13. The Economic Effects on Regional Australia of RUN-Member Universities

See “The Facts on University Funding” (April 2017), available from https://​www.​universitiesaust​ralia.​edu.​au/​Media-and-Events/​submissions-and-reports/​The-facts-on-university-funding.

There is an extensive literature on the benefits of universities to their local economies (see, for instance, Florax 1992, Feldman and Desrochers 2003, Drucker and Goldstein 2007, Fritsch and Slavtchev 2007, Dalziel et al. 2009, and Harris et al. 2011). We confine ourselves to only economic effects. Universities can also have non-market benefits, both of a private and public nature, but we do not deal with those in this study. See McMahon (2009) for a discussion of non-market benefits.

Different industries can have varying regional economic impacts due to the composition and geographical sourcing of their inputs. A characteristic of the education industry is that, like certain other service industries (e.g. those operating tourist attractions and sports and other events), it attracts out-of-region visitors who increase local demand. This attribute is dealt with in items Sects. 13.2.2 and 13.2.3 below.

Giesecke and Madden (2006), Madden (2014), and Hermannsson et al. (2014) all model changes in the sizes of universities, while Madden (2017) examines the local economic contribution of a university.

The enrolment data in Table 13.3 include full-time and part-time commencing and continuing students by campus. These data are converted to equivalent full-time student load (EFTSL) figures using EFTSL-to-enrolment ratios for domestic and international students at each RUN-member university.

Giesecke and Madden (2006) assumed that this intra-state student multiplier was 0.5, but their study looked at students in Tasmania. Since RUN-member campuses are located in much larger states, the proportion of students with permanent residences within a reasonable commute of campus will be smaller, so we use a multiplier of 0.6 instead of 0.5 for these students.

More specifically, “Diploma” includes the award categories Advanced Diploma (AQF), Associate Degree, Diploma (AQF), and Enabling; “Bachelor” includes Bachelor’s Graduate Entry, Bachelor’s Pass, Bachelor’s Graduate Honours, Postgrad. Qual. Prelim and Non-award courses; “GradDip” includes Graduate certificate, Graduate (post) dip. (ext. area) and Graduate (post) dip. (new area); “MAPhD” includes Master’s by coursework, Master’s by research, PhD by coursework and PhD by research.

We assume that the wage premia by award category are the same for each year over 2013–2016.

See https://​www.​education.​gov.​au/​search/​site/​data

Note how the general decrease in the shares made up by “General university funds” in Table 13.6 reflect the trend that Universities are now relying more on external research funds.

Recall that the research productivity effect was calculated using discounted research income back to 1992, a sufficient period to justify the counterfactual of each RUN-member University having had no regional campus for many years.

As a less conservative counterfactual simulation, we also run simulations in which ¾ of the students in RUN campuses would study elsewhere in Australia. Consequently, ¾ of the labour productivity effects accrue to other regions in Australia, and ¾ of the returns to the stock of research knowledge are retained. For directly affected campus regions, the two assumptions generate almost identical impacts. This is so because we effectively assume that of the ¼ or ¾ of the residents who study elsewhere, none returns to their region of origin. That is, when CQU’s Rockhampton campus closes, ¼ of those who studied there go on to study elsewhere in Australia, but none returns to Rockhampton after completing their studies. This is likely to be not too far from observed patterns of youth migration. For example, Hillman and Rothman (2007) note that over the period 1997–2004, “… just under three-quarters (74%) of non-metropolitan young people in 1997 were still in non-metropolitan areas”. If we accounted for the small share of the 26% of those who were not in non-metropolitan areas and returned to Rockhampton (for example), this would slightly mitigate the losses reported for CQU-Rockhampton in Table 13.9, more so if the share of graduates who studied elsewhere was assumed to be ¾. In the rest of Australia, the two assumptions provide proportionally different impacts but they are both small in percentage terms.