My prior would have been brexit-related changes reducing the number of smart but non-rich students coming from the EU, but I don’t actually know how the numbers there changed. Also, what are the rankings based on? When I was applying, I think it was lots of stuff like:

- grades of incoming class (the changes 'pg alleges could lower those grades even if th actual quality of the incoming students don’t change. Balance by subject can affect this too as eg science students tend to have more UCAS points. Private school students may also have more UCAS points because their schools are more likely to do things like putting students in for extra A-levels or GCSEs (taking those exams costs the schools money)). Alternatively, university funding is in a dire state in the U.K. (though less so for Oxford and Cambridge given their endowments?) so maybe they can trade prestige for letting in a larger number of international students who pay full fees but who would have otherwise not met the bar.

- research output metrics, which seem quite unrelated to undergraduate selection – there is a high lead time and if you get the selection wrong you can still hire researchers from elsewhere. These metrics also seem somewhat gameable

- metrics around outcomes for graduates. I wonder how biased these are by subject mix (ie how much is this just a measure of what percentage do courses that lead to good programmer/finance jobs) and how much they are affected by students perusing further education. I think to some extent this can also be affected by class mix because more privileged students may find themselves in better jobs (either because of parental connections or just class filters in hiring though one would hope that the university would train students to be able to pass such filters)

I recall being sceptical of these league tables when I was applying many years ago for reasons like these (not that it stopped me from applying to highly ranked universities).

Though comparing to American schools, I do think there are reasonable advantages to going to the US – you’re much more likely to work in the US (and therefore likely to get paid a lot more) if you go to a North American school. If you’re trying to compare Oxford to Harvard (with offers) and the financing works out either way, it seems to me Harvard would obviously be a better choice today and 10 years ago before the ranking changes. I’m not sure what the quality of US school is where you prefer Oxford.

One other thing: Oxford and Cambridge delegate a lot of admissions to colleges so I’m not sure how much one can claim that it is a global shift in attitude, though there are some ‘second chance’ mechanisms and schools that send many students to oxbridge will have better recommendations for which colleges to apply to, and the policies between colleges can still move in a coordinated way even if each college does its own policy.