The article explores the use of philosophical techniques of abstraction and idealization in the creation of normative theoretical models. It shows the difficulties related to distinguishing abstractions from idealizations in scientific research, as well as their common functional features, the main of which is the representation of the object under study in a form that would meet the goals and objectives of the researcher. The paper also demonstrates the ongoing popularity of ideal normative theories in the social field and explicates the problems that provoke criticism of ideal theories from various theoretical positions. It is concluded that it is necessary to strengthen such normative models by distinguishing between heuristically «good» and «bad» idealizations within the framework of the general ideal theory, and also providing methods of transition from an ideal theoretical model to the one that could be applied in practice.