We are regularly being told that in an era of technology-driven labour market change we will need to reskill and retrain much more than we did in the past. Perhaps we will. But it is hard to find evidence for this in the available data.
Let’s start in higher education. As I have noted before, mature-age undergraduate education is trending down. But domestic postgraduate coursework commencing student numbers are also down on their 2014 peak, as seen in the chart below. Education and business courses are driving the decline. Only health and IT courses have enjoyed enrolment increases since 2014.
Total student numbers are still high by historical standards. But with record numbers of eligible students (people who already have degrees), and undergraduate initial professional entry courses being converted into postgraduate qualifications, we would expect strong growth in this type of qualification. It is not happening.
In vocational education too enrolments are trending down, including for people who already have a Certificate III or above qualification (taking the Certificate III as more clearly a career qualification than Certificates I or II).
In vocational education other things going on could explain a downward trend. Rogue operators have undermined confidence in the industry. Funding cuts and other policy decisions make it harder for prospective students to realise their aspirations. So it’s possible that underlying demand is going up but it is not converting into enrolments.
But there are some interesting trends in the reasons given for undertaking training. Of those who have enrolled, the proportion training to get a better job is fairly stable. But a declining share of students say they undertook training as a requirement of their current job or to gain extra skills for their job. Could existing jobs be changing less, reducing the need for additional training?
Employers sometimes complain about inflexible training package vocational qualifications, so perhaps they (and their staff) prefer shorter, cheaper training courses that more directly meet their needs. But again the available data does not support the hypothesis. The ABS work-related training survey shows that the proportion of people engaging in non-credentialed work-related training declined between 2013 and 2017, as seen in the chart below.
It remains true that the more education you already have the more likely you are to do additional training, but the decline is right across the education levels.
And not surprisingly, given the trends in other data sources, the ATO reports that fewer people are claiming self-education expenses. In 2016-17, 5.1 per cent of taxable individuals claimed a self-education expense, the lowest proportion in a time series going back to 1991-92.
In the vocational education statistics at least, and quite possibly in the other data series, there are things going on that could help explain why the observed trends are contrary to underlying demands or needs. But to me it is still striking that none of them support the theory that there is an increasing need for reskilling and retraining.
So what is going on? Perhaps the theory is right, but free information available on the internet is changing the nature of skills development. Neither workers nor employers need to pay to acquire new skills as often as they did in the past. Current surveys aren’t well-designed to track the skills learned from a Google ”how to..?” search, so we can’t see that the number of hours spent learning might actually be going up.
Or perhaps the theory is right but not yet, that potentially big changes to employment and occupations coming from artificial intelligence and other technology are still in their early stages. We are in the calm before the storm.
Or maybe there is something wrong with the theory. As Jeff Borland and other labour market economists point out, employment trends often do not neatly match claims about the impact of technology.
I do not know which of these explanations, or others I have not considered, might be right. But there are some surprising trends in this data, which need watching and further exploration.