A MATHEMATICAL FORECASTING MODEL TO ESTIMATE TAX TO GDP RATIO

In this research different linear and nonlinear models are applied to estimate tax to GDP ratio of Pakistan. The data from 1991-2013 has been analyzed in this context and forecasting is done till 2025. The efficiency of the models are validated by comparing the values of the forecasting errors i.e. mean absolute deviation (MAD), mean squared deviation (MSD) and mean absolute percentage error (MAPE). It is evident from the analysis that the nonlinear model is suitable for forecasting tax to GDP ratio. Moreover, no substantial difference exists between the actual values with those predicted from the model. Therefore, this analysis may provide the policy makers and government with the opportunity to investigate Pakistan's tax revenue.