The Australia Institute report offers little to improve education standards

The Australia Institute recently published a report called Buying An Education: Where Are The Returns Highest? (pdf, 115Kb). It argues that parents would be wiser to save the money they might spend on non-government schooling, and use it to pay for a full-fee university place.

To support this hypothesis the author, Richard Denniss, showed that the estimated cost of tuition in high-fee independent schools is higher than the estimated cost of a full-fee university place. As estimates of the cost of specific courses in specific institutions they are not in dispute. However, these are present-day figures, and it cannot be inferred that they will be applicable in 12 or more years’ time.

The report’s central argument rests on two claims. First, that schools themselves offer no academic advantage, so non-government schools are not worth the expense. Second, that anyone with enough money will be able enrol in the course of their choice at their preferred university, which is inequitable and will lower standards.

These findings were detailed in the press as early as 25 December 2003, but the report itself was not released until 8 January, 2004. This made it impossible for the report to be scrutinised at the time it was being publicised. This is not a constructive way of advancing public debate on a serious issue. Furthermore, the evidence and logic used to support these claims is demonstrably flawed.

First, not all non-government schools are high-fee schools. The eight schools listed in Buying An Education are not representative of the non-government school sector. The non-government sector consists of Catholic systemic schools and independent schools. Catholic systemic schools form around two thirds of the non-government sector and have very low fees on average: $1400 per year in 2001.

Independent schools are much more varied. The high-fee schools used as examples are not typical of the independent school sector. Average fees in independent schools were $5,300 per year in 2001. This is a third of the high-fee level ($17,000), and half of the "moderate" fee level ($10,000), depicted in the Australia Institute report, and indicates that there are a large number of schools with even lower fees.

Second, schools are not simply pathways to university. School education is important in its own right. Parents in the position to do so choose a school based on a variety of reasons, and academic performance is one factor among many.

Academic performance is, of course, important. Because information that allows individual schools to be compared is not available to the public, comparisons between schools are necessarily broad and highly generalised. This is unfortunate, because the differences within school sectors are as great as the differences between school sectors. This makes comparing school sectors a difficult task but not an impossible one. The Australia Institute report draws on two sets of evidence to support its suggestion that non-government schools have, on average, no academic advantage: Results of the 2003 NSW Higher School Certificate and recent research by the Australian Council for Educational research.

The HSC results quoted are that "74 of the top 130 students were educated in government schools". This statistic shows that government school students were slightly underrepresented, if only 74 out of 130 (57 per cent) of students were from government schools when they fielded an estimated 63 per cent of candidates. This is a crude statistic and is heavily influenced by the success of selective government schools, which Denniss acknowledges.

More sophisticated evidence comes from the Australian Council for Educational Research (ACER). Denniss cites the most recent ACER report, LSAY Research Report 36 (pdf, 509Kb), to support the hypothesis that schools have little effect on student achievement, once the socioeconomic status (SES) of the students is taken into account. This ACER study looked at literacy and numeracy in Year 9, not tertiary entrance scores, so to infer university admission from these findings is a stretch. Even if this is granted, Denniss's account of the report’s findings is seriously flawed.

The relationship between SES and student achievement is well-established, and operates at two levels. LSAY Research Report 36 found that SES had a small but significant effect on student achievement at the individual level but a larger effect at the school level. Buying An Education concludes that "it is not the school that directly affects students’ performance but rather the location of the school in a high SES area and the individual SES of the child’s parents".

This interpretation is not the only one, however. Other research indicates that other factors at the school level, related to SES, affect student achievement much more. One such factor is teacher quality. Ken Rowe found that a large proportion, about 60 per cent, of the variance in student achievement can be attributed to teacher quality. It is therefore likely that SES has a largely indirect effect on student achievement, by the greater likelihood that high-SES students receive high-quality teaching. This is a serious problem and should be treated as such.