An ideological struggle has broken out between two of our leading “happiness thinkers” about whether the claims made for CBT can be justified. On one hand, Tony Blair’s adviser, the economist Professor Richard Layard, is about to launch a health service programme of CBT nationwide employing 10,000 novice recruits. On the other, psychologist Oliver James tells Daily Mail readers that CBT only appeals to Tony Blair because it is “quick, cheap and simplistic” but is seriously lacking in long-term efficacy.

So what does the evidence say? First, that it is reasonable for the Government to turn to the psychological therapies as the frontline treatment for conditions such as anxiety and depression when the routine prescription of benzodiazepine tranquillisers and antidepressants costs more than £11 billion a year and can yield disappointing results.

In CBT’s favour is the fact that, as a talking therapy, it does what it says on the label. To take the simplest example, if you believe nobody loves you then CBT therapists believe they need only to produce evidence that one person does love you for you to be proved wrong and for your behaviour to change.

The fact that, in a relatively short period, CBT has produced an impressively positive research base must be qualified by the observation that because CBT is tasked with “symptom removal”, not “treatment of the whole person”, research has proved relatively easy and cheap to undertake. Setting out to measure whether someone has got rid of a single symptom (such as spider phobia) leads to only two relevant answers: yes or no. It is much more difficult to evaluate a therapy seeking to show whether you have gone from “greater” to “lesser” unhappiness but the experience in itself might prove more life-changing.

Critics also observe that the case for standard CBT has been favoured by the way the guidelines on anxiety and depression, sponsored by the National Institute for Health and Clinical Excellence (NICE), are presented. Much of the pro-CBT information is to be found in headline summaries; significant qualifying remarks about other valid therapies are found in the small print.

This matters because Oliver James is right about research in the longer term. According to the most authoritative sources, at least half those patients receiving CBT for panic disorder had suffered relapse or sought new help after 24 months, which isn’t very cost effective.

Last Monday, at a conference on Practice-Based Commissioning in Manchester, Professor Layard admitted that CBT is appropriate for only about 40 per cent of patients overall. Stunningly, the largest body of evidence into counselling outcomes, the 35,000 cases comprising the CORE Survey, has been totally ignored by NICE and Layard alike. Looking at the figures just for depression, CORE shows there is no significant difference in the long-term success rates for CBT over traditional forms of therapy such as “person-centred” or “psycho-dynamic”: CBT works for 75 per cent of patients; the rest for 76 per cent.

So a summary of the evidence tends to show that alll talking treatments are roughly equal in effectiveness because it is the relationship with the therapist that counts. Patient choice should count, too. I suggest the NHS would be unwise to put all its eggs into a CBT basket.

Phillip Hodson is a Fellow of the British Association for Counselling and Psychotherapy (www.bacp.co.uk)

Mark Henderson is away