self_improvementPsychotherapy

Therapeutic Alliance

The agreement and the bond between the person seeking help and the person helping — one of the most consistent findings in psychotherapy research.

Translated from the Portuguese original. Read the original (Português)

boltIn short

The therapeutic alliance (or working alliance) is the quality of the partnership between client and therapist. It has three parts: agreeing on where to go (the goals), agreeing on how to get there (the tasks), and the bond of trust between the two. Decades of research show that stronger alliances go hand in hand with better outcomes — across any therapeutic approach. It is not, however, just “being nice,” nor does it replace an effective technique: it is the foundation on which the technique works.

Technical definition

The therapeutic alliance is the collaborative quality of the relationship between client and therapist. In the pan-theoretical model of Edward Bordin (1979), it is made up of three interdependent elements: agreement on the goals of treatment, agreement on the tasks that lead to those goals, and the affective bond of trust, acceptance, and mutual respect. Because it is pan-theoretical, the definition applies to any approach — psychodynamic, cognitive-behavioral, behavior-analytic, or other.

Three threads woven together

It's tempting to reduce the alliance to “getting along” with the therapist. But the technical concept is more demanding: it involves agreeing on the direction, agreeing on the path, and trusting the person going with you. The three sustain one another — a strong bond gives courage to face difficult tasks; completing the tasks shows that the goal is attainable; and seeing progress increases trust in the therapist.

Goals where to go Tasks how to go Bond with whom Working alliance
The alliance is not a single thread: it is woven from agreement on goals, agreement on tasks, and the bond of trust. When one thread loosens, the whole rope loses strength.

Explore each thread and see how it shows up in a therapy-room conversation:

Agreement on the goals

Client and therapist need to want the same thing. It sounds obvious, but this is where a lot of therapy derails: the family arrives asking for the child to “stop having tantrums,” and the professional wants to work on functional communication; someone seeks help to sleep better, and the therapy turns toward childhood. Neither direction is wrong in itself — the problem is that they weren't negotiated out loud. Defining goals that are clear, revisable, and important to the person is part of treatment, not paperwork.

Agreement on the tasks

Once the direction is agreed on, the method still needs to be agreed on: what happens in sessions, what is done between them, how long it should take, how you'll know it's working. An exposure exercise, a thought record, a skills-training homework — none of it works well if the person hasn't understood why they're doing it. Explaining the rationale and checking that it makes sense to whoever will carry it out is how the alliance gets built.

The bond

This is the affective tie: trust, acceptance, and mutual respect — feeling heard and taken seriously, trusting that the therapist is on your side. The bond is not friendship or informality: the therapeutic relationship has asymmetrical roles, ethical limits, and a working purpose. But without that tie, it's hard for anyone to expose themselves to what is uncomfortable, and therapy usually demands exactly that. A more recent formulation, by Wampold and Flückiger (2023), adds to this emotional component a distinct cognitive one: the belief that the therapist is competent to choose and carry out an effective treatment.

Where the idea came from

The concept was born in psychoanalysis. Freud, in 1912, already noted that part of the patient's attachment to the analyst was rational and collaborative — not a phenomenon to be interpreted, but a resource to be preserved. In the 1950s, Elizabeth Zetzel coined the term “therapeutic alliance,” and Ralph Greenson, in the 1960s, distinguished the “working alliance” from transference: the patient's capacity to work with purpose even while feeling difficult things.

The generalization of the concept came in the second half of the 1970s, with Luborsky and Bordin — whose paper came out in 1979: the concept left its psychoanalytic cradle and came to describe the working collaboration present in any kind of helping relationship. That generalization opened the way to measuring the alliance and studying it empirically across any approach — which is why today we talk about alliance in cognitive-behavioral therapy as much as in behavior-analytic interventions.

This generalization brought a decisive consequence: it made the alliance measurable. One of the instruments used for this is the Working Alliance Inventory (WAI), built by Horvath and Greenberg in 1989 directly from Bordin's three components — 36 items on a Likert scale, divided into three subscales of 12 (Goal, Task, and Bond), plus an overall score. There is also a revised short version of 12 items, the WAI-SR, and other scales, such as the CALPAS and the Helping Alliance Questionnaire. For children and adolescents there are dedicated instruments, such as the TASC. The alliance can be rated by the patient, by the therapist, or by outside observers.

What the research actually shows

The relationship between alliance and outcome is one of the most replicated findings in psychotherapy research. A 2018 meta-analysis gathered 295 independent studies and more than 30,000 patients, published between 1978 and 2017. It found a correlation of around 0.28 between alliance quality and treatment outcome.

What does that number mean, in plain language? That the alliance is a reliable but modest predictor: it relates to something around 8% of the variation in outcomes. That's a consistent association — and, at the same time, insufficient to support the claim that “the relationship is all that matters.” Dismissing the alliance and treating it as the whole treatment are both exaggerated readings.

  • The association shows up across different theoretical approaches, across various clinical problems, and at different stages of treatment;
  • it shows up whether it's the patient rating the alliance or the therapist;
  • it also shows up in internet-delivered psychotherapy, at a magnitude practically equal to in-person care;
  • with children and adolescents the pattern repeats, a bit weaker: a 2018 meta-analysis of 28 studies found a correlation around 0.19. Here there's an added complication — there are distinct alliances at the same time: with the child and with the caregivers. Research shows they may not move together: within the same treatment, the alliance reported by parents can grow while the one reported by the child stays flat.
verified

Formal recognitionThe APA's interdivisional task force on evidence-based therapy relationships (Norcross & Lambert, 2019) classified the alliance in individual psychotherapy as having “demonstrated effectiveness” — the highest category in its rating system. In other words: tending to the alliance is not a humane extra separate from technique; it is a component of treatment with its own empirical support.

One honest question remains open: does the alliance cause improvement, or is it a consequence of it? People who are getting better tend to like therapy more, after all. The evidence leans toward the alliance having a causal role — alliance measured early on predicts outcome even after accounting for improvement that already occurred, and analyses that track the same person across sessions show alliance fluctuations preceding symptom reductions. Even so, part of the effect likely runs in both directions, and research keeps refining this distinction.

When the alliance breaks

Alliances aren't stable: they rise, fall, and sometimes rupture. A rupture is any tension or break in the collaboration — and it tends to show up in two forms. In a withdrawal rupture, the person pulls away: becomes monosyllabic, misses sessions, changes the subject, agrees with everything without engaging. In a confrontation rupture, the dissatisfaction comes out in the open: complaints about the therapy, about the therapist, about how slow the results are.

The counterintuitive point is that rupture is not, by itself, a bad sign. What distinguishes outcomes is the repair — naming the tension, taking in the criticism without defensiveness, understanding what happened, and recombining goals and tasks. A 2018 meta-analysis found a correlation of around 0.29 between rupture-and-repair episodes and better outcomes.

time start rupture repair the work continues
A repaired rupture is not a failure of therapy: naming the tension and resolving it openly tends to go hand in hand with better outcomes.

Whose job is it to build it?

This is the part that is most uncomfortable — and that matters most. A 2007 study of 331 patients treated by 80 therapists separated two things that ordinary statistics lump together: the variation in alliance between patients of the same therapist and the variation between therapists. The result was clear: what predicted outcome was the therapist's contribution. Professionals who were able to form good alliances with a range of different people got better results; whether a given patient was “easier” to bond with did not predict improvement.

The practical implication is direct: in the study, the patient's contribution to the alliance did not predict improvement — building the alliance is, therefore, a professional task, not a matter of luck or of having an “easy patient.” Talking about a “resistant patient” or a “difficult family” as the explanation for a poor alliance shifts onto the other person a responsibility that belongs to the professional — and, worse, it closes off the inquiry exactly where it should begin.

Myths worth undoing

Tap the cards to see what the evidence says:

In evidence-based practice

There's a false opposition going around: on one side, “relationship therapy”; on the other, “technique therapy.” Evidence-based practice does not accept that divide — it requires both, plus the values and preferences of the person being treated. In applied behavior analysis, this attention to the relationship shows up at several points in the work:

  • Pairing and rapport: before teaching, the professional pairs themselves with good things for the child, becoming someone worth interacting with;
  • Goals negotiated with the family — rather than imposed on it: this is the direct equivalent of Bordin's “agreement on goals,” and it is closely related to the social validity of programs;
  • consent and assent: treating the child's refusal as relevant information, not as an obstacle to overcome;
  • Compassionate care: a 2019 paper argues that empathy, compassion, and collaboration skills need to be part of the behavior analyst's repertoire with the same rigor used to train technical skills — because relationship-building skills strengthen commitment to treatment;
  • cultural humility: recognizing that the goals that seem “obvious” to the professional may not be the family's.
warning

Signs worth talking aboutIf you don't know what the goals of your therapy are, don't understand why you do what you do in sessions, don't feel heard, or feel you can't disagree — that isn't fussiness, it's clinical information. Bring it up in session: a good alliance is built exactly through these conversations. If, after trying, nothing changes, looking for another professional is a legitimate decision.

working alliance agreement on goals agreement on tasks therapeutic bond rupture and repair Working Alliance Inventory (WAI)

Test your understanding

quiz

Quick quiz

Three questions to lock in the concept. You'll see the explanation right away.

Further reading

  1. Bordin, E. S. “The generalizability of the psychoanalytic concept of the working alliance”. Psychotherapy: Theory, Research & Practice, 1979.
  2. Flückiger, C.; Del Re, A. C.; Wampold, B. E.; Horvath, A. O. “The alliance in adult psychotherapy: a meta-analytic synthesis”. Psychotherapy, 2018.
  3. Eubanks, C. F.; Muran, J. C.; Safran, J. D. “Alliance rupture repair: a meta-analysis”. Psychotherapy, 2018.
  4. Karver, M. S.; De Nadai, A. S.; Monahan, M.; Shirk, S. R. “Meta-analysis of the prospective relation between alliance and outcome in child and adolescent psychotherapy”. Psychotherapy, 2018.
  5. Shirk, S. R.; Karver, M. S.; Brown, R. “The alliance in child and adolescent psychotherapy”. Psychotherapy, 2011.
  6. Baldwin, S. A.; Wampold, B. E.; Imel, Z. E. “Untangling the alliance-outcome correlation: exploring the relative importance of therapist and patient variability in the alliance”. Journal of Consulting and Clinical Psychology, 2007.
  7. Wampold, B. E.; Flückiger, C. “The alliance in mental health care: conceptualization, evidence and clinical applications”. World Psychiatry, 2023.
  8. Zilcha-Mano, S. “Is the alliance really therapeutic? Revisiting this question in light of recent methodological advances”. American Psychologist, 2017.
  9. Norcross, J. C.; Lambert, M. J. (eds.). Psychotherapy Relationships That Work. 3rd ed. Oxford University Press, 2019.
  10. Horvath, A. O.; Greenberg, L. S. “Development and validation of the Working Alliance Inventory”. Journal of Counseling Psychology, 1989.
  11. Hatcher, R. L.; Gillaspy, J. A. “Development and validation of a revised short version of the Working Alliance Inventory”. Psychotherapy Research, 2006.
  12. Taylor, B. A.; LeBlanc, L. A.; Nosik, M. R. “Compassionate care in behavior analytic treatment: can outcomes be enhanced by attending to relationships with caregivers?”. Behavior Analysis in Practice, 2019.
info

This entry is educational and does not replace assessment, diagnosis or treatment by a qualified health professional. Every person is unique — for a specific case, seek individual guidance.

Related entries

Looking for professional guidance?

Instituto Walden4 is a psychology and education center in Brasília, Brazil, offering evidence-based assessment and care for autism, psychotherapy and development — in person and online. Our team serves in Portuguese; get in touch and we will find the best way to help.

chat Message us on WhatsApp
chat

Conta Walden4

Entrar

Use o e-mail e a senha da sua conta.