Das Ziel der RISIS Research Seminars ist es, den Austausch von Ideen innerhalb und außerhalb der RISIS-Community zu fördern. In jedem Seminar präsentieren und diskutieren Wissenschaftler:innen sowie Praktiker:innen ihre laufenden Forschungsarbeiten. Die Seminare unterstützen den wissenschaftlichen Austausch in einer kritischen Phase der Forschung zu einem gemeinsamen Thema.
Die Seminare sollen ein breites wissenschaftliches Publikum aus den Bereichen Policy, Hochschulforschung und Innovation einbeziehen. Zur inhaltlichen Vertiefung der Seminare wird zusätzlich eine Diskutant:in aus der wissenschaftlichen Community hinzugezogen.
Allgemeiner Ablauf der Seminare
Die Seminare bieten Raum für die Präsentation und Diskussion laufender quantitativer Forschungsarbeiten von Wissenschaftler:innen und Praktiker:innen zu relevanten Fragestellungen. Die Nutzung von RISIS-Datensätzen oder des European Higher Education Sector Observatory ist erwünscht, jedoch keine Voraussetzung. Die Online-Seminare werden in englischer Sprache abgehalten.
Als Vortragende:r teilnehmen
Wenn Sie Interesse daran haben, Ihre Arbeit im Rahmen der RISIS Research Seminars vorzustellen, wenden Sie sich bitte an Georg Zahradnik.
RISIS Seminar #58
Title: Firms’ disclosure of university ties on their website: An explorative analysis of its role for innovation performance
Date: 23 September 2026
Presenter: Bastian Krieger, ZEW, Leibniz Centre for European Economic Research
Discussant: tba
Abstract: This paper explores a novel web-based indicator to examine how firms' disclosure of university ties on their websites shapes their innovation performance. First, using data from the German Community Innovation Survey 2023 and the Tenders Electronic Daily database, combined with firms' disclosure of university ties on their website provided by ISTARI.AI, we investigate the indicator's properties by comparing the most frequently disclosed types of university ties: innovation collaborations, university customers, and employee education, with firms' survey responses and their procurement contracts. Second, we analyze how website disclosure of university ties relates to firms' revenues from new or significantly improved products or services, applying Ordinary Least Squares, a Control Function, and a Lewbel Instrumental Variable approach. In sum, the website disclosure of ties with universities is significantly associated with their related survey items and procurement contracts. Moreover, website disclosures show no consistent association with revenues from innovations new-to-the-firm. A consistent statistically significant relationship emerges only for small firms, where website disclosures are associated with higher revenues from market novelties. These findings suggest that our web-based indicator captures ties between firms and universities and that disclosing these ties on firms' websites may influence the market success of their novel products.
RISIS Seminar #59
Title: Doctoral labor market returns in Switzerland
Date: 14 October 2026
Presenter: Kevin Schönholzer, Università della Svizzera Italiana (co-authored with Benedetto Lepori)
Discussant: tba
Abstract: Most doctoral graduates do not enter or stay in academic positions, despite a sustained policy concern about doctoral overproduction (Skakni et al., 2025; OECD, 2023). Existing evidence on the labour-market returns to the doctorate is dominated by pooled Master-versus-PhD wage comparisons and by STEM-heavy register studies. Neither approach can examine within-destination earnings differentials, and ISCO classifications further obscure the picture because they encode professional field rather than work content and contain no research-occupation code. Here we use the Swiss Graduate Survey (EHA) for four cohorts surveyed between 2017 and 2023 to estimate within-destination PhD-Master wage gaps across four labour-market segments. The regression-adjusted gap falls in a 17 to 20 percent band: Academic 18.0 percent, Industry-research 16.7 percent, Public-non-research 19.7 percent, Private services 19.1 percent. None of the PhD-by-destination interactions reaches conventional significance, and the 95 percent confidence intervals overlap, so the data do not reject equal within-destination gaps. The PhD median sits at the 68th, 67th, 86th, and 71st percentile of same-destination Masters. The within-destination band exceeds the pooled t+5 PhD premium of 14.5 percent because PhDs are over-represented in Academic, where the Master comparison group is depressed by Gymnasium teachers on part-time pensums. To compute these comparisons we construct a research-occupation indicator from EHA self-reports that shows convergent agreement across independent classifiers. The analysis is observational and the within-destination gap admits credentialist, productivity-selection, and non-cognitive-trait selection readings that the data cannot arbitrate.
RISIS Seminar #60
Title: Re-examining the Cost-Employability Link in Higher Education: A Simulation-Based Sensitivity Analysis
Date: 11 November 2026
Presenter: Rosario Scandurra, Universitat Pompeu Fabra - Barcelona School of Management
Discussant: tba
Abstract: Despite widespread assumptions that higher university costs reflect better graduate outcomes, this study finds consistently weak correlations between university expenditures and employability metrics. Using a comprehensive Monte Carlo simulation and global sensitivity analysis, we assess whether various input parameters, including tuition fees, living expenses, housing costs, and different measures of employability, influence the correlation between overall university cost and QS employability ranking. However, heterogeneity analysis reveals that the relationship becomes stronger for Anglophone higher education systems, showing that in highly hierarchical stratified higher education system. The results reveal a correlation ranging between r=-0.03 and r=0.07. Heterogeneity analysis focusing on Anglophone universities report a strong relationship ranging between public funding and employability (r=-0.012) metrics European universities show that a higher share of public funding in university budgets does not immediately translate into better employability outcomes. This suggests that neither increased student expenditures nor greater public intervention guarantees superior graduate labour market performance. Sensitivity analyses, conducted through Random Forest feature importance, Sobol' variance decomposition, and Simulation Decomposition (SimDec), show that methodological choices in constructing employability metrics, including handling missing data, weighting components, and selecting or excluding indicators, are the primary drivers of almost 70% of the variability in the observed correlation. In contrast, tuition fees, living expenses, housing costs, and public funding levels have minimal impact. The findings challenge common widespread assumptions that resource-intensive education directly translates into better labour market outcomes. The results highlight critical measurement validity issues in university rankings and suggest that higher education employability metrics depend more on methodological choices than substantive differences in educational investment. Current employability metrics may mislead prospective student choice and private and public investment in higher education: a standardization of measurement approaches is strongly needed.
