CBT was built for a specific problem: distorted thinking in an otherwise functional brain. For that problem, it works. This book is about the gap between that design and current clinical practice.
The Research Nobody Quotes at the Conference
CBT’s effect sizes have declined steadily since the 1970s. This is not a fringe finding. It appears in major meta-analyses, has been discussed openly in the clinical research literature, and has a well-documented explanation: researcher allegiance bias. Studies conducted by CBT researchers produce larger effects for CBT. When active control conditions are used, the apparent advantage shrinks or disappears.
None of this makes CBT ineffective. It makes its effectiveness more conditional than four decades of institutional endorsement imply.
The more specific finding, and the one with the most direct clinical implications, is this: therapists who deviate from the CBT manual, adapting their approach to the client rather than following the protocol, tend to outperform those who adhere to it faithfully. That result has replicated across multiple studies. It raises a basic question about what the manual is actually contributing.
Who CBT Was Designed For
The original CBT protocols were developed with a specific population: motivated, psychologically minded adults with discrete symptom presentations and a willingness to complete structured homework between sessions. Clients who cooperate with a logical model applied to their thinking.
That population exists, and CBT works for them. The problem is what happens when the same protocol gets applied, without adjustment, to populations it was never designed for.
Three groups sit outside CBT’s design parameters. Resistant clients, where the protocol has no mechanism for working with noncompliance other than motivational interviewing, which itself requires the client to cooperate. Psychosomatic presentations, where locating the problem in the client’s thinking is both clinically inaccurate and relationally harmful. And relational problems, where the individual cognitive model cannot address the structural dynamics that maintain the symptoms.
These are not edge cases. They represent a substantial portion of the caseloads most practitioners carry.
What Works Instead
The clinical frameworks this book covers, strategic therapy and related systemic approaches, were not developed as alternatives to CBT. They were developed independently, for different populations and different problems. They operate on different assumptions: that the presenting pattern, the relational structure that maintains it, and the function the symptom serves are the problem. For clients where logic is not the obstacle, that distinction determines whether the intervention works.
The book covers the evidence behind these approaches, the specific techniques, and what clinical application looks like in practice. The argument is not that CBT is useless. The argument is that CBT is a tool with a defined scope, and the field has systematically overapplied it for institutional reasons that have little to do with client outcomes.