Opinion & Commentary
What’s fair about defining disadvantage?
Under draft policies, for the first time a ‘low SES’ classification would deliver significant benefits to individuals. ‘Partnerships’ funding would oblige universities to actively seek low SES students. University marketing and admissions requirements would favour those people classed as ‘low SES’. According to proposed guidelines released late last year, equity ‘participation’ funding of up to $126 million a year is to be spent on low SES students, to provide services such as academic preparation, mentoring, and tutoring. Performance funding policies offer additional rewards for higher pass rates and lower drop-out rates for low SES students.
These policies give the issue of how ‘low SES’ is defined a new dimension. Measures that may be satisfactory for identifying broad patterns and trends are not necessarily suitable for allocating individual benefits. We need high levels of accuracy if a classification means the difference between being encouraged to enrol or not encouraged, admitted to university or not being admitted, or given additional academic assistance or denied it. Given what is at stake, the policy must be fair to individuals, treating people in like circumstances in similar ways.
The government’s SES definition discussion paper, released in December 2009, focuses on links between SES and educational disadvantage. However, it lacks data on the actual relationship between common SES indicators and academic performance. The results of the NAPLAN (National Assessment Program Literacy and Numeracy) school testing program are a useful source of this information. Its 2009 Year 9 results give us a preview of 2013’s potential first-year university class.
Table 1: Year 9 bands 8 to 10 NAPLAN 2009 achievement, by parental education (%)Source: 2009 National Assessment Program Literacy and Numeracy
For Year 9 students, the minimum national standard is reaching ‘band 6’ level for writing, spelling, numeracy, reading and grammar and punctuation. The vast majority of students from all backgrounds achieve band 6 or better. However, students whose parents are high SES by education are much more likely to achieve good results. For bands 8 to 10, as seen in table 1, students with a university-educated parent are two to three times as likely to do well in NAPLAN tests than students whose most-educated parent has a Year 11 education or below. For bands 9 and 10, covering the approximately 20% of students most on track for university entry, the socioeconomic differences are even larger. On most indicators, students with a university-educated parent are four or more times as likely to be high achievers.The NAPLAN results show that while students whose parents have a Year 11 or below education have the worst results, students whose parents have a Year 12 only or a vocational certificate education also have a good claim to be considered educationally disadvantaged. Relative to the children of university-educated parents, they are still well behind. The gap is larger for band 9 to 10 results than it is for band 8 to 10 results.
Relative SES measures such as the current lowest 25% of the population would disadvantage the many young people whose parents have Year 12 or certificate-level education. Statistics in the Australian Bureau of Statistics (ABS) publication Education and Work show that the 25% cut-off would exclude them from the low SES indicator. Because so many Australians fall into these categories, even a lowest 50% SES measure would exclude families with certificate III or IV education.
Victorian Year 12 results ranked according to postcode measures of SES—the ABS Index of Education and Occupation currently used—also suggests that any relative measure should be at least the lowest 50%. Using 2008 figures, students living in the second, third, fourth and fifth deciles by postcode all received fairly similar Year 12 results.
Figure 1: Victorian school results by SES postcode decile 2008Sources: VTAC; ABS, SEIFA data cube. Unadjusted for population
A broad SES definition would match parental status and student educational advantage fairly well in a statistical sense. However the two are not perfectly correlated. Even in the most disadvantaged group by parental education, between 6% and 11% of children, depending on the indicator, receive very good band 9 or 10 NAPLAN results. In the Victorian 2008 Year 12 results, 7% of students from lowest 25% postcodes received ENTERs of 90 or more. At the other end of the SES scale, some students struggle despite coming from advantaged homes. Depending on the indicator, 5% to 14% of university-educated parent’s children have NAPLAN results at or below the minimum standard. In the Victorian schools analysis, more than 10% of students living in top SES decile postcodes achieved ENTERs below 50. Students from top private and selective schools, with no parental or school-level educational disadvantage, nevertheless sometimes struggle at university.
Given an only approximate connection between SES and educational disadvantage, broad SES categories should not be used to allocate individual entitlements. Funding policy and service delivery policy have different requirements. Funding policy needs easy-to-collect indicators that will deliver resources to universities according to the likely levels of educational disadvantage across their entire student body. Service delivery can be far more complex and nuanced, with frontline university staff making individual assessments of which students need assistance and how much help they need. Universities should also be allowed to spend money on improvements that would benefit all students, such as better educational technology and facilities.
Funding policy needs to take into account that SES is on a continuum, and small SES differences should not lead to large funding differences. One way to do this would be to copy Victorian schools funding, which gives weights to different SES groups. Victoria uses parental occupation, but weights could be constructed for parental education. For example, Year 11 or below might get a weighting of 1, certificate a weighting of .75, Year 12 a weighting of .6, diploma of .4, and university education 0. This would help avoid an excessive focus on recruiting the lowest SES group at the expense of those who are only marginally better off.
The government is rightly concerned about poor educational outcomes for people from low SES backgrounds. However the policy documents released late last year suggest that the government is looking at an overly narrow definition of low SES. If carried through into recruitment and service delivery, this could result in significant unfairness to individual students. A broad definition of low SES, combined with flexibility in how equity funding is spent, will avoid the irony of an unfair equity policy.
