Raising school attainment and informing parent choice
The Welsh Government's abolition of school league tables in 2000 provided an opportunity to test whether such information is important to school performance.
League table improvements
Professor Simon Burgess from the Centre for Market and Public Organisation and his colleagues found that the abolition of school league tables in Wales significantly damaged pupil attainment. Widespread public reporting of his results contributed to the reversal of this policy in 2011. His work also influenced changes in the content of school school tables in 2001.
"This change improves pupil attainment directly and indirectly. The new form of tables allows parents to make a better-informed decision on where to send their child to school. The new measure also gives schools more of an incentive to focus across the ability distribution,” explained Burgess.
He and his colleagues used data on all secondary school pupils over a decade, before and after the reform, in both Wales and England.
The data included GCSE performance, prior attainment, pupil demographics, school expenditure, and neighbourhood characteristics. The analysis involved taking each secondary school in Wales and comparing its performance to a very similar school in England, both before and after the reform.
“We found significant and robust evidence that the policy change substantially reduced school effectiveness in Wales, on average by about two GCSE grades per pupil.
“The strongest effect was on schools in more disadvantaged neighbourhoods. So the policy resulted in both lower overall school effectiveness and greater inequality.” Professor Simon Burgess, Centre for Market and Public Organisation
These findings were widely publicised in the press as well as broadcast media and also discussed on the floor of the Assembly.
The findings were described as a “slap in the face with good data” by the Senior Policy Advisor to the Welsh Assembly, Professor David Reynolds, and in February 2011, the Welsh Minister for Education announced a major change of policy to reverse the decade-long policy of not publishing performance information.
Reynolds confirmed that Burgess’s research was influential in this decision: “I can testify to the major impact that the research…had on our educational thinking in general and specific policies in particular.... The research helped to prepare for the introduction of the Banding system we launched in 2012 which of course involved the publication of school results. It was for us quite seminal research.”
Burgess’s research suggests that this policy reversal will have a major impact on pupil attainment and life chances in Wales.
For pupils on the margin of achieving 5 good passes, the reform will raise them above this crucial threshold, bringing them an estimated 25% earnings increase.
Informing parent choice
Around the same time as the policy changes in Wales took effect, Burgess and his colleague Rebecca Allen, a Reader at the Institute of Education, University of London were investigating which metric in school performance tables was most useful to parents trying to decide which school their child would do best in academically.
“Until we looked into it, the debate about the best metrics to use in school performance tables has largely been conducted by assertion,” said Burgess.
In 2010, Burgess and Allen used the entire cohort of more than half a million pupils who chose secondary school in 2003 and tracked their progress. The results showed that some school performance metrics were very useful, significantly and substantially producing better choices than choosing at random, while others were much less helpful.
Initial discussions between Burgess and Allen and officials at the Department of Education took place in February 2010 on reforming school league tables. Graham Stuart MP, Chair of the House of Commons Education Select Committee, requested a meeting with Burgess and Allen in November 2011.
In November 2011 new league tables were released that adopted key components of their proposal.
The new tables included a measure showing the GCSE performance of students with differing levels of initial ability. For each school the new tables reported the percentage of pupils attaining at least five A* to C grades separately for low-attaining pupils, high attainers and a middle group.
This gave a measure of how much a school was able to raise a pupil's attainment during the time they were at the school.