Published in The Age, Tuesday 5 June 2012

In higher education, students learn to carefully analyse evidence and arguments to reach well-justified conclusions. Those shaping higher education policy, however, often go for thinking shortcuts: historical precedents, what other countries are doing, and numbers that sound about right.

Unfortunately, cutting intellectual corners often leads to bad policy.

The government’s equity policies are a case of policymaking by not very meaningful numbers. They turn on a definition of who counts as a ”low socio-economic status” student eligible for equity assistance. The definition matters, because universities have government recruitment targets for low SES students, and special funding to meet their needs.

Since the early 1990s, low SES has meant living in a postcode the Australian Bureau of Statistics classifies as in the lowest 25 per cent, according to indicators such as education and occupation. For years, this measure was criticised as a blunt instrument. Postcodes are too diverse. Public housing high rise apartments tower over $1 million-plus homes in parts of Melbourne. So rich people can end up in poor postcodes, and poor people can end up in rich postcodes.

Recognising this problem, the government fine-tuned its geography. It now classifies students according to much smaller areas, reducing the risk of mistaken socio-economic identity. But the old 25 per cent cut-off survived. Unfortunately, there is no basis for it.

People living in the next 25 per cent of areas by SES have educational disadvantage that is just as great, as measured by multiple sources including school results and university attendance rates.

People who are equally disadvantaged will be unequally treated by universities as the result of a wrong number.

Commonwealth-supported student funding uses numbers that have more to do with the past. The most important driver of funding rates is an expenditure study carried out more than 20 years ago. Real information, such as the funding now necessary to meet a set standard of education, is not considered. Setting and costing higher education standards would not be easy, but regulators in other industries including health and energy find ways to price complex services.

There is even less science to how much students pay. The first HECS charge was the inflation- adjusted 1988 value of average tuition fees before they were abolished in 1974. It was equivalent to about 20 per cent of average per student costs. That most OECD countries had private contributions of 20 per cent or less was seen as a useful ”overseas precedent”. Alas, two weak and lazy arguments are not more convincing than one. In 2005, universities were allowed to increase Commonwealth- supported student charges by 25 per cent in most subjects. There was no basis to the 25 per cent, other than that it was what the government could get through the Senate. How much students might want to invest in their own education was not considered. For the 40 per cent of students not in Commonwealth-supported places, we use markets to test how much they are willing to pay.

Only one element of Commonwealth-supported student charges has a real rationale. In 1997, HECS charges were adjusted in light of the likely earning potential of graduates in each field. This minimised the chances that applicants would react negatively to price increases, and meant that the affluent paid more, which seems fair. So in 1997 there were clear principles, though the charges themselves came from back-of-the-envelope calculations.

Dubious numbers are also used in setting attainment goals. In 2009 the government set a target of 40 per cent higher education attainment among those aged 25 to 34 by 2020. The general assumption that demand for university-qualified workers will grow is almost certainly right. But the 40 per cent target itself owed more to looking at targets in other countries than any Australian reality.

Skills Australia, the government’s labour market forecasting agency, predicts that just a third of all jobs would require degrees by 2025. A 40 per cent target risks increasing the already large number of graduates working in jobs that do not require degrees. Many surveys can tell us how many graduates the Australian labour market might need, without looking to other countries with very different economies.

Policymaking is inevitably an imprecise art. Often policy- makers will not have all the information needed to make good decisions, and educated estimates are necessary. However, in making Australian higher education policy we rely too heavily on intuition, history, and overseas guides when better information is available. Ironically, the intellectual skills and practices of good policy are taught in our institutions, but too often not applied when making higher education policy.