That's an interesting study, especially the BAME aspect. It's a very high percentage, proportionally
Could this be because the % is being compared with a national % demographic instead of a local %
If an urban area has a higher BAME population and the areas with the highest numbers of Covid cases or deaths are also urban. Then it would skew the data when compared to the whole country and its national % ethnic demographic.
So if London, or Birmingham, has a high number of Covid cases, compared with a more rural area (same population for the sake of the maths) then you would expect more BAME cases compared to the national ethnic demographic, simply because more BAME people live in urban areas, where there is more covid19
Hope that makes sense, maths is inst my strong point 