Power Dynamics in Therapy: Authority, Expertise, Boundaries, and Therapeutic Alliance
Author: Ukrainian Psychological Hub · Published: September 24, 2026 · Editorial Policy
Power dynamics in therapy are the patterns through which therapist and client shape decisions, interpretations, goals, disclosure, boundaries, and the conditions of treatment. The relationship is structurally asymmetrical: the therapist occupies a professional role, has specialized training, controls parts of the therapeutic frame, and may hold institutional or legal responsibilities that the client does not. Yet therapy also depends on the client’s consent, knowledge of their own life, participation, feedback, and continuing willingness to work toward agreed goals.
That asymmetry is not the same thing as domination. A therapist can have legitimate professional authority without claiming authority over the client’s identity, values, memories, relationships, or life choices. The central psychological question is therefore not whether power exists in therapy. It is how professional power is bounded, explained, shared where possible, challenged when necessary, and used in a way that strengthens rather than replaces the client’s agency.
This article owns the therapy-specific intent. For the broader relational construct, see Power Dynamics: Psychology of Influence, Dependence, Status, and Control. For the psychology of legitimate authority across settings, see Authority in Psychology: Legitimacy, Expertise, Trust, and Obedience.
What Are Power Dynamics in Therapy?
In psychotherapy, power dynamics can be defined as the changing distribution and use of influence, authority, expertise, choice, dependency, and control within the therapeutic relationship. Some of this power comes from the formal role. Some comes from knowledge. Some comes from the therapist’s access to records, referrals, diagnoses, institutional systems, or decisions about the therapeutic frame. Some comes from the client’s dependence on the relationship for help, validation, documentation, or access to other services.
At the same time, clients are not passive recipients of professional power. They decide what to disclose, what goals matter, whether an interpretation fits, whether a task is acceptable, whether to continue treatment, and—outside settings where choice is constrained—whether to consent at all. Even in highly asymmetric situations, power is enacted through interaction rather than exhausted by job titles.
This is why power in therapy should be separated from several neighboring concepts. Social power concerns the capacity to affect another person’s outcomes or options. Authority concerns a recognized right to direct or decide within a domain. Expertise concerns knowledge and skill. Influence is the process through which attitudes or behavior change. Personal control concerns influence over one’s own circumstances. These phenomena can overlap in therapy, but they are not interchangeable.
Why Therapy Is Structurally Asymmetrical
Psychotherapy is a professional helping relationship rather than a relationship between interchangeable peers. The therapist is paid or institutionally assigned to occupy a role with specific competencies, ethical duties, record-keeping obligations, confidentiality rules, and boundaries. The client arrives because the therapist is expected to know something about psychological assessment, change processes, clinical risk, treatment methods, or how to organize a therapeutic conversation.
That creates legitimate and expert forms of influence. A client may give considerable weight to a therapist’s interpretation because the therapist is perceived as knowledgeable. The therapist may also determine practical conditions such as session length, cancellation policy, communication outside sessions, and what kinds of services the practice can provide. In some systems, therapists or related clinicians may contribute to diagnosis, treatment planning, disability documentation, referrals, or risk-management decisions. These functions increase the practical consequences of what the professional says and does.
The asymmetry is intensified when the client has few alternatives, is in acute distress, depends on a service for access to other resources, or is receiving treatment under institutional or legal pressure. Dependence is a major source of social power generally, and it matters in therapy because the cost of disagreeing, leaving, or seeking another provider can vary dramatically across contexts. The broad psychology of those relational mechanisms is covered in Power Dynamics.
Yet professional asymmetry does not justify unlimited authority. A therapist’s expertise is domain-specific. Knowing psychotherapy does not make the therapist a superior judge of every moral, cultural, relational, occupational, political, or existential choice in the client’s life. Professional authority gains legitimacy through competence, transparency, ethical limits, accountability, and a treatment purpose—not from a general right to govern the person.
Authority, Expertise, Influence, and Control Are Different
Professional authority
Professional authority is role-based. It concerns what the therapist is authorized to do within a legitimate therapeutic role: offer psychological treatment, maintain a clinical frame, apply methods within competence, set practice policies, document care, and meet ethical or legal duties. The client can recognize that authority without surrendering general self-determination.
Expertise
Expertise is knowledge and skill. It can make advice more credible and can create expert power, but expertise is not infallibility. A clinically responsible use of expertise includes explaining the basis of recommendations, acknowledging uncertainty, distinguishing evidence from interpretation, and revising a formulation when new information changes the picture.
Influence
Influence is what happens when one person’s communication, behavior, or position changes another person’s thinking or action. Therapy deliberately involves influence: therapists ask questions, reframe events, invite experiments, challenge patterns, model skills, and direct attention. Ethical therapy does not eliminate influence. It makes the purpose, boundaries, and accountability of influence visible.
Control
Control is stronger. It concerns determining options, restricting alternatives, or directing outcomes. Some control is built into any professional service—the therapist controls the office schedule, for example—but control over the client’s broader life requires a separate justification and is usually outside the ordinary therapeutic role. For the distinction between power over others and control over one’s own life, see Power and Control: Why Power Over Others Differs From Personal Control.
The distinction also clarifies why power and authority should not be treated as synonyms. A therapist may have authority to maintain a boundary without having the right to dictate a client’s private values. A client may accept a therapist’s expertise while rejecting a particular interpretation. A therapist may influence a client without coercing them.
The Therapeutic Alliance: Collaboration Across Unequal Roles
The therapeutic alliance is one of the strongest research traditions for understanding how a professional relationship can remain collaborative despite unequal roles. Edward Bordin’s influential formulation described the working alliance in terms of agreement on goals, agreement on therapeutic tasks, and the development of a bond (Bordin, 1979). Later scholarship has retained this three-part structure as a widely used way to conceptualize alliance (Wampold & Flückiger, 2023).
A large meta-analysis by Flückiger and colleagues synthesized 295 independent studies involving more than 30,000 patients. In face-to-face psychotherapy, alliance quality was associated with outcome at about r = .28 (Flückiger et al., 2018). This is a robust association across a large literature. It should still be described as an association rather than as proof that alliance alone causes improvement; better early progress can also strengthen alliance, and client and therapist characteristics contribute to both.
The collaborative elements of alliance show similar patterns. A meta-analysis of 54 studies found goal consensus associated with outcome at r = .24, while 53 studies found patient–therapist collaboration associated with outcome at r = .29. The authors explicitly noted limitations that prevent simple causal conclusions (Tryon, Birch, & Verkuilen, 2018).
These findings matter for power because alliance is not the abolition of professional difference. A therapist can remain trained, responsible, and directive where a method requires direction while still treating goals and tasks as matters that require meaningful client participation. Healthy collaboration does not require pretending that therapist and client occupy identical roles.
Healthy Power in Therapy
Healthy therapeutic power is accountable power. It is exercised for a defined professional purpose, constrained by consent and ethics, open to questioning, and proportionate to the therapist’s actual competence. The therapist can say, “Here is what I recommend and why,” while leaving room for the client to answer, “That does not fit,” “I am not ready,” or “I want another option.”
A healthy relationship therefore makes several things easier rather than harder: asking why an intervention is being recommended; discussing alternatives; correcting the therapist; expressing discomfort; revisiting goals; naming cultural or relational differences; requesting clarification about confidentiality; declining a specific exercise; seeking a second opinion; and discussing ending therapy.
This does not mean every preference can always be followed. A therapist may decline to provide an intervention outside their competence, refuse an ethically inappropriate request, maintain a safety-related duty required by law, or end a service that is no longer clinically appropriate. The important distinction is whether limits are explained as professional limits and negotiated as transparently as the situation allows, rather than converted into personal dominance.
Empathy also matters, but it should not be romanticized as a substitute for accountability. A therapist can be warm and still misuse authority, while a firm boundary can be clinically responsible. The relationship between power and accurate attention to other people is examined more broadly in Power and Empathy: Perspective Taking, Social Distance, and Emotional Accuracy.
Informed Consent Makes Professional Power More Accountable
Informed consent is one of the clearest institutional mechanisms for limiting therapeutic power. It turns important parts of the treatment frame into information the client can evaluate rather than rules discovered only after conflict occurs. The American Psychological Association’s ethics materials, for example, require psychologists to address the nature and anticipated course of therapy, fees, third-party involvement, and limits of confidentiality early enough for the client to ask questions (APA Ethics Code).
APA’s public guidance also identifies practical topics such as record keeping, the psychologist’s expertise and training, services the psychologist cannot or will not provide, estimated duration, alternatives, fees, and the client’s right to terminate treatment (American Psychological Association, Potential ethical violations). These examples are profession- and jurisdiction-specific, but the underlying function is general: clients should understand what relationship they are entering and what important limits apply.
Consent is not a one-time signature that neutralizes every later power issue. Treatment changes. New interventions are proposed. Confidentiality questions arise. Goals shift. A client who initially agreed may later withdraw from a particular task or want to reconsider the treatment plan. Consent works best as an ongoing communicative process in which changes that materially affect participation are made explicit.
Boundaries: Structure Without Domination
Therapeutic boundaries define the professional frame: what kind of relationship this is, what it is for, what belongs inside it, and what the therapist will not turn it into. Boundaries can involve time, contact, self-disclosure, physical touch, gifts, financial arrangements, social media, dual roles, confidentiality, emergency communication, and relationships outside therapy.
A 2026 systematic review and meta-synthesis of 25 qualitative studies found that therapists described boundary work as an active relational process rather than a simple checklist. Themes included contracting versus intuitive boundary management, rigidity versus flexibility, different forms of self-disclosure, therapist unease, self-restraint, caretaking versus bearing witness, and collaboratively analyzing boundary dilemmas. The authors also reported variability in methodological quality, so these findings describe patterns of professional experience rather than proving which boundary style produces the best outcomes (Lawson, Schröder, & Russell, 2026).
The useful distinction is between a boundary and a controlling demand. A boundary defines what the therapist will do, what the service can provide, or what conditions are required for the professional relationship. Control attempts to govern the client’s independent behavior beyond what the therapeutic role legitimately requires. This distinction is developed in Boundaries vs Control: What Is the Difference in Relationships?.
Professional boundaries can sometimes feel disappointing, especially in a relationship involving trust, attachment, and disclosure. That feeling does not by itself show that a boundary is harmful. The stronger question is whether the limit has a coherent professional rationale, is applied consistently enough to be understandable, respects the client’s dignity, and avoids using dependency for the therapist’s personal, financial, sexual, social, or ideological benefit.
Multiple Relationships and Boundary Crossings
Ethics codes often distinguish the existence of an additional relationship from the risk that it creates. APA guidance on multiple relationships emphasizes that an overlapping role is not automatically unethical; the central concern is whether it could reasonably impair the psychologist’s objectivity, competence, or effectiveness, or create a risk of exploitation or harm (American Psychological Association, 2025).
That nuance matters in small communities, specialized populations, training settings, and online life, where complete separation of social worlds may be unrealistic. It also prevents a simplistic rule in the opposite direction: because some boundary crossings can be benign, a therapist does not gain permission to ignore foreseeable conflicts of interest or exploitation.
Sexual relationships with current therapy clients occupy a different ethical category and are prohibited in major professional codes. Financial entanglements, business partnerships, secret social relationships, or arrangements that make the client responsible for meeting the therapist’s personal needs likewise require particular scrutiny because the therapeutic role already gives the professional privileged access to vulnerability, trust, and influence.
When Therapeutic Power Becomes Harmful
Therapeutic power becomes harmful when professional authority is converted into domination, exploitation, retaliation, or control that serves the therapist rather than the therapeutic purpose. The problem can be obvious, as in sexual exploitation or financial manipulation, or subtle, as when disagreement is repeatedly reinterpreted as evidence that the client is irrational, resistant, disordered, or incapable of understanding their own experience.
A pattern is more concerning when the client’s ability to question, leave, compare alternatives, or seek outside support is systematically narrowed. Examples include discouraging second opinions without a sound clinical reason, treating all criticism as pathology, threatening abandonment to obtain compliance, making the client responsible for the therapist’s emotions, pressuring the client into ideological or personal beliefs unrelated to treatment goals, or using confidential knowledge to gain leverage.
Such conduct should be distinguished from ordinary therapeutic challenge. Effective therapy can include disagreement, exposure to avoided situations, behavioral commitments, discussion of painful patterns, or limits that the client dislikes. The difference lies in purpose, transparency, proportionality, evidence, consent, and accountability—not in whether therapy always feels comfortable.
It is also useful to distinguish harmful therapeutic power from psychological manipulation and coercive control. Those concepts have their own canonical intents. In therapy, they become relevant when influence relies on deception, intimidation, restriction, dependency, or punishment rather than transparent professional reasoning.
Disagreement Is Not the Same as Therapeutic Failure
A strong alliance does not mean permanent harmony. Clients and therapists can disagree about goals, interpretations, pace, diagnosis, homework, what happened in a session, or whether an intervention is helping. Alliance research treats such disruptions as potentially meaningful rather than as proof that either party has failed.
An alliance rupture is commonly described as deterioration in collaboration on goals or tasks, or strain in the emotional bond. Eubanks, Muran, and Safran synthesized 11 studies involving 1,314 patients and found that successful rupture resolution was associated with better treatment outcome at about r = .29. Their separate meta-analysis of rupture-repair training or supervision did not find a statistically significant overall outcome effect, underscoring that the evidence should not be reduced to a simple technique guarantee (Eubanks, Muran, & Safran, 2018).
From a power perspective, rupture repair is important because it tests whether the client can disagree without the therapist using role authority to win. Repair may involve acknowledging a misunderstanding, inviting the client’s account, clarifying intent, apologizing when appropriate, changing the plan, or deciding together that the therapeutic fit is poor. A therapist who can revise their own formulation demonstrates that expertise is compatible with corrigibility.
Shared Decision-Making: Collaboration Without Pretending Expertise Disappears
Shared decision-making offers another framework for distributing choice. In mental health care, it generally means that clinician and service user exchange information, discuss options, and incorporate the person’s preferences and values into decisions. A 2022 Cochrane review of 15 randomized trials found that shared decision-making interventions may increase people’s perceived involvement immediately after encounters, while much of the evidence for broader effects remained low or very low certainty (Aoki et al., 2022).
An umbrella review published in 2023 similarly found a heterogeneous evidence base and noted that much mental-health shared decision-making research had centered on psychopharmacological decisions rather than psychotherapy itself (Chmielowska, Zisman-Ilani, Saunders, & Pilling, 2023). It is therefore reasonable to apply the collaborative principle to therapy while remaining careful about claiming outcome effects that have not been established for every psychotherapeutic context.
Shared decision-making does not mean that therapist and client must contribute the same kind of knowledge. The therapist can contribute evidence, clinical reasoning, experience with interventions, and awareness of risk. The client contributes goals, preferences, values, context, tolerability, lived experience, and knowledge of what is feasible outside the consulting room. The decision becomes stronger when neither knowledge source is treated as disposable.
Client Preferences Are Part of the Power Structure
Treatment preferences matter because choice affects whether clients remain engaged and whether treatment fits their expectations and needs. A meta-analysis of 35 studies found that adults matched to their preferred therapy conditions were less likely to drop out prematurely (odds ratio .59) and had better outcomes on average (d = .31) than those who were not matched (Swift, Callahan, & Vollmer, 2011).
Preference-sensitive care does not require offering every possible intervention. Availability, evidence, competence, cost, risk, and setting all constrain choices. But when more than one reasonable option exists, treating the client’s preference as clinically relevant information reduces the temptation to equate professional expertise with unilateral decision authority.
Cultural Power, Social Position, and Therapeutic Humility
Therapy does not occur outside social life. Race, ethnicity, language, gender, sexuality, disability, class, age, migration status, religion, professional status, and other social positions can affect how authority is perceived and how safe disagreement feels. These factors do not mechanically determine the relationship, but they can change what a therapist’s certainty, silence, curiosity, or misunderstanding means to a particular client.
A 2025 meta-analysis of 13 studies found therapist cultural humility positively correlated with therapeutic alliance (r = .66) and psychotherapy outcomes (r = .39). These are sizable associations, but the evidence is correlational and should not be read as proof that cultural humility alone causes those outcomes (Orlowski, Moeyaert, Monley, & Redden, 2025).
Recent psychotherapy scholarship has also emphasized training clinicians to recognize power, privilege, and intersectionality in the therapeutic relationship rather than assuming that a neutral professional role erases social location (Wright et al., 2025). The practical implication is not that every session must become a discussion of social power. It is that therapists need enough awareness to notice when social difference is shaping trust, interpretation, disclosure, or the client’s willingness to challenge them.
Cultural humility is especially relevant to expert power. A therapist can know a great deal about psychopathology or treatment and still know little about what a particular cultural practice, family obligation, identity, or historical experience means to the person in front of them. Responsible expertise therefore includes disciplined curiosity.
Diagnosis and the Power to Define
Diagnosis can organize care, support communication, guide evidence-based treatment, and provide access to services. It can also carry institutional consequences and shape how later professionals interpret a person’s behavior. That gives diagnostic language a form of practical and interpretive power.
Good clinical use of diagnosis keeps several distinctions clear. A diagnostic formulation is a professional judgment made under defined criteria and with uncertainty that varies by case. It is not a total description of the person. A client’s disagreement with a diagnosis does not automatically validate or invalidate it, but disagreement is clinically relevant and can reveal missing information, different explanatory models, or a rupture in trust.
The same principle applies to interpretations made within therapy. An interpretation should function as a testable contribution to understanding, not as a rhetorical device that makes the therapist impossible to contradict. If every objection becomes evidence for the interpretation, the therapist’s theory has become insulated from feedback.
Confidentiality, Records, and Information Asymmetry
Confidentiality is central to the therapeutic frame because clients disclose information they might not share elsewhere. But confidentiality has limits, and those limits can vary with law, jurisdiction, setting, age, third-party payment, court involvement, and risk circumstances. The therapist often knows these rules better than the client, which creates an information asymmetry.
That asymmetry is best managed through clear explanation before the information becomes consequential. Clients should know, as applicable, who can access records, what billing or insurance processes disclose, what the limits of confidentiality are, how electronic communication is handled, and how information may be used in institutional or legal contexts. APA’s public ethics guidance explicitly emphasizes explaining confidentiality limits and record-keeping practices (American Psychological Association).
Transparency here is more than administrative courtesy. It changes the client’s ability to make informed choices about disclosure. A person cannot meaningfully decide what to reveal if they misunderstand who may later receive the information.
Mandated, Court-Ordered, and Institutional Therapy
The power structure changes when therapy is not fully voluntary. Court orders, probation conditions, school requirements, employer programs, child-protection systems, inpatient settings, or other institutions can limit the client’s practical ability to refuse, leave, or control who receives information.
In such settings, the therapist should not simulate a degree of voluntariness that does not exist. The more constrained the client’s choice, the more important role clarity becomes: who required the contact, who the client is, what information may be shared, what decisions the therapist can affect, what choices remain available, and what consequences follow from participation or nonparticipation.
External constraint does not make alliance impossible. It does, however, change what collaboration means. A client can still have preferences, goals, boundaries, and meaningful choices inside a mandated framework even when the larger decision to attend is not theirs.
Dependency: Therapeutic Resource and Therapeutic Risk
Therapy often becomes important precisely because the client experiences the relationship as reliable, attentive, and useful. Trust and emotional significance are not signs that therapy has gone wrong. The risk begins when ordinary therapeutic dependence is deliberately intensified so that the therapist becomes harder to question or leave.
A therapist should not need the client to remain dependent in order to preserve status, income, admiration, or emotional gratification. Treatment goals should generally move toward greater capacity to function, choose, relate, regulate, or seek support—not toward making the professional relationship indispensable for decisions the client can increasingly make elsewhere.
Termination can therefore carry power dynamics of its own. Ending therapy may be clinically appropriate when goals have been met, treatment is not helping, another service is needed, or the fit is poor. It can also be experienced as rejection. Clear discussion of reasons, options, and continuity of care helps keep termination a professional transition rather than a punishment for disagreement.
Directive Therapy Is Not Automatically Authoritarian
Some therapies are more structured than others. Exposure-based treatments, behavioral activation, skills training, trauma-focused protocols, and some cognitive-behavioral methods may include explicit exercises, planned tasks, or a strong rationale for sequencing interventions. Structure can be therapeutically useful.
Directive method and authoritarian relationship are different constructs. A therapist can strongly recommend an evidence-based task while explaining its rationale, checking consent, adapting pace, monitoring effects, and inviting feedback. Conversely, a nondirective style can still conceal power if the therapist subtly rewards agreement, withholds important information, or treats their own interpretations as unquestionable.
The quality of power use is therefore not captured by a simple directive-versus-nondirective scale. More informative questions concern whether the intervention is within competence, linked to agreed goals, supported by evidence or a clear rationale, proportionate to risk, open to revision, and conducted within ethical boundaries.
How Different Therapy Models Organize Power
Cognitive and behavioral therapies
Structured cognitive and behavioral therapies often make tasks, goals, monitoring, and between-session practice explicit. This can reduce ambiguity about what the therapist is asking, but structure can become coercive if adherence is treated as obedience rather than as collaborative experimentation. A formulation works best as a shared model that can be tested and revised.
Psychodynamic and psychoanalytic therapies
Psychodynamic approaches may give particular importance to transference, unconscious conflict, defenses, and the meaning of what happens inside the therapeutic relationship. This creates a distinctive interpretive authority: the therapist may propose meanings that are not immediately obvious to the client. Such interpretations become safer when uncertainty remains visible and disagreement can itself be explored without being automatically pathologized.
Person-centered and humanistic therapies
Person-centered traditions deliberately reduce therapist directiveness and emphasize empathy, congruence, and unconditional positive regard. That can redistribute conversational authority, but the professional role does not disappear. The therapist still controls aspects of the frame, holds confidential information, and exercises judgment about competence and ethics.
Systemic, couple, and family therapies
When several people participate, power becomes more complex because the therapist must consider influence not only between therapist and clients but among family members or partners. Role clarity—who is the client, what confidentiality means, how information is handled, and whether the therapist is being pulled into an alliance with one person against another—becomes especially important. This article does not absorb the separate relationship-power intents reserved elsewhere in the Psychology of Power cluster.
Therapist Self-Disclosure and the Use of Personal Presence
Therapist self-disclosure can humanize the relationship, clarify cultural position, normalize experience, or support a specific therapeutic aim. It can also shift attention toward the therapist, burden the client, invite caretaking, or blur roles. The power question is not whether disclosure is always good or bad, but whose need it serves and what it does to the client’s freedom inside the relationship.
The 2026 boundary meta-synthesis found that therapists described different forms of self-disclosure as part of boundary negotiation rather than as a single uniform practice (Lawson, Schröder, & Russell, 2026). A useful standard is functional: disclosure should have a defensible therapeutic purpose, be proportionate, and leave the client free from pressure to reciprocate or care for the therapist.
Feedback Is a Power-Sharing Mechanism
Feedback changes the informational structure of therapy. Without it, the therapist may rely heavily on their own interpretation of whether the client feels understood, whether the pace is tolerable, whether goals remain relevant, or whether treatment is helping. Explicitly inviting feedback gives the client a recognized channel for influencing the process.
The invitation must be credible. A therapist who asks, “Was anything unhelpful today?” but becomes defensive whenever the client answers honestly teaches the client that feedback is ceremonial. A therapist who can receive criticism, clarify misunderstandings, and alter the plan where appropriate makes collaboration behaviorally real.
This is one reason rupture-repair research matters: the relationship can become stronger not by eliminating disagreement, but by making disagreement discussable. The therapist’s response to criticism is itself evidence about how power is functioning in the room.
Signs of Constructive Power Dynamics
Constructive power dynamics usually have a recognizable pattern. The therapist’s role and limits are clear. Recommendations come with reasons. Uncertainty can be admitted. The client can ask questions without being shamed. Goals can be revised. Cultural differences can be named. Boundaries are understandable. Disagreement does not automatically become pathology. The therapist distinguishes clinical expertise from personal preference.
Constructive dynamics also preserve alternatives. The client can discuss referrals, second opinions, changes in treatment, or termination. When an option is unavailable, the reason is transparent. The therapist does not monopolize interpretation by implying that only one account of the client’s experience is psychologically legitimate.
Finally, constructive power tends to expand the client’s practical agency over time. The client may still value the therapist deeply, but treatment supports greater capacity to make choices, tolerate uncertainty, set boundaries, understand patterns, and use support systems beyond the therapy room.
Warning Signs That Power May Be Misused
No single uncomfortable moment proves misuse. Patterns matter. Concern increases when a therapist repeatedly discourages questions, claims special insight that cannot be challenged, pressures the client to adopt the therapist’s personal values, uses diagnosis as a threat, retaliates after criticism, demands secrecy about the therapeutic relationship, cultivates exclusivity, or frames all attempts to leave as evidence that the client must stay.
Other warning signs include unclear financial arrangements, boundary changes that primarily benefit the therapist, inappropriate gifts or business proposals, sexualized behavior, unnecessary physical intimacy, disclosure that makes the client responsible for the therapist’s emotions, or using confidential material to gain leverage. Professional codes differ, but exploitation of the therapeutic role is a central ethical concern across the field.
A difficult therapist is not necessarily an unethical therapist, and a warm therapist is not necessarily a safe therapist. The more reliable test is whether power is being used transparently for a defensible therapeutic purpose under meaningful ethical constraints.
What Clients Can Do When the Power Dynamic Feels Wrong
If something feels controlling, confusing, or unsafe, one useful first step—when it is safe and realistic—is to make the process itself a topic of therapy. Questions such as “What is the clinical reason for this?”, “What alternatives do I have?”, “What happens if I disagree?”, “Is this a requirement or a recommendation?”, and “What are the limits of confidentiality here?” turn implicit power into something that can be examined.
Clients can also ask for the informed-consent document, clarification of fees and records, the therapist’s credentials or supervision status, an explanation of the treatment model, or a referral for a second opinion. If the concern involves serious misconduct, exploitation, sexual behavior, threats, or misuse of confidential information, consultation with an independent licensed professional, relevant clinic, professional body, regulator, or legal resource may be appropriate depending on jurisdiction.
Leaving therapy is also a legitimate option in many outpatient contexts. There are exceptions where participation is legally or institutionally mandated, and abrupt termination can carry practical or clinical risks in some situations. The key is to understand what constraints actually apply rather than assuming that a therapist’s preference is itself a binding rule.
What Therapists Can Do to Use Power Responsibly
Responsible use of therapeutic power begins by recognizing that professional influence exists even when the therapist feels collaborative. The client may attribute more certainty, authority, or moral weight to a therapist’s words than the therapist intends. Naming uncertainty and separating recommendation from requirement can therefore matter.
Therapists can make power more accountable by explaining rationales, revisiting informed consent, inviting dissent, asking whose goals are being pursued, distinguishing evidence from personal opinion, checking whether the client feels pressured, monitoring dependency, clarifying role conflicts, and seeking consultation or supervision when boundaries become complicated.
They can also watch for a subtle professional temptation: interpreting discomfort with the therapist as a problem located entirely inside the client. Sometimes client criticism reflects transference, avoidance, fear, or a recurring interpersonal pattern. Sometimes the therapist was unclear, insensitive, defensive, culturally uninformed, or simply wrong. Good clinical reasoning keeps both possibilities available.
What the Evidence Supports—and What It Does Not
Several parts of this topic have relatively strong empirical support. The alliance–outcome association has been replicated across a large psychotherapy literature, and goal consensus and collaboration are consistently associated with outcomes. Client preference matching also has meta-analytic support for lower dropout and modestly better outcomes.
Other parts of the topic have a different evidential status. Research on how therapists negotiate boundaries is currently dominated by qualitative studies; the 2026 systematic review synthesizes professional experiences and processes rather than establishing a universally optimal boundary style. Cultural humility is associated with alliance and outcome, but much of that evidence is correlational. Shared decision-making interventions in mental health show promise for perceived involvement, while evidence for broader clinical outcomes remains uncertain and is often drawn from psychiatric medication decisions rather than psychotherapy.
The broad claim that therapy contains a power asymmetry is a conceptual and ethical description of the professional relationship, not a single experimentally verified effect size. How much power a therapist has varies with setting, legal authority, scarcity of alternatives, client vulnerability, professional role, institutional rules, and the client’s ability to refuse or leave.
That distinction matters. Evidence does not support a simple formula in which more therapist power is always harmful or less therapist power is always better. Psychotherapy requires expertise, structure, and influence. The more defensible conclusion is that professional power works best when it is proportionate, transparent, bounded, collaborative, and corrigible.
A Practical Model: Bounded, Transparent, Collaborative Authority
The healthiest therapy relationship can be summarized as bounded, transparent, collaborative authority. Bounded means the therapist’s authority applies to a professional role with ethical limits rather than to the client’s entire life. Transparent means recommendations, rules, confidentiality limits, conflicts, and uncertainty are explained rather than hidden. Collaborative means goals and tasks are negotiated to the extent the setting allows and feedback can change the work.
Authority remains real in this model. Expertise remains real. Boundaries remain real. What changes is the relation between professional power and client agency. The therapist uses their position to make effective treatment possible while preserving the client as an active participant whose values, choices, interpretations, and consent matter.
This is also why healthy therapeutic power is compatible with strong boundaries. A therapist who says no to a dual relationship may be protecting the client from exploitation. A therapist who says yes to a client’s disagreement may be protecting the alliance from authoritarianism. The psychological task is to know which kind of limit serves treatment and which kind of control serves the professional.
Frequently Asked Questions
Are power dynamics inevitable in therapy?
Yes, in the ordinary descriptive sense that therapist and client occupy different roles with different forms of knowledge, responsibility, influence, and control over the therapeutic frame. The degree and meaning of that asymmetry vary greatly. The goal is not to pretend power is absent but to make its use ethical, transparent, and collaborative.
Does the therapist always have more power than the client?
The therapist usually has greater professional and institutional authority inside the therapeutic role, but power is multidimensional. Clients retain important forms of agency: they provide or withhold information, evaluate fit, shape goals, give feedback, accept or decline many interventions, and in most voluntary outpatient settings can leave or seek another provider. Mandated and institutional contexts can substantially narrow those options.
Is therapist expertise the same as authority?
No. Expertise is knowledge and skill; authority is a recognized right to act or decide within a role. A therapist may be highly expert without having authority over a client’s personal values or life choices. Conversely, a professional role may carry formal authority even when the therapist’s expertise is limited to a particular domain.
Can I disagree with my therapist?
In ordinary psychotherapy, disagreement is compatible with treatment and can be clinically useful. Goals, tasks, interpretations, pace, and fit can all be discussed. A therapist may still maintain professional limits or strongly recommend an intervention, but disagreement should not automatically be treated as disobedience.
What is a therapeutic alliance rupture?
A rupture is a deterioration in collaboration or bond, such as disagreement about goals, tension over therapeutic tasks, withdrawal, confrontation, or a felt loss of trust. Ruptures are common enough to be a major research area. Their significance depends partly on whether they can be recognized and repaired.
Are strict boundaries always healthy?
No single degree of strictness is universally optimal. Some boundaries are essential; others require contextual judgment. The 2026 systematic review of boundary negotiation found themes of both rigidity and flexibility and emphasized the relational process of managing dilemmas. Healthy boundaries protect the purpose and safety of therapy without becoming arbitrary instruments of control.
What if my therapist says my wish to leave is part of my disorder?
That may sometimes be a clinically relevant hypothesis, but it should remain discussable rather than functioning as a trap in which every attempt to leave proves that leaving is impossible. Ask what evidence supports the interpretation, what alternatives exist, and whether a second opinion or referral is available. Legal or institutional constraints should be distinguished from a therapist’s recommendation.
Can a therapist be both caring and powerful?
Yes. Care, expertise, authority, and influence can coexist. The important question is how power is used. Warmth does not erase structural asymmetry, and authority does not erase the possibility of empathy or collaboration.
When should I seek outside help about a therapist’s conduct?
Independent consultation is especially reasonable when there are concerns about sexual behavior, exploitation, threats, major confidentiality breaches, financial entanglement, retaliation, coercion, or boundary changes that appear to serve the therapist rather than treatment. The appropriate professional body, regulator, clinic, or legal channel depends on the therapist’s profession and jurisdiction.
Related Articles
References
American Psychological Association. (n.d.). Ethical principles of psychologists and code of conduct. https://www.apa.org/ethics/code
American Psychological Association. (2010). Potential ethical violations. https://www.apa.org/topics/psychotherapy/potential-violations
American Psychological Association. (2025). A former patient wants to join my peer consultation group. What should I do? https://www.apa.org/ethics/resources/asked-answered/multiple-relationships
Aoki, Y., Yaju, Y., Utsumi, T., Sanyaolu, L., Storm, M., Takaesu, Y., Watanabe, K., Watanabe, N., Duncan, E., & Edwards, A. G. K. (2022). Shared decision-making interventions for people with mental health conditions. Cochrane Database of Systematic Reviews, 11, CD007297. https://doi.org/10.1002/14651858.CD007297.pub3
Bordin, E. S. (1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research & Practice, 16(3), 252–260. https://doi.org/10.1037/h0085885
Chmielowska, M., Zisman-Ilani, Y., Saunders, R., & Pilling, S. (2023). Trends, challenges, and priorities for shared decision making in mental health: The first umbrella review. International Journal of Social Psychiatry, 69(4), 823–840. https://doi.org/10.1177/00207640221140291
Eubanks, C. F., Muran, J. C., & Safran, J. D. (2018). Alliance rupture repair: A meta-analysis. Psychotherapy, 55(4), 508–519. https://doi.org/10.1037/pst0000185
Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy, 55(4), 316–340. https://doi.org/10.1037/pst0000172
Lawson, J., Schröder, T., & Russell, K. (2026). How do therapists negotiate boundaries: A systematic literature review and meta-synthesis. Journal of Contemporary Psychotherapy. https://doi.org/10.1007/s10879-026-09750-9
Orlowski, E. W., Moeyaert, M., Monley, C., & Redden, C. (2025). The effects of cultural humility on therapeutic alliance and psychotherapy outcomes: A systematic review and meta-analysis. Counselling and Psychotherapy Research, 25(2), e12835. https://doi.org/10.1002/capr.12835
Swift, J. K., Callahan, J. L., & Vollmer, B. M. (2011). Preferences. Journal of Clinical Psychology, 67(2), 155–165. https://doi.org/10.1002/jclp.20759
Tryon, G. S., Birch, S. E., & Verkuilen, J. (2018). Meta-analyses of the relation of goal consensus and collaboration to psychotherapy outcome. Psychotherapy, 55(4), 372–383. https://doi.org/10.1037/pst0000170
Wampold, B. E., & Flückiger, C. (2023). The alliance in mental health care: Conceptualization, evidence and clinical applications. World Psychiatry, 22(1), 25–41. https://doi.org/10.1002/wps.21035
Wright, A. J., Bergkamp, J., Williams, N., Garcia-Lavin, B., & Reynolds, A. L. (2025). Privilege in the room: Training future psychologists to work with power, privilege, and intersectionality within the therapeutic relationship. Psychotherapy, 62(1), 82–89. https://doi.org/10.1037/pst0000563