Chapter 18: Does Therapy Actually Work? — The Honest, Complicated Answer

Does therapy work?

The short answer: yes. For most mental health conditions, particularly anxiety and depression, evidence-based psychotherapy produces meaningful improvement. The evidence is robust — meta-analyses spanning thousands of studies and hundreds of thousands of participants consistently show that therapy outperforms no treatment and, for many conditions, performs comparably to medication.

That's the reassuring part. Now for the complications.

Because the longer answer includes: not all therapy is equally effective, not all therapists are equally skilled, the match between the specific approach and the specific condition matters, the research doesn't support every therapeutic approach currently practiced, the popular version of therapy (as represented on TikTok) often bears little resemblance to the evidence-based version, and "does therapy work?" is actually several different questions depending on what you mean by "therapy," "work," and "for whom."

This chapter attempts to answer the question honestly — which means acknowledging both the strong evidence that therapy helps and the significant nuances that the popular version consistently glosses over.

Before You Read: Confidence Check

Rate your confidence (1–10) that each statement is true.

1. "Therapy is effective for most mental health conditions." ___
2. "All types of therapy are equally effective (the 'dodo bird verdict')." ___
3. "The therapist's skill matters more than the specific therapeutic approach." ___
4. "Therapy as promoted on TikTok represents evidence-based practice." ___
5. "If therapy isn't working, the problem is with the client, not the therapy." ___

What the Evidence Shows: Therapy Works

The Meta-Analytic Evidence

The evidence base for psychotherapy is enormous. Here are the key findings:

Therapy vs. no treatment. Meta-analyses consistently show that psychotherapy is more effective than no treatment for depression, anxiety, PTSD, OCD, eating disorders, and many other conditions. Effect sizes are typically in the medium range (d ≈ 0.5–0.8), comparable to many medical treatments.

Therapy vs. medication. For mild to moderate depression, therapy (particularly CBT) performs comparably to medication. For severe depression, the combination of therapy and medication often outperforms either alone. For anxiety disorders, CBT frequently outperforms medication.

Long-term outcomes. Therapy (particularly CBT) may have advantages over medication in preventing relapse. When you stop medication, the risk of relapse increases; the skills learned in therapy persist after treatment ends.

Specific approaches for specific conditions: - CBT for depression and anxiety: Effect sizes of d ≈ 0.5–0.7 compared to waitlist controls. One of the most studied treatments in all of psychology. - CBT for OCD (with exposure and response prevention): Very effective, with effect sizes of d ≈ 1.0+. - Prolonged Exposure and CPT for PTSD: Strong evidence, effect sizes d ≈ 0.7–1.0. - DBT for borderline personality disorder: Reduced self-harm and suicidal behavior by approximately 50% compared to treatment as usual. - Behavioral activation for depression: Effective and simpler than full CBT; particularly useful when access to trained therapists is limited.

What "Works" Means

It's important to be specific about what "therapy works" means:

It means symptom reduction. Most therapy outcome studies measure reduction in symptoms (depression scores, anxiety scores, PTSD symptoms). "Working" typically means a clinically significant reduction in symptoms — not necessarily complete elimination.

It means improvement relative to no treatment. Therapy outperforms waitlist controls and, in some designs, outperforms placebo controls. This is the standard evidence threshold.

It does not mean 100% success. Even the best therapies don't work for everyone. Response rates vary, but roughly 50–65% of patients show clinically meaningful improvement with evidence-based therapy. That means 35–50% don't improve substantially — or don't improve with that particular approach.

It means for specific conditions. "Therapy works" is not a blanket statement about all therapy for all conditions. The evidence is much stronger for some condition-treatment pairings than others.

The Dodo Bird Verdict: Are All Therapies Equal?

The Claim

In 1975, Lester Luborsky and colleagues published an influential paper whose title referenced the dodo bird from Alice in Wonderland: "Everybody has won, and all must have prizes." Their conclusion: different psychotherapy approaches (psychodynamic, cognitive, behavioral, humanistic) produce roughly equivalent outcomes. This became known as the dodo bird verdict — the claim that the specific type of therapy matters less than the common factors shared by all therapies (the therapeutic relationship, hope, a structured framework for change).

The Evidence For

Meta-analyses do show small differences between approaches. When head-to-head comparisons of different therapies are analyzed in aggregate, the differences between established approaches are often small. A 2014 meta-analysis by Wampold and Imel found that different therapy types accounted for only about 1% of outcome variance — suggesting that what they share (common factors) matters more than how they differ.

The therapeutic alliance predicts outcomes. The quality of the relationship between therapist and client — warmth, trust, agreement on goals, collaboration — is one of the strongest predictors of therapy outcome, regardless of the therapeutic approach used. This supports the common factors argument.

The Evidence Against

For specific conditions, specific approaches excel. The dodo bird verdict breaks down when you examine specific condition-treatment pairings: - OCD responds much better to CBT with exposure than to general supportive therapy - PTSD responds better to trauma-focused therapies (PE, CPT, EMDR) than to non-trauma-focused approaches - BPD responds better to DBT than to most alternatives - Specific phobias respond dramatically well to exposure therapy and poorly to talk-only approaches

Not all therapy has evidence. The dodo bird verdict applies to established, evidence-based approaches. It does not mean that any therapeutic approach works equally well. Approaches without evidence (some forms of energy therapy, rebirthing therapy, past-life regression) are not equivalent to CBT simply because they take place in a therapy room.

Researcher allegiance effects. Some of the apparent equivalence between approaches may reflect researcher allegiance — therapists studying their own approach tend to produce better outcomes, regardless of the approach. When allegiance is controlled for, some differences between approaches emerge.

The Balanced View

The honest answer to "are all therapies equal?" is: mostly, but not entirely. For many conditions, different established approaches produce similar outcomes — and common factors (particularly the therapeutic alliance) matter enormously. But for some conditions, specific approaches are substantially more effective, and the claim that "it doesn't matter what kind of therapy you do" is dangerous if it leads people to pursue ineffective treatments for conditions that respond to specific interventions.

Therapist Skill: The Underappreciated Variable

The Finding

One of the most consistent findings in therapy research is that therapist effects are larger than treatment effects. In other words, the specific therapist you see matters more for your outcome than the specific type of therapy they practice.

Wampold and colleagues have estimated that therapist effects account for approximately 5–8% of outcome variance — which sounds small but is 5–8 times larger than the variance attributable to the specific therapy type (~1%). Some therapists consistently produce better outcomes across patients; others consistently produce worse outcomes.

What Makes a Good Therapist?

Research on therapist effectiveness has identified several factors:

Interpersonal skills. Warmth, empathy, genuineness, and the ability to form a strong working alliance are consistently associated with better outcomes.

Flexibility. Effective therapists adapt their approach to the individual client rather than rigidly applying a manual.

Expertise. Therapists with more training and experience tend to produce somewhat better outcomes — but the effect is modest, and some novice therapists outperform experienced ones.

Deliberate practice. Therapists who seek feedback on their outcomes, review their sessions, and actively work to improve produce better results over time. Most therapists don't do this.

The Quality Problem

Here is an uncomfortable truth: not all therapists are good at their job. As in any profession, there is a distribution of skill. Some therapists are excellent; some are adequate; some are ineffective; and a small percentage may be actively harmful.

The therapy industry does not routinely measure or report therapist outcomes. Unlike medicine (where surgical outcomes are tracked) or education (where student outcomes are measured), therapy often operates without systematic outcome monitoring. Most therapists believe they are above average — which, statistically, cannot be true for all of them.

This means that the answer to "does therapy work?" depends significantly on which therapist you see. The population-level evidence shows therapy is effective on average, but your individual experience depends heavily on the specific therapist and the match between their skills and your needs.

TikTok Therapy vs. Evidence-Based Therapy

The therapy that dominates social media looks very different from the therapy with the strongest evidence:

This is not to say that all TikTok therapy content is bad. Some of it is excellent psychoeducation. But the overall emphasis — on validation over challenge, on insight over behavior change, on understanding the past over addressing the present — doesn't align with the strongest evidence base.

The evidence-based approaches that work best (CBT, exposure therapy, behavioral activation) often involve doing things that are uncomfortable: facing fears gradually, challenging unhelpful thoughts, changing behavioral patterns. The TikTok version of therapy — which emphasizes being "validated" and "held" — may inadvertently promote therapeutic relationships that feel good but don't produce change.

Verdict: "Therapy is effective for most mental health conditions" ✅ SUPPORTED — Meta-analyses consistently show therapy outperforms no treatment for depression, anxiety, PTSD, OCD, eating disorders, and other conditions. Effect sizes are medium and comparable to many medical treatments. Evidence is particularly strong for CBT. Evidence: Hundreds of meta-analyses spanning thousands of trials. Cuijpers et al. (multiple reviews), Hofmann et al. (2012), Barlow et al. (2017).

Verdict: "All types of therapy are equally effective (dodo bird verdict)" ⚠️ OVERSIMPLIFIED — For many conditions, different established approaches produce similar outcomes. But for specific conditions (OCD, PTSD, BPD, specific phobias), specific approaches substantially outperform alternatives. Common factors (especially the therapeutic alliance) matter enormously, but the claim that "it doesn't matter what therapy you do" is dangerous for conditions that respond to specific treatments.

Verdict: "The therapist matters more than the therapy type" ✅ SUPPORTED — Therapist effects account for approximately 5–8% of outcome variance, compared to ~1% for therapy type. The specific therapist you see is one of the strongest predictors of your outcome.

When Therapy Doesn't Work: What Then?

If therapy isn't working after a reasonable trial (typically 8–12 sessions), the problem may be:

1. The wrong approach for the condition. General talk therapy for OCD when exposure therapy is indicated. Non-trauma-focused therapy for PTSD when PE or CPT would be more effective.
2. The wrong therapist. Poor therapeutic fit, inadequate skill, or a therapist who isn't tracking outcomes.
3. An undiagnosed condition. What looks like depression might be ADHD, a thyroid condition, or substance use.
4. Life circumstances. Therapy cannot fix poverty, abuse that is ongoing, or environmental stressors that remain unaddressed.
5. The approach needs more time. Some conditions (personality disorders, complex trauma) require longer treatment.

The important message: if therapy isn't working, it's not necessarily your fault. It may be the wrong type, the wrong therapist, or the wrong time. Discussing this openly with your therapist — or seeking a second opinion — is appropriate and encouraged.

Fact-Check Portfolio: Chapter 18

If any of your 10 claims involve therapy, mental health treatment, or therapeutic approaches: - Does the claim distinguish between evidence-based and non-evidence-based approaches? - Does it account for therapist effects and condition-specific treatments? - Does it conflate TikTok therapy content with evidence-based practice? - Does it present therapy as universally successful or acknowledge that outcomes vary?

After Reading: Confidence Revisited

1. "Therapy is effective for most conditions." — What is the evidence, and what does "effective" mean specifically?
2. "All therapies are equally effective." — When does the dodo bird verdict hold, and when does it break down?
3. "The therapist matters more than the approach." — What do therapist effect estimates show?
4. "TikTok therapy represents evidence-based practice." — What are the key differences?
5. "If therapy isn't working, the problem is the client." — What other explanations exist?