Educational reform and Junior Cert reform in particular has been getting a bit of coverage lately. One observation which struck me was by Tom Collins, formerly Professor of Education at NUI Maynooth. He mentioned anecdotal evidence claiming that primary school teachers could predict eventual secondary school educational outcomes from as early as six years of age. Some recent work I did is consistent with this observation (http://www.ucd.ie/t4cms/WP14_20.pdf) . Using Growing Up in Ireland data, I looked at the scores of nine year olds in the Drumcondra maths and reading tests. Children were partitioned into four groups on the basis of their mothers’ education. Rather than looking at gaps in average outcomes for each group, I looked at gaps for selected percentiles. This is in the spirit of John Romer’s analysis of inequality of opportunity, whereby between-group gaps calculated at the same percentile are regarded as having to some degree controlled for “effort” and the gaps can be regarded as reflecting ex post inequality of opportunity. At the limit the gaps are about one standard deviation and are pretty constant across percentiles. I also perform quantile decompositions for pairwise gaps and find that in general about one third to one half of the gap is accounted for by observable characteristics (including school, class-size and teacher characteristics) with most of this “explained” gap arising from income and the number of books in the house. Unobserved factors and “returns” to observed factors thus account for more than half the gaps.
The paper provides evidence that gaps in educational outcomes, on the basis of parental education, kick in at remarkably early ages. It is also consistent with the idea that returns to education apply across as well as within generations. There is a good recent review of this area by my colleague Paul Devereux here http://wol.iza.org/articles/intergenerational-return-to-human-capital.