EMTA (Eesti Maksu- ja tolliametile) is the Estonian Tax and Customs Board. The Estonian Tax and Customs Board’s activities include administration of state revenues, implementation of national taxation and customs policies and protection of the society and legal and economic activities.
EMTA uses big data and data analytics technology for fraud detection and evaluation purposes. Through data analytics, they redefined their strategy towards the identification of cases to verify. They moved from an “unstructured approach” to this "case selection towards data-driven methods” based on an algorithm identifying risk coefficient for each case with the overall objective of increasing tax compliance and prevent frauds. For this purpose, EMTA analyses a large amount of structured data coming from government sources, mainly such as business registers and tax declarations.
EMTA has several systems harmonised in order to meet risk management expectations. The current study is focused on the system used by the tax audit unit (KOKE – Control Environment) that receives the list of high-risk cases automatically calculated by the VAT risk model.
How KOKE works
In Estonia, taxpayers submit tax declarations to have a tax return on the e-MTA portal. Every taxpayer has an ID card that automatically logs the citizens in the system for presenting the declaration. Each application is automatically evaluated by risk models in order to identify risks. The cases with higher risks are stored and elaborated in KOKE. Once the cases are collected in the system, auditors can review their tax declarations and integrate them with more information (there is a space on KOKE where the auditors can add information manually).
The process for the selection of the audit cases can be summarised in three main steps: the taxpayer submits VAT declaration and the appendix to the value-added tax declaration in the KMD form (value-added tax return) on the e-MTA portal; the automated risk models (the most important is the VAT risk model) calculate the high-risk cases and send them to the auditors (KOKE); KOKE receives the list of the cases with a higher risk of fraud, and the auditors can start to analyse them.
KOKE significantly supports EMTA in fraud detection. Since the system is very flexible, it is possible to assess different risk models easily and quickly, addressing different types of frauds. Currently, KOKE is mainly used for auditing and reused by other service departments, but create a new environment can provide to EMTA the possibility to increase the audience, including also additional services. The main idea is to have a more flexible framework that provides the possibility to change content and logics autonomously.
Moreover, KOKE has increased the efficiency of the organisation. Indeed, thanks to the use of analytics solutions for fraud detection, the Estonian Tax & Customs authority has reduced the number of people in the organisation substantially as they were able to optimise their working processes by involving innovative technology solutions.