Metoda instrumentálních proměnných v průřezové analýze

In this thesis, the effect of so-called weak instrumental variables on estimates in applied econometric research is examined. Presented within are possibilities of how to detect and handle weak instruments. In the first section, it is shown how the consistency of the OLS estimator depends on the assumption that the regressors are independent of the disturbance term. Concrete examples of a violation of this assumption are considered, i.e. the omission of a relevant variable, an error in the measurement of a regressor and autocorrelation and reverse causality. With instrumental variables estimation, a solution to this inconsistency problem is introduced. The advantages and disadvantages of this estimator are discussed, as are the assumptions which have to be made in order to use it. Based on recent research, the nature of weak instrumental variables is explained. The most important results are verified through a small Monte Carlo simulation in the final section preceding the conclusion.