Sorry to take so long to come back to the thread. I have not been idle!
Firstly, I’ve been looking at the available statistical data to see if we can draw some inferences about what’s significant and what probably isn’t.
For context, there are some fairly blunt statistics available about missing episodes in the UK compiled by the charity Missing People, based on National Crime Agency data. For 2019–20, there were over 350,000 missing incidents, more than 137,000 involving adults and nearly 215,000 involving children. That number of episodes is generated by nearly 98,000 unique adults and 70,000 unique children. From that we can infer that children (strictly defined as under 18s) are more likely on average than adults to go missing multiple times. The most vulnerable individuals are Looked After Children (in care or with foster families), who are more than 20 times more likely to go missing than children as a whole, and on average each go missing six times.
Key information and statistics about missing - Missing People
So that’s a lot of missing people, and a lot of work for the police and other agencies to do, and you would think that it might be helpful to be able to break those numbers down by something more than adult/child, unique/repeat, but in fact managing to do that is another matter.
@Angleterre is completely right in saying that comprehensive statistics about missing persons are not standardised between the various different forces. In England and Wales, there are 43 different forces, using a variety of different computerised systems. A number of forces use a system called Compact, which uses data gathered to generate actions or tasks to be done, which (in theory at least) keeps an investigation on track. But many forces use different systems and the data gathered by each is different, and is more, or less, complete in terms of the sort of demographic information we might be looking for in order to establish if resources are being allocated evenly.
Or at least, we might guess so, because as well as issues around what data is input, there are also issues around what data comes out of the system. If you want figures relating to missing cases for any given force, you can submit a FOIA request – a request made under the Freedom of Information Act. But FOIA requests have an upper limit of 18 hours’ work for a police force, and if they estimate your request will take longer to process then they can refuse it – and in practice they often do.
Whether this is because any demand on force time is in effect a debit on operational time and is resisted reflexively, or whether it’s because data is collated in ways that aren’t amenable to efficient extraction, is another thing we just don’t know. So the one thing that comes out of my research loud and clear is that there’s a yawning gap for the police and the various agencies and charities dealing with missing persons cases to have a comprehensive, responsive, standardised data handling system, which, at present, they just don’t. There is a national missing persons database supposedly in development for use by the police, but when it will be up and running is anyone’s guess.
In the meantime, then, trying to find statistics that shed any light on the role of race in investigative priorities or efficiency is basically impossible; likewise lots of other important variables, such as socioeconomic status to name but one. The idea that we might be able to request a list of deaths and disappearances involving water broken down by age, sex and race or ethnicity, and then match that data to investigative costs or outcomes would appear to be the stuff of fantasy at the moment. I was very surprised by that, when you consider the range and sophistication of data capture and analysis in the commercial sector, but there it is.
Various initiatives are trying to plug, or sidestep, that gap. For instance, Missing People has a survey open at the moment to try and generate qualitative data about people’s personal experiences of racial discrimination in the course of a missing persons enquiry, so that they can campaign to improve matters despite the lack of hard data.
Experiences of discrimination - Missing People
Then in addition there’s a certain amount of academic research in process that’s trying to establish whether and to what extent and effect bias is in play in missing persons cases.
According to a paper presented by a team from Goldsmiths London at the International Conference for Missing Adults and Children at Portsmouth Uni last week, we know that around 14% of the UK’s missing population are BAME, compared with around 3% of the population as a whole. We also know that BAME people are less likely to seek help with mental health problems, which play a key role in missing episodes. Research on media coverage has established that missing white people get around three times more media coverage than missing BAME people (‘missing white woman syndrome’) and a study of the social media feeds of the Met and the GMP has shown that posts about missing white people attract ‘significantly’ more public engagement. So that prior knowledge points to some of the known issues around disappearances of BAME people.
The Goldsmiths research is particularly concerned with racial bias in resource allocation in police investigations of missing people. Racial bias in police decision-making and resource allocation generally has been studied to some extent, and has shown that crimes with BAME victims are less likely to be solved, that BAME rape victims are more likely to be blamed for poor case outcomes, and that high status victims are more likely to have resources allocated to them
and that BAME victims are less likely to be considered high status. However, the specific phenomenon of missing people hasn’t been studied through this prism at all, and, as we’ve seen, hard statistical data that would allow that study isn’t available. So the Goldsmiths team designed a vignette study, which involved showing police decision-makers summaries of missing person scenarios that were identical except for the names and ethnicities of their subjects, in order to gauge whether and how responses were liable to change as a result.
What they found was that participants were poor at deciding which subjects were at high risk, which is significant to resource allocation because the perception of high risk correlates to the application of resources, but what they did
not find was any racial bias in that decision-making. I think it’s important to remember also (and this is me speaking now, not the researchers) that in missing persons cases particularly the heavy application of resources may well not correlate to satisfactory outcomes.
A vignette study involving a comparatively small number of participants is hardly definitive in addressing the quite complex concerns expressed on this thread (and elsewhere, obviously) and it’s clear that more research with harder data is needed to begin to understand the role of racial bias in cases such as these, but the reality is that that research, and indeed that data, just doesn’t exist currently, or not in the UK anyway.
So that’s the best I could do on that front. I have also looked at each of the cases mentioned on the thread individually, some more closely than others, and will come back to the thread to update as and when I have time over the next few days.
Apologies for the epic post. Hopefully it’s of some use or interest.
I'll see what I can find out.