As always the elephant in the room is aptitude differences. The UK has an enormous data set for IQ of secondary school pupils, from a test called the CAT (Cognitive Assessment Test):
The CAT shows that non-white pupils, taken as a whole, are slightly less intelligent than white pupils; that is, the white IQ is slightly above the mean of 100. However, there is enormous variation between different groups of non-whites. Indians perform about the same as whites, blacks and Pakistanis underperform whites, and Chinese outperform whites (note that the quantitative and non-verbal reasoning scores, in which non-whites do better than in verbal reasoning, are probably more representative because they tend to wash out the effect of English not being some pupils' first language).
Now the government is implying that the racial differences in admissions are due to prejudice rather than bulk differences in the candidates' actual performance. This hypothesis predicts that all these diverse groups of pupils with different aptitudes but all without traditional British names (including "white other groups"
should do worse in admissions than pupils with traditional British names (including non-whites) to a similar degree. If in fact aptitude is the deciding factor, we would expect pupils with Chinese names to do better than pupils with traditional British names, who in turn should do better than pupils with Indian names (due to the small penalty they suffer on the verbal reasoning tests, probably due to ESL), who should in turn do better than pupils with Pakistani/Bangladeshi names, who should in turn outperform pupils with African and distinctively Afro-Carribean names.
In fact Afro-Carribeans, who have the worst aptitude scores, are probably no more likely to have non-British names than white pupils. Are they represented on-par with white pupils, or are they underrepresented? This is a strong test of whether universities are prejudiced based on racial hatred, or whether they discriminate correctly based on aptitude; in this case, underrepresentation of Afro-Carribeans would show that universities are
less likely to be prejudiced.
It is so easy to conclusively prove or disprove whether universities are discriminating based on prejudice (e.g. names), or whether they are discriminating entirely appropriately based on aptitude differences, that I am not sure why these tests have not been carried out and the results published. If it turns out that universities do discriminate based on race (e.g. if Chinese are underrepresented relative to white British) then this seems like a limp response to a real problem. If in fact the government has performed this analysis and not found evidence of such discrimination, it's difficult to see why pursue this policy at all.