Modeling and forecasting value added tax revenue in Uganda (July 2011 – June 2016)

Abstract

This study investigated the relationship between Value Added Tax Revenue and inflation, exchange rate and government expenditure in Uganda and forecasted within sample forecasts. The period of study was from July 2011 to June 2016 and Vector Auto Regression frame work (VAR) and the Vector Error Correction Model (VECM) were used for analysis. Results of the unit root test using Augmented Dickey and Fuller (ADF) test revealed that the series were integrated of order one after taking their first differences. And there existed a long run equilibrium relationship between the series based on the computed number of trace statistics. The cointegrating of VAT had a significant negative sign (a coefficient value of -0.13). This implied that in each month, it adjusted by about 13 percent to close the gap between the current level and long run equilibrium level. From the VECM results, there was a significant positive long-run relationship between Government Expenditure with VAT and between real GDP with VAT in Uganda at 5 percent level. In addition, a significant negative long run relationship was established between inflation and VAT, and between real effective exchange rate with VAT in Uganda at 5 percent level. On the other hand, a significant positive relationship was established between previous VAT revenue and real effective exchange rate (coefficient=1.01, t=2.73) at 5 percent level in the short run. It is recommended that government should regulate the devaluation of the foreign currency in order not provoke price instability and at the same time maintain a significant level of government expenditure to foster improvement in the revenue generation in the country. Government needs to ensure relatively low inflation to boost consumption hence enhancing tax revenue.