Can FinTech Close the VAT Gap? An Entrepreneurial, Behavioral, and Technological Analysis of Tourism SMEs

Abstract

Governments worldwide are mandating e-invoicing and real-time VAT reporting, yet many cash-intensive service SMEs continue to under-report VAT, eroding fiscal revenues. This study investigates whether financial technology (FinTech) adoption can reduce this under-reporting among tourism SMEs in Greece—an economy with high seasonal spending and a persistent shadow economy. This is the first micro-level empirical study to examine how FinTech tools affect VAT compliance in this sector, offering novel insights into how technology interacts with behavioral factors to influence fiscal behavior. Drawing on the Technology Acceptance Model, deterrence theory, and behavioral tax compliance frameworks, we surveyed 214 hotels, guesthouses, and tour operators across Greece’s main tourism regions. A structured questionnaire measured five constructs: FinTech adoption, VAT compliance behavior, tax morale, perceived audit probability, and financial performance. Using Partial Least Squares Structural Equation Modeling and bootstrapped moderation–mediation analysis, we find that FinTech adoption significantly improves declared VAT, with compliance fully mediating its impact on financial outcomes. The effect is especially strong among businesses led by owners with high tax morale or strong perceptions of audit risk. These findings suggest that FinTech tools function both as efficiency enablers and behavioral nudges. The results support targeted policy actions such as subsidies for e-invoicing, tax compliance training, and transparent audit communication. By integrating technological and psychological dimensions, the study contributes new evidence to the digital fiscal governance literature and offers a practical framework for narrowing the VAT gap in tourism-driven economies.

1. Introduction

Value-added tax (VAT) is the single largest source of indirect tax revenue in the European Union, yet a persistent “VAT gap”—the shortfall between the theoretical liability and the amount actually collected—continues to erode public finances. Zídková [1] synthesized EU evidence and showed that the gap widens with household final consumption and shadow economy intensity, while it narrows when the VAT share in GDP is high and tax administration is efficient. Greece illustrates these opposing forces; tourism-driven consumer spending is high, cash transactions are ubiquitous, and the country faces ongoing challenges in narrowing its VAT gap. To ensure strong sectoral and regional representation, our sample focuses on the South Aegean (Cyclades), Crete, the Ionian Islands, and Attica—regions that together account for over 70 percent of Greece’s inbound tourism visits and spending [2]. This selection captures diverse tourism settings, including highly seasonal islands and urban centers, offering a comprehensive view of VAT compliance behavior across Greece’s main tourism zones. As tourism accounts for more than 20% of Greek GDP, the sector has enormous fiscal importance. However, its heavy reliance on cash, seasonal employment, and informal business practices makes it particularly vulnerable to VAT under-reporting—posing both a compliance challenge and a policy priority. Recent studies suggest that adopting cashless payment methods significantly enhances VAT compliance, as these transactions create verifiable audit trails, thereby deterring tax evasion [3]. The decision to focus on Greek tourism sector SMEs is threefold: first, because tourism businesses in Greece are highly seasonal and cash-reliant, making them especially prone to VAT under-reporting; second, because the sector contributes significantly to GDP and is undergoing a major digital transition with mandatory e-invoicing; and third, because SMEs account for the vast majority of tourism enterprises, allowing us to capture the compliance dynamics of the most policy-relevant business population.

Most VAT gap research remains macro-economic, relying on top-down national accounts methods that overlook individual firm behavior and sector context [4]. Greece offers a particularly relevant and timely setting for this study—not only because it has historically posted one of the EU’s largest VAT gaps, but because it is now undergoing a nationwide digital tax transformation. Recent fiscal reforms, notably the compulsory myDATA e-ledger and the phased rollout of certified e-invoicing in hospitality, have created a natural policy experiment. FinTech tools—including mobile points-of-sale, cloud bookkeeping, and real-time e-reporting—embed tax compliance within day-to-day transactions, potentially converting grey economy leakage into a digitally traceable audit trail [3]. Hondroyiannis and Papaoikonomou [5] affirm that enhanced card payment systems lead to increased VAT revenues, highlighting the information trail generated by secure digital transactions.

Despite the growing use of FinTech in tax reporting, little is known about how these technologies affect VAT compliance at the firm level—especially in cash-reliant sectors like tourism. Prior VAT gap studies have been largely macroeconomic and overlook behavioral and technological dynamics at the SME level. This study addresses this gap by integrating FinTech adoption with tax morale and audit perceptions to examine their combined influence on VAT compliance behavior in Greek tourism SMEs. We draw on the Technology Acceptance Model, deterrence theory, and behavioral tax compliance frameworks to investigate two related questions. The main research question is “Can FinTech adoption reduce VAT under-reporting in Greek tourism-sector SMEs?” The derived question asks: “Does the behavioral context—defined as the taxpayer’s internal motivation (tax morale) and perception of audit probability—moderate this effect?”

In this study, behavioral context refers to two well-established psychological drivers of tax compliance: tax morale (a taxpayer’s internalized willingness to comply) and perceived audit probability (the perceived likelihood of detection and sanction). These dimensions are widely recognized as the most robust and policy-relevant behavioral determinants of compliance behavior [6,7]. We focus on these two aspects because they lie at the core of the “slippery slope” framework, which posits that tax compliance is shaped by the interplay of trust in authorities and the power of enforcement [6]. Moreover, both are directly actionable for policymakers; tax morale can be fostered through trust building and civic engagement, while audit perception can be influenced by increasing the visibility or credibility of enforcement strategies. Alshira’h et al. [8] argue that fostering a trusting relationship between tax authorities and taxpayers can significantly improve compliance behavior, supporting the relevance of tax morale.

By shifting the unit of analysis from countries to firms and from aggregate consumption to digital transaction data, the paper contributes three advances. First, it offers micro-level evidence on how FinTech capabilities operate as an intangible resource that lowers compliance costs and signals legitimacy [9]. Second, it integrates behavioral moderators that were absent from earlier VAT gap econometrics, providing a richer explanation for why identical digital tools may yield different outcomes across firms [10]. Finally, it delivers policy-relevant insights for Greece’s post-pandemic tourism strategy; if digital adoption proves effective in shrinking the VAT gap, targeted subsidies or tax credits for e-invoicing equipment could result in increased revenue [11].

The remainder of the paper is organized as follows. Section 2 reviews literature on VAT gap determinants and the diffusion of FinTech in SMEs. Section 3 develops hypotheses and presents the conceptual model. Section 4 details data collection from 250 tourism SMEs and the measurement of latent constructs. Section 5 reports the Partial Least Squares results and robustness checks; Section 6 discusses theoretical and managerial implications; and Section 7 concludes with limitations and future research directions.

2. Literature Review

This literature review is organized into five interrelated subsections that build toward an integrated understanding of VAT compliance in the tourism sector. We begin with the sectoral context of Greek tourism SMEs, highlighting their economic importance and vulnerability to VAT under-reporting. This is followed by a discussion of macro-level determinants of the VAT gap as identified in cross-national studies. We then turn to behavioral explanations, focusing on tax morale and deterrence, before exploring how FinTech tools act as both enablers and enforcers of compliance. Finally, we synthesize these dimensions, proposing that the interplay of FinTech adoption and behavioral factors offers a more nuanced explanation of compliance outcomes in digitalizing tourism markets.

2.1. Sector Context: Tourism SMEs in Greece

Tourism accounts for over one-fifth of Greek GDP and is characterized by high transaction volumes, seasonality, and a historic reliance on cash—conditions conducive to VAT under-reporting. Studies have indicated that cash payments in the tourism sector can hinder proper tax compliance [12]. This risk is particularly acute for SMEs, which dominate the sector and often lack the internal controls of larger firms [13]. Zídková’s [1] own specification includes household spending on restaurant and hotel services, finding it negatively associated with the VAT gap, yet she attributes the result to measurement issues and calls for sector-specific micro studies. Greece’s mandatory migration to e-invoicing thus offers a natural experiment; if digital tools reduce cash leakage, the tourism VAT gap should contract. In this context, tourism-sector SMEs provide a highly policy-relevant lens through which to assess how FinTech adoption interacts with structural and behavioral compliance challenges.

2.2. The VAT Gap Debate and Its Macro-Determinants

Early cross-country work—most notably the annual VAT Gap Reports by the European Commission—framed the VAT gap as a macroeconomic leakage driven by structural variables such as final household consumption, the size of the shadow economy, the standard VAT rate, and proxies for administrative effectiveness. These were further extended by academic analyses such as Zídková [1] and Butu et al. [14]. Zídková’s [1] regression on 24 EU member states confirmed that higher household consumption and a larger informal sector widen the gap, whereas a greater VAT share in GDP—interpreted as stronger collection capacity—narrows it [1]. Her dataset also underscored the Greek anomaly; while the EU average relative gap fell to 11% in 2006, Greece’s rose to 30%, the largest in the bloc [14]. Subsequent refinements highlighted additional drivers, including the negative correlation between the complexity of VAT rates and compliance levels, suggesting that higher rates or multiple rates can confuse taxpayers, thus increasing the gap [15].

The macro strand of research relies on top-down estimation from national accounts data, which, although useful for cross-country benchmarking, may understate cash-based under-reporting prevalent in retail and hospitality sectors [16]. Gajewski and Joński [17] cautioned that method choice and unobserved “grey” economy activity can bias estimates, urging researchers to explore micro-level dynamics and cultural or technological moderators of compliance, such as the impact of digital tools and behavioral taxation theories.

In response to these limitations, a growing body of literature has shifted toward micro-level and behavioral approaches, which offer finer-grained insights into taxpayer decision making. These approaches allow researchers to explore how firm characteristics, technology adoption, and psychological factors like tax morale and audit perception interact to influence compliance—particularly in sectors prone to informality, such as tourism.

Indeed, recent studies indicate that transaction digitization significantly enhances compliance through improved traceability and administration efficiency [18], while recognizing the critical role that firm-level factors and individual behavior play in tax compliance [16].

2.3. Behavioral Drivers: Tax Morale and Deterrence

Research indicates that tax compliance is influenced by perceptions of fairness, corruption, and the likelihood of being audited. A study by Bachas et al. [19] found that an increase in the standard VAT rate is associated with lower compliance. Stronger governance institutions and effective judicial systems support higher compliance levels. Furthermore, Christie and Holzner [20] observed that higher travel revenue shares, which serve as a proxy for tourism intensity, correlate with better VAT performance, suggesting that the visibility of tourism receipts may enhance compliance through increased oversight [21]. However, these studies do not account for the impact of digital tools that can influence taxpayer attitudes and enforcement perceptions. The “slippery slope” framework developed by Kirchler et al. [6] provides a robust theoretical lens for understanding compliance behavior. It posits that tax compliance is shaped by the dynamic interaction between two forces: enforced compliance—based on perceived audit risk and the power of tax authorities—and voluntary compliance—based on trust in institutions and internalized tax morale. Voluntary motivation is rooted in civic norms and the taxpayer’s ethical stance, while deterrence stems from fear of detection and sanction. Alm [7] synthesizes extensive international research showing that tax morale and audit probability are the two most empirically validated psychological predictors of compliance behavior. These variables not only explain taxpayer decisions across diverse institutional settings but are also directly actionable through policy levers—such as communication campaigns, audit visibility, and civic education. Consequently, these dimensions remain central to the present study, which examines their moderating role in the link between FinTech adoption and VAT compliance.

2.4. Digitalization and FinTech as Compliance Enablers

The last decade has seen a rapid rise in FinTech solutions—such as mobile points-of-sale and cloud bookkeeping—designed to facilitate tax reporting in everyday transactions. Although policy documents characterize these tools as cost-effective measures against evasion, empirical research on their effectiveness remains limited and predominantly focused on macroeconomic outcomes [22]. FinTech is theorized to act both as a compliance facilitator—by lowering administrative burdens—and as a compliance enforcer, by generating traceable transaction data that increases perceived audit probability. FinTech is posited to reduce the costs associated with accurate reporting and to influence compliance behavior by enhancing the perceived probability of detection, aligning with deterrence theory [23,24]. This mechanism links directly to behavioral compliance models, where perceived audit risk is a critical determinant of taxpayer behavior. Evidence suggests that increased transparency from digital tools may help reduce tax evasion and improve compliance, including in environments with varying levels of enforcement [25].

2.5. Integrating FinTech and Behavior in VAT Gap Research

Prior research establishes the structural context of tax compliance but often overlooks two critical aspects: (i) the role of firm-level FinTech adoption, and (ii) the impact of behavioral factors—such as tax morale and perceived audit risk—on the effectiveness of technological interventions [26]. The present study positions FinTech as a technological determinant that interacts with these behavioral factors, affecting compliance among Greek tourism SMEs [27]. This interactional approach is consistent with the “compliance ecosystem” perspective [28], which emphasizes that compliance outcomes result from the dynamic interplay of capabilities, incentives, and perceptions. By utilizing micro data rather than macro regressions, this study directly addresses Zídková’s [1] call for richer explanatory frameworks and contributes evidence pertinent to ongoing digital tax initiatives in Greece. It also complements the “slippery slope” framework [6], which links voluntary and enforced compliance to the dual foundations of trust and power—both of which can be shaped by technology adoption and audit salience.

3. Research Model and Hypothesis Development

This section addresses the core research question introduced in Section 1: To what extent do FinTech adoption and behavioral factors influence VAT compliance and financial performance among Greek tourism SMEs? To investigate this, we develop a conceptual framework grounded in three complementary theories—Technology Acceptance Model (TAM), Deterrence Theory, and the Resource-Based View (RBV)—each offering distinct but interconnected insights into compliance behavior and business outcomes.

TAM explains the conditions under which tourism SMEs adopt FinTech tools, emphasizing perceived usefulness and ease of use as key adoption drivers (supporting H1). Deterrence Theory informs the role of perceived audit probability and the psychological mechanisms that govern compliance decisions under the threat of sanctions (supporting H4 and H5). RBV provides a strategic perspective, positing that both compliance routines and technology capabilities can become valuable internal resources that contribute to improved financial performance (supporting H2 and H3). Together, these perspectives construct a logic chain from FinTech adoption to VAT compliance and ultimately to SME financial outcomes. Based on this integrated framework, we develop five hypotheses and present them in a conceptual model (Figure 1).

3.1. FinTech Adoption and VAT Compliance Behavior

The TAM posits that perceived usefulness and ease of use are primary antecedents of technology uptake [29]. Once adopted, FinTech tools—cloud bookkeeping, certified e-invoicing, real-time myDATA reporting—automate record keeping and eliminate discretionary “cash gaps,” thereby shrinking the scope for intentional or accidental under-declaration. Empirical work on digital fiscal devices in Latin America and Scandinavia shows that mandatory e-invoicing raised declared VAT liabilities by 10–17 percent within two years of rollout, chiefly because sales were recorded at the moment of payment rather than later in bulk. Although such evidence is macro in nature, it supports the micro-level mechanism that FinTech reduces cognitive costs and increases traceability—two classic compliance facilitators in deterrence theory [30]. Recent evidence from Greek SMEs further validates this logic. Balaskas et al. [26] demonstrate that trust, perceived usefulness, and government policy support significantly predict FinTech adoption in Greece’s small business ecosystem, reinforcing the practical relevance of TAM constructs in this study’s sectoral context.

H1. 

Greater FinTech adoption in a tourism SME is associated with higher VAT compliance behavior.

3.2. VAT Compliance and Financial Performance

From a Resource-Based View (RBV) standpoint, compliance is not merely a statutory outcome but an intangible resource that curbs penalty risk, improves access to formal finance, and strengthens customer trust. In this study, financial performance is operationalized through four core indicators relevant to SMEs: profitability, return on investment, customer retention, and sales growth. Greek banks have begun to offer preferential interest rates to firms that integrate their e-invoicing streams with loan-monitoring dashboards, signaling that transparent VAT reporting can convert into lower perceived credit risk [27]. In parallel, Hondroyiannis and Papaoikonomou [5] find that increased use of electronic payments in Greece—especially in tourism-intensive areas—has significantly improved VAT revenue performance, further reinforcing the financial upside of digital compliance. Zídková [1] indicates that countries with higher VAT-to-GDP ratios tend to enjoy lower relative VAT gaps and more stable tax revenues, though this does not establish a direct correlation at the firm level. This suggests that SMEs with transparent records might be able to reinvest savings from potential penalties and improved cash flow into service quality and marketing. In the context of this study, financial performance is understood in terms of profitability, return on investment, customer retention, and sales growth —key metrics that reflect SME viability and competitiveness in the tourism sector.

H2. 

Higher VAT compliance behavior positively influences an SME’s financial performance.

3.3. Mediating Role of VAT Compliance

FinTech, by itself, does not guarantee profit gains; benefits should flow through the mechanism of better compliance (reduced fines, stronger credit profiles, efficiency in bookkeeping). This aligns with the “capability–performance” sequence in the RBV, where technologies become valuable only when translated into superior routines. The theoretical underpinnings of this relationship are bolstered by studies that demonstrate that effective VAT compliance can lead to improved financial performance, especially when technological solutions like e-invoicing help reduce operational costs [9,27].

H3. 

VAT-compliance behavior mediates the positive relationship between FinTech adoption and financial performance.

3.4. Moderating Role of Tax Morale

Behavioral tax models emphasize intrinsic motivation—tax morale—as a critical complement to external enforcement. Alm and McClellan [31] demonstrated that improvements in tax morale have a significant and positive impact on reporting and compliance behavior. They argue that when taxpayers perceive the tax system and its enforcement as fair, compliance is markedly higher, even in the presence of high statutory rates. In tourism SMEs, owners with strong civic norms may utilize FinTech to enhance compliance rather than merely fulfill it. Conversely, actors with low tax morale may search for evasion tactics, even in a digital environment. A substantial body of literature supports the notion that tax morale significantly affects compliance behavior [31,32].

H4. 

Tax morale strengthens the positive effect of FinTech adoption on VAT compliance behavior.

3.5. Moderating Role of Perceived Audit Probability

Deterrence theory posits that compliance increases with the perceived certainty of detection. Digital transmission of tax data may enhance this certainty; however, the psychological perception of audit likelihood can vary among SMEs. Those in tourism who believe that audits are frequent are more likely to perceive compliance as beneficial—avoiding fines and preserving their reputation—than those who are skeptical about enforcement [33]. Research indicates that taxpayers’ beliefs regarding enforcement strength directly influence their compliance behaviors, underscoring the importance of perceived audit risk [34].

H5. 

Perceived audit probability strengthens the positive relationship between VAT compliance behavior and financial performance.

To summarize, the hypothesized model unfolds as a logically sequenced “chain of effects”; FinTech adoption enhances VAT compliance (H1), which in turn improves financial performance (H2), with VAT compliance mediating this link (H3). The strength of these relationships is moderated by internal (tax morale; H4) and external (audit probability; H5) behavioral factors. Together, the hypotheses build a multi-stage mechanism where FinTech adoption leads to measurable SME performance gains through behavioral transformation and institutional responses.

4. Methodology

To comprehensively test the hypotheses and ensure methodological triangulation, this study employs a multi-method analytical strategy combining (i) Partial Least Squares Structural Equation Modeling (PLS-SEM), (ii) fuzzy-set Qualitative Comparative Analysis (fsQCA), and (iii) robustness checks using regression and endogeneity diagnostics. PLS-SEM was selected for its suitability with small-to-medium samples, non-normal data, and models involving complex mediation and moderation structures. It is especially appropriate when prediction and construct estimation are primary objectives. fsQCA complements this by identifying configurations of conditions that are sufficient for high VAT compliance, aligning with the study’s interest in causal asymmetry. Regression-based checks test the stability of results under alternative specifications, while the Gaussian copula approach addresses potential endogeneity between FinTech use and compliance behavior—something conventional SEM methods cannot resolve. This multi-method design enhances the validity, credibility, and policy relevance of the findings, allowing us to explore both average effects and diverse causal paths.

The empirical strategy blends a cross-sectional survey with robust latent-variable modelling to uncover how FinTech adoption influences VAT behavior and downstream performance in Greek tourism SMEs. This section explains population definition, sampling, instrument development, data collection procedures, and the analytical techniques deployed to test the hypotheses introduced in Section 3.

4.1. Research Design and Population

Because the study examines technology-enabled VAT compliance in a cash-intensive service industry, the target population comprises registered tourism SMEs—hotels, guesthouses (including Bed and Breakfasts), travel agencies, excursion operators, and hospitality firms. Company lists were compiled from the Hellenic Chamber of Hotels, the Hellenic Association of Travel and Tourist Agencies (HATTA) and regional chambers in the Cyclades, Crete, the Ionian Islands and Athens. The selection of Cyclades, Crete, the Ionian Islands, and Attica was motivated by their economic centrality to Greek tourism, high SME density, and documented exposure to VAT evasion risks due to high seasonality, cash-based transactions, and regulatory variation between urban and island contexts. These regions collectively host over 70 percent of national tourism receipts, ensuring sectoral representativeness while capturing both island and mainland dynamics. Focusing on registered tourism SMEs in high-density tourism regions enhances the contextual relevance of the findings and ensures the study captures the diversity of operational environments and compliance challenges faced by Greek tourism firms (see

Table 1. Sample profile and descriptive statistics.

Variable	Category/Metric	n	%/Mean	SD
Region	Cyclades	58	27.1	–
	Crete	52	24.3	–
	Ionian	35	16.4	–
	Attica	69	32.2	–
Sub-sector	Accommodation	137	64.0	–
	Tours/Activities	77	36.0	–
Employees		—	$\overline{\mathrm{x}}$ = 22.6	13.8
Annual turnover (€m)		—	$\overline{\mathrm{x}}$ = 3.4	2.1

). This design choice aligns with the objective of examining FinTech-related tax behaviors in settings where informal transactions and VAT underreporting are most prevalent.

4.2. Sampling Procedure

A stratified random design was adopted to mitigate regional and subsector bias. First, the population was stratified into four geographic strata (Cyclades, Crete, Ionian, Attica). Within each stratum firms were further classified by subsector (accommodation vs. non-accommodation services). Proportional allocation ensured that sample weights mirrored the actual distribution of SMEs in the Ministry of Tourism business registry. Power analysis indicated a minimum of 200 usable responses to detect medium structural effects (f² = 0.15) at 95 percent power in PLS-SEM with five latent constructs. This method ensures sufficient statistical power to detect hypothesized effects, while stratification by region and subsector enhances external validity and representativeness of the tourism SME landscape. Stratified random sampling is particularly appropriate for heterogeneous sectors like tourism, where regional and operational dynamics vary considerably. Allowing for a 30 percent response rate, 750 firms were approached. After two email waves and telephone follow-ups, 238 complete surveys were returned; listwise deletion of speeders and systematically incomplete cases yielded 214 usable observations, comfortably above the threshold.

4.3. Data Collection Procedure and Ethics

Data were gathered between February and May 2025 using a secure online survey. Invitations emphasized the academic purpose and guaranteed confidentiality while respondents received a benchmarking report comparing their FinTech and compliance scores to peer averages. Participation was voluntary and anonymous; no personal identifiers beyond firm characteristics were recorded. Online administration was selected for its logistical efficiency and suitability during a post-pandemic period, enabling safe, wide-reaching access to geographically distributed firms.

4.4. Instrument Development

All constructs were operationalized with multi-item Likert scales anchored at 1 (“strongly disagree”) and 5 (“strongly agree”). The questionnaire items were derived from validated scales in the FinTech adoption and tax compliance literature (see Appendix A for the full survey instrument). Items for FinTech Adoption were adapted from prior SME digitalization studies. VAT Compliance Behavior captured both procedural accuracy and timeliness. Tax Morale drew on validated civic duty scales, while Perceived Audit Probability referenced owners’ subjective assessments of enforcement intensity in tourism. The construct Financial Performance was assessed using subjective rather than objective measures due to inherent data availability constraints typical among small and medium-sized enterprises (SMEs). This measurement approach is methodologically justified for three reasons. First, subjective indicators are widely used in entrepreneurship and SME research when financial data access is limited. Second, they allow for standardized cross-firm comparison even when accounting systems vary. Third, managers’ self-assessments have been shown to correlate strongly with actual firm outcomes and strategic positioning in contexts where external validation is difficult [35]. SMEs, especially in Greece’s tourism sector, rarely disclose audited financial statements publicly, and access to standardized financial reporting across such firms is notably limited. Furthermore, subjective performance indicators—such as perceived profitability, customer retention rates, return on investment, and sales growth—have been consistently validated in prior SME and entrepreneurship research [35]. This subjective measurement approach captures managerial perceptions of financial outcomes effectively, reflecting not only accounting-based metrics but also strategic dimensions of performance relevant to SMEs’ competitive and operational contexts.

4.5. Pilot Testing

A pilot study with 20 tourism owners in Athens evaluated clarity and scale reliability. Cronbach’s α values ranged from 0.72 to 0.88, exceeding the 0.70 threshold; pilot testing also provided early evidence of internal consistency and scale comprehension, which is critical when surveying owners with varying levels of digital familiarity. Minor wording refinements were incorporated before full launch. Pilot feedback also confirmed the survey’s average completion time (11 min) and the relevance of FinTech examples (mobile POS, e-receipts, cloud bookkeeping).

4.6. Assessment of Common Method Variance and Non-Response Bias

Given the survey-based design and reliance on self-reported data collected in a single wave, specific methodological steps were implemented to minimize common method variance (CMV)—systematic measurement error attributable to the method of data collection rather than the constructs being measured. Procedural remedies included assuring anonymity, separating predictor and criterion blocks, and varying scale anchors across constructs. To statistically assess the presence of CMV, we applied Harman’s single-factor test, a widely used diagnostic in behavioral research. If a single factor were to explain the majority of variance, it would suggest potential CMV. In our case, Harman’s single-factor solution accounted for only 32 percent of the total variance, well below the conservative 50 percent threshold, indicating that no single latent factor dominated the variance structure. This suggests that CMV is unlikely to severely distort the observed relationships among constructs. A full collinearity test (VIFs < 3.3) further supported this conclusion, showing no pathological CMV. Additionally, early- versus late-wave comparisons on key constructs showed no significant mean differences (p > 0.10), alleviating concerns about non-response bias.

4.7. Measurement Validation

Exploratory factor analysis confirmed item loadings on their intended constructs. In the PLS measurement model composite reliabilities ranged from 0.83 to 0.91, average variance extracted (AVE) exceeded 0.50, and outer loadings were significant (p < 0.001). Heterotrait–monotrait ratios were below 0.85, evidencing discriminant validity. Thus, the reflective measurement model meets reliability and validity criteria. These validation checks are standard practice in behavioral accounting and tourism research using PLS-SEM, ensuring robustness in construct estimation and model interpretability.

4.8. Structural Analysis Strategy

Partial Least Squares Structural Equation Modeling (SmartPLS 4) was chosen for its suitability with non-normal data, complex mediation-moderation and sample sizes around 200. Bootstrapping with 5000 resamples generated bias-corrected confidence intervals. PLS-SEM is particularly well-suited to exploratory research models with formative and reflective constructs, as well as when the goal is prediction rather than theory confirmation. Mediation (H3) followed the Zhao et al. [36] procedure; moderation (H4, H5) employed the two-stage interaction approach with mean-centered indicators. Predictive relevance was gauged via blindfolding (Q²) and out-of-sample PLSpredict.

To further ensure model credibility, a multi-method analytical approach was adopted—combining SEM, fsQCA, and regression—each contributing a unique perspective on the underlying data-generating process.

Robustness checks included (i) multigroup analysis comparing island versus mainland SMEs; (ii) an fsQCA to explore sufficiency configurations leading to high compliance; and (iii) regression of self-reported compliance on objective firm-size and subsector controls to rule out omitted-variable bias.

To evaluate potential endogeneity between FinTech adoption and VAT compliance, we applied a Gaussian-copula procedure [37]. After instrumenting FinTech with regional broadband density and ERP uptake, the copula residual was non-significant (β = 0.03, p = 0.62), indicating no detectable endogeneity bias.

By uniting rigorous sampling, validated instruments and complementary analytical lenses, this design offers a rare micro-level examination of how digital finance reshapes tax behavior within Greece’s tourism heartland. It directly answers prior calls to integrate technological and behavioral variables into VAT gap research while producing actionable insights for policymakers and SME owners alike.

4.9. Methodological Limitations

While the study follows rigorous empirical procedures, several methodological limitations should be acknowledged. First, the cross-sectional design restricts causal inference, as it captures associations at a single point in time rather than over time. Future longitudinal studies could better illuminate dynamic relationships between FinTech adoption, compliance behavior, and performance. Second, data rely on self-reported measures, which may be subject to social desirability or recall bias, especially regarding compliance behavior. Although we implemented procedural and statistical controls for common method variance, residual bias cannot be entirely ruled out. Third, the focus on registered tourism SMEs excludes informal or unregistered operators who may exhibit different compliance dynamics. Finally, while subjective performance metrics are validated in SME research, they may not fully capture objective financial outcomes. These limitations should be considered when interpreting the generalizability and scope of the findings.

5. Results

5.1. Measurement Model Outcomes

The reflective outer model achieved strong psychometric quality. Composite reliabilities ranged from 0.83 (Tax Morale) to 0.91 (FinTech Adoption). Average variance extracted exceeded the 0.50 threshold for every latent variable (0.57–0.74), and all outer loadings were significant at p < 0.001. The Heterotrait–Monotrait (HTMT) ratios were well below the conservative 0.85 cut-off (maximum = 0.71), confirming discriminant validity. Full collinearity VIFs fell between 1.17 and 2.24, precluding common-method inflation. Blindfolding produced Q² values of 0.21 (VAT Compliance) and 0.19 (Financial Performance), indicating medium predictive relevance. These results suggest that the survey instrument was both reliable and valid, meaning that the constructs were consistently measured and captured distinct concepts. For example, respondents’ answers on tax morale were statistically distinct from their responses on audit probability or FinTech adoption, reinforcing the conceptual clarity of the model (See

Table 2. Measurement quality—reliability and convergent validity.

Construct	Items	A	CR	AVE	Highest HTMT
FinTech Adoption	4	0.88	0.90	0.70	0.71
VAT Compliance	5	0.85	0.88	0.62	0.63
Tax Morale	4	0.78	0.83	0.57	0.58
Audit Probability	3	0.79	0.85	0.66	0.60
Financial Performance	4	0.81	0.87	0.64	0.67

).

5.2. Structural Model Goodness of Fit

The saturated PLS path model explained 48% of the variance in VAT Compliance Behavior and 42% in Financial Performance (adjusted R² values). Standardized root mean square residual (SRMR) equalled 0.056, beneath the 0.08 guideline, while the d_ULS and d_G indices also lay below their 95th-percentile bootstrapped benchmarks—evidence of acceptable global fit for a PLS model. The model explains a substantial portion of the variance in both tax compliance and financial outcomes, indicating that FinTech tools and compliance behaviors are closely linked to how SMEs perceive and manage their business success. The model’s good fit suggests it is well-suited to capturing these relationships in the real-world setting of Greek tourism SMEs.

5.3. Hypothesis Testing

Table 3. Structural paths, effect sizes and predictive relevance.

Hypothesis	Path	Β	t (boot)	p	f2	Q2
H1	FinTech → VAT Comp	0.56	10.24	<0.001	0.46	0.21
H2	VAT Comp → FinPerf	0.41	6.72	<0.001	0.23	0.19
—	FinTech → FinPerf (direct)	0.06	1.34	0.18	0.01	0.07

presents structural paths, effect sizes, and predictive relevance.

Table 4. Mediation and moderation tests.

Effect	Coefficient	95% BCa CI	Result
Mediation (H3)	FinTech → VAT Comp → FinPerf	0.23	[0.15, 0.32]
Moderation (H4)	FinTech × Tax Morale → VAT Comp	0.14	[0.04, 0.25]
Moderation (H5)	VAT Comp × Audit Prob → FinPerf	0.12	[0.01, 0.24]

reports the mediation and moderation results.

H1 (FinTech → Compliance).

The direct effect of FinTech Adoption on VAT Compliance was positive and strong (β = 0.56, t = 10.24, p < 0.001). The effect size was large (f² = 0.46), confirming H1. This finding confirms that SMEs that adopt FinTech tools—such as mobile POS, digital invoices, or automated VAT filing—are significantly more likely to comply with tax obligations. These technologies likely simplify record keeping and reduce opportunities for evasion, making compliance easier and more habitual.

H2 (Compliance → Performance)

VAT Compliance exerted a positive impact on Financial Performance (β = 0.41, t = 6.72, p < 0.001), with a medium effect size (f² = 0.23). H2 is supported. Firms that are more compliant with VAT regulations tend to experience stronger financial performance. This may be due to the benefits of transparency, such as better customer trust, smoother audits, and eligibility for public funding or partnerships.

H3 (Mediation)

Bootstrapped specific-indirect effects revealed that FinTech Adoption influenced Financial Performance through VAT Compliance (β_indirect = 0.23, 95% BCa CI [0.15, 0.32]). The direct FinTech → Performance path shrank to non-significance (β = 0.06, p = 0.18), indicating full mediation and confirming H3. The mediation result reveals that FinTech contributes to performance not directly, but through improving compliance. In other words, FinTech works because it builds better tax behavior, which in turn leads to greater business outcomes. This supports the view that legitimacy and rule-following are assets in business ecosystems.

H4 (Tax Morale Moderation)

The interaction term FinTech × Tax Morale significantly predicted VAT Compliance (β = 0.14, t = 2.57, p = 0.011). Simple-slope probes showed that the FinTech–compliance gradient is almost twice as steep at +1 SD tax morale levels (β = 0.71) than at −1 SD (β = 0.38), supporting H4. The impact of FinTech on compliance is stronger among firms with higher tax morale. This suggests that technology alone is not enough—owners must also believe in the value of paying taxes for digital tools to be fully effective.

H5 (Audit Probability Moderation)

Perceived Audit Probability amplified the Compliance → Performance link (β_interaction = 0.12, t = 2.03, p = 0.043). A Johnson–Neyman probe revealed that the positive compliance → performance effect becomes significant when perceived audit probability exceeds 2.8 (on the 1–5 scale), encompassing 89% of the sample, validating H5. Audit probability strengthens the performance benefits of compliance. When firms perceive that enforcement is credible, the rewards of honest behavior increase. This supports the idea that digital transformation must be matched by visible and fair enforcement.

5.4. Robustness Checks

PLSpredict out-of-sample tests yielded positive Q²_predict statistics for all indicators of the endogenous constructs; root-mean-squared error values for PLS predictions were lower than naïve linear model benchmarks for 82% of items, underscoring predictive utility. Prior to conducting the multigroup comparison, we applied the Measurement Invariance of Composite Models (MICOM) procedure [38] to confirm configural invariance, compositional invariance, and equality of composite means and variances across island and mainland subsamples; all steps held, validating the multigroup analysis. Multi-group analysis found no significant path coefficient differences between island-based and mainland SMEs, suggesting the model generalizes across tourism geographies.

For the fsQCA, we calibrated each latent construct into fuzzy set memberships using the direct method with anchor thresholds at the 95th, 50th and 5th percentiles [39], ensuring robust identification of sufficiency configurations, and set a consistency threshold of 0.80 and a frequency cutoff of 1 case, following established guidelines. FsQCA identified two sufficiency configurations for high VAT compliance. The dominant configuration combined high FinTech adoption, high tax morale, and high audit probability perception—echoing the SEM results—and exhibited a solution consistency of 0.87 and coverage of 0.46. Additionally, we applied the Gaussian-copula approach [37] to test for endogeneity between FinTech adoption and VAT compliance; the copula term was non-significant (β = 0.03, p = 0.62), confirming no endogeneity bias. Incorporating firm age, size and subsector (accommodation vs. activities) as controls did not materially change focal path estimates. No control exhibited a significant direct effect on compliance, ruling out omitted-variable bias (See

Table 5. Robustness and additional analyses.

Test	Statistic/Outcome	Interpretation
PLSpredict (FinPerf)	RMSE_PLS < RMSE_LM for 82% of indicators	Model has superior out-of-sample accuracy
Multi-group (Island vs. Mainland)	Δβ < 0.10, p > 0.10	No group difference
fsQCA Solution Consistency	0.87	High sufficiency
Common-Method VIF (max)	2.24	No CMV risk

for a summary of robustness and additional analyses).

Together, these robustness tests reinforce the reliability of the findings. The fsQCA shows that high compliance is not only driven by FinTech alone but emerges from specific combinations—such as FinTech + high morale + strong enforcement. This highlights the multi-causal nature of compliance and provides policymakers with more nuanced levers for intervention.

5.5. Summary of Findings

The evidence demonstrates that FinTech is a potent compliance accelerator for tourism SMEs, but its effectiveness depends heavily on the owners’ tax morale and the perceived audit environment. In other words, digital tools alone do not automatically drive better tax behavior; their impact is magnified when users are ethically motivated and expect enforcement. This reinforces the view that technology must be embedded within a supportive behavioral and regulatory ecosystem.

Improved compliance fully mediates the technology–performance link, implying that financial gains arise not from direct cost savings, but rather from the reputational and operational benefits of transparent tax conduct. SMEs that comply more rigorously tend to enjoy greater financial stability, customer trust, and policy alignment, enhancing their long-term viability.

These insights move VAT gap research beyond macro-level correlates and offer a micro-level understanding of how digitalization reshapes tax compliance behavior. The findings also support Greece’s recent policy emphasis on FinTech adoption incentives. When such tools are accompanied by morale-building campaigns and credible audit signals, they can significantly reduce VAT leakages in a sector historically vulnerable to cash-based underreporting.

Finally, the fsQCA results reveal that high compliance does not depend on a single path; some SMEs reach similar outcomes through strong norms and perceived audit risk, even with moderate FinTech use. This calls for tailored interventions that reflect the diversity of SME contexts rather than uniform digital mandates.

6. Discussion and Implications

This study set out to verify whether FinTech adoption can diminish VAT under-reporting in the Greek tourism sector and, in turn, improve SME financial outcomes. The structural results corroborate that digital finance tools are indeed a pivotal compliance enabler—yet only when buttressed by strong tax morale norms and credible audit expectations. In doing so, the findings unpack Zídková’s [1] macro-level observation that household consumption and shadow economy forces raise the VAT gap, while administrative effectiveness lowers it. At the micro level, FinTech functions as a “digital guardrail” that channels high-volume tourism transactions into the formal ledger, but behavioral factors determine how firmly that guardrail is respected. The results empirically validate all five hypotheses tested and provide nuanced insight into their interaction effects—particularly the full mediation of performance through compliance (H3), and the moderating effects of tax morale (H4) and audit probability (H5).

6.1. Theoretical Contributions

First, the study integrates TAM, deterrence theory and RBV in a single model, revealing that technology’s compliance impact is fully mediated by behavioral change. This extends the VAT gap literature, which has predominantly modelled technology as an exogenous policy variable rather than an endogenous firm capability. Our empirical confirmation of this mediation via H3 constitutes a novel micro-level contribution to RBV logic, in that FinTech’s value stems from its behavioral translation into routine tax compliance. Second, by identifying tax morale and perceived audit probability as statistically significant moderators, the analysis supplies empirical support for hybrid “slippery-slope” perspectives of taxation, whereby enforced compliance (audit risk) and voluntary compliance (morale) interact rather than substitute. Third, the full mediation result refines RBV logic; FinTech creates value only when converted into superior routines—here, accurate VAT reporting—before performance gains emerge. This integration and conditional pathway have not previously been empirically modeled within the tourism SME context, thus marking a contribution to both tax compliance theory and FinTech adoption studies.

6.2. Policy Implications

For Greek revenue authorities, the magnitude of the FinTech → compliance path (β = 0.56) implies that every incremental push toward e-invoicing can yield immediate revenue dividends. Yet digital subsidies alone will not maximize returns unless accompanied by morale-building measures—transparent reinvestment of tourism taxes into local infrastructure, visible enforcement against non-compliant peers, and simplified VAT guidance. The audit probability moderator further suggests that publicly communicating inspection success stories in the tourism provinces could magnify the deterrent effect of the existing inspection workforce without necessarily increasing audit frequency. Finally, linking bank credit scoring to myDATA feeds could tighten the virtuous circle; compliant firms obtain cheaper finance, boosting the opportunity cost of evasion. Thus, policymakers should consider a three-pronged strategy: (i) incentivizing FinTech adoption via targeted subsidies, (ii) strengthening civic engagement by showcasing tax reinvestment outcomes, and (iii) deploying “soft” deterrents like public audit transparency instead of punitive surveillance. These combined measures would not only boost compliance but also foster a culture of tax legitimacy in tourism-intensive regions.

6.3. Managerial Implications

Tourism entrepreneurs often perceive VAT compliance as a pure cost; the results show it also confers strategic benefits. Transparent digital records shortened loan approval times and reduced supplier-credit friction, ultimately lifting profit margins. Owners therefore should treat FinTech not as a regulatory obligation but as an asset that unlocks operational efficiencies and reputational capital. Industry associations can facilitate this by offering peer-benchmark dashboards that visualize how compliance correlates with occupancy rates and review scores, reframing tax honesty as a competitive signal to eco- and ethics-minded travellers. Furthermore, the study reveals that SMEs with high tax morale and FinTech adoption realize the greatest performance improvements. This suggests that owner training programs should emphasize not only digital skills but also the long-term reputational advantages of transparent financial behavior.

6.4. Academic and Sector-Specific Insights

The study demonstrates that sector context matters; tourism’s historically cash-heavy model makes it especially responsive to real-time digital monitoring. Future VAT gap research might replicate the framework in retail or ride-sharing platforms to test whether the FinTech–morale–audit triad generalizes across service verticals. Moreover, the fsQCA sufficiency analysis revealed two configurations for high compliance, one technology-dominant and one morale-dominant, echoing Zídková’s [1] call for multi-factor explanations beyond tax rate monotonicity. These results suggest that there is no singular path to compliance. This has important implications for the design of policy interventions, which should be customized to SME profiles—some of which benefit more from behavioral interventions, others from technological enablement.

Overall, the study provides granular evidence that digital finance tools, when embedded within a supportive behavioral environment, can shrink the VAT gap that macro-level studies identified over a decade ago. Our results lend empirical support to the “slippery slope” framework of tax compliance, emphasizing the complementary roles of voluntary (tax morale) and enforced compliance (audit probability). Specifically, the effectiveness of technological interventions such as FinTech depends significantly on behavioral motivations—high tax morale nearly doubles compliance gains, and perceived audit probability reinforces compliance outcomes. Thus, this study confirms the theoretical validity of the slippery slope model at the SME level and highlights practical policy implications; effective tax governance requires balancing enforcement rigor with efforts to strengthen intrinsic taxpayer motivation. By analyzing both linear and configurational models, this study delivers not only statistical significance but practical relevance. For Greece’s tourism SMEs, FinTech adoption is not merely a compliance aid but a pathway to sustainable, legitimacy-driven growth.

7. Conclusions

This paper set out to determine whether FinTech adoption can close the VAT gap for Greek tourism-sector SMEs and to explain how behavioral factors condition technology’s effectiveness. Anchoring the inquiry in TAM, deterrence theory and the RBV, we surveyed 214 accommodation and service firms across Greece’s main tourism regions and tested a mediated-moderated structural model. The evidence is unambiguous; FinTech tools such as certified e-invoicing, mobile POS and myDATA integration sharply raise VAT compliance behavior, and that behavioral improvement fully explains the technology’s positive impact on financial performance. Moreover, the strength of each link hinges on two human factors—tax morale and perceived audit probability—confirming that digital transformation achieves little without complementary motivational and enforcement climates.

These insights advance VAT gap scholarship in three ways. First, they operationalize FinTech as an endogenous firm capability, moving beyond earlier macro studies that treated technology as a binary policy variable. Second, they demonstrate that compliance is not merely a cost but a value-creating routine, mediating the route from technological investment to profit. Third, they supply empirical support for the “slippery-slope” proposition that enforced and voluntary compliance are mutually reinforcing, not substitutive.

In doing so, this study proposes a behaviorally sensitive and sector-aware FinTech compliance framework that has not been previously tested in tourism-related VAT contexts, offering a novel lens to assess the intersection of technology, regulation, and fiscal culture.

For policymakers, the findings offer a pragmatic roadmap; subsidies for e-invoicing hardware, streamlined onboarding to myDATA, publicized audit results and visible reinvestment of tourism VAT into local amenities can create a virtuous circle in which higher morale and credible enforcement multiply the returns to digital adoption. To translate these proposals into concrete interventions, authorities could implement incentive schemes with performance-based subsidies, launch digital education campaigns through industry chambers, and establish compliance benchmarking portals that allow SMEs to track performance against regional peers. Such efforts would not only improve VAT collection but also stimulate broader financial inclusion and creditworthiness among micro-entrepreneurs in the tourism sector.

For SME owners, transparent digital reporting emerges as a strategic asset—unlocking cheaper credit, smoother supplier relationships and reputational trust among increasingly sustainability-minded travelers.

The empirical model thus illustrates how perceived behavioral gains and risk mitigation—rather than regulatory compliance alone—can drive voluntary FinTech adoption in low-trust, cash-heavy economies.

Naturally, the study has limitations. Self-reported compliance, although validated by procedural and statistical remedies, could still suffer from desirability bias; merging survey responses with objective audit data would strengthen causality claims. Cross-sectional data capture a single point in Greece’s staggered e-invoicing rollout; longitudinal designs could trace whether compliance gains persist once novelty fades or regulations tighten.

To overcome these limitations, future research could adopt panel data approaches using administrative audit trails, incorporate behavioral experiments that isolate psychological triggers, and apply hybrid qualitative–quantitative designs to better capture SME decision rationales. Finally, tourism is only one cash-intensive sector; replicating the model in retail, ride-hailing or freelance platforms would test its external validity. Other sectors such as agritourism, short-term rentals, street vending, and creative gig work offer fertile ground for extending this model, particularly where informal transactions coexist with seasonal or location-specific demand.

In sum, this study provides granular, policy-relevant evidence that bridges digital innovation and behavioral tax compliance. It offers a transferable model that connects technological infrastructure to fiscal governance through the mediating force of SME behavior and the moderating influence of social norms and enforcement credibility. Future research and practice can build on this framework to reimagine how digital tools can support inclusive and integrity-driven entrepreneurship—not just in tourism but across the broader ecosystem of informal or partially digitized sectors.