We study whether and when research and development (R&D) activities by foreign multinationals facilitate the formation and growth of new innovation clusters. Combining information on nearly four decades’ worth of patents with socioeconomic data for regions that cover virtually the entire globe, we use matched difference-in-differences estimation to show that R&D activities by foreign multinationals have a positive causal effect on local innovation rates. This effect is sizeable: over a five-year period, foreign research activities help a region climb fourteen centiles in the global innovation ranks. This effect materializes through a combination of knowledge spillovers to domestic firms and the attraction of new foreign firms to the region.

However, not all multinationals generate equal benefits. In spite of their advanced technological capabilities, technology leaders generate fewer spillovers than technologically less advanced multinationals. A closer inspection reveals that technology leaders also engage in fewer technological alliances and exchange fewer workers with local firms abroad than less advanced firms.

Moreover, technology leaders tend to set up their foreign R&D activities in regions with lower levels of economic development than less advanced firms, yet with comparable public-sector research capacity. These findings suggest that technologically leading multinationals face comparatively unfavorable trade-offs between the costs and benefits of local spillovers, underscoring the importance of taking the strategic choices that firms face into account when analyzing innovation clusters.

R. Crescenzi, A. Dyèvre, F. Neffke, Innovation catalysts: How multinationals reshape the global geography of innovation, Economic Geography (2022) 1–29