When I was a child at summer camp, an epidemic of home-sickness broke out. Several campers presented themselves to the resident medic, who talked to them gently and asked who they were missing most. Those who were pining for their mothers got a pink-coated chocolate. Those missing their fathers (an apparently less virulent condition), received a blue one. From what I observed, the treatment was highly effective.

This was my introduction to the placebo effect. Last month, a team led by Irving Kirsch of Hull University reported that trial data submitted to the US Food and Drug Administration showed that antidepressants were, in most cases, only marginally more effective than placebos.

Now a new study shows expensive placebos are more effective than cheaper ones. Dan Ariely, a Duke University economist, and a team from the Massachusetts Institute of Technology, administered light electric shocks to the wrists of 82 volunteers before and after giving them what they said were painkillers but were really dummy pills.

Half the group were told the pills cost $2.50 each and half that they cost 10 cents each. Of those who took the more expensive placebos, 85 per cent reported feeling less pain, compared with 61 per cent in the 10-cent group.

What is striking is not just the difference between the two groups but how many people in both said the placebos worked.

As my colleague Clive Cookson pointed out on this page on March 1, this was also a notable feature of the Kirsch study, which did not show that anti-depressants were useless. They were not – it was just that placebos worked almost as well.

This has led to debate about whether doctors should prescribe placebos, with Stuart Derbyshire, a psychology researcher at Birmingham University, warning it “would undermine the trust between society and medicine”.

Less often noticed is how many doctors are already prescribing placebos. In her superb new book, Suckers: How Alternative Medicine Makes Fools of Us All, Rose Shapiro says 42 per cent of UK general practitioners prescribe homeopathic treatments or refer patients to homeopaths. Nearly half of Dutch family doctors prescribe homeopathy. In Belgium, 85 per cent do.

All these doctors must know that homeopathic remedies have performed no better than placebos in repeated clinical trials. Homeopathic medicines are so heavily diluted that there is in effect nothing left by the time they are administered.

Homeopaths contend the diluted medicines work because the water in which they are dissolved retains a memory of them. As Ms Shapiro says, water must then retain a memory of everything that has ever been dissolved in it. How, she asks, does water distinguish between homeopathic medicine and, say, the soap from Marilyn Monroe’s bath? “Can water have a selective memory?”

The dilemma for doctors is that people want homeopathy. In the US, where it had almost died out by the 1970s, at least 6m people now go to homeopaths. Boiron, the NYSE Euronext-listed homeopathic company, says 300m people worldwide use homeopathy.

The reasons are complex, but among them is a feeling that doctors do not give patients enough time.

In the UK, the average consultation with a family doctor lasts less than 10 minutes (six to eight minutes in London) and the patient speaks for 23 seconds before being interrupted. An initial homeopathic consultation takes an hour or more, giving patients the sense of being listened to and valued. Many doctors say they would love to do the same, but are under too much pressure to meet targets or clear backlogs of patients. Viewed that way, sending people to talk to homeopaths is a form of outsourcing.

Does it matter? While homeopathic remedies may not do any good, beyond the placebo effect, they do not do any harm either. That is in contrast to herbal medicines (which do have active ingredients and are occasionally effective) or conventional medicines, which can have side-effects that are not always disclosed. Last week, for example, the UK’s Medicine and Healthcare Products Regulatory Agency wrote a stiff letter to GlaxoSmithKline about the drugs group’s tardiness in reporting that children taking the anti-depressant Seroxat could experience suicidal feelings.

Yet, for all their occasional lapses, scientific medicine and the pharmaceutical industry have brought enormous benefits, from the disappearance of smallpox to better survival rates from heart attacks.

It cannot be right for doctors to undermine the scientific method that lies at the heart of their training and practice. The money that doctors are now sending to homeopaths would be better spent on in-house counsellors with time to talk. We could then study whether it is the hours of sympathetic chat that make the difference, or whether you have to prescribe a sugar-coated chocolate too.

Send your comments to michael.skapinker@ft.com

More columns at www.ft.com/skapinker