An Empirical Approach to Equal Opportunities
Over the course of the last decade, public discourse relating to equal opportunities (which phrase I use advisedly in place of the more fashionable but problematic “diversity”) has become increasingly polarised and subject to dangerous sectarian political extremism, often cynically motivated, as ably exposed by Helen Pluckrose and James A. Lindsay in their best-selling work, “Cynical Theories; how Activist Scholarship Made Everything about Race, Gender, and Identity — and Why This Harms Everybody”. A more condensed version of their commentary, specifically on the question of race, is available for free on the New Discourses ‘blog, entitled, “How to be not-racist”: reading at the very least the executive summary of that is highly recommended; I will not repeat its analysis in full here.
As Pluckrose and Lindsay explain, the cynical theories (self-titled “critical” theories) are fundamentally anti-intellectual, seeking by dishonest obfuscation and often intimidation to suppress critical scrutiny of the claims that they make. This usually involves seeking to suppress rigorous reasoned inquiry generally, demanding that the claims made by the theories be accepted without evidence and seeking to intimidate those who require evidence by claiming — dishonestly — that the very act of refusing to believe the claims made by the theories without properly controlled empirical evidence is of itself, for example, racist, implicitly inviting others to impose serious and harmful personal consequences on those so deemed (a practice popularly termed “cancelling”). In the context, for example of race, the phrase “white fragility” is cynically used to suppress dissent, and constitutes responding to reasoned criticism with what is no more than thinly veiled personal abuse, made all the more sinister by its racial overtones. As with virtually all attempts to suppress or obfuscate reasoned scrutiny, as nearly everyone who has conducted a cross-examination in court will well know, it is motivated by knowledge that the claims made by the theories are incapable of withstanding such scrutiny.
Fundamental to an effective legal profession, and to the rule of law which it exists to serve, is precisely that rigorous analytic scrutiny that the cynical theories seek to destroy. A properly functioning legal system is one which requires sufficient evidence before reaching a contested conclusion, and allows the evidence for that conclusion and the reasoning leading to it to be subject to open and rigorous testing by those with an interest in that conclusion.
None of this means, however, that there are no equal opportunities problems to solve: far from it. The idea that one must either be an “ally” of the cynical extremists or complicit in, for example, racism, is a false dichotomy convenient to the extremists as a means of intimidating those who propose other means of dealing with what are often real problems the existence of which helps to lend the cynical theories much of their popularity. For this reason, the extremists have a cynical motive to obstruct progress towards solving the real problems, as that would reduce the popular support of their theories from which they often personally profit, whereas those opposed to them have an equally strong motive to make as much progress as possible in resolving the real problems for precisely the same reason, and an equally strong motive to expose and marginalise the genuine bigots of all political persuasions who have no desire to maximise equality of opportunity at all. In this, as with many aspects of life, the extreme right and extreme left of politics are far closer to each other than either are to the centre, which exposes the fallacy of a one dimensional understanding of politics.
As to what the real issues are, data from the Bar Standards Board[1] show that, controlling for class of degree, the proportion of BPTC graduates declaring themselves as “white” on equal opportunities monitoring forms obtaining a first six pupillage was notably higher than those with other self-declared racial/ethnic categories; of those with an upper second class degree, for instance. 45.1% of 2014–2018 BPTC graduates in the “white” category had commenced a pupillage by March 2020 compared to 20.8% in the “black/black British” categories and 22.9% in the “Asian/Asian British” categories. Contrary to the claims of the cynical theories, this does not necessarily mean that pupillage providers are inherently racist, but it is certainly a cause for concern that requires further investigation for anyone who truly believes in equality of opportunity.
One cannot sensibly hope to address an issue successfully without first understanding it and what causes it, and that understanding requires the full rigour of the scientific method. There was once a time when medicine was practised, not on the basis of science, but on the basis of anecdote and untested theory. Physicians would prescribe remedies that had undergone no clinical trials and whose efficacy was often based on no more than a case study with no means of telling whether the patient’s recovery in that case had anything to do with the treatment in question; sometimes, by chance, these remedies were effective, but at least as often they were useless or even positively harmful; indeed, there was a time when, for precisely this reason, deaths in childbirth were much higher among the wealthy, who could afford medical practitioners, than poorer people who could not. Fortunately, medicine (with the exception of the mostly fraudulent “alternative medicine”) has now moved beyond the questionable methods of the past and is largely founded in science. The resulting benefit to humanity has been incalculable. Unfortunately, social sciences have, for the most part, not yet made similar progress, and as many purported remedies to social ills do harm as do good, complicated, in the case of the social sciences more than in the case of medicine, by cynical self-interest in preventing the true position from being understood. To deal with the problems of equality of opportunity, and other social problems, this must change, and change urgently.
In the case of equality of opportunity in recruitment, a scientific approach involves first of all identifying, in the case of each individual recruitment process, which characteristics of candidates that are not themselves indicators of genuine merit affect candidates’ chances of progressing through each stage of the recruitment process (sift, first round interview, second round interview, and so forth). Each round must be treated separately as what is causative in one round may not be causative in another. A mathematical analysis of the data is necessary to calculate the “p-value” of each characteristic in question affecting the chance of progressing through the relevant recruitment round. The “p-value” is the probability, given the data, that any pattern in the data (e.g. that people who declare themselves as “white” in the monitoring section of the form are more likely to progress to the next round) could have occurred by chance. A p-value of 0.05 (i.e. 5%) is by convention considered the threshold of significance; it is important, to avoid any possibility of bias in interpreting the results, to set the threshold of significance before calculating this value.
An analysis showing a p-value of >0.05 does not necessarily mean that the process is free of bias; there may be insufficient data to reach a conclusion either way. For that reason, the data gathering and analysis ought to be continual and cumulative so that every recruitment round increases the accuracy of the data.
If a p-value of <0.05 be found, that suggests a possible problem. If possible, one would control the test to see whether the difference can be explained by, for example, differences in academic results, but the current monitoring data on the Pupillage Portal does not allow this to be done unless chambers implement their own monitoring data gathering system; to enable more robust monitoring, the Pupillage Portal ought to be modified so as to allow controlling of data in this way.
If an issue has been identified, the next step is to formulate a hypothesis and then, critically, to test the hypothesis in a controlled manner: one should never make the mistakes of the physicians of the past by assuming that any hypothesis that is internally consistent and is also consistent with currently known facts must be true; to have reason to be confident in a hypothesis, it must survive the test of falsifiability. For example, if the hypothesis be that candidates who have had sufficient opportunity to engage in interesting-sounding recreational activities whilst at university (e.g. playing sport) might tend to be those candidates from wealthier backgrounds such that taking into account having engaged in these activities in a candidate’s favour disadvantages candidates of equal ability who lacked those opportunities, e.g. by reason of having to study part-time, instead of simply redacting all references to recreational activities in the following year’s application forms, take a randomised control and redact the recreational activities in half the forms, leaving the others unredacted, and test whether there is any statistically significant difference between the two in terms of outcome (using the same analytic method used to find the disparity in the first instance). Only if this measure be shown to be effective by this test can one have confidence in the original hypothesis. (That does not necessarily mean that the measure ought not to be implemented; if the cost of introducing the measure be low and its possible benefits, although highly uncertain, be high, then it might be worth implementing it despite the evidence being inconclusive; one would not refuse a free lottery ticket because one could not be certain of winning; but the analysis would be different for an equally uncertain measure with a higher cost and/or lower potential benefit).
In cases where there is already scientific research on the matter (e.g. showing the efficacy of redacting names in application forms and competency tests in interviews), the measures indicated by the existing research should be taken.
Further, if chambers and other organisations over the whole legal sector, and beyond, were not only regularly to conduct this sort of controlled test, but publish the findings in detail, and allow them to be rigorously peer reviewed, a proper body of scientific knowledge of bias-minimising and equal opportunity and merit maximising recruitment practice could develop that has the potential to transform recruitment practice and equal opportunities generally in the same way that science has transformed medicine, and confine the cynical theories to the past in the same way as the scientific practice of medicine has confined miasma theory to the past.
This sort of process should not be considered as mere burdensome regulatory overhead: quite the contrary — minimising non-merit related discrimination in recruitment is commercially valuable, as any organisation will profit from maximising the merit of those whom it recruits. Contrary to the falsehoods promulgated by the cynical extremists, minimising bias and maximising equality of opportunity is not a zero sum game, but rather is an enterprise to which any rational person ought to be committed.
[1] Bar Standards Board BPTC Key Statistics 2020 4.A.7.1, ethnicity, first degree classification and BPTC overall grade of full-time and part-time BPTC graduates commencing first six pupillage; UK/EU domiciled students — 2014–2018 aggregated.
James E. Petts — The 36 Group
