The psychotherapy research literature has accumulated a large and consistent finding over the past four decades: the quality of the therapeutic alliance, broadly defined as the emotional bond between therapist and client and the agreement on goals and tasks of treatment, predicts treatment outcomes across therapeutic modalities more reliably than the specific techniques employed.
This finding is counterintuitive to the modeldriven way that psychotherapy is taught. Trainees learn cognitivebehavioral techniques, psychodynamic interpretations, motivational interviewing skills, and dialectical behavior therapy protocols. The implicit message is that the technique is what works. The evidence suggests that the technique provides the container, and the relationship does much of the work.
What the alliance actually consists of
Bordin's working alliance model identifies three components: the emotional bond between therapist and client, agreement on the goals of treatment, and agreement on the tasks, the specific activities that constitute the work of therapy. All three contribute to outcome, and problems with any of them predict deterioration or dropout.
The clinical implication is that alliance ruptures, moments when the relationship is strained, when the client feels misunderstood, criticized, or not heard, deserve explicit therapeutic attention. Research on alliance repair demonstrates that therapists who directly address ruptures when they occur have better outcomes than those who proceed without acknowledging them. The repair itself is often therapeutic: it models that conflict can be addressed, that mistakes can be acknowledged, and that the relationship can survive difficulty.
