POLICY IMPORT TAX INCENTIVE UTILISATION DURING THE COVID-19 PANDEMIC

Abstract

The objective of a policy will be achieved if the policy targets are well-defined. However, it often happens that the policy targets do not utilise the public policies given optimally. Therefore, it is essential to predict the targeted group that will utilise the policy provided so that the policy implementation can be more efficient. This study uses the Data Mining method as a policy decision support tool to analyse the best predictive model of the PPh 22 Import Tax Incentive utilisation during the COVID-19 Pandemic. The predictive model can be practical and valuable for Indonesia's tax authority and policy analysis knowledge. This study uses administrative data of 43.547 taxpayers who are already utilising and not utilising the tax incentive, combined with the data mining method. The results showed that with the Random Forest algorithm, the utilisation of the tax incentive could be predicted with an accuracy above 94%. Furthermore, The Total Sales Value, The Export Value, and The Bonded Zone Category are the variables that mostly predict the utilisation of the tax incentive. Therefore, tax authorities should be able to utilize data mining as a tool to be able to implement tax incentive policies more efficiently and accurately.