Our paper was the first in the British Journal of Psychiatry that attempted to answer a simple question that many UK psychiatrists have been asked by their ethnic minority patients — does racial discrimination cause mental illness? (Reference Chakraborty and McKenzieChakraborty & McKenzie, 2002).

Patients know that the rates of psychosis, for instance in Black Caribbeans in the Caribbean, is the same as for White British people in the UK, but that the rates of psychosis in Black Caribbeans in the UK is markedly higher. There has been no plausible biological hypothesis to explain this and all the evidence, including the genetic evidence, points to a social aetiology (Reference Sharpley, Hutchinson and MurraySharpley et al, 2001).

With specific reference to Dr Eagle's comments: although there is no evidence whatsoever of a biological cause or of increased vulnerability in ethnic minority groups, there is cross-sectional evidence of an association between experiencing racial discrimination and both psychotic and non-psychotic illness in ethnic minority groups in the UK. There is also longitudinal evidence of a link between experiencing discrimination and the development of psychotic symptoms in The Netherlands and these associations cannot be explained by other known risk factors (Reference Chakraborty and McKenzieChakraborty & McKenzie, 2002).

We do not invoke charges of political incorrectness. We invoke scientific logic and scientific equipoise. Given the available information and the resurgence of social causation theories of psychosis, it is difficult not to come to the conclusion that racial discrimination is a practical area of investigation.

Dr Eagles is wrong in his assumptions about the paper by Boydell et al (Reference Boydell, van Os. and McKenzie2001). Movement within the London wards that were surveyed was very limited and could not explain the results.

Professor Freeman is correct to cite the high rates of depression in some developing countries and we would support his call for more research in this area. He may not be aware of the methodological flaws in the work of the Manchester group which make their findings very difficult to interpret (Reference McKenzieMcKenzie, 1999).

Qualitative and quantitative research formats are complementary and offer different types of information. They are both scientific techniques, if used appropriately.

Racism is an experience that depends on context. We do hope that we have misunderstood Professor Freeman's suggestion which seems to be to try to establish some sort of league table of distress across different times or continents — this would be a bizarre idea. Phenotypic differences that we mention in our paper are not limited to skin colour and, of course, we accept that discrimination against many different White groups has been rife in the UK. We note the high rates of mental illness in some of these groups, such as the Irish.

Racism remains a major cause of the perpetuation of socio-economic differences between minority groups and ethnic majority groups in the UK and all of those working in the area, including governments, agree on this.

Most ethnic minorities in the UK are not first-generation immigrants, they were born in the UK. The majority of first-generation immigrants were asked to come to the UK to work during post-war labour shortages. Only a minority were fleeing persecution. Immigrants to the UK have always put more into the country than they have taken out. Professor Freeman's comments on the stress hypothesis are thus misinformed.

We agree with Professor Freeman that the ethnic density findings need much more detailed work to help make sense of the situation. In this regard, we point to the fact that qualitative methods are of particular use in investigating complex social systems.

We understand Professor Freeman's call for individualised care. However, we would feel better able to support him if the call was actually for individual choice of different models of care. There are some people to whom race, ethnicity and culture are very important; ignoring this or taking a ‘colour-blind’ approach offers them a poor service.

Professor Freeman states that there is a lack of serious debate on issues of racism in psychiatry and institutional racism. It is difficult to sustain such an argument. Although these issues rarely reach mainstream journals, there has been debate on this subject for decades in the UK, mainland Europe and the USA and there is a rich literature on these subjects (for a UK perspective see Reference BhuiBhui, 2002). Our modest editorial was an attempt to push the work forward and to link the literature to an outline service response.

No one can deny the need for more research but one must always balance the need for research with the problems with delay and the likely positive outcomes. Public health approaches have wide-based outcomes which must always be kept in mind when analysing their impacts. For instance, a public health policy aimed at reducing racist attacks, reducing institutional racism, improving schooling, supporting the family, decreasing the number of Black children in care and offering people skills to deal appropriately with discrimination, could have such a positive impact on society that it would be a reasonable initiative for psychiatrists to support, even if there was only a modest direct decrease in the rates of mental illness in ethnic minority groups. However, we are aware that there is a danger that protracted, scientific attention given to empirical questions might overshadow consideration of the more important error which is the implication that socio-moral tenets can be appropriately derived from science. Concepts such as liberty, justice and freedom from discrimination are neither determined nor justified by scientific results but flow from constitutional and moral principles. Science has a role in social policy but, in this regard, it is less in defining rights and more in developing methods for achieving rights. Cost—benefit analysis of public health measures based on rights is a moral discourse. Given the difficulty in raising research funds for research into racism, waiting for the evidence, or opposing such initiatives because of the lack of ‘evidence’, is not a zero-sum game. We have no initiative to decrease the rates of mental illness in ethnic minority groups in the UK. Is it not time that we did?