During his tenure as the UK government's chief adviser on drug abuse, David Nutt ruffled many a feather with his provocative remarks. Earlier this year, for example, he published an article that called for a wider debate on society's approach to risk and that favourably compared the dangers of the psychoactive amphetamine drug MDMA (ecstasy) to those of horse-riding (D. J. Nutt J. Psychopharmacol. 23, 3–5; 2009).

But it was only on 30 October that Nutt, a professor in neuropsychopharmacology at Imperial College London, was summarily fired from his position as chair of the British government's Advisory Council on the Misuse of Drugs by home secretary Alan Johnson. According to Johnson, Nutt's crime was to muddy the allegedly clear waters of government drugs policy by publicly making statements that questioned it, thereby going beyond his remit as a scientific adviser (see Nature doi:10.1038/news.2009.1053; 2009).

That concern should not be dismissed lightly. Politicians cannot always base their decisions solely on scientific advice, but must also consider such factors as public sentiment. Scientific advisers who publicly attack decisions they consider to be less than ideal, and in so doing provide ammunition for political opponents of those decisions, are entering dangerous territory.

Nonetheless, in this case, the position of the Labour government and of the leading opposition party, the Conservatives, which vigorously supported Nutt's sacking, has no merit at all. It deals a significant blow both to the chances of an informed and reasoned debate over illegal drugs, and to the parties' own scientific credibility.

Nutt's fate was decided following the publication last week of a thoughtful analysis of the challenges in estimating the harm done by drugs (see http://go.nature.com/dPiUAt). In this paper, for example, Nutt notes that public perception is highly influenced by the way the media cover dramatic events such as drug deaths. He goes on to analyse such coverage over a decade. The data show that the media have disproportionately highlighted the comparatively small number of deaths caused by drugs such as cocaine and MDMA compared with the far more numerous deaths caused by other substances, such as alcohol.

The imbalance has convinced the public — and politicians — that some notorious drugs are much more dangerous than they are. Such perceptions heavily influence the government's classification of substances under the Misuse of Drugs Act, which in turn determines the penalty for being found in possession of a given drug.

Nutt goes on to detail efforts to develop an evidence-based scheme that ranks drugs according to “nine parameters of harm”, which range from an individual drug user's ravaged veins to society's extra payments for health care. The harm ranking that results bears only an approximate resemblance to the official classification. For example, alcohol and tobacco, which don't fall under the Misuse of Drugs Act, rank considerably higher than MDMA, which is classified as among the most dangerous drugs.

That scheme has no official standing, unfortunately. But in describing it, Nutt attempted to do what he has done in many other public discussions: portray risk in terms that people can understand, look at the underlying factors that influence scientific and public debate, and thereby highlight how politicized discussions about drug use have become. To Johnson, this was apparently tantamount to “lobbying” against political decisions.

Thus the sacking. As Nature went to press, two members of Nutt's former committee had resigned in protest against the government's intolerance. They were right to do so. The government, meanwhile, badly needs to restore its credibility on this issue. One good way to do that would be to follow Nutt's suggestion to turn the advisory council into an independent body reporting to parliament as a whole, not to any individual official. An independent, scientifically run drug-regulation system would also free politicians from having to politick over who is toughest on drugs, something that would spare them and scientists much unnecessary future trouble.