Firms Partnering for the Twin Transition: The Role of Spatial, Technological and Relational Proximity

This paper investigates the determinants of interfirm agreement formation in the context of the twin digital and green transition. We focus on strategic alliances and joint ventures involving at least one Italian firm, using SDC Platinum data on agreements announced between 2000 and 2025. Digital and green agreements are identified through a keyword-based classification of deal synopses. The empirical analysis is conducted at the dyadic level by comparing realised agreements with potential firm pairs within the framework of rare event logit models, focusing on the role of geographical, technological and relational proximity. The results show that technological proximity is the strongest predictor of agreement formation. Firms operating in connected industrial domains are substantially more likely to collaborate, suggesting that compatible knowledge bases and absorptive capacity are central to partner selection. Geographical proximity also matters, mainly through coordination and interaction costs rather than administrative co-location. The comparison between digital and green agreements shows that both domains require technological compatibility, but they rely on different forms of proximity and complementarity. Digital agreements are especially sensitive to broad network-based technological proximity, consistent with the modular and cross-sectoral nature of digital technologies. Green agreements combine compatible but differentiated capabilities with a stronger spatial and implementation-related component, reflecting their connection to infrastructures, regulation, and local coordination conditions. Prior relational proximity increases the probability of agreement formation in the full sample, while network effects are more exploratory in the digital and green subsamples. The paper contributes to the literature on alliances, proximity, and transition-oriented innovation by showing that twin-transition collaboration is shaped by multiple and partially distinct proximity mechanisms.