Scientific hypotheses should be in principle testable and the onus lies on the one proposing the hypothesis to present supporting evidence. The assertions about institutional racism in psychiatry have never been based on evidence. Instead the onus is placed on services to prove that they are not institutionally racist. We should therefore first agree that the charge of institutional racism is an ideological position rather than a scientific fact.

Patel & Heginbotham illustrate this difference between fact and opinion in the two testable statements they make: ‘we consider that institutional racism in mental health and in wider public services is a contributory factor’ and ‘all the evidence suggests that Black people and many people from other minority ethnic groups are being admitted to and detained in psychiatric hospitals either unnecessarily or at disproportionate rates’. The authors provide no evidence for either of these statements or a conceptual or clinical model whereby these could be tested. Simply asserting one's belief, no matter how strongly held, is not sufficient. Why do they consider that institutional racism is a contributory factor in ethnic differences? How does it contribute? What does it contribute to? If it operates at an institutional rather than an individual level, what processes does it influence? An institution is ultimately a collection of individuals working within certain policies and procedures. If the individuals are not racist, then the policies or procedures must be responsible. So which psychiatric policy leads to higher rates of psychosis among minority ethnic groups? The role of social adversity as a causal factor in psychosis is well-documented and researched. Should we ignore that evidence and instead believe that institutional racism is to blame?

The necessity for psychiatric detention is predicated on the nature and severity of illness, need for treatment, presence of risk and lack of alternatives to hospitalisation. There is some evidence that factors contributing to detention, such as social isolation, delay in help-seeking, stigma, ethnic differences in family attribution, lack of general practitioner and help-seeker, and lack of community support are higher in some minority groups. Where is the evidence that it is institutional racism and discriminatory practice rather than such socioeconomic and clinical differences that lead to differences in detention rates?

The recognition by McKenzie & Bhui that we should put patients first is long overdue and most welcome. Perhaps they will join us in moving the debate away from racism to more relevant underlying sociocultural factors. Their claim that ‘some ethnic groups do not get equitable mental health treatment’ is backed by two references. I would strongly recommend that readers peruse these two citations for scientific robustness before accepting this claim. Few would disagree that where the availability of services, such as interpreting services, trauma therapists, etc., does not meet the needs of the community the situation must improve. So can we therefore move away from charges of ‘eurocentric misdiagnosis’ and ‘racist coercion’, which have been so damaging to trust between minority ethnic communities and services?

I do not misunderstand the concept of institutional racism, nor did I create a ‘straw man’. I simply pointed out that the claim of institutional racism is scientifically hollow, empirically untestable and clinically corrosive in everyday psychiatric practice. The bogey man of psychiatric discrimination has been created by those who seek to ‘ racialise’ all differences between population groups, ignoring the complex socioeconomic and cultural variables for which race and ethnicity are proxies. We should all welcome the search for the reasons underlying ethnic differences, as long as it is based on a scientific paradigm rather than on belief.