Irrationality and unjustified accusations of racism: the fruits of INQUEST’s ideological blindness
INQUEST describes itself as ‘the only charity providing expertise on state-related deaths and their investigation to bereaved people, lawyers, advice and support agencies, the media and parliamentarians. Our specialist casework includes deaths in police and prison custody, immigration detention, mental health settings and deaths involving multi-agency failings or where wider issues of state and corporate accountability are in question. This includes work around the Hillsborough football disaster and the Grenfell Tower fire.’
INQUEST has published figures on deaths in police custody of black and minority ethnic (BAME) people, concluding: ‘The disproportionality in the use of force against Black people adds to the irrefutable evidence of structural racism embedded in policing practices.’ Here, one of DDU’s supporters, writing anonymously, argues that such claims aren’t backed up by the evidence.
INQUEST’s argument to show discriminatory treatment fails at the most basic logical level – that is, assuming at the outset the conclusion that is to be demonstrated. Going no further than comparing outcomes based on simple numbers of people in each ethnic population rules out from the get go any agency, other than police agency, in the outcomes. Without considering other variables, including cultural differences in life outcomes, it is hard to assess whether the actions of one group – that is, the police – are discriminatory; and if they are, what aspect of the other group – those arrested and restrained – is being discriminated against. Ideological commitments need to be held at bay while constructing, undertaking and interpreting data in investigating social problems.
Death in custody and arrest statistics
The chain of events leading to a death in custody begins with an incident and proceeds like this: incident – police assessment – arrest – custody – death in custody.
If one wished to assess whether deaths in custody were disproportionate by ethnicity and in particular, to assess the contributions of police officers when people are in their custody, then it is necessary to start from the beginning of that custody – that is, from arrest.
Making use of two official reports and its own data sources INQUEST draws conclusions of the following kind:
“From this table [INQUEST Table 1], Black people are 2.7 times more likely to die than the proportion of the population they represent.” Similarly – two times more likely from Table 2 data and four times more likely from Table 3 data.
(Tables 2 and 3 involve different combinations of death in, and not in, custody, but the argument from death in custody establishes the kind of analysis that is needed in these cases as well.) 
Looking now at the statistics for arrests we see that black people are 2.4 times more likely to be arrested than the population as a whole or three times more than white people. This is consistent with the 2.7 x, 2 x and 4 x times likelihood of death from INQUEST’s tables 1, 2 and 3 respectively. 
A comparison of arrest statistics with death in custody statistics does not reveal an anomaly. That is, there is no discriminatory treatment.
Death in custody and population level statistics
To extend the analysis back to the crude population figures used by INQUEST the chain of reasoning has two extra steps:
population – selection – incident – police assessment – arrest – custody – death in custody
To begin and end at the level of population omits each and every step necessary to understand deaths in custody and assumes that the population as a whole is homogeneous in respect of any characteristic that may determine an endpoint such as death in custody.
It is easy to show that the populations by ethnicity are heterogeneous – different from each other – in relevant characteristics. The key characteristic is the relative criminality of the various ethnicities – since greater criminality will lead to a greater number of incidents requiring police intervention.
Noah Carl has made an analysis using homicide as a proxy for criminality, concluding that the percentage of police killings and percentage of homicides by ethnicity are very nearly the same. In other words, according to Carl’s analysis, black people are no more likely to die than white people, when we take black and white homicide rates into account. 
From a further dataset, compiled from data that INQUEST obtained from the IOFC, INQUEST concludes a sevenfold increase in the likelihood of death in particular circumstances – but this time not with reference to ‘the proportion of the population they represent’ but compared with the likelihood for white people.
In this case, the chain of events is as follows:
population – selection – incident – police assessment – arrest or detention – police assessment – restraint – death in or following custody
Limiting the population in this way to those cases for which restraint was used, further narrows the selection. The narrower the selection, the more likely a finding is to be both interesting and spurious. There were 23 cases involving restraint and black people over the eight-year period 2012/13 to 2020/21. 
Setting this aside, what factors might make the black and white populations significantly and relevantly different at the restraint stage so far as the subsequent death is concerned? Official statistics show that ‘a significantly higher percentage of Black men (3.2 per cent) experienced a psychotic disorder in the past year than did White men (0.3 per cent)’. That is, Black men are 10 times more likely to suffer a psychotic disorder than White men. It might be the case that people with psychosis are more likely to be restrained and may be more likely to die under that restraint. 
In summary, INQUEST makes the same basic logical error in this analysis – assuming the matter to be proved rather than demonstrating it – as in the earlier analyses.
Mental health examples
The cases of death in custody INQUEST has chosen to highlight are all mental-health cases, sometimes drug induced, and involve restraint. These most difficult cases may be the most effective probes of the robustness of the procedures – and the Coroner’s recommendations quoted include constant supervision of persons in restraint and the training of police and others about the consequences of restraint.
Cultural factors and ideological blindness
INQUEST is right to mention the importance of education. Only with that advantage can children, as they become adults, realise that they have choice over their life trajectories and are not confined to those that may be both readily culturally available and harmful. We can see this in the arrest statistics for children, by ethnicity where black children are 2.7 times more likely to receive a caution or sentence than all children  . Policing is not a substitute for socialising at home and at school.
Striking disparities in crime statistics (and in educational outcomes) resulting from cultural factors need to be recognised. Without recognition of problems how can problems be remedied? Externalising problems to other people is at best a distraction and at worst a generator of racial animus that may not have existed before.
The ideological blindness that does not require going beyond simple population by ethnicity suggests a prior belief in racism, and that no individual has a causative role in their own life outcomes.
[Tables 1 & 2 page 41, Table 3, page 43]
[Figure page 46]
[Section 3. By ethnicity and sex]
[Percentage and number of people within each broad ethnic group that falls into each age group]