The consequences of correctly implementing a normative political theory arguably bear on its acceptability. A theory whose correct implementation permits slavery is highly implausible. We defend a claim about incorrect implementation. We argue that normative political theories that will predictably be put to bad use deserve harsher assessments than theories that will predictably be put to better use. Theories that key political actors will predictably invoke to justify bad policy recommendations are bad theories, even if those recommendations are not logical consequences of the theories (even in conjunction with well-established factual propositions). However, standard discussions of political theories ignore such “misapplications” when it comes to assessing the principles and values proclaimed by those theories. We reject this view because the action guidance that a normative political theory effectively provides is mediated by political actors’ prior beliefs. At the limit, these actors may regard public policies as specifying principles that the theory proclaims, as something different from regarding such policies as causally effective in achieving the ends or enforcing the rights proclaimed by those principles. As a result, political actors see themselves as exempted from the need to undertake careful causal analysis. Morally defensible action guidance requires, then, that those principles be qualified by empirically-grounded policy constraints. Normative political theories qualified in those ways constitute theoretical improvements, that is, improvements on the action guidance that the theories provide on their own.