A new study published in the latest issue of the British Educational Research Journal shows that the socio-economic composition of schools has a large effect on student achievement. A student attending a school with low average socio-economic status (SES) tends to have lower school outcomes than a student from a similar family background attending a school with high average student SES.
The findings of the study contradict the claims by Melbourne University researcher, Gary Marks, that school SES is a statistical artefact and that individual student SES has little effect on student achievement. These claims have been used by elite private school organisations and their advocates to argue that low SES students and schools should not receive additional funding.
The British study found that that an increase of one unit in the PISA index of school composition is associated with an increase of about 69 points on the PISA scale of student mathematics achievement. This is a large effect. It also found that a student’s own SES background also has a strong impact on achievement. An increase of one unit in the PISA index of family socio-economic background is associated with a gain of 21 points in achievement.
These findings confirm those of many other studies showing that school SES affects student achievement over and beyond the effects of individual student SES. Since the path-breaking Coleman Report in the United States in 1966, school composition effects on student achievement have been consistently verified across school systems and academic subjects. Several OECD PISA studies have found that school composition effects even outweigh the impact of individual students family background in many countries. These studies also show that the magnitude of the school SES effect is lowest in comprehensive school systems such as in Finland where there is a greater mix of students from different social backgrounds. Some studies have found that school composition tends to have a greater effect on the achievement of low SES students than high SES students.
Only a few studies have systematically investigated the factors that account for the impact of school composition on student achievement. They show that high SES schools differ from low SES schools in several ways. High SES schools tend to have fewer disciplinary problems, high student and teacher morale, more material, financial and human resources, better student-teacher relationships, higher teacher expectations, and more demanding curricula. These factors are closely related to school achievement.
The new study used multi-level modelling techniques to estimate the direct and indirect impact of school composition on student achievement. In particular, it explored the extent to which school factors mediated the relationship between school composition and school achievement. It analysed the effect of twelve school factors classified into three categories: opportunity to learn (for example, time spent on mathematics), school climate (for example, disciplinary climate, student-teacher relations, teacher morale) and school organisation/management (for example, school autonomy, number of assessments per year).
It found that three school climate factors were significant in connecting school composition to student achievement – school disciplinary climates, students’ positive behaviour and student morale. Each of them was able to account for school compositional effects in a significant way. Together, they accounted for about one-quarter of the school compositional effect on student achievement. As an aside, it is interesting to note that the study found that school autonomy had significant negative effects upon student achievement.
The results suggest that if low SES schools can manage to create a positive school climate through managing discipline, student behaviour and stimulating student morale, they can significantly reduce school composition effects on student achievement. The study noted that effective strategies can be used to improve school disciplinary climate, students’ morale and reduce behavioural problems.
If these aspects of school processes can be improved in some socioeconomically disadvantaged schools, the negative effects of their low school SES composition can be mitigated. In some best-case scenarios, some of the least performing schools, usually low SES schools, may be considerably improved. [p.439]
However, the study also states that it would be over-optimistic to simply rely on improving school climates to solve the problems related to disparities in school composition. High SES schools have several advantages in creating favourable learning climates. It is easier to create favourable climates in these schools because high SES students tend to have fewer disciplinary or behavioural problems and often have high morale. Moreover, high SES parents are more likely to pressure schools to create positive school climates than lower SES parents.
The study also notes that while school processes can be manipulated to mediate the effect of school composition, some part of the effect is beyond the control of schools. It is very difficult to change school composition. Income inequality frequently leads to residential segregation and in turn school segregation because usually students are either required or expected to enrol in schools in their neighbourhood. Where school choice is available, high SES parents tend to enrol their children with peers of similar background while low SES parents tend, indeed have little choice but to enrol their children in the neighbourhood school.
Also, enrolment competitions and selective admissions to some schools also contribute to disparities in school composition as some schools are able to skim off higher achieving students who tend to come from higher SES families.
This has led some education commentators to advocate policies to reduce social segregation in schools as the only approach to reduce school composition effects and to create equal learning opportunities for all. The study comments that there is insufficient political or individual will to integrate schools along socio-economic lines, because advantaged groups and individuals who benefit from socially segregated school systems are unwilling to give up their advantages.
Another consideration is that a significant part of the variance in student achievement is associated with students (and their families and communities), rather than with the schools they attend. As a result, the study says, achieving true equality of learning opportunities entails addressing inequalities in family and community resources, apart from policies that can be implemented by schools.
The study does not explore the implication of its results for school funding policies. While low SES schools can mitigate part of the school composition effect on student achievement by creating better learning environments this largely depends on the human resources available to these schools.
There is a huge gap in the average results of high and low SES schools in Australia, but low SES schools in Australia remain under-resourced in terms of teaching and other support staff. Australia has amongst the largest disparities in the resources available to low and high SES schools in the OECD. The disparity is so great that the OECD has labelled Australia a low equity nation in the resourcing of schools.
The student/teacher ratio of low SES schools is actually higher than for high SES schools. Australia is one of only seven out of 34 OECD countries where this is the case.
Teacher shortages in low SES schools in Australia are high by OECD standards. Only eight other countries have higher teacher shortages in low SES schools than in Australia. Indeed, Australia has the largest difference in teacher shortages between low SES and high SES schools in the OECD; the difference is more than double the average for the OECD.
Almost all the schools with high concentrations of low SES students are public schools. In 2013, 94 per cent of schools with 50 per cent or more students from the lowest SES quartile were public schools.
Increased funding for low SES schools is critical to overcoming the effect of disparities in school composition on student achievement in Australia. Without this, the large gap in school outcomes between rich and poor will continue. Full implementation of the Gonski funding plan is a necessary step to improve the prospects of disadvantaged schools and their students.
Trevor Cobbold
Hongqiang Liu, Jan Van Damme, Sarah Gielen & Wim Van Den Noortgate, School processes mediate school compositional effects: model specification and estimation. British Educational Research Journal, 41(3): 423-427.