The relationship between engineering practice and technology policy is a perennial question. In this paper, we examine it in the context of the spectrum regulation debate. There is a widely held belief that regulatory and policy decisions can be justified, and indeed quantified, based on hard engineering and science facts. Similarly it is often felt that regulatory and policy domain issues should not stifle innovation by dictating the course of technology development for broadband radio systems. In this paper we strongly argue against such views and contend that key policy decisions in communications must be connected to engineering in a more explicit and nuanced way. We hold that policy makers' belief in "technology neutral" regulation is an illusion; engineering results are often used by businesses and politicians to justify their actions and promote their interests, yet the engineering means to achieve social ends are usually far upstream from political outcomes, and their connection is tenuous. Conversely, wireless communications engineers typically focus on the network protocol stack, ignoring the broader socio-economical context of their work. We show that is useful to consider a more holistic framework for radio system design that includes not just the network protocol layers but also the regulatory and commercial aspects of the technology. Given the inevitable complexity of the decision space in spectrum regulation, we believe that transparency and taking a systems view are essential for supporting evidence-based decision making around wireless technology policy and spectrum regulation. We ground our work in a technical case study, considering the extent to which engineering results can (or cannot) be used to argue the benefits of unlicensed spectrum allocation in the TV white spaces. We critique the unsophisticated use of spectrum efficiency metrics by providing a framework that demonstrates the many ways in which metrics can be constructed and interpreted. We emphasize the importance of N-dimensional engineering analysis to properly reason about the complex technical design space of modern wireless systems, and of the disclosure of financial interests by expert interlocutors. We frame our analysis of the relationship between engineering and policy using a technology circle which generalizes the seven-layer OSI network protocol stack by adding Layer Zero, the regulatory rules that constrain what may be implemented at the networking layers, and Layer Eight, the business and social practices that are built on the network, and that inform regulatory decisions at Layer Zero. Finally, we advocate the use of a fail-safe design philosophy to inform engineering work and regulatory decisions. We believe that our recommendations would make the interaction between wireless engineering and radio regulation more effective, thus strengthening the spectrum regulation and policymaking process.