This note sets out simple ways to increase diversity at the English Bar, using the existing setup of the English Bar, but tweaking some small aspects, so without introducing any major structural changes. What this in turn means is that these ways should be easy to implement, given the collective will.
The starting point of this analysis is considering the 2 paths to becoming a practitioner at the English Bar. The primary path involves completion of an approved pupillage (i.e. formal legal apprenticeship) at a set of barristers’ chambers. The secondary path involves transfer across from another approved profession, such as the Solicitors or Legal Executives, plus completion of approved experience and required qualifying exams.
The number of students qualifying into the English Bar each year through the primary pupillage route is less than 600. (This number is currently restricted both through the limited availability of sufficiently well-funded pupillages, and by optimal Bar profession size, ceteris paribus). The total number of practising English Bar barristers is more than 16,000. So, the number of new barristers per practising barrister per year is less than 1 per 30 working barristers, implying an approximate, steady-state, average career length per practising barrister of c.25+ years.
Suppose, for simplicity, that we assume as a standardized model that all barristers work in average size sets of 30 barristers (in reality there is a long tail, power law distribution so most sets are much smaller down to standalone practitioners) and each model set takes one pupil per year for a combined 1st + 2nd 6 pupillage, leading to tenancy. Suppose also that we are considering a diversity subset, say an ethnic group, that makes up 10% of all pupillage candidates. Then one would expect, absent any bias or other factors, each model set to appoint one of the diversity pupils every 10 years on average, and the average waiting time for one diverse pupil to get pupillage at that model set, to be 5 years. So, it would need many years of observations to be able to conclude statistically, even just on the balance of probabilities test, that any individual chambers viewed in isolation was operating a biased (including implicitly or inadvertently biased) policy, if it in fact never, ever, appointed any diversity pupillage candidates. Furthermore, if all sets fell into this category then they could all end up operating in a systematically biased way, without ever actively coordinating, or being held individually accountable for this.
The recommended solution in this example is to pool chambers together up to a sufficiently larger combined scale, say into 20 chambers in a pool. (This essentially applies and exploits aspects of the so-called “law of large numbers”). Each pool would thus be chosen so that it was expected to appoint 2 of the diversity candidates to pupillage per year on average. Normal variation would be 1-3 per year, but zero per pool per year would need to be remedied immediately by an error-correction policy of say appointment of a good diversity candidate from a reserved list, and/or imposition of a substantial fine on that pool.
How the chambers in each pool managed the allocation of the appointments of the diversity pupil candidates between their constituent sets could vary e.g., by new policies, agreement, rota, lottery, etc, as long as the method chosen was within the law. The key would then be to mandate and monitor at the right granular scale that this pool approach was being applied, and getting the desired results.
Under the second method, the perceived diversity shortfall at the Bar could be offset by an increased incentive, perhaps using reduced fees, discounts and scholarships, and increased experience recognition for diversity candidates to cross over more easily from other professions to the Bar. There are of course many dimensions of diversity that could be quickly tackled in this way including across ethnicity, gender and age.
Macey-Dare, Rupert, Simple Ways to Increase Diversity at the English Bar (March 4, 2021).