Over-education vs over-skilling

As regular readers know, I have long been interested in graduate over-education – graduates who have jobs that typically require lesser qualifications. Using ABS definitions of jobs that normally require university education or equivalent experience, about three-quarters of graduates are appropriately matched with their jobs and about a quarter are over-educated.

However the significance of this has always been open to interpretation, given that some degrees are not taken for vocational reasons, some people may be happy with their jobs, or the situation could be temporary. And we would always expect some level of mismatch so it is hard to know what the benchmark figure for ‘too high’ should be.

A recent study by Kostas Mavromaras and colleagues, using HILDA data, looks at this issue in more depth. It uses a different definition of over-education, that the person has more education than the modal level of education in their job. Among full-time employees, they find 13% over-education among graduates – about half the ABS figure (though to what extent this is due to excluding part-timers, and what extent due to classifying FT jobs differently, I can’t say).

The study adds a subjective measure of over-skilling – people disagreeing with the proposition that ‘I use many of my skills and abilities in my current job’. Only 5% of full-time graduate workers are over-educated and over-skilled. So there are graduates who on paper look over-qualified, but believe that they are matched with their jobs, skills-wise. This may be because their degree did not give them skills, or the job uses more skills than are apparent from a general classification.

For males, there are wage penalties relative to well-matched people only for men who are both over-educated and over-skilled. For females, there are wage penalties across all categories, but these are highest for women who are both over-educated and over-skilled.

For both sexes, over-education on its own has only limited effects on job satisfaction. But being over-skilled, or over-skilled and over-educated, does have negative effects

Over-education may, from a human capital perspective, signal inefficient public and private investment in higher education. But for the individuals involved, subjective measures of over-skilling are more reliable indicators of whether or not there is a problem.

6 Responses to “Over-education vs over-skilling

  • 1
    Rajat Sood
    June 15th, 2011 19:44

    Is the reference to wage penalties in a lifetime earnings sense or in an annual earnings sense? If the latter, is this penalty relative to workers of the same age or same experience?

  • 2
    Andrew Norton
    June 16th, 2011 05:16

    Rajat – It is weekly earnings, after adjusting for age, marital status, number of children, SES, unemployment history, country of origin, employment and occupational tenure, union membership, firm size, and industry.

  • 3
    June 16th, 2011 09:12

    Andrew – I subscribe to Trow’s views on over-skilling – embedded in this quote, but applicable to other higher education qualifications: “Moreover, there is a steady educational inflation of jobs – people with doctorates come increasingly to take jobs formerly held by people with lesser qualifications and in the process, they often reshape the jobs they enter.”

  • 4
    Rajat Sood
    June 16th, 2011 09:25

    Thanks Andrew. Just intuitively, it seems hard to control for both age and experience. Presumably, if someone has more years of education, they must be less experienced than a worker of the same age, which would explain the penalty. Of course, the other explanation is that more education is consumed by lower-quality workers to compensate for or offset their lower quality. But it would be hard to disentangle this effect from less experience.

  • 5
    Andrew Norton
    June 16th, 2011 10:47

    Matt – I think there is something to that. We tend to think of supplying graduates to meet demand that does or would otherwise exist, ignoring the entrepreneurial side – that graduates can generate new or improved services that people want to buy. And over-educated graduates do tend to earn more than less-educated workers in the same jobs.

  • 6
    Andrew Norton
    June 16th, 2011 10:57

    Rajat – I am no expert on the technical side of this study, but I think these variables may pick up different things. Tenure in occupation picks up general experience, tenure with current employer picks up firm-specific experience, and age may not be duplicating either but instead dealing with another aspect of their data, that income tends to peak when people are aged in their late 40s.

    Maybe there is some cohort effect factor here, or maybe as people pass the years of raising kids their need for cash declines and they take less demanding jobs.