Psychology › Therapy & treatment › Talk therapy (psychotherapy)
The therapeutic alliance
1979 (Bordin)· Concept
Well replicatedStrong, repeated evidence.
The working partnership between therapist and client: agreed goals, agreed tasks and a bond of trust.
See it on the Psychology map →What it says
Psychologist Edward Bordin defined it in 1979 as three parts: agreement on the goals of therapy, agreement on the tasks used to reach them, and the emotional bond between the two people. Questionnaires such as Adam Horvath and Leslie Greenberg's Working Alliance Inventory (1989) measure it, often from the client's side. A strong alliance early in treatment predicts better outcomes across CBT, psychodynamic, humanistic and other therapies, and even in medication trials. Ruptures, such as a client feeling misunderstood, are common and repairable.
Why it matters
It suggests that the relationship, and how well the pair works together, matters about as much as the method used. Finding a good fit counts.
Think of it like
Like an owner and property manager: the same business plan succeeds or stalls depending on shared goals and trust.
The evidence
Status: well replicated. Hundreds of studies since the 1980s, pooled in repeated meta-analyses; the link holds across therapy types, countries, and in-person and online formats.
Open questions
Whether a good alliance causes improvement or partly reflects it: clients who are already getting better may simply rate the relationship higher.
Connections (2)
▸The therapeutic alliance predicts success in Talk therapy (psychotherapy)Well replicated
The connection
Across many kinds of talk therapy, clients who agree with their therapist on goals and methods and feel a trusting bond tend to improve more. This shared ingredient helps explain why very different therapies often produce similar results.
Who showed it and when
Measured from the late 1970s by Edward Bordin, Lester Luborsky, Adam Horvath and others; pooled in meta-analyses from 1991 to 2018.
The evidence
A 2018 meta-analysis led by Christoph Flückiger, pooling 295 studies and more than 30,000 patients, found a consistent moderate correlation, about 0.28, between alliance and outcome.
Open questions
Correlation is not cause: clients who improve early may then rate the alliance higher. Session-by-session studies suggest the influence runs both ways.
▸The therapeutic alliance was made central by Person-centered therapy (Carl Rogers)Well replicated
The connection
Psychoanalysts coined the term, but Carl Rogers made the relationship itself the healing force. In 1957 he proposed that a therapist's empathy, genuineness and unconditional acceptance are enough to produce change. Later alliance research grew partly out of testing his claim.
Who showed it and when
Rogers' 1957 paper 'The Necessary and Sufficient Conditions of Therapeutic Personality Change'; the word 'alliance' came from psychoanalysts Elizabeth Zetzel (1956) and Ralph Greenson (1965).
The evidence
A 2018 meta-analysis of empathy led by Robert Elliott found a moderate link to outcome, supporting Rogers' emphasis though not his claim that it suffices alone.
Open questions
Scholars disagree on how much modern alliance research owes to Rogers versus psychoanalysis; Bordin's 1979 model drew on both traditions.