One criticism of Graduate Winners was that I should have paid more attention to OECD comparisons. I am wary of ‘OECDitis’– taking OECD averages as normative when they are merely descriptive. In my view, higher per student public spending on higher education in some other countries reflects their overall political and economic systems, and does not make their higher education systems better.
But could higher per student spending make their higher education systems worse? I don’t think this is automatically the case. But I think high per student spending creates a greater risk of what I call the paradox of public spending: it may increase demand, but it also tends to decrease supply. Even big-spending European social democracies have budget constraints (as they are very painfully finding out). So if they can’t control spending by making students pay more, they control spending by reducing the number of students.
The figure below uses figures from OECD Education at a Glance showing average student fees at public institutions and overall higher education attainment rates. The Nordic countries tend to combine low fees and reasonably high attainment, but many other European countries have free or very cheap higher education and relatively low attainment.
Overall the higher fee countries have higher attainment than low fee countries (correlation of .35 between fees and attainment).
I don’t know enough about the particularities of each country to confirm my paradox. But given it is easier politically to control numbers than to increase fees (because the losers are less obvious, and less prone to rioting), it is likely that politicians in many OECD countries have capped supply of higher education, and reduced their rates of higher education attainment.
Andrew Norton 
An email commenter asks if the results would be different if we looked only at younger age groups. Looking only at people aged 25-34 the correlation drops to .29, driven mainly I think by large increases in attainment in Norway, Sweden and Finland. With the exception of the US, the higher fee countries also have substantial increases in attainment but only Korea matches the Nordic growth rate. The US is the only major country with lower attainment among 25-34 year olds than than among 35-44 year olds.
LikeLike
I quite agree on OECDitis – I’ve always found OECD’s Education at a Glance publications can reliably turn up a stat to support any argument.
I’ve finally got around to reading your excellent report and I was wondering whether you’d had much feedback on the difference between calculating student fees and graduate premiums. It seemed to me that your calculations didn’t account for non-completion, thereby overstating the potential premium to starting students. I’m not suggesting that it’s a major flaw in the report (you had to stop adding assumptions somewhere), but it’s something that does bug me about the use of graduate premiums in general.
LikeLike
Dave – The chapter on price sensitivity has a little on this topic, with a pretty strong relationship between school results and completion. However we have no data on returns (if any) to incomplete university education. Though we could not tackle that issue fully, I think a strength of the report is that the ‘breakeven’ point idea is an attempt to assess risk – though we did not get much beyond discipline and not working full time as the main drivers of weak financial results.
My view coming out of this analysis is that overall risks for poor outcomes for school leavers with reasonable resuls and a taking courses with clear career potential are quite low, and the chances of doing very well are high. What we are not so clear on are weaker students, whose drop-out rates are high and we don’t know much about what happens to them if they do complete. Given the overall graduate employment rate their risk of no job at all is probably reduced, but they may not get much of a premium. However, risk reduction may still be an ok outcome for some people.
LikeLike
Gary Marks’s 2007 paper (based on data from the Longitudinal Survey of Australian Youth) looked at school leavers’ short term outcomes (to about age 23) and found that those who started uni and then dropped out had lower unemployment rates than those who hadn’t gone to uni at all, but that their wages, job satisfaction etc were about the same as those with Year 12 only (after controlling for differences in labour market experience and literacy/numeracy).
LikeLike
Back to the original post, my instinct would be that budget constraints would mean that countries would cram more students into the lecture halls (decreasing per student funding) rather than reducing supply? Is there data on which OECD countries are contracting supply?
Also, are the comparative uni participation results skewed by the role of the polytechnic sector in some countries?
The main thing I take from the figure is that there really doesn’t seem to be much pattern at all. There are plenty of outliers (not good for correlation). Rich countries have been successful in raising participation through different pathways. Northern Europe has high participation in uniformly good universities, USA has high participation with high fees and great diversity in university quality, Australia is somewhere in the middle.
LikeLike
Thanks Andrew and Mark – my concern is mainly about how graduate premiums are sold to the marginal student, when the marginal student may well have a negative return to higher education. That is, of course, a marginal issue to Andrew’s latest report!
We’re starting to get some better data in NZ on ex-student earnings (matching national student numbers and tax numbers) so hopefully we can add to the discussion soon.
LikeLike
Peter – I agree with your last statement. The correlations are modest, but if those who are saying public funding is essential are right they should be negative rather than positive. There are good publicly funded systems, but public funding brings risks of under-investment as budgetary factors will always rule in the end. A pure market system minus a loan system also risks under-investment, but there are no examples of this in the OECD.
I suspect that there could be some classification issues for some countries – especially strong performers like Germany, which has a very good vocational system.
An article on cuts in Sweden: http://www.universityworldnews.com/article.php?story=20120824152811615
LikeLike
[…] that the level of public funding is not the key variable in higher education systems. As I showed in an earlier post, there is no evidence that lower fees result in higher attainment. Indeed, the data suggests the […]
LikeLike