Send 7 people to therapy, and about 1 of them gets better because of the therapy itself. The other 6 either improve for reasons that have nothing to do with the room, or they do not improve at all. That number comes straight from the field’s own research, once someone bothers to clean it up.<\/p>\n\n\n\n
For 50 years, researchers tested therapy in more or less the same way. Take a group of people who are struggling. Give half of them therapy. Put the other half on a waiting list. Wait a few months. Compare. Therapy won those tests almost every time, and it should have, because sitting on a waiting list is one of the worst things you can do to a person in trouble. You are told help is coming, so you stop looking for it anywhere else. You go on hold. Beating a group of people who have been told to wait is a very low bar to clear.<\/p>\n\n\n\n
Studies that found therapy did not work often never got printed. Journals like winners. A trial with a disappointing result would sit in a drawer while the trials with good results filled the record. So the public picture of therapy was built from the successes and quietly missing most of the failures.<\/p>\n\n\n\n
In 2018, a team led by Pim Cuijpers<\/a>, who is about as respected as anyone alive in this field, went back and redid the math for depression. He threw out the waiting-list comparisons. He kept only the carefully run studies. He estimated the missing failures and added them back in. When he was done, the measured benefit of therapy had fallen by more than half. What was left is real. Therapy does beat a sugar pill. But the honest size of the win is small, and it works out to about 1 person in every 7 who walks in<\/a>. The rest of the people who feel better would have felt better anyway. Time passed. A job came through. A season changed. Humans heal on their own more often than the brochures admit.<\/p>\n\n\n\n Antidepressants, the thing therapy is often sold as a gentler alternative to, score almost exactly the same<\/a>. And 50 years of trials on therapy for children and teenagers<\/a> show the results have not budged. Hundreds of new studies, new methods, new letters after new names, and the outcomes for kids sit right where they sat in the 1970s. A field that was actually learning would show a line that climbs. This line is flat.<\/p>\n\n\n\n So the first bet is whether you are the 1 or one of the 6.<\/p>\n\n\n\n Therapy is sold in brands. CBT. EMDR. Psychodynamic. Each has its own books, its own trainings, its own believers who will tell you theirs is the one that works. Researchers have put these brands head to head for 40 years, and the finding is so stubborn it earned a nickname. They call it the dodo bird verdict, after the dodo in Alice in Wonderland who runs a race and then announces that everybody has won and all must have prizes. The gap between the brands is tiny<\/a>. Which method you pick barely moves your odds.<\/p>\n\n\n\n What moves your odds is the specific human sitting across from you. Two therapists in the same clinic, with the same training and the same license on the same wall, get reliably different results<\/a>. The person matters far more than the method. And you cannot shop your way to a good one. A therapist with 20 years behind them gets about the same results as one fresh out of school. In fact, one large study that followed therapists over time<\/a> found their results drifted slightly worse as the years went on. Every single thing you can check before you hand over your money and your months tells you almost nothing. The brand, the credentials, the years, the reviews. The one thing that decides your outcome is who this particular stranger is, and that is the one thing you cannot see until you are already at the table.<\/p>\n\n\n\n A therapist never meets your mother. Never sits with your husband. Never watches the actual fight from the corner of the actual room. They hear your version, told by you, colored by whatever you happened to be feeling the week you told it. That would be survivable if human memory worked like a recording that you play back unchanged. It does not. In the lab, in animals, scientists have shown that a memory does not sit safely on a shelf. Pulling it up makes it soft again for a few hours, and the version that gets filed afterward can be edited by what you were told and how you felt while you held it. Whether the same rewriting runs as cleanly in human memory for the story of your own childhood is still argued over by the people who study it. But nobody argues that the story stays perfectly still. Tell the story of your childhood 40 times in a paid room built to hunt for wounds, and the story will move.<\/p>\n\n\n\n Sometimes it moves far enough to end a relationship. Two studies<\/a> followed clients whose therapists said things that cut against a marriage or a family. A partner named as toxic. Distance from a parent framed as growth. The more of those statements a therapist made, the more often the relationship later broke<\/a>. This spring a survey reached about 7,000 parents who had been cut off by an adult child<\/a>, and roughly a third of them said a therapist had recommended or pushed the child toward the cutoff. Those parents found the survey and signed themselves up for it, which is exactly the kind of group that would blame the therapist, so hold the precise number loosely. Hold the direction firmly, because it matches everything else on the table. Some of those cutoffs are the right call. Abuse is real, and walking away from it is health. But a stranger signing off on a one-sided, self-editing story is a strange way to make a decision that a family has to live with forever, and that decision lands on people who were never in the room and never agreed to play.<\/p>\n\n\n\n Every drug you have ever swallowed went through trials that counted the harm. Side effects tallied, listed, printed on the label in type too small to read but printed all the same. Therapy has no label. When researchers recently went looking for therapy studies that even tracked whether the treatment hurt anyone, they found 57 out of 1,430<\/a>. The field has run for a century on a comfortable assumption that talking can only help. It can hurt. A 2024 review<\/a> found serious harm in about 1 of every 21 people across therapy trials. Dependence on the therapist. New fears that were not there before. Relationships that got worse under advice. And when the therapy mindset gets rolled out to everyone at once, the same shadow appears.<\/strong> The largest school mindfulness trial ever run<\/a> taught thousands of teenagers to monitor their own mental states and found they were no better off for it. A closer look at the same trial<\/a> found the kids who were already at risk came out slightly worse on depression and well-being. Teaching people to scan themselves for damage turns out to be a decent way of teaching them to find some.<\/p>\n\n\n\n A therapist with 20 years behind them gets about the same results as one fresh out of school.<\/h1>\n\n\n\n
The field has run for a century on a comfortable assumption that talking can only help. It can hurt. <\/h1>\n\n\n\n