This study looks at the impact of latent individual ability on educational attainment,
specifically the decision to acquire tertiary education. The analysis aims to determine
whether ability is significant over and above socio-economic status in determining
educational attainment. The Cape Area Panel Study (CAPS) (2002-2005) and the
School Register of Needs Survey (2000) provide data at the individual, household and
school level, which are used to model the decision to apply for tertiary education for a
sample of respondents who have completed matric. The CAPS data-set provides a set of
literacy and numeracy test scores, which are regarded as measures of manifest ability.
After purging out the effects of age, education level, as well as household and schoollevel
characteristics from these test scores, a latent measure of ability is obtained, which
is relatively independent of socio-economic status. Probit regressions are then used to
determine the impact of this ability measure on the decision to apply for tertiary
education. Population groups are analysed separately due to their wide disparities in
household and school resources in the South African context. Furthermore, a household
fixed effect analysis is used to control for any other unobservable household
characteristics that may affect educational decision-making. The results indicate that
overall, individual ability is significant in explaining the decision to pursue tertiary
education over and above contextual factors. However, for the Black sample, ability is
not significant in explaining this decision, and is significantly less important in
explaining the attainment of a matric exemption compared to the Coloured and White
groups. This suggests that a lack of resources and unfavourable learning environments,
which are characteristic of the Black population, lead to a 'crowding out' of individual
ability.