The "common agency" model, introduced by Bernheim and Whinston in 1986 combines the fundamental challenge of the principal–agent model with the challenges of coordinating multiple principals. In "Incomplete information VCG contracts for common agency," Alon, Talgam-Cohen, Lavi, and Shamash show that the class of common agency settings for which there exists a contract that guarantees truthfulness of all principals, welfare maximization, and the two standard properties from contract theory—limited liability for the agent and individual rationality for the principals—is identifiable by a polynomial-time algorithm. Furthermore, for these settings, the authors design a polynomial-time computable contract: given valuation reports from the principals, it returns, if possible for the setting, a payment scheme for the agent that constitutes a contract with all desired properties. We study contract design for welfare maximization in the well-known "common agency" model introduced in 1986 by Bernheim and Whinston. This model combines the challenges of coordinating multiple principals with the fundamental challenge of contract design: that principals have incomplete information of the agent's choice of action. Our goal is to design contracts that satisfy truthfulness of the principals, welfare maximization by the agent, and two fundamental properties of individual rationality (IR) for the principals and limited liability (LL) for the agent. Our results reveal an inherent impossibility. Whereas for every common agency setting there exists a truthful and welfare-maximizing contract, which we refer to as "incomplete information Vickrey–Clarke–Groves contracts," there is no such contract that also satisfies IR and LL for all settings. As our main results, we show that the class of settings for which there exists a contract that satisfies truthfulness, welfare maximization, LL, and IR is identifiable by a polynomial-time algorithm. Furthermore, for these settings, we design a polynomial-time computable contract: given valuation reports from the principals, it returns, if possible for the setting, a payment scheme for the agent that constitutes a contract with all desired properties. We also give a sufficient graph-theoretic condition on the population of principals that ensures the existence of such a contract and two truthful and welfare-maximizing contracts, in which one satisfies LL and the other one satisfies IR. Funding: Funding is by the European Union (ERC, ALGOCONTRACT) [Grant 101077862]. [ABSTRACT FROM AUTHOR]