We are surprised that Professors Singh and Murray and Dr Fearon (pp. , this issue) refuse to engage with the concept of institutional discrimination. We agree that institutional racism is not something that can be tested empirically or investigated readily using the standards of positivist scientific proof. It is therefore not unsurprising that we do not have the kind of ‘evidence’ that Professor Singh calls for. However, the lack of previous research should be a call to action, not a reason to ignore the issue.

We know that many Black people who have received mental healthcare perceive discrimination in services (Sainsbury Centre for Mental Health, 2002; Norfolk, Suffolk and Cambridgeshire Strategic Health Authority, 2003; Mental Health Act Commission, 2006). A sociological framework is required to understand better the way in which this occurs. Without an acceptance that there are differing ways of knowing and understanding the world, further dialogue is limited by emotional and anxious fears.

Using a wider ‘sociological lens’ allows us to appreciate the ways in which particular approaches to knowledge are used to sustain race-neutral conceptions of certain fields of endeavour and suppress the experience of other voices (Reference NkomoNkomo, 1992). The logical conclusion from Singh, Murray and Fearon's argument would be the adoption of a colour-blind approach to mental health service provision. Yet it is exactly this kind of approach that has resulted in the faulty premise that because institutional racism has not been assessed in a randomised control trial it does not exist. There should indeed be more research on institutional racism as it is a largely neglected issue in the field of organisational studies.

Solomos (Reference Solomos1999) criticises the definition of institutional racism given by MacPherson and cautions against ‘using such terminology… loosely and rhetorically’. The discourse must be given a context within the particular setting being considered, as only then can we begin to understand how the discourse itself determines our conceptions about race or ethnicity.

We do agree with McKenzie & Bhui that it is important to focus on the needs of patients. Notwithstanding our concern that institutional discrimination may be one factor in the over-representation of Black groups in mental health services, we believe that much more must be done to tackle the problem. The disproportionate admission and detention rates for Black groups, especially (young) Black British men and women, suggests either serious problems with public services (including mental health services) or an epidemic of mental illness in these groups. Whichever it is, whether it is a mixture of factors, it demands urgent attention, with the recognition that this is not solely a health service issue but requires concerted cross-government action.