Wednesday, June 02, 2010

Measuring regional disadvantage in higher education

The Australian Government wishes to increase the proportion of young Australians attending university. It also wishes to increase the proportion of young Australians from poorer economic backgrounds attending university.

The two are connected in that the Government's overall participation target cannot be achieved without also achieving the second.

In today's Australian, education writer Gavin Moodie reports on a new Commonwealth study that attempts to measure some of the factors affecting participation at university by regional students. Gavin's piece is worth reading, although I disagree with aspects of his analysis. The second link above will take you to the full study.

The report abstract states:

The report shows that regional participation in university among 19-21 year-olds students increased from 18 per cent in 1996 to 21 per cent in 2006. However, university participation among metropolitan students increased faster from 28 per cent to 35 per cent, so that the gap between regional and metropolitan participation increased from 10 percentage points to 12 percentage points.

The report finds that lower socioeconomic status, as measured by education and occupation levels, explains most of the gap in participation between regional, outer metropolitan and inner metropolitan areas. While proximity to campus matters, the report finds that access to university appears to have less influence on university participation than socioeconomic status.

If you look at the numbers in the first paragraph, you can see why the Government faces a problem in achieving its overall participation target.

The relationship between educational participation and socio-economic status is well established, although the linkages within it (the relative importance of and relationships between different variables) are less well established.

Each part of the education chain is affected. The gap that emerges at primary school affects participation at secondary school, affecting participation at university in turn. Since regional and outer metropolitan areas in general have lower socio-economic status than inner metropolitan areas, you would expect a difference in university participation rates, and indeed the study shows this to be the case. Further, the linear regression analysis suggests that variation in socio-economic status is the overwhelming determinant in statistical terms.

None of this should be too surprising. However, there is an immediate kicker in all this.

Australia's higher education system is highly competitive, with institutions competing for students on one side, funding on the other. Within this competition, regionally based institutions argue that their expansion will help close the regional participation gap by improving access. The study suggests that access as such is a minor determinant. This lead Gavin to head his article, Regional argument won't hold.

Both Gavin and the report recognise that regionally based universities bring other benefits beyond participation. However, the conclusion is still an apparent blow.

Before going on, let me make one point clear. This is a useful report. Appendix Five, for example, provides useful statistical data at a modified LGA level. However, the difficulty is that the actual policy content in the report is quite low. By this, I mean that the conclusions that might be drawn from it for policy purposes are limited without a fair bit of further work.

Perhaps the most important policy point is simply that socio-economic status and university participation are related. We know that, but it is helpful to have it restated. I have tried to make the point before that current Government participation plans are likely to fail because they do little to overcome the financial barriers impeding participation by kids from poorer families.

Beyond this, the constructs used in the report themselves impose a barrier when it comes to specific analysis and planning. The tripartite subdivision into inner metropolitan, outer metropolitan and regional is an example. This provides broad indicators, but is not very helpful when it comes to detailed analysis.

Cessnock, for example, is classified as regional, so is Moree Plains. A student in Cessnock is one hour by road from Newcastle University, a student in Moree is over three hours by road from the nearest University in Armidale. A Cessnock student can live at home while studying, a Moree student has to board away.

Cessnock has a lower socio-economic status than Moree Plains, and a lower university participation rate despite the closeness to university. Both measures are below the national average in both towns. Still, despite geography, Cessnock, seems to fit the pattern. But let's dig down a little.

With Cessnock, I do not know how good public transport to Newcastle is, nor do I know the cost. I do know that in Sydney a weekly concession public transport pass can cost 10% of weekly Youth or New Start allowance. In the past, kids used to buy an old bomb, but this is more difficult now. It may be that the frequency and cost of public transport is a major impediment in Cessnock. In that case, improved public transport may be a necessary condition for improved university participation rates.

In Moree Plains, 24% of the Shire population is Aboriginal. I have not done the analysis to check, but it may be that the Moree averages combine an Aboriginal group with well below average participation rates with a non-Aboriginal group with average or above average participation rates. If so, improving the Moree average may depend upon action targeting the Aboriginal community.

My point in these examples is one I have made before. While general studies are useful, you have to test the general conclusions by digging down into specific cases. Otherwise, the results are likely to mislead when it comes to action.

No comments: