We examine how innovation policies can better address development challenges by focusing on countries' existing capabilities and the complex and tacit nature of knowledge. We argue that effective policies should be place-based and multidimensional, rather than imitating solutions from elsewhere, to support sustainable economic growth.

We analyze global innovation patterns using data on scientific publications, patents, and trade, finding that economic complexity in each of these domains is closely linked to future growth, patenting, and publishing. Our results show that a country's embedded capabilities shape both its future diversification opportunities and the path-dependent nature of innovation.

We measure innovation capabilities across science, technology, and production, using data from publications, patents, and trade. Our results reveal that higher innovation complexity is linked to economic growth and that existing capabilities predict future innovation, emphasizing the need for nuanced policy indicators and strategies.

This chapter analyzes global innovation clusters using geo-localized patent and publication data, emphasizing the impact of high-skilled migrants. It finds that migration-driven personal connections are often as important as, or more important than, organizational collaborations in cluster success.

This paper uses global patent and publication data to show that knowledge production has expanded to more countries beyond traditional innovation hubs, driven in part by the rise of Global Innovation Networks. However, within countries, knowledge activities remain concentrated in a few hotspots, leading to greater connectivity among these areas and potential distributional challenges.

We examine the educational backgrounds and career trajectories of U.S. faculty from 1880 to 1990, to understand the reliance of U.S. universities on foreign-educated and foreign-born faculty. Our findings reveal that over a century, U.S. universities transitioned from relying U.S.-born, foreign-educated to foreign-born, U.S.-educated faculty.

We analyse the effects of geopolitical tensions on international research and innovation collaborations. We find that belonging to different geopolitical blocs significantly reduces the likelihood of collaboration, and this association is even more pronounced in critical technologies.

A paper describing stylized facts and listing potential research avenues for geographic content metrics. It extracts geographic content from titles and abstracts using simple NLP models.

We study how migration changes the geographic focus of research by Colombian scientists. We find that scientists who initially studied Colombia begin to study other countries after leaving Colombia.

We study how technologies combine in a single individual and in collectives of individuals and firms. We find that although a single individual combines a median of two technologies, and a country combines 155, both produce a similar number of technology-technology combinations. However, the strength and significance of these associations are highly heterogeneous.

We study the historical origins of current prestige hierarchies and patterns of inequality in US faculty hiring. We find that, since 1880, the supply (education) of US faculty has been highly concentrated in a few universities. Prestige hierarchies have remained persistent, even though the most prestigious universities in 1880 differ from those of today.

A paper proposing an empirical methodology to map the linkages between science, technology, and production capabilities. Our findings reveal that having strong scientific and production capabilities in cross-related domains predicts subsequent technological development.

A paper studying the role of ethnicity, inferred from name analysis, in inventor collaborations. After controlling for different types of proximity measures, we find that inventors from the same ethnicity have a larger probability of collaborating to file a patent.

A paper describing a methodology to define geographic cluster boundaries from millions of georeferenced data.