Stochastic input-output modeling

Input-output data (IO data) are compiled by survey methods, which are based on a sample. It is well known, that IO matrices are not stable overtime. Therefore different actualizations methods have been developed. Econometric methods play important role in the realization of input-output matrices. These methods, on the basis of the relevant data, can facilitate the formulation of IO value forecasts. Established parameter and random term estimators are a source of uncertainty for IO parameters. In a historical perspective the author discusses stochastic methods in IO theory as well as their applications.