The Race Report 2021
For those who are unaware, the Conservative Government commissioned a race report in the UK to examine the presence of institutional racism in the UK. This is in relation to the protests this side of the Atlantic which were triggered by the murder of George Floyd, May 2020 and subsequent protests in the USA.
The vast majority of ethnic minorities in the UK would have predicted that the report would have, without a doubt, found a presence of institutional racism in the UK. However… it didn’t.
This post is not exploring the evidence of institutional racism in the UK, but is looking at the methodology of the report which is, in my opinion, riddled with bias. Bias we can define as
“disproportionate weight in favor of or against an idea or thing, usually in a way that is closed-minded, prejudicial, or unfair”
The issue is that bias gets in the way of truth. I argue that the bias in the Race Report is so strong that it gets in the way of the truth. I.e. due to poor methodology, from an objective standpoint we don’t know if there is institutional racism in the UK. However, due to the publicity that the race report has gained, as Doreen Lawrence stated, it has potentially knocked UK race relations back ‘20 years or more’.
Quality of public argument
The most disappointing thing for me is the low quality of public opinion on both sides. The psychological phenomena of “authority” leads people to take information at face value when they are proposed a notion which is supported by long technical jargon they don’t understand. For example they might say:
“Just trust the experts… the report is conclusive…”
Individuals who argue in this manner have been baffled by the length of the 256 page report and the technical jargon it proposes. They don’t have the statistical literacy to understand the report and are therefore forced to take it at face value.
Then we have the other common narrative which is, for ethnic minorities…
“I disagree with the report… it doesn’t match my lived experience”
The problem with this report is that it relies on anecdotes (subjective personal experiences) which is far weaker than data. It would be like relying on Donald Trump for medical advice after you have tested positive for coronavirus, just because bleach and hydroxychloroquine worked for him… I wouldn’t advise it.
The issues with both reactions to the race report is that they are focused on the outcome. But the outcome is predicated on the methods. If flawed, the outcome is rendered useless because it is unreliable. The race report, in my opinion is riddled with methodological flaws, rendering the outcome unreliable for public use.
If I can teach one person to critique the report in a way which enables them to argue against its findings in a logical and effective way then I consider that this post has been successful. The overarching fear that I have is that this report could become weaponised in supporting passivity in the face of racial injustice.
Critiquing is the process we go through to analyse the effectiveness of a chosen subject. Bias gets in the way of finding these underlying truths and disrupts effectiveness. We’re going to critique the report to discover the underlying biases.
Remember how so many individuals were protesting at the beginning of the pandemic. Many of them denied the existence of COVID-19. If I were to ask one of these deniers to write a report on whether COVID-19 exists, do you see how I may have chosen poorly?
This is due to confirmation bias. Regardless of what this individual finds when examining the evidence to support or oppose the hypothesis that COVID-19 exists, they will subconsciously try to support their bias, which is that COVID-19 doesn’t exist.
So, when Tony Sewell, who has denied the existence of systemic racism for over a decade is asked to write a report on whether systemic racism exists or not, we can suspect to see some confirmation bias. Tony Sewell has also been accused historically of simplifying the cause-effect relationships of race and sexual orientation without providing data to support his theories.
“We heteros are sick and tired of tortured queens playing hide and seek around their closets. Homosexuals are the greatest queer-bashers around. No other group of people are so preoccupied with making their own sexuality look dirty.”
“What we now see in schools is children undermined by poor parenting, peer-group pressure and an inability to be responsible for their own behaviour. They are not subjects of institutional racism. They have failed their GCSEs because they did not do their homework”.Tony Sewell
Confirmation bias reduces the efficacy of the report.
If I wanted to sell you an idea, I may want you to look past the realities that may discourage you from not buying into that idea. For example, if I wanted to sell you a pint of milk, it is unlikely that I will show you an image of lines of cows, festering together, puss oozing from their udders as they’re pumped for milk for 8 hours a day. Instead I will show you a cow in a large green field with a proud plump farmer right by its side.
You would expect to see this in an advertisement but not a government sanctioned report which should be providing information which is as close to the truth as possible. This means balancing both sides of the coin. No persuasion. Just an open and honest depiction of the reality or at least as close to this as possible.
Page 28 tries to paint a utopian view of Britain during the Olympic ceremony, recalling memories of racial unity as we see the generation of West Indians emerging from the Windrush ship. It does fail to mention the gentrification that followed, forcing the descendants of those West Indians out of east London.
Page 8 went further than this, promoting the notion that the slave era should be narrated as the foundation from which Africans transformed themselves into British Africans. A novel experience almost, or at least a worthy evil for the production of these transformed Africans.
I don’t know many of the black diaspora who would agree with the points made on page 8 and 28. I do however know that these statements will cause many people to abandon their rationality as they look forward in the report to justify this utopian picture that the report has created, increasing their bias as they do.
In short, this is cognitive priming and has no place in a government report. For the unaware reader it will project bias throughout their reading of the race report. It depicts an image which isn’t true, nor is it reflected in the experiences of those black people that they are referring to.
Cognitive priming manipulates the reader to disregard untruths.
Imagine if you wanted to investigate my garden because you were concerned that some flowers weren’t growing as fast as they should do. You are interested in equality so you want to make sure that all of the flowers are growing at a similar rate.
I encourage you not to worry and that the flowers are only 2.3% behind the expected growth. You’re happy, however on closer examination, you see that the yellow flowers are 30% ahead of schedule and that the blue flowers are 33% behind schedule.
This is how I could use selection bias to fool you. I have selected a part truth to relay to you a generalised reality because it fits the point that I’m trying to make; that point being that the flowers are fine and on schedule.
Page 28 argues that the pay gap has reduced to 2.3% between BAME and non-BAME people. But then we look deeper at the study referenced and we see that black people earn 13-15% less than their white peers. This generalised report of the data allows the race report to hinder a deeper look at the discrepancies. In fact despite the inter-ethnic discrepancies being found in the same study they referenced, they fail to elaborate.
This is also similar to publishing bias where they only show you what they want to.
To demonstrate the harm of selection bias or publishing bias, a large pharmaceutical company used selection bias to pass a drug in west Africa to treat HIV which was no more effective than paracetamol.
Selection bias is dangerous and sabotages the uncovering of fundamental truths.
Publication Bias / Poor critique
The following issue is a bit of a grey area because it’s not quite publication bias and is more just plain bad reporting. The reporting however is so bad that I would assume it is bad due to intentional bias. Ie someone felt the need to report on this study but because the study didn’t show them what they wanted to, they decided to disregard the outcome.
Studies which have very low variability and large sample sizes which produce results that can be replicated again and again are considered to be very reliable.
This would be the equivalent of me giving 1000 people a drug but unbeknown to them, half are fake (placebo) and half contain medicine (variant 1). It would be easy for me to measure the success of this study because the individual taking the drug is unaware of whether they’re taking a placebo or not. If I were to replicate this study over and over again, I can then be assured that the results did not occur down to luck, hence why it would therefore be far more reliable.
Page 121 refers to a study with this level of efficacy yet disregards its utility. Why? I have no idea.
The two studies quoted involved the scientists submitting 2961 and 3200 job applications respectively, only changing the surname. They found that an individual with non-English surname would have to submit their CV between 50-90% more times to gain the same number of interviews. Both studies reached the same outcome.
Normally, findings such as this would promote more qualitative research to balance the quantitative and qualitative data. Ie a study to investigate the personal experiences of ethnic minorities trying to gain employment. Lo and behold studies such as this were quoted and disregarded just as these two studies were. Surprisingly a shallow look at the Olympic opening ceremony is relevant but the findings of these two studies with high replicability aren’t. Something is fishy here, in my opinion. It is blatant bias.
The BMJ, internationally renowned for its reliability in interpreting medical data also identified “cherry picking” throughout the report and argues that institutional racism is indeed a factor within the healthcare system.
“It lacks the scientific credibility and authority to be used for major policy decisions. Its methodology and language, its lack of scientific expertise, and the well known opinions of its authors make it more suitable as a political manifesto rather than an authoritative expert report”BMJ, British Medical Journal
Amanda Parker from the Financial Times also wrote of cherry picking in the race report, ignoring the University of Aberdeen’s research, which concluded that whites out-earn certain ethnic minorities irrespective of education. Ie white without degrees earn more than some ethnic minorities with degrees.
The report also minsconstrudes data by mixing different narratives. Ie on page 29 without reference, the report states that “numerically largest disadvantaged group is low income White boys, especially those from former industrial and coastal towns, who are failing at secondary school and are the people least likely to go to university”.
This is a per-population narrative, ie they’re elevating the importance of white working class boys because there are more of them. I do not disagree that there is an argument to be made about the disenfranchisement of the white working class. The problem is that they have swapped from a per capita narrative (ie likelihood per person) to a per-populous narrative (ie the size of all individuals added) which are completely different arguments.
We saw the same issue with popular views on coronavirus deaths last year. The notion that “it only kills the elderly” started to become commonplace. This was the majority prioritising their wellbeing over the minority. Ethically the issue with this is clear, however this shares similarities with the conclusions made in the race report.
This choice in perspective can also be seen on page 122 where the race report explains that a chosen study shows that 11% of white employees vs 29% of black employees feel that their race has held them back in the workplace. The race report states goes on to say “however this is not the majority” of black people. Equally, someone could say that almost triple the number of blacks report that their race prevents them from excelling in their career. The shift in perspective creates a different argument. I would say the latter is a more accurate reflection on the data
Their methods of reporting are biased leading to a disingenuous interpretation of the date.
To conclude, does institutional racism exist?
The answer is, due to the missed opportunity in this government report, we don’t know. This was an opportunity to review current literature in an holistic way, however due to the obvious bias we are left none the wiser. We can rely on our own interpretation of the data, however the likelihood that another report which has had 10 minds and multiple hundreds of thousands invested into it will come into fruition in the near future is unlikely.
My fears are that this report could be used in a very harmful way for our community. If that does happen, we must unite. We cannot wallow in despair. We must bring the best of our academic persons together to fight the narrative which works against us.