Dual Relationships

home  »  psychology »  boundaries  »  avoiding exploitives

Avoiding Exploitive Dual Relationships:
A Decision-Making Model

Michael C. Gottlieb

NOTE: For related articles on this site that discuss multiple relationships & boundaries, please see:

ABSTRACT: Ethical principles provide general guidelines for professional behavior. Unfortunately, these principles are not adequate for practical decision-making. One ethical principle which generates frequent consternation is that of avoiding dual relationships. Some models have been developed to address this problem, but they are typically general and not especially helpful when specific ethical dilemmas arise. The principle of avoiding dual relationships is briefly reviewed, and problems with the principle are noted. This article presents a specific decision-making model to avoid exploitive dual relationships; the model's uses and limitations are critically examined. Avoiding dual relationships is an ethical injunction which frequently generates dilemmas for psychologists. Ethical principles provide general guidelines for professional conduct, but often provide little or no specific guidance for practical decision-making. This article reviews the relevant literature and presents a specific decisionmaking model for avoiding exploitive dual relationships. Examples of its use and pertinent limitations are noted.

View citation and copyright.


Dual Relationships

Professions use ethical principles (Beauchamp & Childress, 1983) to develop ethics codes to guide the conduct of their members. All major mental health professional codes (American Psychological Association, 1990; National Association of Social Workers, 1979; American Psychiatric Association, 1986; American Association for Marriage and Family Therapy, 1985) contain a proscription against dual relationships.

Problems with dual relationships may be explained by role theory. Social roles contain inherent expectations about how a person in a particular role is to behave as well as the rights and obligations which pertain to that role. Role conflicts arise when the expectations attached to one role call for behavior which is incompatible with that of another role (Kitchener, 1988).

A dual role relationship exists when an individual simultaneously or sequentially participates in two role categories (Kitchener, 1986). Carroll, Schneider & Wesley (1985) concluded that a dual relationship exists when, in addition to the professional relationship, there is some other relationship with the individual: friend, family member, student or business partner.

The rule that mental health professionals avoid dual relationships is aspirational in nature. It is a goal we strive to reach, but one which is impossible to avoid completely on a daily basis (Adleman & Barrett, 1990; Haas & Malouf, 1989; and Kieth-Spiegel & Koocher, 1985). For example, university faculty members serve graduate students in numerous capacities simultaneously, such as research director, teacher, employer and supervisor. On the other hand, because of their potentially serious consequences, some dual relationships have been specifically prohibited, such as sexual relationships with psychotherapy clients (APA, 1977).

The American Psychological Association ethical principles (APA, 1992) recognize "multiple relationships." The code acknowledged that in certain situations,". . . it may not be feasible or reasonable for psychologists to avoid social or other nonprofessional contact with persons such as patients" (p. 1601). Nevertheless, it cautions against entering into such relationships if, "it appears likely that such a relationship reasonably might impair the psychologist's objectivity or otherwise interfere with the psychologist's effectively performing his or her functions as a psychologist, or might harm or exploit the other party" (p. 1601).

When a psychologist is confronted with an ethical dilemma regarding dual or multiple relationships, where is she or he to turn? In an effort to address this problem, some writers have developed ethical decision-making models to assist professionals in these situations.

Ethical Decision-Making Models

Two models have addressed ethical decision-making problems at a general level. Rest (1983) proposed a Four-Component Model of moral behavior which encompasses the entire moral action process. Component I recognizes that a moral problem exists. Component H requires reasoning about the problem, and Component III involves choosing a moral course of action in the face of competing values. Component IV entails carrying out the action. Woody (1990) proposed a five-dimensional model for defensible decisions. The bases are: theories of ethics, professional codes of ethics, professional theoretical premises, the socio-legal context, and personal! professional identity.

Two other models offer assistance by asking more specific questions. Handelsman (1991) developed a series of eleven questions which the professional must answer to make a decision in a particular ethical dilemma. Haas & Malouf (1989) present a five-stage decision-making model. The process begins by asking what is the relevant professional, legal, or social standard, and second, is there a reason to deviate from the standard. Third, can one identify the ethical dimensions of the dilemma and establish one as primary? If so, the fourth step is to generate a list of possible actions and assess whether the new course of action satisfies the needs of the affected parties. Finally, one must assess whether the course of action is ethical and possible to implement. If so, the chosen course of action may be implemented.

Kitchener (1988) addressed dual relationship problems specifically when she offered three guidelines for differentiating between relationships that have a high probability of leading to harm and those that do not. The first guideline states that as the incompatibility of expectations increases between roles, so will the potential for harm. Second, as the obligations associated with different roles diverge, the potential for loss of objectivity and divided loyalties increases. Third, as the prestige and power differential between the professional's and the consumer's roles increase, so does the potential for exploitation. Kitchener (1988) assumed that as the risk of harm increased, so should the ethical prohibitions against engaging in the relationships.

Roll & Millen (1981) developed guidelines for psychologists who must respond to requests for psychotherapy from acquaintances. They included: avoiding doing so if possible, remaining mindful of possible transference relationships, obtaining consultation, maintaining boundaries, being aware of one's own values, being prepared to lose the friendship, remaining mindful of confidentiality, recognizing when treatment should be terminated, and insuring that the therapist's personal needs are met to avoid abusing the psychotherapy process.

Gonsiorek & Brown (1989) were the first to propose highly specific rules with regard to sexual relationships with former psychotherapy clients. They make a distinction between two types of therapy. Type A therapy is one in which the transferential relationship plays a primary role in the process. On the other hand, Type B therapy is short-term and offers little opportunity for transferential relationships to develop. Based on this distinction, they offer six rules. First, sexual contact with former clients who have received Type A therapy is always and forever prohibited. Second, post-termination sexual contact with severely disturbed clients will always be considered unethical and improper, regardless of whether Type A or B therapy was rendered. Third, it is never permissible for the therapist to initiate posttermination romantic contact. Fourth, sexual contact with a client who has received Type B services always will be considered unethical if it occurs within two years after termination of therapy. Sexual contact with any client who has received Type B therapy will continue to be prohibited if the therapist at some time has told the client to return to therapy at a later date, or if any other than incidental social contact occurred during the two-year period. Fifth, any mental health services that cannot clearly and easily be defined as Type B shall be considered to be Type A for purposes of this model. Sixth, rules for ethical action in nonsexual post-therapy overlapping or dual relationship situations should be formulated apart from rules regarding sexual contact.

Finally, The American Psychological Association clarified its position on the issue of sexual intimacies with former psychotherapy patients. Because such behavior is, "so frequently harmful. . . and undermines public confidence . . . ." (p. 1605), it is specifically prohibited for two years after termination, and it remains prohibited even after two years "except in the most unusual circumstances" (p. 1605). If a psychologist engages in such behavior, the burden of demonstrating that there has been no exploitation rests with the psychologist. The code lists seven factors which must be considered; the time since termination, the nature and duration of therapy, the circumstances of termination, the patient's personal history, the patient's current mental status, the likelihood of adverse impact and any statement made by the psychologist during therapy suggesting or inviting a post-termination relationship.

In summary, the models of Rest (1983) and Woody (1990) offer general guidelines, but do' not specifically address the issue of dual relationships. Handeisman (1991) and Haas & Malouf (1989) have made contributions by developing decision-making models based on general guidelines. Unfortunately, their models do not offer the specific assistance that is required. The recommendations of Roll & Millen (1981) are helpful but can be applied only in very specific circumstances. Gonsiorek & Brown (1989) recommend specific actions which apply only to the psychotherapeutic relationship and leave no room for professional judgment. The APA acknowledges that multiple relationships may be acceptable, but only provides specific guidelines for sexual intimacies with current and former psychotherapy patients.

What is needed is a specific decision-making model which delineates the dimensions relevant in all types of dual or multiple relationship problems yet provides latitude for professional judgement. The model offered here is intended to fulfill those expectations and is best understood as an extension of Kitchener's (1988) work.

A Decision-Making Model

The model presented below offers three advantages over the existing models. First, it is designed specifically to address potential dual relationship ethical dilemmas. Second, it is neither so broad as to leave the professional without guidance, nor so specific as to 'dictate behavior. Third, the model encompasses all potential dual relationship problems that may arise, regardless of the professional context in which they occur.


Seven assumptions are required to use the model. First, the model is applicable to all professional relationships in which psychologists engage. It is not restricted to relationships with clients, students, or supervisees, and applies to anyone who is a consumer of, psychological services, regardless of the type of service provided. The model assumes that a psychologist's social role is that of a professional, regardless of the context in which the relationship with the consumer takes place.

Second, the aspirational goal of striving to avoid all dual relationships (APA, 1990) is unrealistic in many circumstances. This assumption is consistent with Kieth-Spiegel & Koocher (1985) and Haas & Malouf (1989), who agree that such relationships cannot be completely avoided. The assumption is also consistent with the APA Ethical Principles (APA, 1992) and the notion of overlapping relationships offered by the Feminist Therapy Institute's Code of Ethics (1987). The purpose of the model is to assist colleagues in managing these relationships more sensitively and effectively, when they cannot be avoided.

Third, due to the inherent high risk, all additional relationships with consumers should be evaluated to assess potential harm.

Fourth, the model assumes that not all dual relationships are exploitive per se. It presumes that in some circumstances, dual relationships may be pursued with low risk and may be beneficial; dual relationships must, however, be avoided whenever there is a reason to believe that they may prove harmful.

Fifth, the model is intended to sensitize the psychologist to the relevant issues, and make recommendations for action.

Sixth, the model assumes that the professional's dilemma is the result of contemplating the addition of a second relationship to an existing one. It is not intended for situations where multiple relationships are already in existence.

Seventh, the dimensions below must be assessed from the perspective of the consumer, not from that of the professional. Since the psychologist generally does not have access to the consumer's feelings in these situations, decisions must be made on the most conservative basis to insure consumer welfare.

The Model

The model is based upon the use of three dimensions (Gottlieb, 1986) which are believed to be basic and critical to the ethical decision-making process.

The first dimension is Power. It refers to the amount or degree of power which a psychologist may have in relation to a consumer. Power can vary widely. The psychologist who gives a speech at the local PTA on childrearing practices has relatively little power over members of the audience when compared with a therapist's influence over someone in long-term, insight-oriented psychotherapy.

Second, Duration of the Relationship, an aspect of power, is important because it is assumed that power increases over time. Power is lower when relationships are brief, such as in a single assessment session for referral, and increases as relationships continue, such as that of a student and teacher.

Third, Clarity of Termination refers to the likelihood that the consumer and the psychologist will have further professional contact. Performing a psychological assessment of a job applicant involves an unambiguous termination, with little chance of further contact. On the other hand, some family psychologists assume that their obligation to a family never ends. How does one decide when the professional relationship has been terminated? In this model one must assume that the professional relationship continues, as long as the consumer assumes that it does, regardless of the amount of time elapsed or contact in the interim. When the psychologist does not know the consumer's feelings, the ethical choice is to assume that the consumer always has the right to renew the professional relationship in the future.

Table 1 - Dimensions for Ethical Decision-Making
Low Power Mid-Range Power High Power
Little or no personal relationship
Persons consider each other peers (may include elements of influence).
Clear power differential present but relationship is circumscribed. Clear power differential with profound personal influence.
Brief Duration Intermediate Duration Long Duration
Single or few contacts over short period of time. Regular contact over a limited period of time. Continuous or episodic contact over a long period of time.
Specific Termination Uncertain Termination Indefinite Termination
Relationahip is limited by time externally imposed or by prior agreement of parties who are unlikely to see each other again. Professional function is completed but further contact is not ruled out. No agreement regarding when or if termination is to take place.

Using the Decision-Making Model

When a psychologist is contemplating an additional relationship, the model is to be used as follows. -

Step 1. Assess the current relationship according to the three dimensions. From the consumer's perspective, where does the relationship fall on each? How great is the power differential, how long has the relationship lasted, and has it clearly ended? If the relationship falls to the right side on two or three of the dimensions (i.e., higher power, longer duration and no termination), the potential for harm is high, and no other relationship should be contemplated.

For traditional individual psychotherapy, group, marital, and family therapy the case is clear. The power differential is great, the duration of treatment can be long, and termination is not always clear cut. Furthermore, the consumer may presume it their right to return for service at any time in the future. For example, families may assume they have a family psychologist, just as they have a family physician, who will always be available to them. In such cases, the presumption that the professional relationship never ends is appropriate.

If the relationship falls to the left side on the three dimensions (i.e., low power, short duration, and clearly terminated), one may move down to the next level. When the relationship falls at mid-range on the three dimensions, some types of additional relationships may be permissible, and the psychologist may move down to the next level.

Step 2. Examine the contemplated relationship along the three dimensions, as was done for the current relationship. If the contemplated relationship falls to the right side of the dimensions (i.e., it would involve great power over a long time with an uncertain termination), then it should be rejected if the existing relationship also falls to the right. If the proposed relationship falls either in the mid-range or to the left side of the dimensions, it may be permissible, and the psychologist should move down to step three.

For example, a psychologist might consider initiating a friendship with the parents of a child whom she once assisted through a painful medical procedure which need not be repeated. In this case, the psychologist had great power, but for only a brief and clearly-defined time period, and termination of the professional relationship was unambiguous. The new relationship, although of indeterminate duration and uncertain termination, involves little or no power differential.

Another possibility is that the first relationship may fall to the left side on the dimensions, but the contemplated relationship might fall to the right. In this case, the new relationship may be acceptable. For example, a psychologist could consider treating a child in psychotherapy after he gave a speech on children's reactions to divorce which the mother had attended.

Step 3. Examine both relationships for role incompatibility if they fall within the mid-range or to the left side of the dimensions. According to Kitchener (1988) role incompatibility increases as a function of greater differences in expectations of the two roles, greater divergence of the obligations of the two roles, and an increase in the power differential. If the two different roles are highly incompatible, then the contemplated relationship should be refused. For example, a psychologist should not accept an employee as a brief psychotherapy patient. Faculty members should not initiate business relationships with students.

If both relationships fall within the mid-range, or to the left side of the dimensions, and the incompatibility is low, the psychologist might proceed. For example, a faculty member may consider having a student in one of his or her classes work as a research assistant under his or her direction. A psychologist who has treated a man for smoking cessation may consider treating him and his wife for marital problems.

Step 4. Obtain consultation from a colleague. Consistent with the seventh assumption, the new relationship must be assessed from the standpoint of the consumer, and decisions should be made on the most conservative basis. Consultation with a colleague should be considered a routine matter when making such decisions. A colleague familiar with the circumstances, the consumer, and the decision-maker is the ideal choice. For example, a colleague might consider it inadvisable that a recently divorced, distressed, male internshiptraining supervisor accept a social invitation from one of his single female trainees.

Step 5. Discuss the decision with the consumer if the psychologist chooses to proceed with the additional relationship. He or she must review the essence of the decision-making model, its rationale, the pertinent ethical issues, available alternatives, and potential adverse consequences as a matter of informed consent. If the consumer is competent, and chooses to engage in the second relationship, the psychologist may proceed, once the consumer has had adequate time to consider the alternatives. If the consumer is unable to recognize the dilemma or is unwilling to consider the issues before deciding, he or she should be considered at risk, and the contemplated relationship rejected.

Case Illustrations
Case 1

Dr. X was a clinical psychologist in private practice. A single woman in her early twenties consulted him for career and adjustment issues. After working together for six months, the patient felt that the issues were resolved, the psychologist agreed, and treatment was terminated. Two years later, the psychologist attended a social gathering and coincidentally met his former patient. They had a lengthy conversation. Toward the end of the evening she asked the psychologist if he would be interested in establishing a friendship. He told her he would enjoy such a relationship, but noted that he was not free to do so because of their pre-existing professional one. In explaining the dilemma, he specifically mentioned the possibility that a social relationship would preclude any future professional consultation with him. She appeared to understand the issue, waived her right to consult him in the future, and agreed to accept a referral from him if she desired service in the future. They subsequently met for meals, but the frequency of their meetings decreased, and eventually contact ended. More than a year after their last social engagement, she called the psychologist requesting service. He declined, citing the conversation at the party and volunteered to refer her. She immediately became infuriated with him and hung up the telephone. There has been no contact subsequently.


Many would argue that Dr. X used good judgment. He was aware of the potential role conflict that might arise by engaging in the friendship with his former patient. He was even mindful of informed consent procedures in the midst of a social gathering. If all this is true, what went wrong?

The model suggests that Dr. X had a relationship with high power of intermediate duration and an apparently specific termination. The model also reveals great role incompatibility when therapists become friends with former psychotherapy clients. Also, Dr. X should have paid more attention to the client's need for this relationship. He accepted what she told him at face value, and did not fully examine the proposed relationship from her perspective. The model also prescribes a waiting period and consultation with a colleague. Had Dr. X followed the model, he might have reconsidered.

Case 2

Dr. Y, a tenured professor in a large psychology department, was having an informal conversation with a current graduate student, a female of similar age, who was leaving for her internship within the year. In the course of the conversation, Dr. Y mentioned missing having a man in her life; she had been widowed some years previously. Some weeks later the graduate student called Dr. Y at home, reminded her of their conversation, and offered to introduce her to a man whom she believed Dr. Y would find interesting. Dr. Y agreed to the meeting, but she consulted a trusted colleague the next day. As a result of the consultation, she called the student declining to meet the friend.


Some might feel that Dr. Y's decision was needlessly conservative. The graduate student was advanced in her studies, of similar age and doing her dissertation with another faculty member.

The model shows that the power differential was mid-range, of indefinite termination and pos sibly of long duration. Dr. Y realized that as long as the power differential continued, the role incompatibility would remain great. The consultation revealed additional information critical to her decision. Dr. Y realized that if she went out with the man and developed a relationship with him, she might feel indebted to the student and vulnerable to potential manipulation. Had the date gone poorly, she might have displaced hostile feelings onto a well-meaning student.


Avoiding exploitive dual relationships is an ethical principle that frequently generates dilemmas for psychologists. Ethical principles have provided general guidelines but little or no specific guidance when practical decision-making is needed. This article delineates the relevant dimensions one must consider when such decisions are required, and describes a decision-making model to assist psychologists in making these professional judgments. Although the model may help in numerous situations, some questions remain unanswered.

What should be done about social relationships with former psychotherapy clients? Even if the treatment course is short and termination is clear, it should be assumed that a power differential continues (especially if the former client reserves the right to return for further treatment), making such relationships very inadvisable. Other social contact may be less problematic. For example, one certainly could consider accepting an invitation to the bar mitzvah of a former adolescent psychotherapy patient, if the issues were discussed in the manner described above.

Another most nettlesome situation is faced by academics and trainers who have intermediate to long-term, personal relationships with students and trainees. Initially, the power differential is great, and relationships may continue for years, often developing into peer, colleagial, platonic or romantic ones. In this situation, psychologists must remain mindful of the assumption noted above that the dimensions must be assessed from the viewpoint of the consumer. It is not enough that the formal professional relationship is at an end.


The decision-making model has not been empirically verified as yet. Therefore, if one chooses to use the model at this time, the most sensitive professional judgement will be required.

Second, most dual relationships continue to be precluded. The model is designed to allow dual relationships only under the most specific situations and only after the most careful consideration.

Finally, it is assumed that the model is being used by a competent professional who is able to make rational decisions independently of his or her own needs. The author realizes that such an assumption is not always realistic, and that psychologists' feelings are frequently compromised. For this reason, consultation has been included at every step in the decision-making process. There is still no substitute for a trusted colleague.



[Back to Top]