Big differences in GCSE grades
First, some caveats.
We use data from the Millenium Cohort Study, and sample sizes are relatively small – our dataset only includes just under 400 “bright” children from poor backgrounds.
The tests we use to identify “bright” children america rcs data also have limitations, being based on assessments taken at ages 3 and 5. We discuss this issue of measurement error at length in our paper and how we attempt to adjust for it in our analysis.
Our central estimates suggest that – during primary school – “bright” 5-year-olds from poor backgrounds largely manage to keep pace academically with their equally bright but rich peers.
But the same does not hold true with respect to achievement at the end of secondary school. The chart below provides best estimates of the difference in obtaining the equivalent of a GCSE A-grade in English and mathematics amongst “bright” 5-year-olds from rich and poor backgrounds (some of the MCS cohort would have taken their GCSEs prior to the alphabetic to numeric grade reforms).
We find that bright-but-poor 5-year-olds are around 26 percentage points less likely to achieve an A-grade in GCSE mathematics than bright 5-year-olds from the richest backgrounds. The analogous difference for GCSE English is around 21 percentage points.
What may be driving this result?
While it’s hard to say for sure what is driving these differences, the data does provide some clues.
For instance, Key Stage 3 – between the ages of 11 and 14 – seems to be a critical period. This is when we see sizeable differences in the cognitive skills of early high achievers from rich and poor backgrounds emerge, which is consistent with previous research.