There is nothing like second hand limited anecdotes to support a generic extrapolation.

I just started reading it – see https://www.sra.org.uk/sra/research-publications/potential-causes-differential-outcomes-legal-professional-assessments#exeter I cannot see a way to download it as a pdf which would be a lot better. In a sense it sums it all up – we get a mickey mouse summary only, very hard to get the details and then they treat readers like children who need tiny snippets page by page; and then they point out that exams are too hard for some people if they closed book and that perhaps we should change the rules presumably to give “prizes for all” even if they cannot pass the exams/put in the work…. anyway I will read on with an open mind.

I agree. My sister’s best friend is a lawyer. She is black with an accent. Slightly over 30.
She shared quite a few stories from her TC and from client meetings. Nothing direct but a few clients, particularly white men aged around 40-50, seemed shocked or disgusted than she dared to speak during the meeting.

The headline statement says “White students more likely to secure law firm sponsorship”. Is this correct or does the research instead say that of the people who are on a course, X% from Y demographic have also got (or not got) sponsorship?

The suggestion is that proportionately, if a black or other ethic minority student were to apply for sponsorship, they would be less likely than the same white student in obtaining the sponsorship.

Would be interesting to, e.g., look at the statistics of the top 15 law firms in the City and publish the following data:
1. Of all the applications made to your firm, are you proportionately more or less likely to offer a place to a white student than to a minority student? If a bias does emerge from this data, in whose favour does it lean?
2. Is the make up of incoming and current trainees representative of UK demographics? Do you disproportionately have more white students? Is there balance or is there a disproportionate bias towards particular demographics?

I think this would be really helpful in assessing whether there are biases prejudicial to particular groups.

Put simply, differences between ethnic minorities are more significant and frankly more interesting than those between the white majority and the various ethnic minorities. The failure to properly granularise these findings is at best lazy and at worst disingenuous.