The Iranian Government's Budget Deficit Analysis By Artificial Neural Network Simulation

The main purpose of this studyisinvestigated the impact of fluctuations in oil revenues and effects of tax revenues in the budget deficit simultaneously. Also in this study we has been used a multilayer neural network (MLP) as a powerful tool for the simulation of nonlinear andfluctuations behavior to simulate the impact of changes in oil and tax revenues on the budget deficit. Results showthat 13 to 26 percent increase in tax rate can improve budget deficit, in a manner that 26 percent increase in the tax rate will have the greatest improvement in the budget deficit. The budget deficit gets worse for upper than 26 percent increase in tax rate. For example 31 percent increase in tax rate can increase budget deficit in 30 percent. In other scenario we investigate that30 percent decline in oil revenues must be at best 60 percent tax rate increase to 80 percent the size of the budget deficit wills improves.This reflects theheavydependenceof budget deficit on oil revenues and the low effects of tax revenues on budget. Also in the other scenario the results show that if the oil revenue decreases 30 percent, even with a 10 percent increase in tax rate, the government deficit will increase into twice. This is due to reduced government revenues resulting from declines in oil revenues due to decreased investment and lower total production. In addition to thesefindings, the results ofthis study indicate that, the improvement in the budget deficit by growth rate increases until a threshold point. After this threshold point, the rate of improvement in the budget deficit is reduced by growth rate increases.The reason for thiscan be government expenses behavior, government spending increases withincreased in income. Assuming no changes in tax rates and oil revenues,the economic growth rate of 7.8percent (threshold point), budget deficit, 70 percent will improve.