even in the role of advocate, supervisors must not lose sight of the essential common ground between the staff and the administration.

Availability to Staff
The issues of job stress and job manageability are crucial ones for both frontline social workers and supervisors. A number of studies have pointed to stress and job manageability as important factors. In a study of workloads for supervisors in a public welfare agency, Galm (1972) found that supervisors simply did not have enough time to supervise. In Kadushin's (1974) study 53 percent indicated that not having time to supervise was one of the strongest sources of their job dissatisfaction. A supervisor's stress can have a powerful impact on his or her availability to social workers as well as on the supervisor's capacity to provide support (Shulman, 1993).

Availability to social workers is a particularly important variable in those arenas of practice that generate significant stress and lead to social worker burnout. The term burnout has come to be used to describe a syndrome exhibited by workers who deal with intense stress over a period during which little support is available. Although most commonly noted in the child welfare literature, burnout also is seen in reference to workers in any high-stress field of practice, particularly in large government agencies (see Borland, 1981; Copans, Krell, Gundy, Rogan, & Field, 1979; Daley, 1979; Falconer, 1983; Freudenberger, 1974; Riggar, Godley, & Hafer, 1984).

Issues of Caseload Size and Trauma
Another source of stress for front-line social workers is caseload size. Social workers in both public and private agencies have increasingly been asked to do more with fewer resources. Increased caseloads, the growing complexity of the problems facing clients, and limitations on the availability of other resources combine to make social workers' jobs more difficult.
Social workers also experience high levels of burnout because of the powerful nature of the events and the emotions in the lives of their clients. The impact of the death of a child on one's caseload, the emotions associated with sexual and physical abuse, counseling a grieving family in a hospital or a suicidal client, and working with people with acquired immune deficiency syndrome are examples of work that exacts an emotional toll on the caring professional. A particularly traumatic incident—for example, a suicide on a psychiatric ward—may have effects that are felt by the social worker involved as well as by all of his or her colleagues. The impact of trauma can be exacerbated if the initial agency response is to investigate to determine blame, rather than to provide support to the helpers.

Loughlin (1991) focused on the trauma of child sexual abuse and its impact on those who work with it as well as their institutions and networks. She stressed the potential for the agency itself to become dysfunctional, responding to the social worker involved in ways that parallel the abuse in the family and for boundaries to become blurred. She suggested that “The emotional impact of such work should not be underestimated and that good supervision which considers the emotional impact of this and similar life crisis interventions is vital for everyone working in this field” (p. 111). Greene (1991) made a similar case for the importance of support for social workers in geriatric practice because they also face emotional stress and inherent value conflicts.

Kurland and Salmon (1992) argued that teachers, field instructors, consultants, and supervisors must recognize the intractability of many of the problems faced by clients and to help prepare their supervisors for the difficulty of influencing real change.

As the social work profession has increasingly turned its attention to issues of diversity in practice, similar issues have emerged in the context of supervision and management. One major issue has been the underrepresentation of women in management roles (Fanshel, 1976; Shulman, 1993; Shulman et al., 1981). Chernesky (1980), reviewing the research in this area, described how women tend to be located and to remain in direct-service positions. Similar concerns have been raised about the number of members of racial and ethnic groups who assume management roles. Affirmative action programs, designed to assure that front-line staff and management become diversified and begin to look like the clients they serve, have been initiated in federal, state, municipal, and private agencies.

As increasing numbers of women and minorities have assumed supervisory positions, interest has grown in the nature of the experience of these populations once management positions have been assumed (McNeely, 1983). Wright, King, and Berg (1985), pointing out that most studies of job satisfaction have included samples predominantly of white males, examined factors that affected job satisfaction of black females in management positions. Their tentative findings, with a few exceptions, replicated earlier findings with male populations that had pointed to organizational variables as most predictive of job satisfaction. The findings suggested that

Black female managers who receive positive job performance evaluations, who have some degree of authority over their job performance, who have clearly specified responsibilities and functions, who are in positions that are commensurate with their experience, training, and education and who occupy positions previously occupied by females will—all other things being equal—be more satisfied with their jobs. (p. 71)

Wright et al.'s (1985) job-stress index, which included four variables that specifically identified stress related to being black and female, did not predict job satisfaction for this admittedly small sample.

Increased interest in cross-cultural supervision also has been noted in the recent literature. In their study of field supervision, McRoy, Freeman, Logan, and Blackmon (1986) noted that cross-cultural supervision was both desirable and problematic. In another study that examined the experiences of emotional support, social undermining, and criticism of African American practitioners, Jayaratne et al. (1992) found that social undermining had substantial and negative effects on the social worker, which were not significantly diminished by social support. Another report (Gant et al., 1993) on the impact of undermining an African American social worker's perceptions of coworker and supervisory relationships found a pattern of social worker perceptions that were governed by the gender and race of the supervisor. The relatively recent growth of women and minorities in supervisory positions and the tentative and exploratory nature of most of the recent studies on the effects of gender and race on supervision suggests that additional research is needed.

Consultation is an interaction between two or more people in which the consultant's special competence in a particular area is used to help the consultee with a current work problem (Caplan, 1970). Social work consultation may encompass the functions described for the supervisor; however, there is one major difference. Consultants usually do not carry administrative responsibility and accountability. Their authority is derived from their perceived expertise in a subject rather than from formal sanction by the agency or setting. Miller (1987) alluded to this difference when he suggested that consultation “consists of structured advice giving and problem clarification about clients and professional [practice; it] becomes more or less equivalent to supervision, with all the pleasures and few of the headaches” (p. 749).

Social workers have increasingly gained recognition for their knowledge in a wide range of settings; the two major areas of consultation include case consultation and program consultation. Case consultation usually involves a social work consultant's working with line staff to assist them in providing direct services to clients. Program or organizational consultation usually involves work with administrative staff and may focus on agency policies, programs, and procedures.

Drisko (1993) described one example of case consultation with special education teachers. In this model, social workers meet with teachers to detail jointly three single-spaced profiles that focus the consultation on student strengths. “The profiles aid the teacher in understanding the student from a psychological perspective while offering new knowledge and skills in assessment and self-awareness” (p. 19).

Social workers also may be consultants to teams in interdisciplinary settings (Abramson, 1989; Kadushin, 1976). Abramson pointed out that many social work skills can easily be adapted to the consultation role. These include “(1) problem assessment; (2) problem definition; (3) mediation and negotiation; and (4) contract development” (p. 57).

In a final example, periodic case consultation may be provided to private practitioners. Kaslow (1991) described general guidelines for marital therapy consultation, which, because of its private and voluntary nature, leaves the trainee “free to utilize or disregard the consultants' ideas and recommendations” (p. 133). She pointed out that clinicians may seek consultation for a number of different reasons including “licensure and/or organizational membership requirements for a certain number of documentable hours of supervision” (p. 133). They also may seek consultation because they

“may desire supervision for a provocative professional interaction that stretches their knowledge and skills, because they are expanding their areas of therapeutic practice and want guidance and affirmation, or because they have some particularly difficult case and would like a `second opinion' from a senior and respected colleague.” (p. 132)

Lawrence Shulman, EdD, is professor, Boston University, School of Social Work, 264 Bay State Road, Boston, MA 02215.

For further information see
Case Management; Clinical Social Work; Continuing Education; Direct Practice Overview; Ethics and Values; Interdisciplinary and Interorganizational Collaboration; Management Overview; Organizations: Context for Social Services Delivery; Planning and Management Professions; Private Practice; Professional Conduct; Professional Liability and Malpractice; Program Evaluation; Purchasing Social Services; Quality Assurance; Social Work Education; Social Work Practice: History and Evolution; Social Work Profession Overview; Volunteer Management.

Key Words
administration
consultation
field instruction
supervision
teaching