Anchoring talent to regions: the role of universities in graduate retention through employment and entrepreneurship

ABSTRACT

Drawing on the concept of human capital externalities, this paper investigates universities’ contribution to regional economies by analysing two types of graduate retention: labour retention (graduates employed in the region where they studied) and entrepreneurship retention (graduates starting businesses in the region where they studied). Using a panel of English universities (2010/11–2015/16), the paper examines the extent to which the specialization and diversification of universities’ subject mix influences graduate retention rates across urban and non-urban areas. Findings show that agglomeration dynamics affect labour and entrepreneurship retention differently, and that universities’ knowledge offer (subject specialization) matters across diverse geographical contexts.

KEYWORDS:

• graduate start-ups
• knowledge spillover
• human capital externalities
• urban
• rural
• STEM

JEL:

• I23
• J16
• L26

ACKNOWLEDGEMENTS

The authors thank the editorial team of the special issue and the anonymous referees for valuable comments and suggestions. Earlier versions of the paper were presented at the Regional Studies Association Winter Conference (London, UK, 2018), the Regional Innovation Policies Conference (Florence, Italy, 2019), and the 5th Geography of Innovation Conference (Stavanger, Norway, 2020). The authors appreciate the insightful comments received at these events. Any remaining errors are responsibility of the authors alone.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. Throughout, we employ the term ‘region’ to denote the subnational surrounding environment of the university, while we acknowledge the differences in the level and type of engagement of universities in their regions and localities, and implications for the spatial agglomeration. We identify the specific definitions of the terms and data sources in the empirical analysis section.

2. Recent figures (2017/18) in the UK show that 15 months after graduation, 66.3% of UK graduates of known domicile and location of employment were working in their home region, and 53.7% in the region where they had attended university (AGCAS/Prospects Luminate, 2020).

3. A recent study shows that when an STEM graduate exhibits a specialized university curriculum, entrepreneurial entry is more likely (Colombo & Piva, 2020).

4. The European Tertiary Education Register provides the geographical location of the universities based on the location of the main campus. In our analysis, there are 82 PUA and 51 non-PUA areas.

5. Given the scope of this paper, we collected information only on full-time graduates domiciled in the UK.

6. The DLHE distinguishes graduates ‘starting a business’ from those registered as ‘self-employed’. When we computed ‘graduate entrepreneurship retention’, we considered only graduates starting a business and excluded individuals who registered as ‘self-employed’.

7. The response rates of the six academic years analysed here are, respectively: 78.8%, 79%, 78%, 78.4%, 78.7% and 77.7% (see https://www.hesa.ac.uk/data-and-analysis/publications/destinations-2015-16/introduction).

8. The DLHE provides information on the workplace location of graduates from each university. This location is defined at the NUTS-1 level. All other variables concerning university and regional characteristics are built with location information at the NUTS-2 level.

9. Table A1 in Appendix A in the supplemental data online presents detailed definitions of the variables. Table A2, also online, includes summary statistics.

10. The HEBCI survey is a compulsory survey taken annually by the population of universities in England.

11. Regional absorptive capacity is linked to both high-tech manufacturing and KIBS employment; however, due to collinearity issues between high-tech manufacturing employment and wages in manufacturing, we only include the latter in the analysis.

12. In the UK, the larger the city, the more universities are likely to be located there. Generally those are a combination of pre- and post-1992 HEIs (Goddard et al., 2014). In our data, London (UKI3 and UKI4) hosts 28 HEIs, followed by the next biggest cities, Birmingham (UKG31) with five and Manchester (UKD33) with three, respectively.

13. Results for random effects panel model are presented in Table A4 in Appendix A in the supplemental data online.

14. Table A4 in Appendix A in the supplemental data online includes as an additional robustness check a spatial model that accounts for the spatial autoregressive process of the dependent variable, that is, a spatial error regression model (SEM). The results remain unchanged with respect to the ones presented here (see Appendix A in the supplemental data online).

15. The selection of this distance is based on other studies concerned with changes in employment location among cohorts of young adults who have recently graduated (Abreu et al., 2015). In order to test the robustness of our results, we additionally estimated our models using 50 and 100 km distance thresholds, and the results remain unchanged (the results are available from the authors upon request).

16. See Figure A1 in Appendix A in the supplemental data online for more details on the geographical patterns of retention.