Determinants of Patent Applications across European Regions: Evidence from Panel Estimators, K-Means Clustering and Predictive Validation

Europe measures the innovative performance of its regions largely by counting patents, and allocates public resources accordingly. This paper asks what regional patenting actually reflects, drawing on Regional Innovation Scoreboard data for 245 European regions observed annually between 2016 and 2023. Three candidate drivers are considered: the research effort of firms, the intensity of formal collaboration between the research base and industry, and the propensity to protect intangible assets through trademarks. Business research effort emerges as the dominant correlate throughout, science–industry collaboration as a weaker but consistent one, and trademark activity as a positive one, suggesting that firms which protect brands are not forgoing patents but exercising a single appropriation capability across several instruments. Two findings carry implications beyond measurement. Regional innovative capacity proves remarkably immobile: differences between regions account for roughly 95 per cent of the variation in the data, and differences within a region over the eight years for the remainder, so the short-run movements on which policy evaluation typically relies carry very little information, and the returns to innovation investment should be sought over horizons far longer than a programming cycle. And when the 245 regions are grouped into four innovation profiles rather than treated as a single population, the relationship that holds on average holds almost nowhere in particular: it is strong among leading and lagging regions, statistically absent in the largest group, and displaced by brand-led appropriation in a fourth group of 27 regions whose innovation is real but largely non-technological. Uniform innovation policy prescriptions and uniform managerial benchmarks are correspondingly difficult to justify.