How far is a therapist’s empathy tied to the success of therapy?

What the client felt, more than what the therapist got right. Pooling 6,138 clients gave a correlation of 0.28 between empathy and outcome, a moderate link with a therapeutic outcome that held across methods of therapy and kinds of problem. Ratings by clients and by observers predicted better than technical measures of empathic accuracy, although the link is not strong.

The first difficulty is defining what is measured. Therapeutic empathy is not being nice but an attempt to understand a client’s experience precisely and return it so they feel understood. That is hard to measure from outside, so the review assumed no single correct method and compared several: reports by the client, by the therapist and by an observer.

That comparison produced the interesting finding. What the client or an outside observer perceived as empathic predicted the outcome better than technical measures of empathic accuracy, meaning how correctly the therapist identified the content. The feeling of being understood matters more than the accuracy itself.

The size deserves precision. Pooling eighty-two independent samples and more than six thousand clients gave a correlation of 0.28, a real and moderate link. It appeared across different approaches to therapy and different problems, so it is not a property of one method.

And here is the large caveat, which is about the direction of cause. Most of the data are correlational, and therapy that is going well can itself make both sides feel more understood, in which case the empathy is a result rather than a cause. Empathy is also not sufficient on its own: diagnosis, technique, the fit between therapist and client, and safety still decide whether therapy helps.