There is a running joke featured on the BBC programme Little Britain which involves a series of questions being put to an unemotional respondent. On each occasion the questions are greeted with the reply “computer says no”.

There is a certain unyielding quality to the relationship between the Government and its Advisory Council on the Misuse of Drugs that would not be out of place on a comedy show. The Government keeps asking its advisers whether, by any chance, it might just possibly be a good idea to reclassify cannabis as a more harmful drug, and keeps getting the reply “computer says no”.

However wryly amusing this repeated scene may have been to observers, it has taken its toll. Alan Johnson did not sack Professor David Nutt because of one speech that “crossed the line” into politics. He sacked him because, over the years, the refusal of the Government’s advisers to fall into line with its political position had become intolerable to the Home Office. Mr Johnson has blundered badly.

The Government reclassified cannabis to a less serious category of harm in 2002 after the recommendation of the advisory council. By the end of 2004 the decision had proven sufficiently controversial that the Government returned to the council and asked it to consider the matter again. It did and the computer said no — the advice remained the same. Charles Clarke, the then Home Secretary, rather bravely decided to accept their advice.

When Gordon Brown became Prime Minister, wishing as Mr Clarke puts it “to appear tougher than Tony Blair”, the issue was put again to the council. This time, it was made clear that the Government was determined to reverse its 2002 decision regardless of the council’s advice. Once again the computer said no, but the Home Secretary (Jacqui Smith) overrode them. On both sides, relations deteriorated.

It is quite a feat to fall out with the genial Mr Johnson, someone who has shown a high level of tolerance for even the most difficult of people, and this does raise questions about the way Professor Nutt conducts himself. He clearly has little patience for politics and politicians, and this is odd in a person required to deal with both. The classification of drugs involves many issues besides its harmfulness — the social context is important. There is clearly a difference between the way government will, and should, look at a substance already legal and one that is currently illegal. So political subtlety is essential in a chairman of the council and Professor Nutt’s lack of it is a deficiency.

Yet Mr Johnson’s decision to sack him was an extraordinary and unforced error. Professor Nutt and his colleagues are unpaid advisers, not civil servants. They hold office precisely because they have expertise and it is absurd to expect professors to accept unpaid advisory jobs that prohibit them from giving their professional judgments in public in their specialist areas.

More than this, it is valuable to public debate for the council chairman to express himself in forthright terms. It might force governments to provide a better explanation of its policy than this one has on cannabis. No government should be in thrall to its experts, and there is much more to drug policy than scientific judgment, but no politician should disregard scientific policy lightly.

Perhaps out of Mr Johnson’s mistake something good might come. The way government deals with scientific advice — often rejecting it without saying why — and the status and civil service support of its independent advisers has long been unsatisfactory. Nothing brings change faster than a nice, big fiasco.