The DGT, Big Data, and Predictive Modeling: “Too Big To Fail”

Big Data is a collection of data generated from the activities and habits of digital users (example: demographic information and consumer psychographic, reviews and suggestions of products, blogs and content on social media). The advantages of Big Data processing and analysis have been proven by the success of large world enterprises such as Google, Apple, Facebook, and Amazon (“GAFA”).

These companies managed to use Big Data for efficient market research, and identification of consumer needs. The National Oceanic and Atmosphere Administration (“NOOA”), a United States government institution, also uses Big Data and routinely manages 30 petabytes of new data per year.

Besides NOOA, International Revenue Service (“IRS”), a United States federal agency, has also implemented Big Data analysis in its operational activities. Processing and analysis of Big Data (widely known as data mining) will produce statistics with a high level of validity and contribute to work efficiency for the organization.

Through the e-filling system, Indonesian government had succeeded in gathering information about Taxpayers for the digital tax system. The Directorate General of Taxes (“DGT”) also collects information from other sources, such as Bank Indonesia (BI), Directorate General of Customs, social media, and ORBIS database. These Taxpayers data collected will be gathered in a system called the Integrated Data Warehouse (“DAWET”). With this procurement of Big Data, DGT aim for high tax collection targets, keeping in mind the large amount of budget support issued amounting to 1.5 trillion Rupiah.

The government targets related to the use of Big Data are to fight fraud & tax avoidance. By using Big Data, the DGT will find it easier to view Taxpayer’s profiles such as relations and/or contact between Taxpayers, life habits, and activities between Taxpayers who routinely conduct financial transactions. Furthermore, based on these data, DGT will conduct quantitative and qualitative analysis to view any potential tax avoidance or fraud by Taxpayers.

The advancement in the Indonesia’s taxation system is indeed commendable, but before initiating major changes, it is best to be thorough in the preparation process. For example, the preparation process may consist of building a long-term cooperation between parties who will help in succeeding the processing of Big Data and supplying accurate end products. These end products that will be generated from Big Data analysis is the prediction of the model of potential tax compliance deviations in Indonesia.

There are several techniques for making predictions in statistics, one of which is the Lasso and Ridge regression and classification models. However, without the accuracy of the model, the technique will not produce valid information.

Data that will be processed to produce a model should not only be quantitative data but also qualitative. In this case, difficulties will arise in collecting accurate qualitative data. The DGT needs to conduct long-term cooperation with the Ministries and other Departments to collect information in the form of qualitative data. In addition, the DGT also needs to engage experts in Behavioral Analytics and Artificial Intelligence to determine which control variables are appropriate for creating predictive modeling in Indonesia.

Experts in both fields, coupled with the Financial Data Analyst (Finance Researcher), will produce information in the form of data that can be interpreted by tax practitioners. This collaboration between practitioners and researchers is a requirement to find the accurate tax fraud predictive model.

The Predictive Modeling produced by the DGT is expected to resolve tax avoidance cases while maintaining the confidentiality of Taxpayers information. Therefore, it is necessary to pay attention to legal issues that may arise, such as audit ethics violations and lack of transparency in the algorithms used. Notification of data usage, Taxpayer confidentiality, and strict control of personal Taxpayer information must be ensured by the DGT.

This is to avoid misuse of Taxpayer data, one of which is the use of data for political purposes. To avoid unwarranted incidents, the DGT needs to conduct pre-trials/ experiments with the predictive modeling results. It is expected that the use of Big Data will produce satisfactory results. A false prediction modeling is too great of a sacrifice for Indonesia that may result in failure.

 

Disclaimer:
The information contained on this article is intended solely to provide general guidance on matters of interest for the personal use of the reader, who accepts full responsibility for its use. The application and impact of laws can vary widely based on the specific facts involved. Given the changing nature of laws, rules and regulations there may be delays, omissions or inaccuracies in information contained on this article. Accordingly, the information on this article is provided with the understanding that the author(s) and publisher(s) are not herein engaged in rendering professional advice or services. As such, it should not be used as a substitute for consultation with a competent adviser. Before making any decision or taking any action, the reader should always consult a professional adviser relating to the relevant article posting. Foresight Consulting as one of the best tax consultant in Jakarta region Indonesia will be happy to help find best solution to your tax issues.

 

Written by : Anita Rosaria Siahaan

Editor : HU / MS
Translation : RP / GP

 

Regulation of the Minister of Finance Number 34/PMK.010/2017 regarding The Collection of Article 22 Income Tax In Relation to Payments for the Delivery of Goods and Activities in the Importation Sector or Business Activities in Other Sectors
Regulation of the Minister of Finance Number 34/PMK.010/2017 regarding The Collection of Article 22 Income Tax In Relation to Payments for the Delivery of Goods and Activities in the Importation Sector or Business Activities in Other Sectors

No Comment

Leave a Reply

Your email address will not be published. Required fields are marked *