Causal Bayesian networks: Modelling an Impact of Interventions in Economic Models

The presented Thesis is focused on the identification of causal effects in causal models. Knowledge of causal relationships among variables allows the decision-maker to make predictions when interventions are employed. In the case of some unmeasurable variables being involved in the model, making predictions is complicated. But even in the presence of unmeasurable variables, the causal effect (an impact of an intervention) can be identified (provided that certain conditions are met).It is shown in the Thesis that, in the case that the structure of original model does not allow to infer the causal effect, certain alterations in the structure of the model can be applied in order to enable the causal effect identification. This problem is presented on a set of example economic models. These examples are created based on the idea of using such models as a support tool for decision-making - thus, in order to show the applicability of the presented approach, the selected set of example models is built with the aid of simulated data, and concrete causal effects are quantified.