Stellenbosch Working Paper Series No. WP08/2017
Publication date: September 2017
This study evaluates the relative importance of family, neighbourhood and school quality in explaining variation in standardized test results, reaching and passing matric, university attendance and labour market earnings. It adds to the literature, by using a spatial approach to link a neighbourhood wealth index from the Census 2011 community survey to a unique administrative school data set from the Western Cape. For the long-term perspective the household and school information from the National Income Dynamics Study are explored. The results from administrative school data show how student wealth and differences in school quality produce vastly different outcomes for a cohort of grade 6 to 12 learners in Cape Town. It shows how grade 6 children going to the richest 20% of all schools are 30% more likely to pass matric in time, furthermore by grade 9 the learning gap is approximately four grade-levels worth of learning in comparison to children going to the poorest 20% of schools. However, this study also demonstrates that even children from the poorest neighbourhood would perform well if they go to one of the richest 20% of schools. Yet, given the limited number of quality schools, the segregated location of quality schools, financial as well as transport constraints, only very few children from the poorest 60% actually attend a top quintile schools. These results can be replicated for the national data set and show that in order to achieve more equal education outcomes, the quality of schools in the poor neighbourhoods have to be drastically improved. In addition, using the new school wealth index as an instrument for school quality, there seems to be a significant premium for quality education in labour markets earnings regressions, which show the long-term implications of the schooling system.

JEL Classification:
D10, I24, I26, J15

South Africa, Education, Spatial analysis, Neighbourhood effects, Family effects