Ali Asghar Banou'i; zahra zabihi; parisa mohajeri; elham tabrizi
Volume 15, Issue 59 , January 2016, , Pages 95-124
Abstract
Errors which occur in the process of collecting and compiling databases and developing symmetric input-output tables are inevitable. The issue of stochastic data contained in the input-output tables has been one of the key issues discussed in input-output economic literature. Foreign researchers have ...
Read More
Errors which occur in the process of collecting and compiling databases and developing symmetric input-output tables are inevitable. The issue of stochastic data contained in the input-output tables has been one of the key issues discussed in input-output economic literature. Foreign researchers have demonstrated in theoretical studies that if the matrix of technical coefficients is stochastic, the Leontief production multipliers will be positively biased. Although the findings of applied studies (that include the complier's and practitioner's approach) confirm the above observation, but they also show that this bias is trivial and therefore can be ignored. In this study, the approaches of practitioner and complier to the analysis and estimation of the bias of stochastic input-output multipliers and the estimation of multiplier bias are explained. For this purpose, we use Monte Carlo simulation method from the Complier’sview to estimatethe bias of production multipliers and the effect of sample size on the bias. The findings of this study suggest that, first, the greater the sample size the less the amount of bias of multipliers and second, the greater the sample size, the higher percentage of elements in matrix of production multipliers demonstrate a positive bias. Third, in a sample with large size, all multipliers have significant positive biases that is in line with the findings and results of analytical studies, however this bias is very small.
