UCLA

Abstract: Research Summary: We use a novel longitudinal dataset, constructed from 16 downloads of VentureXpert records collected over 20 years, to characterize the international migration of venture‐capital‐backed startups. We find that: (i) 1078 firms in our sample (1.4%) migrate; (ii) countries with high levels of in‐migration also have high levels of out‐migration; (iii) migrating firms move to places with more investors; (iv) pre‐move investors and their connections most strongly predict migration patterns; and (v) movers raise more money than non‐movers, primarily from investors at their destinations. Overall, these patterns appear inconsistent with those expected if startups move primarily in search of talent or customers. Instead, the flows across countries look more like international trade, with startups seeking capital, and social connections between investors defining the shipping lanes. Managerial Summary: Although many high‐profile startups have relocated their headquarters from one country to another, systematic information on this phenomenon has been scarce. How frequently do these moves happen? Why do startups move? Over 20 years, we have built a database that can begin to answer these questions. International moves appear rare. When startups do move, they tend to move to places with more venture capital, particularly when their existing investors have connections in those places. Movers, moreover, raise more money than non‐movers, mostly from investors in their destination countries. Capital availability, rather than access to talent or proximity to customers, appears to be the strongest predictor of startup migration.