The debate over cannabis raises two key questions.

The first is: to what extent does cannabis increase the risk of psychotic illness? And secondly, would reclassifying it as a Class B drug reduce that risk, if it is real?

The first question should be amenable to evidence, while the second is a matter of judgment. But in reality the evidence is never going to be that clear-cut.

For about a decade, psychiatrists have been worried that some of the young patients they see with schizophrenia have developed that condition as a result of addiction to cannabis. The findings of control studies in the US, Germany, the Netherlands and Sweden, which have sought to discover if young people with schizophrenia are more likely to have been cannabis users, have been summarised in The Lancet. This found that cannabis users were 40 per cent more likely than nonusers to suffer mental illness and found “a consistent association between cannabis use and psychotic symptoms, including disabling psychotic disorders”.

Such an association cannot prove cause and effect. It may simply mean that young people attracted to cannabis have a genetic predisposition to psychosis. Cannabis addiction, in this case, is simply a marker for a psychotic personality.

Martin Frischer and Ilana Crome, of the University of Keele, who looked at trends in Britain for the Advisory Council on the Misuse of Drugs, found that between 1996 and 2005 there was a significant fall in cases of schizophrenia, and from 2000 in psychoses generally. Cannabis use has also fallen over roughly the same period, but the two academics concluded that the epidemiological evidence is not consistent with the view that increasing cannabis use in earlier decades is associated with increasing schizophrenia or psychoses since the mid1990s.

It could, of course, be an artefact of diagnostic practice. Maybe psychiatrists have simply decided to classify people as psychotic less often than they used to? Maybe better drugs are curing them? Maybe all the data is wrong? And maybe today’s skunk is a more potent trigger of psychosis than yesterday’s hash.

But assuming the analysis is right, there is little reason to believe that cannabis or skunk is contributing to an epidemic of schiozophrenia. There is no such epidemic.

The evidence of the case-control studies and the incidence of psychosis contradict each other, which is why judgments have to be made. This is even more true of the political decision over whether cannabis should be reclassified.

Since it was downgraded, consumption has fallen. The Commons Select Committee on Science and Technology, which examined drug classification in the 2005-06 session, has criticised the Government for using classification as a means of “sending out signals” rather than simply classifying drugs on the basis of harm. It said that there was no evidence that reclassifying the drug acted as a deterrent. If the advisory council has opted to recommend no change it may be wrong: but that decision would not be perverse.

Times Recommends

• Don't despair: great players learn from defeat
• Pope clears way for Cardinal Newman sainthood
• Ecclestone: despots are underrated