Prof. Andrea Antonio Caragliu (p), Prof. Roberto Dellisanti

The current geopolitical landscape is highly fragmented and complex. It is characterized by rising tensions, shifting alliances, and by a largely uncertain global governance. This turbulence has dire consequences, especially at territorial level: it may contribute to widening interregional disparities.
Prior ex ante assessments of the way these instabilities may find a solution suggest that a scenario with the EU deepening its internal integration, coupled with tighter global economic linkages (Integration scenario), would yield stronger economic performance, associated with smaller increases in cross-country disparities. By contrast, a scenario with looser internal integration within the EU, coupled with a fragmented global economy (Fragmentation scenario), would bring about comparatively larger within-country disparities, as slower economic expansion would keep internal inequality more contained.

In this paper we propose a set of simulations aiming at disentangle the role played by pre-market, in-market, and post-market factors in driving regional inequalities. Simulations are based on the fifth generation of the MAcroeconometric Social Sectoral Territorial (MASST5) model.

Our results suggest a mosaic of often contradictory effects. Should either state of the world depicted in the two scenarios prevail, our simulation results provide a detailed picture of the mechanisms driving overall results, thereby informing policy interventions aiming at minimizing the negative effects of specific development patterns.

Ms Anabela Santos (p), Francesco Molica, Andrea Conte

The establishment of more research networks in Europe is considered critical to achieve more strategic autonomy and technological sovereignty in strategic area. There is however, little empirical research on how such networks can advance these objectives and under what conditions. The paper assesses the impact of technology-specific EU-funded research networks on patent activity, using artificial intelligence as a case study. By combining the entropy balancing matching approach and weighted least squares regression, together with regional data, we found a robust positive effect of AI-funded research networks on AI-patenting activities between 2013 and 2021. While the network strength exhibits a monotonic relationship with patent activities, the effect of network centrality, capturing quality and prestige, follows an inverted-U-shaped pattern highlighting the need for more widespread networks beyond major hubs. The level of development, sectoral concentration and the degree of innovativeness further amplifies the magnitude of these effects. In addition, we find evidence of knowledge and technological spillovers from AI-funded research networks to other areas of patenting, notably biotech and environmental innovation. Several policy implications can be drawn from these findings. First, EU programmes aimed at fostering interregional innovation collaborations, most notably Horizon Europe, should be further strengthened as part of the broader policy agenda to advance strategic autonomy. In this respect, the European Commission’s proposal to increase the future budget of Horizon is consistent with the need to reinforce cross-border knowledge linkages in strategic technological domains. Second, policy efforts should not only increase the intensity of collaborations but also encourage a wider diffusion of network linkages beyond the current core of highly central hubs. Evidence about diminishing marginal returns related to excessive centrality suggests that overly concentrated network structures may limit additional innovation benefits.