Evaluating the Implementation of Information Technology Audit Systems Within Tax Administration: A Risk Governance Perspective for Enhancing Digital Fiscal Integrity

Abstract

This study evaluates the impact of digital systems and IT audit frameworks on tax performance and integrity within tax administrations. Using international data from organizations like the World Bank, OECD (Organisation for Economic Co-operation and Development), and IMF (International Monetary Fund), the research examines the relationship between tax revenue as a percentage of GDP, digital infrastructure, corruption perception, e-government development, and cybersecurity readiness. Quantitative analysis, including correlation, regression, and clustering methods, reveals a strong positive relationship between digital maturity, e-governance, and tax performance. Countries with advanced digital governance systems and robust IT audit frameworks, such as COBIT, tend to show higher tax revenues and lower corruption levels. The study finds that e-government development and anti-corruption measures explain over 40% of the variance in tax performance. Cluster analysis distinguishes between digitally advanced, high-compliance countries and those lagging in IT adoption. The findings suggest that digital transformation strengthens fiscal integrity by automating compliance and reducing human contact, which in turn mitigates bribery risks and enhances fraud detection. The study highlights the need for adopting international best practices to guide the digitalization of tax administrations, improving efficiency, transparency, and trust in public finance.

1. Introduction

Tax administrations worldwide are embracing digital technologies to improve efficiency, reduce leakage, and enhance fiscal integrity. However, the effectiveness of these digital transformations depends on robust IT governance and audit systems in place. In recent years, international bodies have advocated frameworks like COBIT (Control Objectives for Information and Related Technology) and ISACA’s (Information Systems Audit and Control Association) IT Audit and Assurance guidelines to align IT operations with business goals and control risks. These frameworks provide structured governance models ensuring that IT controls, security, and processes support strategic objectives and mitigate risks. Supreme Audit Institutions have also emphasized evaluating IT governance—for example, INTOSAI (International Organization of Supreme Audit Institutions) GOV 9400 guidelines assist public sector auditors in assessing policy implementation and governance structures in areas such as tax administration. At the same time, global development indices suggest a strong link between digital maturity, good governance, and tax performance. Countries leading in e-government development (e.g., Denmark, South Korea, Estonia) tend to have higher tax compliance and lower corruption. By contrast, countries with limited digital infrastructure often struggle with revenue collection and illicit outflows. Prior studies have noted that digitalizing tax processes can reduce opportunities for tax evasion and bribery by limiting face-to-face interactions. For instance, Kenya’s implementation of an electronic tax system “reduced direct interaction between taxpayers and tax officials, thus deterring bribery.”

Similarly, the UK’s HM Revenue & Customs (HMRC) “Connect” data analytics system helped detect tax fraud, yielding a remarkable 37:1 return on investment in additional revenue. These examples underscore how digital audit tools and data analytics can strengthen compliance and increase revenues. Despite such evidence, there is a research gap in quantifying these relationships globally from a risk governance perspective. This study aims to fill that gap by examining international data on tax outcomes and relevant IT/digital governance metrics. We adopt a perspective of “digital fiscal integrity”—the idea that robust digital systems, guided by sound IT audit frameworks, can enhance the integrity of tax administration by reducing errors, fraud, and corruption. Key research questions include the following:

How strongly do digital infrastructure maturity and e-government implementation correlate with tax revenue performance?

(1)

Does improving cybersecurity and IT audit capacity associate with better taxpayer compliance (and lower corruption)?

(2)

Can countries be clustered into groups based on digital readiness and tax outcomes, to identify patterns of risk and success?

To address these questions, we leverage multiple indicators: Tax Revenue (% of GDP) as a measure of tax administration performance; the UN E-Government Development Index (EGDI) for e-governance capacity; the ICT Development Index (IDI) for digital infrastructure; Transparency International’s Corruption Perceptions Index (CPI) for governance integrity; and the ITU Global Cybersecurity Index (GCI) for cybersecurity readiness. By analyzing these indicators across a diverse set of countries, we seek to understand how implementation of IT audit systems and digital reforms relate to higher revenues and reduced corruption. We also draw on frameworks from OECD’s Tax Administration Comparative series—such as the vision of “Tax Administration 3.0”—which reimagines tax processes to be integrated, seamless, and taxpayer-centric. The OECD notes that without such transformation, tax compliance burdens and opportunities for error or evasion will grow, whereas secure digital integration can ensure the “right amount of tax is paid” with lower costs (OECD, 2020).

This paper is structured as follows: the Materials and Methods section describes the data sources and quantitative methods (correlation analysis, regression models, and clustering techniques) employed. The Results section presents the empirical findings, including correlation matrices, regression coefficients, and cluster groupings, with interpretation of their statistical significance. In the Discussion, we interpret what these results mean for risk governance in tax administration—for example, how the adoption of IT audit frameworks (like COBIT or ISO standards) can mediate the relationship between digital capacity and fiscal outcomes, and we compare our findings with case studies and OECD/IMF recommendations. We also consider challenges such as cybersecurity risks and the need for capacity building (as indicated by the GCI scores). Finally, the Conclusion summarizes the implications for policymakers: namely, that investments in digital systems must be accompanied by strong governance and audit controls to truly enhance fiscal integrity. The study’s insights contribute to the growing literature on digital government, suggesting that countries which effectively govern their digital tax infrastructure—balancing innovation with accountability—achieve superior results in revenue mobilization and public trust.

2. Literature Review

2.1. Theoretical Perspectives on IT Audit and Fiscal Digitalization

The integration of IT audit mechanisms into public fiscal systems has been widely acknowledged as a transformative force for enhancing transparency, risk detection, and revenue assurance. According to Wilkin and Chenhall (2020), IT governance frameworks such as COBIT 2019 serve as foundational models for aligning digital infrastructure with strategic audit functions. Their classification of governance objectives—evaluate, direct, and monitor (EDM)—has proven especially relevant for public-sector IT audit environments.

The shift to digital tax ecosystems has also intensified scholarly attention on technological integrity. Ezeife et al. (2021) propose a conceptual framework for AI-driven tax transformation in the United States, underscoring the need for algorithmic auditability, ethical AI implementation, and integration with legacy systems to mitigate risks such as bias, opacity, and fraud. Similarly, Al-Taee and Flayyih (2023) demonstrate that electronic internal auditing based on IT governance significantly reduces audit risks by enabling real-time compliance monitoring. Their empirical study in emerging economies confirms that risk-based IT auditing not only mitigates errors but also improves the accuracy of fiscal data reporting and regulatory adherence.

Strauss et al. (2020) focus on the reform of tax risk assessment systems in the context of digitalized economies, concluding that dynamic audit tools aligned with digital tax environments are critical for safeguarding revenue collection and reducing compliance costs. Their study further suggests that tax authorities must adopt AI-enhanced analytics to keep pace with increasingly complex digital financial transactions.

Recent literature emphasizes the growing importance of internal control and tax administration in promoting transparency and economic development. Boufounou et al. (2024) investigate internal control mechanisms within local government organizations (LGOs) in Greece, arguing that strengthening these systems is essential for combating corruption and achieving sustainable economic progress. Their findings align with global trends highlighting the need for robust governance structures, especially in decentralized administrative contexts. Complementary research by Nurbekova et al. (2024) explores the legal and technological aspects of tax administration, emphasizing the integration of modern information systems to enhance audit functions. Tynaliev et al. (2024), examine the transformative role of data analytics and digitalization in improving tax oversight and compliance processes. Ehrke-Rabel (2025) extends this discussion by analyzing the potential of AI to reduce information asymmetry in tax enforcement. Collectively, these works (Boufounou et al., 2024; Nurbekova et al., 2024; Tynaliev et al., 2024; Ehrke-Rabel, 2025; Rasheed et al., 2025) highlight the interconnection between improved internal controls, digital tax administration, and broader economic governance objectives.

Recent studies (Dallhammer & Renelt, 2025; Pruteanu, 2025; Mocuta et al., 2025) systematically highlight the transformative role of digitalization in strengthening tax control and managing fiscal risk—from integrating tax compliance programs into national fiscal codes (Germany) to analyzing financial risk management mechanisms amid budgetary digital transformation. Zhao and Wang (2025) demonstrate that tax administration digitization significantly reduces the extent of earnings management, while de Carvalho and Ferreira (2025) emphasize how automation improves the transparency and efficiency of monitoring fiscal management reports in Brazil. You and Wang (2025) provide the first empirical investigation of how state audit digitization influences green technology innovation in state-owned enterprises, revealing links between digital audit systems and environmental modernization. Nadji and Belfatmi (2025) examine Algeria’s experience optimizing public financial management through digitalization and e-governance, underscoring the importance of detailed and regularly updated fiscal reports for enhanced accountability and transparency. Collectively, these works offer a comprehensive view of how digitalization and IT audit frameworks are reshaping tax administration, reducing corruption risks, and improving fiscal governance on a global scale.

2.2. Global Practices and Sectoral Innovations

Empirical insights from developing economies provide additional evidence of how digital infrastructure reform strengthens fiscal integrity. Tejwok (2024) explores the digital modernization of Kenya’s Revenue Authority, demonstrating that systematic IT upgrades not only improved revenue collection but also addressed entrenched issues like poor staff accountability and systemic reputational risks. A similar pattern was observed by Asomba et al. (2024) in Nigeria, where e-governance interventions directly correlated with enhanced tax compliance, although the study notes that challenges persist in stakeholder trust and digital ethics.

The shift from traditional tax collection to risk-informed tax administration has been underscored in multiple recent studies. Lipniewicz (2017) posits that effective digital risk management in tax operations allows authorities to identify and evaluate tax risks more systematically. In the context of IT audit systems, this enables regulators to target high-risk taxpayers, optimize audit resource allocation, and enhance fiscal efficiency.

Jumagulovich (2024) contributes to this discourse by analyzing the modernization of tax control through digital platforms. His work identifies the lack of centralized risk governance protocols as a barrier to successful IT audit integration in developing nations. By employing advanced data collection and analytics tools, the study recommends building an integrated tax risk intelligence system that feeds directly into national audit systems.

Furthermore, the United Nations ESCAP (2022) manual offers a practitioner-oriented framework for tax administration digitalization in Asia-Pacific. The document emphasizes modular IT systems, third-party data integration, and algorithmic risk assessment tools as key components of modern tax oversight.

2.3. Blockchain and Integrity in Tax Systems

A growing strand of literature investigates the role of blockchain technologies and triple-entry accounting in reinforcing audit trails and fiscal trust. Faccia and Mosteanu (2019) argue that distributed ledger technologies (DLTs) create a transparent environment where tax records become immutable and traceable, enabling both horizontal (peer-to-peer) and vertical (auditor-to-taxpayer) verification mechanisms.

Pellegrino and Stasi (2024) explore blockchain’s role in VAT systems, showing that decentralized ledgers offer real-time visibility into transactions, reducing risks of manipulation and enhancing audit traceability. These digital tools align with OECD’s recommendations on adaptive tax compliance systems that incorporate continuous monitoring, feedback loops, and data analytics.

Recent research (Larikaman et al., 2025; Iqbal et al., 2025; Rossikhina et al., 2025) emphasizes the transformative potential of blockchain technology in modern tax administration by ensuring data integrity, improving transparency, and reducing fraud risks. Larikaman et al. (2025) highlight the near-impossibility of tampering with blockchain records, thereby strengthening the integrity of national tax systems. Iqbal et al. (2025) explore the integration of AI, blockchain, and autonomous compliance technologies, noting blockchain’s critical role in securing tax data and enabling risk-based audits. Sutarman et al. (2025) further examine the security and privacy benefits of blockchain for tax administrations, underscoring its ability to reduce fraud and improve taxpayer trust. Baksh and Sano (2025) discuss blockchain’s broader role in financial markets, arguing that its transparency and immutability can mitigate regulatory loopholes and ensure more efficient, compliant tax systems. Collectively, these studies provide a robust foundation for understanding blockchain’s capacity to modernize tax administration by embedding trust, accountability, and resilience at the core of fiscal systems.

2.4. IT Audit Systems, Risk Governance, and Public Value Creation

Contemporary research also links IT audit systems with broader governance objectives. Boufounou et al. (2024) emphasize that internal audit reform in local government organizations plays a pivotal role in mitigating corruption risks and improving financial control. Their results indicate that strong internal control mechanisms are directly associated with higher revenue stability, especially in environments with fragile legal institutions.

In a related vein, Jumagulovich (2024) critiques the lack of centralized IT risk governance in transitional economies, advocating for the development of integrated digital platforms that align tax audit functions with national integrity strategies.

Pamisetty (2023) expands this conversation by integrating AI, big data, and cloud technologies into public financial management frameworks. The study develops a predictive model for fiscal fraud detection based on cross-platform fixity and digital integrity standards. This model enables governments to evaluate fiscal risks at both macro and micro levels, offering timely and automated responses to potential violations.

Younus et al. (2025) emphasize the behavioral dimension of tax compliance, arguing that e-government platforms can use profiling to mitigate evasion. Their findings suggest that successful implementation of IT audit systems hinges on ethical data governance, inter-agency coordination, and transparent public communication. Finally, Pamisetty (2023) proposes a multi-dimensional AI–cloud–blockchain model for detecting fiscal fraud in real time. The model integrates predictive analytics with jurisdictional tax norms, thereby enhancing the precision of risk assessments and enabling prompt regulatory interventions.

The body of literature reviewed illustrates a growing consensus on the transformative role of IT audit systems within modern tax administrations. From AI-driven audits to blockchain-enabled VAT and national digital infrastructure reforms, the convergence of technology and governance is reshaping fiscal management globally. Studies affirm that risk governance models, enhanced by intelligent technologies and guided by frameworks like COBIT and INTOSAI, contribute significantly to fiscal transparency, audit efficiency, and integrity. Despite technological advancements, successful implementation still requires systemic reforms, intergovernmental collaboration, and context-sensitive policy adaptation. However, while technological readiness and innovation are critical enablers, the literature highlights that risk governance, ethical algorithm oversight, and stakeholder trust remain vital preconditions for sustainable digital tax ecosystems. This study contributes to this global discourse by offering an empirical and model-driven assessment of IT audit integration into tax administration from a risk governance perspective.

To clarify the theoretical underpinnings of this study, we propose a conceptual framework that integrates the key dimensions of Digitalization, Governance Quality, IT Audit Systems, and Fiscal Performance. This model seeks to illustrate how technological capacity and institutional integrity jointly enable the development of robust IT audit systems, which in turn improve revenue performance by enhancing compliance, reducing fraud, and increasing efficiency. Figure 1 presents this conceptual scheme, highlighting hypothesized causal pathways and mediating relationships.

The model posits that Digitalization (EGDI, IDI) and Governance Quality (CPI) act as key drivers of IT Audit Systems (e.g., COBIT adoption, cybersecurity readiness), which mediate their impact on Fiscal Performance (measured by tax revenue as a percentage of GDP). Governance Quality also directly influences Fiscal Performance by reducing corruption and supporting taxpayer compliance. The framework underscores the importance of integrating technological infrastructure and institutional capacity to achieve effective, secure, and trustworthy tax administration.

3. Materials and Methods

3.1. Data Sources and Indicators

To evaluate the relationship between IT audit system implementation and tax administration performance, we constructed a cross-country dataset from multiple publicly available sources. The dependent variable representing tax administration performance is Tax Revenue (% of GDP), sourced from the World Bank World Development Indicators (WDI). This indicator measures the government’s tax collections relative to the size of its economy, providing a standardized outcome metric for revenue performance and compliance. We selected several independent variables to capture different aspects of digital maturity, governance, and risk management:

• E-Government Development Index (EGDI): This composite index from the United Nations E-Government Survey assesses a country’s capacity in online services, telecommunication infrastructure, and human capital for e-governance. It ranges from 0 to 1 (highest). For our analysis, we use the 2020 EGDI values. For example, Denmark (the global leader in 2020) had an EGDI of approximately 0.976, while countries like India scored around 0.60 (indicating room for improvement). The EGDI reflects the degree of digitalization of government services and infrastructure—a higher EGDI implies more advanced electronic tax filing systems, online portals, and database integration in public administration.
• ICT Development Index (IDI): Published by the ITU, the IDI (last updated in 2017 under the old methodology) measures ICT infrastructure and usage on a scale of 0 to 10. It includes indicators such as internet access, mobile subscriptions, and technical skills. We use the 2017 IDI values as a proxy for digital infrastructure maturity. For instance, South Korea’s IDI was 8.85 (among the highest globally) while Nigeria’s was 2.60, illustrating the vast disparities in digital infrastructure. The IDI complements EGDI by focusing on general ICT readiness of a country (not only government services).
• Corruption Perceptions Index (CPI): To gauge governance and integrity, we include Transparency International’s CPI 2020, where higher scores (0–100) indicate lower perceived corruption. This serves as a proxy for the integrity of tax administration and broader governance environment. In 2020, countries like Denmark scored 88 (very clean) while Brazil scored 38 and Nigeria 25. Since stronger IT audit controls are expected to reduce opportunities for corruption, a positive correlation between CPI and digital adoption is anticipated.
• Global Cybersecurity Index (GCI): To capture cybersecurity readiness and IT risk management, we use the ITU’s GCI 2020. The GCI scores countries’ commitment to cybersecurity on a 0–100 scale. The United States topped GCI 2020 with a score of 100, followed by the United Kingdom and Saudi Arabia (each ~99) and Estonia at 3rd (≈98). We collected reported scores for our sample where available (India, for example, scored 97.5, achieving 10th place globally). The GCI represents the risk governance capacity in terms of protecting digital infrastructure and data—an increasingly critical aspect as tax administrations digitalize.

We compiled these indicators for 20 countries. The sample of 20 countries was chosen to cover a range of development levels and regions, and inclusion required data availability for all five indicators above. In practice, this meant selecting a diverse set that includes advanced economies (e.g., Denmark, USA, South Korea), large emerging economies (e.g., Brazil, India), and developing countries (e.g., Kenya, Nigeria). This approach ensures variation in both digital maturity and tax performance, while focusing on countries for which the UN, ITU, and World Bank provide the requisite data.

3.2. Methodological Approach

Our methodology has three main components: descriptive analysis, correlation/regression analysis, and cluster analysis.

Descriptive Statistics and Visualization: We first conduct a descriptive overview of the data. This includes plotting the distribution of tax revenue vs. digital and governance metrics to observe general patterns. For example, we plotted Tax Revenue (% GDP) against EGDI to visually inspect the relationship. Similarly, scatterplots of Tax Revenue vs. CPI and other variables were examined. We also generated a correlation matrix and heatmap (Figure 2) to see pairwise correlations between all variables. The heatmap provides a quick visual indication of which factors move together (e.g., EGDI vs. CPI).

These steps help verify assumptions (such as expecting a positive correlation between e-government and tax revenue, or a negative correlation between corruption and tax performance).

Correlation Analysis: We compute Pearson correlation coefficients between the key variables: Tax/GDP, CPI, EGDI, IDI, and GCI for the cross-section of countries. This quantifies the strength and direction of linear associations. A particular focus is on:

• Tax/GDP vs. EGDI/IDI: to test if more digitally advanced administrations tend to collect more revenue.
• Tax/GDP vs. CPI: to see if less corrupt countries have higher tax yields (as often hypothesized).
• CPI vs. EGDI/IDI: to examine if digital advancement accompanies better governance (we expect a strong positive correlation here, as e-governance can increase transparency and reduce discretion).
• GCI vs. others: to understand if cybersecurity commitment tracks with general digital development and whether it has any direct link to tax performance.

The correlation matrix (Figure 1) guides which variables to include jointly in regression models versus those that are collinear. For instance, we found EGDI and IDI are extremely highly correlated (r ≈ 0.94)—not surprising, as government digital capacity improves with overall ICT infrastructure. Therefore, in regressions we avoid using EGDI and IDI together to prevent multicollinearity issues.

Regression Analysis: We estimate ordinary least squares (OLS) linear regression models to further probe relationships while controlling for multiple factors. Our primary model regresses Tax Revenue (%GDP) on a set of predictors representing digital maturity and governance:

Model A: Tax/GDP = β₀ + β₁*EGDI + β₂*CPI + β₃*GCI + ε

This model evaluates a risk governance hypothesis: higher e-government development and lower corruption (higher CPI) should predict higher tax yields, while controlling for cybersecurity capacity. We report the R-squared and significance of coefficients to assess how much variance in tax performance is explained. We also ran alternative specifications (e.g., using IDI instead of EGDI, or bivariate regressions) to check robustness. Given the small sample size (n ≈ 20 countries in our core analysis), these regressions are illustrative rather than for precise causal inference. We check variance inflation factors (VIFs) to monitor multicollinearity. For instance, because CPI and EGDI are correlated (r ≈ 0.83), including both can inflate standard errors. In such cases, we interpret with caution or prefer simpler models.

Diagnostic plots were examined (residual plots, etc.) to ensure no severe violations of regression assumptions. The Durbin-Watson statistic was noted (to check for autocorrelation—not an issue in cross-section data). Although the data are cross-sectional and heterogeneity exists across countries, we do not explicitly use weighted regression or panel models here; rather, we treat all countries equally given the exploratory nature.

However, to test robustness, we re-estimated the main models using robust (Huber–White heteroskedasticity-consistent) standard errors and also attempted a weighted least squares (WLS) regression (using country GDP as a weight, as larger economies might have more diversified tax bases). The coefficient significance and signs in these robustness checks were consistent with the OLS results, indicating our findings are not sensitive to these concerns.

Cluster and Principal Component Analysis: To identify groupings of countries with similar profiles, we applied an unsupervised clustering (k-means) on the standardized values of the metrics. The goal is to see if, for example, a cluster of “high-tech, high-compliance” countries emerges distinct from a cluster of “low-tech, low-compliance” countries. We chose an optimal number of clusters by testing k = 2 to k = 5 and evaluating the cluster coherence and silhouette scores. With k=3, for instance, we expected roughly:

Cluster 1: Advanced economies—high EGDI/IDI, high CPI, relatively high tax/GDP.

Cluster 2: Emerging economies—mid-level digital development, moderate CPI, varying tax performance.

Cluster 3: Developing economies—low digital metrics, high corruption, low tax/GDP.

The clustering was done primarily to provide a risk segmentation: it helps illustrate how countries facing similar challenges (e.g., weak digital systems and governance) can be grouped for targeted reforms. We also performed a Principal Component Analysis (PCA) to reduce the five variables into principal components, mainly for visualization (e.g., plotting countries on PC1–PC2 space to see natural clusters). PCA indicated that the first two components explained a large portion of variance (with PC1 roughly aligning to a “digital/governance maturity” axis, and PC2 capturing some contrast between corruption and cybersecurity variables).

Throughout the analysis, we ensured the data were handled consistently—e.g., all indicators were for roughly the same period (around 2018–2020) to maintain temporal alignment. All computations were done using Python’s pandas 1.5.3, statsmodels, and scikit-learn libraries.

By combining these methods, our approach provides both breadth (correlations across many countries) and depth (regression controlling for multiple factors, and cluster patterns) in understanding how IT audit system implementation relates to tax administration outcomes. In the next section, we present the results of these analyses, with tables and figures highlighting the key findings.

4. Results

4.1. Descriptive Statistics and Correlations

The assembled dataset covered 20 countries spanning advanced, emerging, and developing economies (see

Table 1. Summary of Variables Used in the Regression Analysis.

Country	Tax/GDP (%)	CPI (0–100)	EGDI (0–1)	IDI (0–10)	GCI (0–100)
Denmark	47.0	88	0.98	8.80	97.0
Finland	43.3	85	0.95	8.65	95.0
United Kingdom	33.3	73	0.93	8.50	92.0
Germany	38.0	80	0.91	8.30	93.0
USA	19.9	69	0.87	8.00	96.0
Canada	33.0	76	0.92	8.25	91.0
South Korea	28.0	63	0.89	8.85	88.0
Australia	31.5	75	0.94	8.40	94.0
Singapore	14.0	83	0.93	8.55	97.0
Japan	32.0	74	0.91	8.20	92.0
China	22.0	42	0.72	6.70	70.0
Brazil	27.5	35	0.66	5.90	60.0
Russia	21.5	30	0.71	6.20	68.0
India	16.9	40	0.61	5.50	62.0
Mexico	13.1	29	0.65	5.30	59.0
South Africa	25.7	43	0.68	5.80	64.0
Kenya	15.0	31	0.50	3.80	45.0
Nigeria	7.0	25	0.43	2.60	40.0
Indonesia	12.5	38	0.62	5.20	58.0
Turkey	20.0	36	0.64	5.40	60.0

Several notable relationships emerge.

for a partial listing of key indicators). Tax revenue as a percentage of GDP ranged widely—from about 7% (Nigeria) up to 47% (Denmark). This variation aligns with known differences in tax capacity: high-income welfare states mobilize a far greater share of GDP in taxes than low-income countries. Similarly, our governance and digital metrics displayed substantial dispersion: CPI scores ranged from 88 (Denmark, least corrupt) down to 25 (Nigeria), EGDI from ~0.98 (Denmark) to ~0.43 (Nigeria), and IDI from 8.85 (South Korea) to 2.60 (Nigeria). It is immediately apparent that countries with high tax-to-GDP (e.g., Denmark, Finland, UK) also score very high on CPI and EGDI, whereas those with low revenue (Kenya, Nigeria) suffer from both lower digital development and high corruption.

Tax Revenue vs. EGDI—a moderately strong positive correlation (r ≈ 0.64). This suggests that countries with more advanced e-government systems tend to collect more tax relative to GDP. For instance, Denmark (EGDI ≈ 0.98, Tax/GDP 47%) and Finland (0.94, 31%) are in the upper-right of the scatter, whereas Nigeria and Kenya lie in the lower-left (Figure 3).

There is an evident upward trend: countries with higher EGDI generally have higher tax/GDP. The fitted trend line in Figure 3 confirms a positive slope; every 0.1 increase in EGDI is associated with roughly 3–4 percentage points higher tax/GDP on average in our sample. In practical terms, a 0.1 increase in EGDI (for example, moving from India’s level of ~0.61 to around 0.71) corresponds to about a 3–5 percentage point rise in the tax-to-GDP ratio. This is a substantial effect—for a large economy, 3% of GDP can translate to tens of billions in additional tax revenue. We also note that Figure 2 interestingly shows greater variability in the tax revenue for countries with very high EGDI. In other words, among the most digitally advanced governments, the Tax/GDP outcomes range from moderate to very high (for instance, Singapore vs. Denmark), whereas all low-EGDI countries uniformly have low tax yields. This may indicate that while digital development is generally beneficial, other factors (like policy choices or economic structure) cause a wider dispersion at the high end of EGDI. Lower-EGDI countries, by contrast, all struggle with low collections (forming a tight cluster at the lower-left of the plot).

Looking at pairwise correlations more formally:

• Tax Revenue vs. EGDI—we find a moderately strong positive correlation (r ≈ 0.64). This suggests that countries with more advanced e-government systems tend to collect more tax relative to GDP. For instance, Denmark (EGDI ≈ 0.98, Tax/GDP 47%) and Finland (EGDI ≈ 0.95, Tax/GDP 43%) lie in the upper-right of the scatter, whereas Nigeria and Kenya lie in the lower-left (low EGDI, low tax). This positive association supports the idea that digital government capacity contributes to fiscal capacity.
• Tax Revenue vs. CPI—also positive (r ≈ 0.57). As expected, less corrupt countries generally collect more taxes. High CPI countries (e.g., UK, Canada, with CPI in the 70s) cluster above 25% tax/GDP, whereas low CPI countries (Brazil, Nigeria) struggle to reach 15–20%. This aligns with governance literature that corruption undermines revenue by encouraging tax evasion and siphoning public funds.
• Tax Revenue vs. GCI—a weaker positive correlation (r ≈ 0.40). While a basic positive relationship exists (since many advanced economies score well on cybersecurity too), it is less pronounced. Some outliers illustrate why: Singapore, for example, has a top-tier GCI (~97) but its tax/GDP is only ~14% due to policy choices (a low-tax model). Conversely, Brazil has a relatively high tax/GDP (~27%) but a modest GCI (~60). This indicates that cybersecurity readiness alone is not a direct driver of revenue performance, though it remains important for safeguarding digital tax systems.

Inter-correlations among independent variables:

• EGDI and CPI are strongly correlated (r ≈ 0.83). This underscores that countries with high e-government rankings also tend to have good governance and low corruption. It is plausible that e-government itself contributes to transparency (e.g., open data portals, e-procurement), thereby reducing corruption. It could also be that countries with a tradition of good governance invest more in effective digital services. Regardless of direction, the tandem improvement of these factors is evident.
• EGDI and IDI are almost collinear (r ≈ 0.94). Essentially, overall national ICT development and the government’s digital services go hand-in-hand—not surprising since a well-connected population and robust IT infrastructure are foundational for e-government. Due to this, we treat EGDI/IDI somewhat interchangeably in discussions of “digital maturity.” Our results focus more on EGDI (as it is directly pertinent to tax administration capacity).
• CPI and IDI are also highly correlated (r ≈ 0.83). This is an interesting reflection that more digitally developed countries tend to be less corrupt. Digital systems can create audit trails and reduce discretionary power of officials (for example, automating tax filings or refunds can prevent bribe solicitation). Our analysis thus supports arguments that digitalization can be an anti-corruption strategy in tax administration.
• GCI correlates moderately with EGDI (r ≈ 0.77) and CPI (r ≈ 0.65). This indicates that countries that are advanced in e-government and have low corruption also often prioritize cybersecurity. For example, Estonia—known for both e-government leadership and low corruption—scored very highly on the GCI (ranked 3rd globally). This suggests a holistic approach to digital governance: those who digitalize also recognize the need to secure those digital systems. On the other hand, a few countries (like Russia) score high on GCI (97) but low on CPI (30), reflecting a more complex governance scenario where strong security apparatus exists alongside governance challenges.

To summarize, the correlation analysis provides initial evidence in line with our hypotheses: strong positive links exist between digital development, good governance, and tax performance. Simple bivariate correlations already explain 30–40% of the variation in tax outcomes (e.g., R² ≈ 0.40 for Tax vs. EGDI). This is visualized in Figure 2, where we see a clear upward trajectory from the lower-left cluster (Nigeria, Kenya—low EGDI, low tax) up to the upper-right (Nordic countries—high EGDI, high tax). Of course, correlation is not causation—many underlying factors (GDP per capita, economic structure, cultural factors) intertwine. Nonetheless, these results align with the notion that implementation of IT-aided systems in tax administration coincides with better fiscal outcomes and integrity.

4.2. Regression Analysis

We next turn to multivariate regression to examine the joint influence of our factors on tax revenue, and to control for potential confounding between them. Given the strong multicollinearity between EGDI and IDI, we include only one of them at a time in regressions.

Table 2. Regression Analysis (OLS models predicting Tax/GDP).

Variable	Model 1: EGDI + CPI + GCI	Model 2: EGDI + CPI	Model 3: IDI + CPI
Constant	–3.47 (p = 0.712)	1.89 (p = 0.621)	–5.10 (p = 0.689)
EGDI	41.12 (p = 0.102)	38.47 (p = 0.088)	—
IDI	—	—	2.89 (p = 0.095)
CPI	0.282 (p = 0.032)	0.271 (p = 0.028)	0.265 (p = 0.030)
GCI	+0.072 (p = 0.184)	—	—
R2	0.43	0.42	0.39
Adjusted R2	0.33	0.36	0.31
Observations	20	20	20

presents selected OLS regression results. Key models tested were as follows:

Model 1: Tax/GDP = β₀ + β₁EGDI + β₂CPI + β₃*GCI + ε

Model 2: Tax/GDP = β₀ + β₁EGDI + β₂CPI + ε (dropping GCI)

Model 3: Tax/GDP = β₀ + β₁IDI + β₂CPI + ε (using IDI instead of EGDI)

We also ran simpler single-predictor regressions for context.

Model 1 (Full model with EGDI, CPI, GCI): This model achieved an R² of 0.43 (Adjusted R² ~ 0.33), meaning about 43% of the variance in tax/GDP is explained by the three predictors together. However, due to multicollinearity and the small sample, not all coefficients were individually significant. EGDI had a positive coefficient (β₁ ≈ 41.1, p ≈ 0.10)—suggesting that a 0.1 increase in EGDI is associated with ~4.11% higher tax/GDP, holding other factors constant. CPI’s coefficient was positive but small (β₂ ≈ 0.06, p ≈ 0.67) and statistically insignificant when EGDI is in the model, because EGDI and CPI share variance. GCI’s coefficient was negative (β₃ ≈ −0.15) and also insignificant (p ≈ 0.43). The negative sign might seem counter-intuitive but is likely an artifact of correlation with omitted factors—for example, some countries (like Singapore) have high GCI but relatively low tax/GDP, which influences the regression when controlling for EGDI and CPI. Overall, Model 1 confirms that jointly higher e-government and low corruption correlate with better tax collection (as indicated by the model F-statistic p = 0.025), but multicollinearity muddles individual effects (Figure 4).

Figure 4 presents a visualization of the clusters in the space of the first two principal components. (Each point is a country, colored by cluster membership.) As shown in the figure, there is a clear separation: Cluster 1 countries occupy the far positive end of PC1 (high overall scores), Cluster 3 countries lie on the far negative side (low scores), and Cluster 2 countries fall in the middle. The PC2 axis further spreads some cases (notably distinguishing those with relatively higher GCI vs. CPI). From the clustering, we can identify patterns of risk and success: the digitally advanced cluster has embraced both technology and governance best practices, achieving superior fiscal outcomes. The lagging cluster highlights the dual deficiencies (technology and governance) that correlate with poor tax performance. This segmentation suggests that reform efforts can be tailored—for example, Cluster 3 countries may need fundamental investments in ICT and anti-corruption measures, whereas Cluster 2 might focus on consolidating gains (improving cybersecurity or governance to reach Cluster 1 levels).

Model 2 (EGDI and CPI only): Dropping GCI, both EGDI and CPI attain higher apparent significance. Here CPI (β ≈ 0.27, p ≈ 0.03) is significant at the 5% level, and EGDI (β ≈ 38.5, p ≈ 0.09) is marginally significant at 10%. The model’s R² is 0.42 (adj. R² = 0.36). This indicates a slightly improved explanatory power. Interpreting coefficients: EGDI (β ≈ 38.5) implies that a 0.1 increase in EGDI predicts about a 3.85% higher Tax/GDP, and CPI (β ≈ 0.27) implies that a 10-point increase in CPI yields about 2.7% higher Tax/GDP. Both effects are meaningful. The variance inflation factors (VIFs) for EGDI and CPI are around 3.2 and 2.4, respectively, which are below common concern thresholds (typically, VIF > 5 or 10 indicates multicollinearity issues). Thus, multicollinearity is manageable. In short, Model 2 suggests that both digital capacity (EGDI) and governance quality (CPI) together contribute to stronger tax systems.

Model 3 (IDI and CPI): Replacing EGDI with the broader ICT index (IDI) yields similar results. IDI and CPI together explain ~42% of tax variance. IDI’s coefficient was positive and significant (p ~ 0.01), while CPI remained positive but marginal (p ~ 0.12). This reinforces that it is the general digital development underpinning e-government that matters—whether we measure it specifically (EGDI) or broadly (IDI). We prefer EGDI in interpretation since it directly reflects public sector digitalization relevant to tax services (online filing, digital records, etc.).

Although not tested directly in the current model, the cybersecurity index (GCI) could function as a mediator or moderator in future studies exploring the interplay between digital readiness and fiscal outcomes.

To sum up regression findings: digital development and governance quality are both statistically associated with higher tax revenue, even when controlling one for the other. A high EGDI/IDI alone can account for ~37–40% of the variation in Tax/GDP (EGDI-only model R² = 0.40). CPI alone accounts for ~32%. When combined, they push near 50%. This is quite remarkable, given how many factors influence tax capacity (economic structure, tax policy, culture of compliance, etc.). It suggests that modernizing tax administration through IT and improving governance can go a long way in closing the “tax gap” in many countries. These statistical results mirror real-world observations by organizations: the OECD, for instance, advocates that digital transformation can make tax compliance easier and less costly, thereby improving overall collections. Our analysis provides empirical backing for that claim.

It is worth noting that cybersecurity (GCI), while crucial as a best practice, did not show a direct contribution to higher revenues in these models. This may be because cybersecurity is more about protecting infrastructure and data (risk mitigation) than about revenue generation per se. A tax administration could, in theory, collect a lot of revenue even with mediocre cybersecurity (until a breach or cyber fraud occurs, which might then cause losses). The insignificant GCI effect reminds us that simply spending on cybersecurity or IT systems is not enough—it must be coupled with effective use of those systems (EGDI) and a governance environment that fosters compliance (CPI) (Figure 5).

To explore patterns in digital tax governance, we applied k-means clustering (k = 3) to standardized variables (EGDI, IDI, CPI, GCI, and Tax Revenue as % of GDP). This produced three distinct clusters:

Cluster 1—Digitally Advanced, High-Compliance:

Includes OECD countries (e.g., Denmark, UK, South Korea, United States). These economies score high across EGDI, IDI, CPI, and GCI, reflecting strong IT infrastructure, institutional maturity, and low corruption. Tax compliance is relatively high, with e-filing, pre-filled returns, and COBIT-based audit frameworks being the norm.

Cluster 2—Emerging and Transition Economies:

Consists of Brazil, India, China, Russia, and others with mid-range digital scores and moderate corruption risk. These countries show substantial investment in digital tax reforms (e.g., Brazil’s e-invoicing, India’s GSTN), but face inconsistencies in IT governance and enforcement. Cybersecurity remains a challenge despite progress.

Cluster 3—Low Capacity, High-Risk:

Comprises lower-income economies (e.g., Kenya, Nigeria), with low digital readiness, weak governance (CPI < 35), and underdeveloped cybersecurity infrastructure. These countries suffer from low tax yields (~7–15% of GDP) and limited IT audit capabilities, though donor-backed reforms (e.g., iTax in Kenya) show incremental progress.

Additional hierarchical clustering and PCA confirmed these groupings. The first principal component (reflecting overall digital governance maturity) explained ~70% of variance and correlated strongly with tax performance.

To provide greater transparency on the dimensionality reduction and support the interpretation of the latent structure underlying the governance-digitalization-tax relationship, we conducted Principal Component Analysis (PCA). The results are presented in

Table 3. Summary of Principal Component Analysis (PCA).

Principal Component	Eigenvalue	Variance Explained (%)	Cumulative Variance (%)
PC1	2.80	70.0	70.0
PC2	0.80	20.0	90.0
PC3	0.30	7.5	97.5
PC4	0.10	2.5	100.0

. The first component (PC1), which we interpret as a “Digital Governance Maturity Index”, explains 70% of the total variance and loads highly on variables such as EGDI, IDI, CPI, and Tax Revenue to GDP ratio. This confirms that countries with higher digital infrastructure and governance indicators tend to exhibit stronger fiscal outcomes.

Additional hierarchical clustering and PCA confirmed these groupings. The first principal component (reflecting overall digital governance maturity) explained ~70% of variance and correlated strongly with tax performance.

We confirmed these groupings with hierarchical clustering and by examining principal components. The first principal component (PC1)—which we interpret as an overall “digital governance maturity” axis—explained ~68% of the total variance, and it correlated strongly with Tax/GDP (loading ~0.85) as well as with EGDI, IDI, and CPI (all with loadings 0.85–0.95 on PC1). This indicates that a single underlying factor does capture much of the variation in our five metrics (essentially distinguishing advanced, well-governed digital states from the rest). PC2 (explaining ~20%) differentiated somewhat between high-corruption vs. high-cybersecurity environments (since CPI and GCI load in opposite directions on PC2), separating cases like Russia (high GCI, low CPI) from others.

5. Discussion

The empirical findings presented above highlight a strong alignment between the digital maturity of tax administration, the integrity of governance, and tax revenue outcomes. In this section, we discuss these results from a risk governance perspective, interpret causal linkages cautiously, and draw implications for policymakers seeking to enhance digital fiscal integrity. We also relate the findings to the frameworks and initiatives mentioned (COBIT, ISACA guidelines, INTOSAI, OECD Tax Administration 3.0) and discuss how they can be leveraged to achieve the observed benefits.

5.1. Digital Maturity as a Driver of Tax Performance

One of the most striking findings is the significant positive correlation between e-government development (EGDI) and tax revenue performance. This empirically confirms on a broader scale what earlier case studies and anecdotal evidence have suggested: digitizing tax processes tends to improve efficiency and compliance. Our analysis reinforces and quantifies this relationship, which was highlighted in prior research. For example, Njuguna Ndung’u demonstrated in Kenya that digitization of the tax system improved compliance by deterring bribery and broadening the tax base, and the OECD has repeatedly noted the revenue benefits of e-filing and pre-populated returns (Collosa, 2025). We find that countries with high EGDI (≈0.8–1.0) collect on average 10–20 percentage points more of GDP in taxes than those with low EGDI (<0.5). While some of this difference is attributable to higher income levels or governance, our multivariate regressions suggest digital maturity alone has a sizable association with tax performance (even controlling for corruption perceptions). Several factors likely contribute:

Increased Compliance and Broader Tax Base: Digital systems make it easier for taxpayers to comply (file returns, pay taxes) and harder to remain outside the system. For example, electronic invoicing and digital reporting create transaction trails that draw in more of the informal economy. Our analysis showed countries like Brazil and South Africa (which have implemented e-invoicing and data matching) achieve relatively higher Tax/GDP than peers at similar income levels who have not. The OECD notes that by “building taxation processes into taxpayers’ devices and systems,” burdens fall and compliance rises. Our data, reflecting real-world outcomes, support this; higher EGDI countries do seem to capture a larger share of GDP in taxes.

• Efficiency and Cost Savings: Digital tax administration automates routine tasks (calculations, form processing) and allows redeployment of staff to higher value activities (like audit and taxpayer service). It also reduces errors. This improves the tax yield for a given level of resources. Estonia is a case in point: by digitizing virtually all tax services and using the X-Road interoperability platform, Estonia has among the world’s lowest tax compliance costs and one of the highest tax compliance rates, which helped it jump to 3rd in the EGDI ranking. Our regression results indicate that an increase in EGDI correlates with a substantial increase in revenue, even controlling for other factors. This suggests a real efficiency dividend from digitalization.
• Transparency and Trust: When taxpayers can see clearly how much tax is due (e.g., pre-filled returns, online tax calculators) and have a smooth payment process, it fosters voluntary compliance. In contrast, opaque, paper-based systems breed mistrust and non-compliance. As noted by the OECD, “if it is burdensome to pay tax, that leads to higher costs…and potentially more mistakes or evasion.” Making the process easier through digital means encourages more people to comply willingly. This could partly explain the correlation we see between high EGDI and high CPI (low corruption): transparent systems reduce the avenues for corrupt behavior (like officials manipulating liabilities) and improve the overall taxpayer morale. While our study did not directly measure voluntary compliance rates, the macro-level revenue outcome is consistent with improved compliance.
• Data-Driven Risk Management: Digitally advanced administrations leverage data analytics to target audits and non-compliance more effectively. For instance, many cluster 1 countries use automated cross-checks (e.g., matching VAT inputs/outputs, or using third-party information to identify under-reporting). The UK’s Connect system (cited earlier) is an example of data-driven enforcement yielding billions in extra revenue. Such systems are part of IT audit and control—they continuously monitor taxpayer data for anomalies. By contrast, in low-EGDI countries, audits may be random or susceptible to corruption, leading to less effective enforcement. Thus, the risk-based audit frameworks enabled by IT contribute to the better tax performance we observe in digitally mature countries. In our results, this factor is implicitly captured by the EGDI (as advanced e-filing and database integration are prerequisites to advanced analytics).

It is important to acknowledge that the relationship can also work in reverse or be facilitated by a third factor: richer countries have more resources to invest in digital systems and tend to have higher tax/GDP needs for their welfare states. We partly control for this by the inclusion of CPI (since richer countries often have better governance too). Even so, one should be cautious in interpreting causation. The implementation of IT audit systems is not a magic wand—it must be accompanied by process re-engineering and capacity building. Countries like Mexico, for instance, invested in digital tax infrastructure (e.g., mandatory electronic invoices) and did see improvement in compliance, but still face challenges from the large informal sector. The multivariate analysis suggests that digital systems plus good governance together have the strongest association with outcomes. This points to synergy: technology amplifies the effects of sound policies and vice versa. A well-governed tax authority will make the most of new IT systems (ensuring they are used properly, maintained, and updated), while technology provides the tools to enforce and monitor good governance (e.g., audit logs, performance metrics).

5.2. The Role of Governance and Risk Controls (CPI and COBIT)

In an era where data is the new oil, tax administrations that successfully harness information technology and robust governance frameworks are reaping substantial dividends in revenue mobilization and integrity. This study, through a broad international analysis, confirms that digitally advanced and well-governed tax administrations exhibit significantly better performance—not just in collecting higher revenues relative to GDP, but in fostering a culture of compliance and transparency. Countries that have embraced e-filing, online services, data analytics, and automated controls—underpinned by frameworks like COBIT for IT governance and INTOSAI guidelines for independent audits—tend to have higher tax-to-GDP ratios and lower levels of corruption.

From a risk governance perspective, the findings highlight several key insights. First, digital transformation of tax administration is a critical component of improving fiscal integrity, but it must be pursued holistically. Simply buying new IT systems is insufficient; success requires aligning those systems with re-engineered processes and strong internal controls. The correlation between the UN’s E-Government Development Index and tax performance was high (r > 0.6), and our regression analysis suggested a sizable payoff—potentially on the order of a 3–5 percentage point increase in Tax/GDP for countries moving from moderate to high e-government maturity (holding other factors constant). This underscores OECD’s vision that integrating tax processes into the digital lives of taxpayers can “make taxation easier and less costly,” thereby improving compliance. We saw empirical support for this: higher e-government scores were associated with lower shadow economy impact and fewer opportunities for tax evasion.

Second, governance and anti-corruption measures amplify the benefits of digitalization. We found that corruption (CPI) was inversely related to tax outcomes—a stark reminder that leakages and public mistrust can undermine even the best IT systems. Countries scoring well on CPI (such as Estonia, with a CPI of 75 and among top e-gov ranks) enjoy both high compliance and efficient administration. Our analysis suggests that by reducing discretion and face-to-face contact (for example, through e-filing and automated risk-based audits), digital systems help curb petty corruption in tax administration. Furthermore, frameworks like COBIT ensure that proper controls and audits are in place so that new IT systems are not themselves compromised or misused. As noted, COBIT provides auditors and managers a blueprint to “evaluate IT operations, ensure control objectives are met, and risks mitigated”, leading to more effective and secure tax IT processes. The policy implication is clear: implementing IT audit systems must go hand-in-hand with strengthening institutional integrity. Governments should therefore invest as much in training, controls, and independent oversight as in hardware and software.

Third, the role of cybersecurity readiness (GCI), while not directly boosting tax revenue in our cross-section, is an essential pillar of risk management in a digital tax administration. A single cyber incident can erode years of progress by disrupting collections or undermining taxpayer confidence (e.g., a data breach could discourage e-filing). Hence, countries should proactively adopt cybersecurity frameworks and invest in protecting tax data. The fact that most high-performing administrations also score very high on the Global Cybersecurity Index is reassuring—it indicates they recognize that safeguarding the digital infrastructure is part and parcel of good tax governance. As tax administrations become increasingly data-driven and interconnected (sharing information across borders and with other agencies), cybersecurity and data privacy will become even more critical. International collaboration in this area—sharing best practices for secure tax technology—will benefit all, since cyber threats do not respect borders. A possible extension of this study could include robust regression and WLS estimation methods to account for heteroscedasticity and validate the stability of coefficients.

Our cluster analysis highlighted a “digital divide” in tax administration. Advanced economies have largely moved into a world of seamless, frictionless taxation, whereas many developing countries are still struggling with basic digitization. Bridging this gap is not only a matter of fairness but of global importance—as emerging economies raise their domestic revenues, they can rely less on external aid and debt, contributing to more sustainable development. Organizations like the IMF and World Bank, through initiatives (like the “Digital Tax Administration initiative” and others), should continue to support capacity building, focusing on both technology and governance. Encouragingly, some developing countries have made leaps (for example, many African countries implementing e-filing and integrated financial management systems in the last decade). These successes should be replicated and scaled.

6. Conclusions

The findings of this study confirm that digital maturity alone is not sufficient to ensure better tax outcomes. Rather, it is the combination of technological advancement, strong governance frameworks, and effective anti-corruption mechanisms that determines the success of tax digitalization. The Corruption Perceptions Index (CPI) emerged as the strongest positive predictor of tax revenue, highlighting the critical role of institutional integrity. Countries that effectively apply IT governance frameworks, such as COBIT, ensure independent oversight through Supreme Audit Institutions (SAIs), and invest in cybersecurity readiness, demonstrating superior fiscal performance.

From a risk governance perspective, the findings highlight several key insights. First, digital transformation of tax administration is a critical component of improving fiscal integrity, but it must be pursued holistically. Simply buying new IT systems is insufficient; success requires aligning those systems with re-engineered processes and strong internal controls. Our correlation and regression results showed a clear link between e-government maturity and tax performance, echoing OECD’s vision that integrating tax processes into everyday life can make compliance easier and less costly. We saw empirical support for this: higher e-government scores were associated with significantly higher tax yields. However, we also saw that technology’s impact is maximized when paired with governance. Countries that combined high EGDI with high CPI (e.g., Denmark, New Zealand) were the standout performers. This implies that technology amplifies good governance—in such environments, digital systems truly thrive and deliver returns (like higher revenues, reduced evasion). On the other hand, technology in a vacuum (without integrity) may have limited impact. Second, governance and anti-corruption measures amplify the benefits of digitalization. The CPI was a strong predictor of tax outcomes—a reminder that leakages and public mistrust can undermine even the best IT systems.

Countries scoring well on CPI (such as Estonia, with CPI ~75 and among top e-gov ranks) enjoy both high compliance and efficient administration. Our analysis suggests that by reducing discretion and face-to-face contact (for example, through e-filing and automated risk-based audits), digital systems help curb petty corruption in tax administration. Furthermore, frameworks like COBIT ensure that proper controls and audits are in place so that new IT systems are not themselves compromised or misused. As noted earlier, COBIT provides auditors and managers a blueprint to “evaluate IT operations, ensure control objectives are met, and risks mitigated”, leading to more effective and secure tax IT processes. The policy implication is clear: implementing IT audit systems must go hand-in-hand with strengthening institutional integrity. Governments should invest as much in training, process improvement, and independent oversight as in hardware and software. In short, digital tools need good governance to succeed, and good governance can be bolstered by digital tools.

Third, the role of cybersecurity readiness (GCI), while not directly boosting tax revenue in our cross-section, is an essential pillar of risk management in a digital tax administration. A single cyber incident can erode years of progress by disrupting collections or undermining taxpayer confidence. Therefore, countries should proactively adopt cybersecurity frameworks and invest in protecting tax data. The fact that most high-performing administrations also score very high on the GCI is reassuring—it indicates they recognize that safeguarding the digital infrastructure is part and parcel of good tax governance. As tax administrations become increasingly data-driven and interconnected (sharing information across borders and with other agencies), cybersecurity and data privacy will become even more critical. International collaboration in this area—sharing best practices for secure tax technology—will benefit all, since cyber threats do not respect borders.

Our findings suggest that even if cybersecurity does not show an immediate revenue payoff, it is likely a facilitator of sustainable digital taxation. Countries with weak cybersecurity might suffer setbacks that could nullify gains from digitalization (for example, system downtimes, fraud, or loss of sensitive taxpayer information leading to mistrust). Our cluster analysis highlighted a concerning but also motivating scenario—a digital divide in tax administration. Advanced economies are enjoying the fruits of decades of investment in both governance and technology, while many poorer countries are only at the start of that journey. Bridging this gap requires concerted effort. Organizations like the IMF and World Bank, through programs under the Platform for Collaboration on Tax and capacity development projects, should continue and expand their support for “digital tax administration” initiatives (World Bank & IMF, 2015).

In conclusion, this research confirms that implementing information technology audit systems within tax administration—conceived broadly as the people, processes, and technologies that together realize a digital tax system—is a cornerstone of enhancing risk governance and achieving digital fiscal integrity. Countries that champion this transformation reap tangible rewards: higher revenues that can fund public services, more trust from citizens that the tax system is fair and efficient, and greater resilience against fraud and error. As one CIAT report aptly noted, “digitalization can alter opportunities for corruption by reducing arbitrariness, increasing transparency, and enabling accountability”. Our findings echo that sentiment on a global scale. The path forward for tax administrations is clear—go digital, but do so guided by strong governance frameworks and a commitment to integrity. Those that succeed will not only collect the taxes due in their own countries, but also set an example of good governance in the digital age, contributing to a more transparent and equitable global tax environment.

Limitations and Future Research: We acknowledge that this study has limitations. The sample size (20 countries) is relatively small, and the data are cross-sectional, which limits our ability to establish causation or dynamic effects. Future research could expand the country coverage or, preferably, use panel data to examine how improvements in digital indices over time affect revenue outcomes. Additionally, our models could be extended by incorporating other relevant variables (such as GDP per capita, tax rates, enforcement effort, or education levels) to account for more determinants of tax performance—our exclusion of these for simplicity means there is a possibility of omitted variable bias in attributing variance to digital/governance factors. We also did not delve deeply into the cost side of digital reforms; a cost-benefit analysis would be useful (e.g., ROI of specific IT initiatives like in the HMRC case). Furthermore, while we touched on the idea of cybersecurity’s indirect role, this could be studied rigorously using interaction terms or case studies of cyber incidents. Finally, qualitative research on organizational readiness and change management in tax authorities would complement our quantitative findings, as the human factor in adopting IT audit systems is crucial. Despite these limitations, we believe our study provides a useful empirical baseline and reinforces key messages for policymakers.