On what basis is it justifiable to treat people with mental health problems differently to those with physical health conditions? Szmukler and Holloway's views will be well-known to readers of the Bulletin (see, for example, Psychiatric Bulletin, November 1998, 22, 662-665), but this discussion paper (written before the Government's White Paper was published) gives the clearest account yet of their position: not only is mental health law discriminatory, the situation is actually getting worse. Moreover, there is no logical or factual justification for it.

The authors start by showing that mental health law has historically oscillated between the two poles of a medical ‘best interests’ model on the one hand and a legalistic ‘dangerousness’ approach on the other; our present law represents a slight retreat from the paternalistic 1959 Act. They then look at the ‘Bournewood gap’ and the Richardson Committee's proposals, largely disregarded by the Government. They consider the plans for compulsory treatment and find them wanting. So far, so familiar.

The most intriguing part of the paper is the thesis that current legislation poses two key questions the wrong way round. The questions are, first, does the person have a mental disorder? Second, is he or she dangerous? Addressing them in this order inevitably leads to the treatment of people with mental health problems on a different basis from the rest of the population. However, if the questions are reversed, the question of dangerousness can be approached in the same way for all; the disorder then becomes a factor in determining how to address the dangerousness. It follows that, if the first question is answered in the negative, there is little justification for compulsory intervention. This analysis is a most helpful contribution to the current debate.

The authors finish by considering compulsory treatment in the community: they have no objections to its use where the patient lacks capacity and the treatment is in his or her best interests. If it is applied as an alternative to hospital admission, or to facilitate earlier discharge, the order must be time-limited, based on recovery of capacity. It should not be used for the protection of others.

The paper concludes with a plea for placing compulsory treatment on a firm ethical basis. As we now know, the White Paper, with its stress on best interests, represents a move back towards paternalism. Let us hope that Parliament will bring some ethical rigour to its own discussions in due course.