New Political Economy has something very important in common with welfare economics: its focus on static, technical efficiency criteria to judge the rationality of a social, political or economic order. This often leads theorists to perceive their objects of research as well-defined problems to which clear-cut solutions can be found and prescribed as policy proposals, addressed at the policymaker or the democratic sovereign. This perspective frequently excludes important empirical phenomena from the research agenda. Although, for example, certain well-defined informational asymmetries are frequently modeled, fundamental knowledge problems such as ignorance of the true model of the economy are usually ignored. In the present paper, this approach is criticized from a Hayekian point of view, with an emphasis on the problems of "scientism" (i.e., the inappropriate transfer of methods from the natural to the social sciences) and irremediably imperfect knowledge, troubling both the agents in the theoretical model and the theorist. Furthermore, it is argued and illustrated with two examples that an extension of Public Choice's research agenda along Hayekian lines may be fruitful, because it leads to a fresh perception even of such problems that already have been extensively researched within the traditional framework.