What is the optimal design of the international corporate tax system? We revisit this classic question in a multi-country general equilibrium model that incorporates three key features of the modern globalized economy: multinational production; intangible capital; and international profit shifting. Our model’s competitive equilibrium is inefficient due to an externality that arises from international spillovers in intangible investment. In the absence of profit shifting, there is little, if anything, a Ramsey planner can do with corporate income taxes to improve the allocation of intangible investment across countries. However, profit shifting allows the planner to use corporate income taxes to internalize the externality and achieve an efficient allocation of intangible investment. To quantitatively investigate the properties of the Ramsey planner’s optimal policy in a more realistic setting, we extend our model to an environment with firm heterogeneity and selection into multinational production. Without spillovers, it would be optimal to shut down profit shifting as much as possible. With spillovers, it would be optimal to allow MNEs to continue to shift profits, and if the planner is restricted to Pareto-improving policies, it would be optimal to allow even more profit shifting than under the status quo.

国际企业税制的优化设计是什么？我们在多国一般均衡模型中重新审视这个经典问题，该模型融合了现代全球化经济的三个关键特征：跨国生产；无形资本；和国际利润转移。由于无形投资的国际溢出效应产生的外部性，我们模型的竞争均衡是低效的。在没有利润转移的情况下，拉姆齐规划者几乎无法利用企业所得税来改善无形投资在各国的配置。然而，利润转移使得规划者可以利用企业所得税将外部性内部化，实现无形投资的有效配置。为了在更现实的环境中定量研究拉姆齐规划者最优政策的属性，我们将我们的模型扩展到具有固定异质性和跨国生产选择的环境。如果没有溢出效应，最好的办法就是尽可能地阻止利润转移。考虑到溢出效应，最优的做法是允许跨国企业继续转移利润，而如果规划者仅限于帕累托改进政策，则最优的做法是允许比现状更多的利润转移。