The ABS has an interesting new publication out today on the financial benefits of higher education.
ABS anlayst Hui Wei uses data from the 1981, 1986, 1991, 1996, 2001 and 2006 censuses to provide estimates of rates of return for investment in higher education. In the figure below, the rates are based on post-tax earnings of graduates compared to someone who finished their education at year 12 (say a Year 12 completer earned $800 a week and a graduate $1,200 a week – the graduate premium would be $400). The graduates are aged through the census of the stated year (eg it assumes that a 1996 graduate would at age 40 earn what a 40 year old graduate earned in 1996).
The costs are assumed to be the opportunity cost of four years out of the workforce with no earnings in that time, plus direct costs such as HECS.
As student labour force participation has increased over this time, the opportunity costs are increasingly over-stated. However the graduate premium is also over-stated, as those who go on to university are not a random group of Year 12 finishers – they are on average of higher ability. A further difficulty is that the highest census income category ($2,000 a week in 2006 ) understates the earnings of many graduates.
With these caveats, the following observations:
* As Wei notes, the rates of return are much higher for male persons than male employees. This is because a significant part of the benefit of being a graduate is protection against unemployment or not being in the labour force.
* For employees, the rate of return is quite stable over time at around 9-11%. The one exception is 2006. What seems to be happening there is that the booming labour market was employing a lot of lower-skill people, who earned significantly less than graduate employees (and so pushed up the latter’s premium) but also reduced the graduate persons premium because the Year 12 only group were increasing their average earnings.
* The massive increase in the number of graduates has not reduced their rate of return.
* HECS has not reduced the rate of return (though for most of this period it was so low that as a cost it was minor compared to foregone wages)
These results are generally positive for the next wave of higher education expansion – past expansions and their assumed consequential decline in graduate quality have not had an obvious impact on rates of return (though it would be interesting to see if there has been an increase in earnings dispersion over time).
Andrew Norton 
“The massive increase in the number of graduates has not reduced their rate of return”
.
I’ll just wait for the comments denying this and saying that it’s all credentialism etc., despite other data (like the level of education needed for most news jobs that have been created) supporting it.
.
“However the graduate premium is also over-stated, as those who go on to university are not a random group of Year 12 finishers – they are on average of higher ability.”
.
People always use this as an argument, even though I imagine it would be possible to estimate the size of this effect — do you know if anyone has done it in Australia?
LikeLike
Conrad – Wei has this to say on the ability issue: “However, despite painstaking efforts to identify and measure the returns to abilities that are independent of education but potentially correlated with schooling levels, the dust has not settled over the best way to correct for ability biases.”
The double difficulty here is that going to uni covers non-cognitive attributes that are valuable in the labour market, such as persistence and future-orientation, as well as academic intelligence.
However in a credentialist labour market people with these attributes but not a degree may never get noticed. So it is worth gonig to uni even if the actual value add is modest.
People of my generation – say left school 1975-1985 – would be an interesting test, with higher education becoming more common but still a reasonable number of able people not continuing to uni.
The guy at my school who received the highest HSC result did not go to uni. He’s now a senior journalist.
LikeLike
So this suggests that my first 4 years at uni weren’t a complete waste of time.
What about PhDs? Should I have taken my honours degree and escaped to a real job?
The one big advantage I have experienced from having a PhD is that it opens doors in areas where you don’t have the direct experience. I presume this is because it signals a combination of being smart enough and persistent enough to finish a thesis.
LikeLike
Conrad, no-one suggests that there isn’t a real and positive rate of return to education. It’s just that calculations based on the raw wage premiums are a big overestimate of that return, mainly because of the selection effects Andrew has pointed to (although both signalling effects – aka “credentialism” – and cohort biases may also lead to an overestimate).
Labour economists have been trying to seperate out how much of the higher wages is *caused* by the education from how much is merely *associated* with it for over 40 years. The US consensus is that roughly half the raw rate of return to college education there is a causal effect, with the rest due mainly to selection effects. For early education, though, the selection effects are less important and so the aggregate rate of return is much less of an overestimate of the actual benefit.
LikeLike
M – If you download the report, at p.16 you can see estimates of the additional earnings associated with higher degrees. These appear to be a much better deal for women than men – not I think because women with p/g degrees are paid more than men, but because women at the lower levels of qualification are paid less than men with similar qualification levels.
LikeLike
There’s higher education and there’s higher education. I’d like to see a study of the financial benefits of degrees from different universities. For example, (apparently) the big laws firms in Sydney only hire graduates from the University of Sydney and UNSW. Presumably law graduates from other universities go on to have careers in suburban law firms where the returns are much less.
LikeLike
Andrew- could it also be that women with higher qualifications are more “career focused” than men, where as women with bachelor level qualifications are happier to take smaller salaries as part of the work-family trade off?
LikeLike
Shem – That sounds plausible but I would need far more of the underlying data to see. It could also be disciplinary differences too, with female bachelor degrees perhaps having higher shares of lower-paying disciplines such as arts and nursing.
LikeLike
“I’d like to see a study of the financial benefits of degrees from different universities.”
.
Look in the graduate careers survey. What you’ll find is that it doesn’t make much difference (indeed, even the effect of discipline is surprisingly small).
LikeLike
Some limited more research on this issue here.
While there is limited research on earnings differences based on uni attended, what we have suggests that in the early years after graduation there is not a significant Group of Eight premium.
This is consistent with employers placing little weight on university attended and instead waiting to see how employees perform.
LikeLike
“employers placing little weight on university attended”
This must vary by employer. Try getting employed at Mallesons with a law degree from the University of Ballarat.
LikeLike
“These results are generally positive for the next wave of higher education expansion”
.
Yes, generally, but can’t we do better than that? You also need to know about particular courses. If people who did a B.A. and majored in comparative religion, or Greek, or whatever, are earning no more than the uneducated, it would add meaning to this ABS stuff.
LikeLike
S of R – Especially as Ballarat does not have a law school:) I know it is counter-intuitive and runs against anecdotal evidence such as you have cited. But I am just reporting what the available data says. One factor may be that Group of Eight students are younger on average than students from elsewhere.
Russell – BA earnings are much lower than other disciplines. This ABS study is purely an average. It would be interesting to see it by discipline and at different points in the income distribution.
LikeLike
This would seem to work against the signalling explanation of higher ed wouldn’t it? With so many more people doing it, the signalling benefits of a degree should have declined (and been pushed into honours and higher degrees).
I guess this could be offset if more employers began looking for signals of discipline and intelligence over that period.
LikeLike
No, the signalling benefit would only decline if it was no longer as accurate.
LikeLike
Not all of the returns to education would be captured by the labour market though. There may be returns in the “marriage-market” (I use quotations because I really mean the market for romantic partners in general) where having university education can mean obtaining a higher quality partner.
LikeLike
Having a degree is good for men’s romantic prospects. For women it is neutral or a disadvantage.
LikeLike