Fuzzy Regression

This article examines the effect of government size on the high, medium and low thresholds of the Gini coefficient in Iran. For this purpose, the auto regression model of soft fuzzy logistic transfer (FLSTAR) has been used for the period of 1997-2019. One of the reasons for using this model is flexibility in its application. The main focus of this paper is to calculate the Gini coefficient bands according to the size of government in the economy. Hence, we calculate the bands (high, middle and low) of the Gini coefficient. The study show that the threshold size of the government is equal 0.499. Findings of this research are applied in a real case which reveal that with increase of government share in economy the Gini coefficient increases as well. Therefore, the government should seriously pursue privatization policies.

Eliminating poverty and reducing income inequality is the most difficult task of economic policy making when considered with an economic growth. Therefore, the impacts of economic growth and development on income distribution largely depend on the growth model. Thus, determining an effect of economic growth on poverty is uncertain and relies on its country's growth pattern. The basic question of this paper which we are looking for is that have economic growth and income distribution had an impact on the poverty in Iran? To respond the question, we employed fuzzy regression for analyzing the variables relationship. The article also under the fuzzy logic presents an analytical framework at the aim of appraising the relationship between poverty, income distribution and economic growth. The results show that there is a positive relationship between the economic growth and poverty, in other words the economic growth has had negative impacts on the poverty reduction in Iran. Computational results based on fuzzy logic in Iran implies that the most ideal conditions come when the economic growth is 8 percent on overage and Gini coefficient is nearly 38.5 percent.