Gendered Returns to the Study of High Level Mathematics and English for the Australian Tertiary Admission Rank

To enter a university program of their choice most applicants in Australia need an Australian Tertiary Admission Rank (ATAR) calculated by the Universities Admission Centre from a variety of inputs, including Year 12 high stakes tests. This study combines the NSW Education Standards Authority administrative records for 46,000 students in New South Wales (NSW) who completed secondary education in 2011 with the data for the same cohort from the Longitudinal Survey of Australian Youth. We demonstrate how students’ individual choices interact with their institutional contexts to reproduce horizontal gender segregation in field-of-study specializations.

We focus on a previously unexamined stratification mechanism: ATAR-related returns to the study of advanced mathematics and advanced English courses taken towards the end of secondary education. First, we find, when holding prior achievement in mathematics constant, that girls boost their university entry rank (ATAR) more than boys by studying advanced mathematics in Year 12. However, the relative gain they obtain through the study of advanced mathematics over their same-sex peers in lower level courses is four times smaller than the advantage of boys over their peers. Using rational choice and vocational identity theories we propose two explanations of this pattern and discuss its potential to effectively discourage high-achieving girls from opting for higher level mathematics. Second, we examine the gendered patterns of ATAR returns to the study of basic and advanced English courses. The paper closes with two discussions. The first concerns the potential that within-gender differentials in ATAR returns to the study of advanced mathematics and English contribute to the reproduction of gender segregation at university.

The second reviews the reasons why understanding gendered returns to ATAR-related secondary study is impossible without the use of administrative data, which, in turn, on its own, is insufficient to consider the full range of sociologically relevant factors.