Quantitative analyses may aim to provide actionable answers to policy questions and to generate tools or insights for decision-making. Given the deep uncertainties involved in any realistic reckoning of policy questions, this study argues that only the second of these goals is achievable. Here, this argument is illustrated by considering analyses of how the electrification of an activity changes the damage from the air pollution emissions that occur because of that activity. The sources of uncertainty in such an analysis include the long life of the technologies being studied. Consequently, the structure and operation of the electricity grid might change because of the new technology and independent of it. Analysts must make subjective choices about what to include in their analysis and what to exclude. For example, policies modeled in isolation may, in reality, be bundled with other policies; interactions between technologies may be missed if the analysis focuses on only one technology; and certain benefits or costs may be neglected because they lie outside the scope of the analysis and the expertise of the analyst. Quantitative policy analysis must aim to be part of the broader discussions in society that ultimately determine what policies get implemented.