The national applications data shows how this produces a counter-intuitive outcome: applicants with ATARs above 90 are persistently less likely to get an offer than students with ATARs of 80-90 or 70 to 80. The 2012 offer rates were, respectively: 91%, 97%, 96%.
It means that a few per cent of the 90+ students take all or nothing gambles. Even though there are hundreds of courses that would accept them, they only apply for one or a small number, and some of them end up missing out. Presumably most of them learn a lesson about hubris and put in a more realistic application the following year.
Andrew H also comments that “the progression of elite students into a narrow range of courses and universities arguably has a distortionary effect on the workforce and society.”
While I partly agree with this, data I received from DIICCSRTE (I hope this name will soon change) on the ATARs of first-year students suggests that the top-performing school leavers are more spread across the disciplines than the applications data might suggest.
The figure below has the 2011 ATAR for a student admitted at the 90th percentile of everyone taking that subject, or in other words with just enough to put them in the top 10%. It just shows those where the 90th percentile is at 98 or above. Surprisingly, radiography is at the top and maths is second (though there are few specific maths courses). From the arts, language and literature is there. If we drop down to 97, politics, history and the performing arts are all there.
Teacher education has a 90th percentile of ATAR 90, with a median of 75. Not spectacular, but far from the very low cut-offs at some unis that attract so much attention.