Abstract: We document that 28% of European-origin unicorns relocate to the US, representing 40% of the value of all European-origin unicorns. We study the aggregate effects of this “business drain” in a two-region endogenous growth model in which heterogeneous startups choose where to locate and develop their product, and rely on venture capital (VC) funding to cover R&D costs upfront. VCs compete for startups under location-specific institutional frictions, so that VC funding depth is endogenously higher in the US. The model generates positive selection into relocation. Using a novel dataset of eleven thousand European VC-backed startups with hand-collected headquarters histories, we document that relocators are positively selected on pre- move funding and innovation, and that after relocating they raise twice as much venture capital and produce 17% more citation-weighted patents than matched stayers. We calibrate the model to these facts and use it to run counterfactuals. Startup relocation has substantial aggregate and distributional effects on the EU economy, reducing EU innovation by 6%, and moves innovation, wages, and equity  stakes to the US, but it also raises world innovation by moving the most productive startups to the ecosystem that provides more funding and is more efficient. Lowering EU VC market frictions to the US level retains most startups and raises EU innovation, but closes neither the US–EU innovation gap nor the funding gap of the most productive EU startups, because they still scale in a less efficient R&D ecosystem and are therefore less valuable to fund.

for information patrizia.pellizzari@unibocconi.it