This Thesis advances Economic Geography by applying causal inference and applied microeconomics methods to research questions in Evolutionary Economic Geography. It provides new empirical insights into the geography of innovation, focusing on how technological dynamics shape manufacturing outcomes at the regional level. A core contribution is to address limitations of the Relatedness Framework, improving its robustness and expanding its applications. The novelty of this work lies in bridging two traditionally separate literatures: the geography of innovation and labor economics. By adopting identification strategies typical of causal inference, it uncovers how innovation and labor institutions interact in shaping regional trajectories. The Thesis comprises an introduction and three main chapters. The Introduction outlines the theoretical background and research questions, focusing on the Relatedness Framework and the relatedness density metric, which ground the analyses in Chapters 2 and 4. Chapter 2 — Turning Technological Relatedness into Industrial Specializations: The Effects of Smart Specialization in Europe — evaluates the productivity impacts of Smart Specialization strategies. Using the relatedness-entry model and patent-based instruments, it shows that EU regions where strategies aligned with technological strengths experienced stronger labor productivity growth. This chapter, published in Economic Geography (2025), is one of the first policy-evaluation works on S3 policies. Chapter 3 — Exposure to Innovation and Labor Market Dynamics: Evidence from the German Manufacturing Sectors — studies the effects of innovation on wage distributions. Drawing on German administrative labor data and patent-based measures of exogenous innovation shocks, it shows that innovation disproportionately benefited routine rather than abstract workers. The chapter argues that this counterintuitive result reflects the role of collective bargaining institutions in German manufacturing. Chapter 4 — Related or Exposed? A Shift-Share Approach to Relatedness Density — introduces a Shift-Share Instrumental Variable (SSIV) design to address endogeneity in relatedness density. The instrument is validated using European and U.S. patent data. Results show that OLS estimates understate the true effect of relatedness on the emergence of new regional specializations, while the SSIV framework delivers more credible causal estimates. Overall, the Thesis makes three contributions: (i) policy evaluation of Smart Specialization strategies; (ii) revealing how labor institutions mediate the impact of innovation on wage outcomes; and (iii) advancing methods in Evolutionary Economic Geography by applying shift-share designs to relatedness analysis. Together, these findings enrich our understanding of the geography of innovation and strengthen the empirical foundations of the Relatedness Framework.

2026-05-25 JG: Author's and/or other hand signatures removed for GDPR compliance