Advanced data analytics in tax administration The case of Independent Authority for Public Revenue (IAPR)

Giannouzi, Xanthippi; Γιαννούζη, Ξανθίππη

Προηγμένη ανάλυση δεδομένων στη φορολογική διοίκηση Η περίπτωση της Ανεξάρτητης Αρχής Δημοσίων Εσόδων (ΑΑΔΕ)

Doctoral Thesis

Author

Giannouzi, Xanthippi

Γιαννούζη, Ξανθίππη

Date

2025-07

Abstract

Τax administration has long been a critical domain of academic inquiry, as governments aim to maximize revenue collection to fund public policies while minimizing compliance burdens on taxpayers. In recent decades, the study of tax compliance has evolved beyond traditional economic models to encompass interdisciplinary approaches that integrate insights from sociology and psychology. A particularly promising development in this field is the application of advanced data analytics, especially when combined with primary behavioral research. Such integrative approaches offer powerful tools for identifying non-compliance risks and improving the strategic effectiveness of tax administration. This thesis explores how advanced data analytics can transform the operations of tax authorities by enhancing the understanding of taxpayer behavior, enabling more effective targeting, and aligning institutional practices with international standards. Using the case of the Independent Authority for Public Revenue (IAPR) in Greece as a focal point, the research provides findings and policy recommendations that are also applicable to other tax administrations globally. The central research question addressed is: How can advanced data analytics enhance the effectiveness and efficiency of tax administration by improving the understanding of taxpayer behavior and supporting the implementation of targeted compliance strategies? To answer this, the study adopts a triangulated methodology: ▪ Primary Research: A semi-structured survey (n = 175) targeting self-employed individuals and business owners, analyzed econometrically using Stata to assess behavioral drivers of compliance. ▪ Secondary Data Analysis: Administrative data from 2018–2023 were analyzed and visualized in Power BI to uncover regional disparities in audits, tax compliance, and revenue collection. ▪ Institutional Benchmarking: The IAPR’s operations were assessed against international frameworks (TADAT and OECD Maturity Models) to evaluate institutional readiness and identify areas for reform. This was complemented by an elite, unstructured interview with the Governor of IAPR. The research addresses three key dimensions: A. Institutional-Level Questions ▪ What is the current level of data analytics maturity within the Greek Tax Administration? ▪ How can interactive data tools and international benchmarking enhance strategic tax governance? B. Behavioral-Level Questions ▪ What are the primary factors influencing taxpayer compliance behavior? ▪ How do audit probability, penalty severity, and business suspension fect compliance intentions? ▪ How do psychological traits—such as need for closure, assessment, and locomotion—interact with deterrence mechanisms? C. Strategic Application & Policy Questions ▪ How can behavioral segmentation and data analytics be integrated into targeted compliance strategies? ▪What practical implications does this research offer for tax administration practitioners and policymakers? Key findings indicate that audit probability, penalty severity, and the threat of business suspension significantly increase taxpayers’ compliance intentions—especially when tailored to psychological profiles. Behavioral segmentation based on cognitive and motivational traits provides a more refined understanding of compliance behavior than traditional deterrence models alone. Visual analysis through Power BI highlights pronounced geographical and income-based asymmetries in audit activity, tax refunds, and revenue distribution. Moreover, international benchmarking reveals significant room for institutional improvement in areas such as risk analysis, data governance, and strategic targeting. Based on these insights, the thesis introduces the Compliance Intelligence Framework (CIF)—a scalable, reform-oriented model that integrates psychological profiling, geographic risk analysis, and differentiated policy levers into a unified approach to strategic compliance management. The CIF offers practical guidance on who to target, where to intervene, and how to respond—representing, to the author's knowledge, a novel contribution to the international literature. In conclusion, this thesis bridges theory and practice by combining social science insights with technological innovation. It contributes to both academic discourse and applied policymaking by presenting a concrete roadmap for advancing toward Tax Administration 3.0—a paradigm shift from data as a tool for enforcement to data as a foundation for prevention, intelligence, and evidence-based governance. The findings are intended to inform researchers, practitioners, and decision-makers seeking to modernize tax administration systems in a data-driven and behaviorally informed manner.

Department

Σχολή Οικονομικών, Επιχειρηματικών και Διεθνών Σπουδών. Τμήμα Οικονομικής Επιστήμης

Number of pages

298

Language

English

URI

https://dione.lib.unipi.gr/xmlui/handle/unipi/18073

Collections

Τμήμα Οικονομικής Επιστήμης

Show full item record

Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού 3.0 Ελλάδα