Delegation of decision authority is a fundamental characteristic of systems engineering problems. Engineers and managers at higher levels within the organization allocate responsibility and resources to other individuals through requirements flowdown and other processes. The prevalence of schedule slippages and budget overruns on systems engineering projects raises questions about the adequacy of and potential to improve existing methods. However, at present the community lacks a foundational understanding of these processes that would be valuable in identifying and validating candidate improvements to the systems engineering process. In this paper, we develop a conceptual modeling framework for delegation in systems engineering based on the principal-agent problem, a game-theoretic model of agent interaction across hierarchical levels. Several variations on the basic model are possible. We study the model and its variations on an illustrative example of a passenger vehicle engineering process. Numerical results highlight the impact of various assumptions, including whether engineers act normatively or according to proposed behavioral decision models. Implications and extensions are discussed, including the need for behavioral validation of engineering decision models and the potential to use the modeling framework to evaluate newly proposed delegation schemes.