Determinants of Tax Avoidance Intentions in Tourism SMEs: The Mediating Role of Coercive Power, Digital Transformation, and the Moderating Effect of CSR

Abstract

Tax compliance and avoidance are critical issues for governments and businesses worldwide, especially as businesses often use legal methods to minimize taxes, which can impact public revenue and equity within the tax system. This study focuses on understanding the factors influencing tax avoidance behaviors among SMEs in Greece’s tourism sector, a sector that has received limited research attention. To this end, a quantitative cross-sectional design was employed, using a structured questionnaire to explore potential factors influencing tax avoidance behavior. Data were collected from 534 SME managers and analyzed using Structural Equation Modeling (SEM) to assess the impact of key factors and their interrelationships, including coercive power, digital transformation, tax knowledge, firm performance, and perceived fairness, on tax avoidance. In addition, corporate social responsibility (CSR) was included as a moderator variable, while coercive power and digital transformation were assessed as mediators. Furthermore, Multi-Group Analysis (MGA) was conducted to explore the differences between small and medium enterprises, as well as different ownership structures. The results indicate that all key determinants, except perceived fairness, are significantly and positively related to tax avoidance intention. Additionally, it was revealed that coercive power increases tax avoidance through firm performance and tax knowledge, while digital transformation mediates the influence of firm performance on tax avoidance by curtailing avoidance intentions. While CSR mitigates the negative influence of coercive power, digital transformation has a dual role: that of promoting transparency and strategic efforts to reduce the tax burden. These findings have important policy implications, as policymakers seek to promote digital adoption and enhance CSR engagement while formulating specific regulatory strategies to reduce tax avoidance among SMEs.

1. Introduction

In the international literature, tax avoidance is characterized as a legal activity that does not follow the principal purpose of the tax system and is operated contrary to its primary intent [1,2,3,4]. Indicatively, the phenomenon of tax evasion refers to practices such as not issuing documents when making sales, issuing fictitious invoices, and misrepresenting assets in the respective tax returns. Conversely, tax avoidance refers to practices such as transferring intra-group profits to tax havens or structuring business operations to minimize tax liabilities [5,6]. Tax avoidance centers on legislative “ambiguities” and tax-specific arrangements that offer the opportunity for taxpayers to exploit them while engaging in aggressive tax planning, with the intention of progressively reducing tax burden [7,8]. While both aim to reduce tax liabilities, they have the potential to effectively violate ethical and moral principles but not the letter of the law, while tax evasion violates both. Scholars have examined tax evasion and tax avoidance from a range of perspectives, including legal, ethical, and economic [4,9,10]. Furthermore, as a fine line can distinguish the two practices, what constitutes legal or ethical behavior can be challenging [11,12]. However, despite the controversy over whether tax avoidance is illegal, both practices have detrimental implications and outcomes on public revenue and economic fairness [4].

The phenomena of tax evasion, tax avoidance, and money laundering are long-standing problems that plague, among others, Greek society [13,14]. Addressing such problems requires the identification of the factors that enhance tax awareness and thus tax compliance within the existing legal and regulatory framework. Existing methods of tax enforcement derived from classical economic theories often fail to explain low compliance rates, thus leading to questions regarding their effectiveness [4,13,14,15]. There is general agreement that nations with lower levels of tax awareness have higher rates of tax evasion and/or avoidance. In fact, according to Bello [4], tax awareness has significant effects on taxpayer behavior and, consequently, on the informal economy. It has also been highlighted that tax consciousness depends, among other factors, on political and institutional settings, the state’s treatment of its citizens, their trust in the political and legal systems, and their degree of participation in fiscal decisions. The magnitude of the revenue foregone is determined according to the economic theory of criminal behavior based on the degree of deterrence of tax evasion as a product of the probability of detecting offenders and the size of the fine imposed [4,16]. Considering the importance of tax awareness, several studies have focused on the factors that shape both the establishment and maintenance of tax literacy. According to Hamid et al. [17], in achieving high levels of tax awareness, institutional and cultural factors are at least as influential as financial incentives [15,18,19,20,21]. Considering the relevance and nature of tax awareness among citizens, the role of tax education in relation to current laws, perceived fairness, and perceived risk emerges as extremely crucial in shaping and maintaining these desired attitudes.

In the context of Greece, tax avoidance poses specific challenges for the tourism sector, which plays a vital role in the economy [9,22,23]. Heavy reliance on cash-based settlements and package deals makes businesses more vulnerable to the underreporting of revenues and the manipulation of financial statements for tax avoidance purposes. Other tourism businesses are organized in cooperation with foreign companies, thus enabling the concealment of revenues through bank accounts in offshore jurisdictions. In these situations, company managers could also be playing a crucial role in hiding or misrepresenting activities to avoid scrutiny by the tax authority [13,14,24]. Although fines are implemented by regulatory bodies in an attempt to prevent this activity and protect public revenue, numerous situations arise where the ineffectiveness of such measures is evident [23,25]. Less severe penalties are imposed if the detection is perceived as likely, undermining its deterrent effect and failing to secure consistent compliance. Beyond the economic penalties, tax avoidance can also bring reputational damage to tourism SMEs, affecting their standing with a variety of stakeholders and the wider community [26,27,28]. Understanding these dynamics is important to formulate effective policies that balance enforcement with support in order to guarantee compliance by tourism SMEs [9,13,14,22].

Our research examines the key drivers influencing tax avoidance intentions among SMEs in the tourism sector of Greece. This study primarily explores the impact of coercive power, digital transformation, tax knowledge, firm performance, and perceived fairness on tax avoidance behavior by drawing from the contemporary literature on tax compliance and organizational behavior [3,29]. A quantitative cross-sectional design allowed the detection of these tax avoidance determinants among the tourism sector SMEs and the assessment of how they shape compliance intentions based on a sample of 584 SME managers. To the best of our knowledge, no previous study has specifically examined the key determinants of tax avoidance intention among tourism sector SMEs, especially those operating in Greece. While there have been various studies on general determinants of tax avoidance behavior in other industries, the peculiar characteristics and difficulties faced by tourism SMEs, such as seasonality, intense cash flows, and pressures due to regulatory requirements, have not been adequately addressed in the existing literature [14,22]. Furthermore, the literature concerning digital transformation and corporate social responsibility (CSR) and their functions in tax compliance is still underdeveloped with regards to tourism-based SMEs. Thereby, the current literature is not only limited but also inconclusive [14,22]. It often fails to provide a comprehensive understanding of how internal and external factors influence tax avoidance behavior in the given sector. To address this gap, the current study attempts to develop a fine-grained understanding of tax compliance in tourism SMEs and provides actionable insights for policymakers and business managers. The present study, therefore, surpasses the identification of subtle roles that coercive power and digital transformation can play as mediating factors in shaping the behavior of tax avoidance and further investigates how CSR can act as a moderator by showing how different levels of CSR practices can moderate the adverse impact of regulatory pressures on avoidance intentions. By integrating these dimensions, the research develops a comprehensive understanding of the interrelationship between technological advancement and ethical business practices in the way tax behavior occurs among tourism SMEs. Thus, such an integral approach ensures that not only are the gaps in the literature being addressed, but substantive insights are also derived, which would help balance enforcement mechanisms with support mechanisms for developing a resilient and compliant tourism sector.

The results indicate that coercive power has a significant influence on tax avoidance in medium-sized enterprises and sole proprietorships, hence suggesting the need for selective regulatory policy measures. The findings also reveal that digital transformation has a significant mediation on the relationship between firm performance and tax avoidance, hence reducing the intentions of tax evasion, especially in partnership and medium-sized firms, underscoring the importance of digital tools for ensuring compliance. It can be observed that CSR weakens the influence of coercive power on tax avoidance, as firms with strong CSR have less involvement in avoidance activities; thus, CSR is an effective tool for internal governance.

The article is structured as follows: Section 2 presents an overview of the relevant research on the factors that influence tax avoidance intentions, as well as the intricacies and reasoning behind SME tax avoidance, specifically in the tourism sector. We focus our interest on how SME managers/owners perceive current tax laws and legislation by studying the impact and magnitude of tax evasion, as well as the developmental dynamics that lead to avoidance behaviors among tourism SMEs. Section 3 contains a detailed description of the model we developed and tested for our research purposes, followed by a data analysis in Section 4. Section 5 examines and interprets the main findings and indicates relevant limitations. Finally, Section 6 and Section 7 conclude the article and make recommendations for policymakers, managers, and further research.

2. Literature Review

2.1. Determinants of Tax Avoidance

In a recent systemic literature review, Sritharan et al. [30] highlighted research gaps in the tax avoidance literature, including the limited studies in informal institutional environments and a lack of comparative studies. Over the years, research has extended beyond the traditional methods of detecting and identifying tax avoidance through economic measures, but contemporary attempts have focused their attention on behavioral patterns and influences. Fuadah, Dewi, Mukhtaruddin, Kalsum, and Arisman [2,29] explored the influence of sustainability practices, e-commerce, and organizational culture on tax avoidance, highlighting context-specific dynamics in countries such as Indonesia. The findings revealed that political costs, monitoring mechanisms, financing decisions, and financial distress significantly influenced tax avoidance under stable conditions. During the pandemic, only political costs and financial distress were impactful. No mediating effects were found in these relationships [29]. In an earlier study, Zhang et al. [31] investigated the impact of corporate tax avoidance on the financial performance of firms in China, focusing on the unique context of its economic reforms. Using structural equation modeling (SEM), the research revealed a significant, negative direct relationship between tax avoidance and market value, suggesting that the opacity of China’s stock market allows managers to exploit tax avoidance for rent-seeking activities, ultimately harming shareholders’ value.

Adam et al. [32] investigated the impact of e-commerce tax awareness and technology optimism on tax compliance intentions among university students in Nigeria. Despite the significant growth in e-commerce transactions and the potential for substantial tax revenue, tax compliance remains low, partly due to the complexity of legal tax provisions, limited tax awareness, and varying levels of technology optimism. Additionally, Li, Al-Sulaiti, Dongling, Abbas, and Al-Sulaiti [5] investigated how technology-driven employee behavior and corporate social responsibility (CSR) practices influence tax avoidance and firm performance in small and medium enterprises (SMEs) in Pakistan. It specifically examines the moderating role of CSR sustainable practices on the relationship between employee behavior and tax avoidance to achieve sustainable business performance. The study finds that tax avoidance, employee behavior, and CSR positively impact firm performance. In addition, the study by Ayuba et al. [33] of tax literacy demonstrated its role in building trust in tax authorities, which in turn supports compliance behaviors among SMEs.

This study extends these insights into addressing, in particular, how coercive power, digital transformation, and CSR bear on the intentions of Greek tourism SMEs to avoid taxes. While previous studies have considered factors such as e-commerce and tax literacy across different regions, our investigation focuses on an industry typified by a high cash flow and seasonal dynamics, a perspective that deepens the insight into the Greek context. It also attempts to fill the gaps in the existing literature that relate to the application of integrated modeling approaches, such as SEM, which test direct and mediating influences of digital transformation and hence provide fresh insights into sector-specific tax compliance challenges.

2.2. Behavioral Factors Influencing Tax Compliance and Behavior

Bornman and Ramutumbu [34] explored the relationship between tax knowledge and tax compliance, particularly among small business owners. They emphasized that overall education levels, alongside specific tax education—including technical knowledge and awareness of compliance motivations—are crucial for enhancing voluntary compliance. In addition, Bhalla et al. [35] examined the impact of tax knowledge on the performance of Indian micro, small, and medium enterprises (MSMEs) based on a primary survey of 450 registered businesses. The findings suggested that greater tax knowledge can lead to timely compliance and reduce tax evasion, avoidance, and scams, offering valuable insights for policymakers, governments, and businesses. The crucial element of VAT evasion was also studied by Lutfi et al. [36], who investigated how socio-economic factors influence VAT compliance. The results showed that all proposed factors significantly impact VAT compliance, with tax knowledge playing a crucial moderating role. These studies emphasize the importance of tax knowledge, social values, and fairness in improving compliance, suggesting that policymakers should focus on educational and awareness programs to enhance tax knowledge.

Similarly, Faridy et al. [37] explored the relationship between VAT complexity, compliance costs, and non-compliance among small and medium enterprises (SMEs) in Bangladesh. It was the first study to estimate VAT compliance costs in this sector using a large sample of non-complying SMEs. The study highlights that VAT law complexity and high compliance costs contribute significantly to non-compliance. The findings offer valuable insights for policymakers and tax authorities, suggesting the need for a simplified VAT system to reduce compliance burdens and improve voluntary compliance in Bangladesh and other developing countries. Nyantakyi et al. [38] and Appiah et al. [39] highlighted that tax knowledge and trust in tax authorities play a critical role in promoting compliance among SMEs. They found that social factors, such as perceptions of fairness and trust in government, significantly impact voluntary compliance. The studies of Trifan et al. [40] and Alsyouf et al. [41] similarly emphasized the importance of tax system fairness and trust in authorities, demonstrating that these factors are crucial in shaping compliance behavior. Rosalita Agusti and Rahman [42] identified tax literacy as being key in shaping trust and perceptions of authority, helping SMEs understand the tax system and government privileges. Trust in tax authorities significantly boosts voluntary compliance and reduces tax evasion, while the power of authorities alone has little impact on evasion, being seen as an obligation rather than as legitimate.

Furthermore, studies such as those by Zhang, Cheong, and Rasiah [31] and Mu et al. [43] highlight the role of tax education and technology in improving compliance. These studies showed that advancements in technology, such as e-filing, can simplify tax processes, though psychological factors may moderate these effects. In another study, Moisescu et al. [44] analyzed corporate fairness concerning public authorities in Romania. The results showed that the perception of fairness significantly enhances customer loyalty across key industries, a finding that underscores the importance of transparency in corporate social responsibility communication behaviors. Finally, Allam et al. [45] assessed IEQ in the European Union, which reported that countries with specific aspects of culture, in terms of high power distance, high uncertainty avoidance, and high collectivism, have higher levels of tax evasion. However, an increase in institutional quality, for instance, the rule of law or government effectiveness, could offset the negative influences of culture on tax evasion.

In the context of tax compliance, Mbilla et al. [46] investigated how social factors, culture, tax education, attitude, and equity impact tax compliance behavior among SMEs in Ghana. The findings show that these social drivers significantly influence tax compliance and highlight that tax compliance in Ghana has a strong social dimension. In another attempt, Bani-Khalid et al. [47] applied the extended Theory of Planned Behavior (TPB) to assess the intentions of owner-managers in Jordanian manufacturing SMEs regarding sales tax compliance. The results showed that attitudes towards behavior, subjective norms, perceived behavioral control, and patriotism significantly influence intentions to comply with sales tax. In addition, Hauptman et al. [48] evaluated tax knowledge and perceptions of tax fairness among Slovene taxpayers, focusing on gender and settlement size differences, revealing disparities in tax knowledge between male and female taxpayers, while settlement size does not significantly affect tax knowledge. In another attempt, Utama et al. [49] investigated the influence of religiosity, perceived risk, and attitude on tax-compliant intention, considering e-filing as a moderating factor. The findings revealed that both religiosity and perceived risk significantly shape attitudes towards tax compliance, which in turn positively affect tax-compliant intention. On the other hand, Anisykurlillah et al. [50] indicated that legitimate power, tax morale, and religiosity positively influence tax compliance, whereas coercive power and perceived fairness do not. Additionally, tax literacy was found to moderate the relationship between perceived fairness and tax compliance, reducing the influence of perceived fairness. In the study by Amani [51], the objective of this study was to examine the impact of corporate legitimacy on tax compliance intentions among small- and medium-sized enterprises (SMEs) in Tanzania. This research uniquely addressed the issue of tax avoidance and leakages in Sub-Saharan Africa by focusing on corporate legitimacy.

While past studies have identified factors such as tax knowledge, socio-economic conditions, and trust as influences on tax compliance, the present study shifts attention to understanding avoidance behavior in the specific context of Greek tourism SMEs. Thus, the limited literature has taken into consideration peculiar characteristics, including the seasonal nature of tourism businesses and their heavy reliance on cash flow transactions [13,14,22]. In doing so, we provide novel insights into how such diverse factors influence tax avoidance intentions and offer practical implications for policymakers and industry stakeholders alike.

2.3. Risk of Detection and Tax Penalties

A study from 2002 explored the interplay between personal ethics, social norms, and deterrence in tax compliance, suggesting that strong personal ethics reduce the need for deterrence measures such as legal sanctions [52]. In the context of detection, a study examined the impact of tax system components—the tax penalty, the tax rate, and the tax audit—on tax compliance behavior among small- and medium-sized enterprises (SMEs) in Yemen’s manufacturing sector, considering the tax compliance cost a mediating factor [53]. The findings reveal a strong positive relationship between the tax rate, the tax penalty, and the tax audit with tax compliance behavior. In contrast, tax compliance costs negatively affect tax compliance. Additionally, tax compliance costs mediate the relationship between the tax rate and the tax penalty with the tax compliance behavior but do not mediate the relationship between the tax audit and tax compliance behavior [53].

Both Lima, Cunha, and Nascimento [12] and Nkundabanyanga et al. [54] emphasized the role of institutional pressures, such as the risk of detection, the severity of punishments, and transparency, in shaping tax morale and compliance behaviors. On the contrary, Kayaoglu and Wıllıams [8] challenged the conventional view of tax non-compliance as solely a rational economic decision, proposing instead that tax morale plays a crucial role. Using neo-institutionalist theory, they examined how citizens’ behaviors are influenced by the normative, cultural-cognitive, and regulatory aspects of their institutional environment. The study found that higher tax morale is associated with trust in government, a sense of national belonging, and perceptions of risk and punishment severity. The findings underscore the significance of tax morale in understanding and addressing tax non-compliance in Turkey. Ya’u et al. [55] examined how detection probability and penalties influence corporate tax evasion, highlighting challenges in multinational contexts. The findings indicated that the tax rate, detection probability, penalties, and environmental regulation significantly influence corporate tax evasion. Eneh et al. [56] focused on tax compliance among SMEs in Bayelsa State, Nigeria, based on tax simplicity, service quality, and penalties, indicating that tax simplicity and service quality significantly enhance tax compliance, while penalties have less impact. The study recommended simplifying the tax system and increasing penalties to improve compliance among SMEs. Another interesting study from de Sousa and Rezende [57] evaluated the impact of Brazil’s SPED (Public Digital Bookkeeping System) on the ICMS (Tax on Circulation of Goods and Services) tax gap from the perspective of tax auditors. The research found that SPED significantly influences the ICMS tax gap by enhancing inspections, improving access to taxpayer information, and affecting taxpayers’ decisions regarding withholding amounts. The findings support the need for policies that optimize tax collection and address the determinants of the tax gap, aiding legislators and tax authorities in their efforts.

In this regard, our research tries to fill this gap by investigating how perceptions of risk detection and penalties influence the propensity for tax avoidance behavior in Greek tourism SMEs. In this way, taking into consideration those peculiar regulatory and economic features that characterize Greece will provide a deeper understanding of more focused compliance strategies for this sector [9,14,22].

2.4. The Mediating Roles of Coercive Power, Enforcement, and Digital Transformation

A study by Al-Rahamneh, Al Zobi, and Bidin [11] examined how tax transparency affects sales tax evasion among Jordanian SMEs, emphasizing the role of moral obligation while increased tax transparency significantly reduces sales tax evasion. The findings suggest that integrating tax transparency and moral obligation into tax policies can help Jordanian authorities effectively combat tax evasion, improving overall revenue collection. Djajanti [21] challenged the assumption that the fairness of the tax system automatically improves compliance, showing that it can have a negative effect in this context. By demonstrating that the power of tax authorities not only influences enforced compliance but also voluntary compliance, the study provided new insights for tax policy and administration, particularly in regions with strong tax revenue potential, such as Jakarta. In 2016, Ayuba, Saad, and Ariffin [33] studied the influence of economic and psychological factors on tax compliance among Nigerian SMEs. Involving 321 SME owners/managers, the findings revealed that the probability of detection, incentives, and public governance quality significantly enhance tax compliance, while tax complexity negatively affects it.

In regard to digital transformation, Lutfi et al. [58] and Bugarčić and Slavković [59] investigated and highlighted the critical role of digital readiness in SMEs, showing that factors such as organizational readiness and management support drive the adoption of technologies such as big data and digital project management tools. In addition, Yakubu et al. [60] examined the factors influencing electronic tax compliance among small- and medium-sized enterprises (SMEs) in Vietnam, highlighting its significance in the country’s tax reform strategy. Data from 402 SMEs were analyzed, identifying four key factors: Taxpayer Awareness (TA), Perceived Ease of Use (PTE), Vietnamese Tax Administration (VTA), and Efficiency of Vietnamese Tax Policy (VTP). Taxpayer awareness was found to be the most significant factor. The study recommended that the Vietnamese government enhance SMEs’ understanding of tax obligations through workshops and training, providing insights for policymakers and practitioners in tax reform efforts.

Another study in Malaysia focused on the antecedents of cashless payment systems among businesses in the country, identifying key factors influencing their adoption, which could reduce cash handling costs and enhance transaction speeds. Agusti and Rahman [42] found that technology competence is key to adopting digital payment systems among SMEs. Finally, a study by Night and Bananuka [61] explored how adopting an electronic tax system mediates the relationship between attitudes toward the system and tax compliance among Small Business Enterprises (SBEs) in a developing country. A survey of 214 owner-managed SBEs revealed that the adoption of the electronic tax system partially mediates this relationship, suggesting that authorities should promote the benefits of electronic tax systems to improve taxpayer attitudes and compliance.

While there is related literature on how enforcement and digital transformation influence tax compliance either in general or with an emphasis on specific technologies, our study further develops this by establishing the mediating role played by enforcement and digital readiness in shaping the behaviors of tax avoidance within Greece’s tourism SMEs [9,14,22]. The approach gives focused insights into the level of balance between enforcement and digital adaptation that should improve compliance and allows subtle guidance for policy-making.

2.5. Corporate Social Responsibility in the Context of Tax Avoidance

Ghadakforoushan et al. [62], analyzed CSR’s role in reducing tax avoidance behaviors, mediated by ethical leadership. The analysis revealed that CSR significantly influences tax avoidance, and work meaningfulness strengthens the negative relationship between CSR and tax avoidance with ethical leadership further enhances this dynamic. Abed et al. [63] discussed the challenges of transparency in financial reporting. Through a systematic literature review, the research established a framework that connects corporate governance to the quality of financial reporting. The study advocated for further empirical research to measure the implications of these determinants on financial practices, aiming to improve transparency and accountability in commercial banks. Farooq et al. [64] investigated how CSR activities can enhance transparency and accountability. The findings reveal that CSR positively affects both organizational identification and knowledge sharing, while the effectiveness of CSR activities varies based on individual employee characteristics, highlighting the need for tailored CSR strategies within organizations. So et al. [65] explored factors influencing sustainability reporting in corporate contexts. Using data from the Indonesia Shariah Stock Index (ISSI) for 2019, the findings revealed that HG and ICG significantly influence sustainability reporting disclosure (SR), with firm size and leverage also being relevant factors. Profitability is not significantly related to SR, and ICG has a negative impact on SR. ITU is significant only when HG is not present. The study concluded that human governance is the most important predictor of sustainability reporting disclosure.

2.6. The Case of the Tourism Sector and Greece

Greek tourism has a decisive competitive advantage over other countries due to the attractiveness and uniqueness of its tourism industry [23,24,25]. Greece is widely known and valued as an extremely popular and safe destination of international importance. The reason for this consists of the state’s cultural wealth, historic heritage, scenic beauty, and variety of the natural environment. The comparative advantages of Greece attribute specific and high rankings in the international ranking of tourist destinations [13,14,22]. However, it also provides the opportunity to develop and expand the portfolio from the mass tourism model into specialized and innovative forms of tourism. Tourism is an economic sector integrated with many interconnections with various aspects of productive and social life [14,23,26]. The rapid development of the tourism sector and its role in the national economy has prompted public authorities to investigate how to best approach the economic dimension of tourism and its impact on public finances. By extension, the national government aims to boost public revenues through tourism in order to cover any shortfalls in national budgets, to make corrective market moves, and enhance tourism and public infrastructure. Therefore, any taxation of tourism activity needs to be guided by the principles of efficiency, stability, simplicity, and effectiveness, and to be subject to considerable scrutiny both at the time of its introduction and during its implementation [14,22,26].

In Greece, the tax system in the tourism sector is based on direct and indirect taxation [66]. The burdens on the tourism market and the tourism industry are divided into taxes and fees. The first category of taxes includes taxes paid directly by the tourist, shown on the invoices they receive, while the second category includes taxes on hotel income, such as VAT, and taxes on profits. In addition, there are also the fees that airlines are charged for the use of airports and the services they provide to the consumer [66,67,68]. The excessive tax burden, both in terms of direct and indirect taxes, continues to negatively affect the growth of tourism and the economy in general. The significant competitiveness gains of Greek tourism due to internal devaluation are significantly offset by this excessive tax burden. In general, Greek tourism enterprises are taxed significantly compared to other countries. Despite the reduction of the tax rate from 29% to 24%, Greece still maintains one of the highest corporate tax rates in Europe, which affects the ability of Greek tourism enterprises to compete internationally [13,14,24].

The recent literature on the tourism and hospitality industry supports a myriad of factors affecting tax compliance, performance, and strategic adaptation. Amani [51] addressed the issue of tax compliance intentions amongst Tanzanian tourism SMEs, corporate legitimacy, and public trust that deter tax avoidance. In the same way, how digital transformation in tourism enterprises, especially those in China, could be utilized to enhance internal controls and reduce aggressive tax behavior has been explored in studies such as that by Tiantian et al. [69]. Chen [70] analyzed the effect of firm characteristics, including leverage, on tax avoidance among Indonesian tourism and hospitality firms, with a special focus on the mediating roles of the firm size and the financial strategy, which proved to be significant. Cao et al. [71] pointed out that the use of tax haven strategies could turn out to be counterproductive to tourism development by citing, as an example, the situation in Panama’s tourism sector. In particular, Widuri et al. [72] examined how board gender diversity and sustainability initiatives affect Indonesian and Malaysian hospitality firms’ tax compliance, indicating that diversity indeed correlates positively with tax compliance. Khelil and Khlif [73] discussed a cross-country comparison of tax avoidance behavior in family-owned tourism firms, considering the effects of corporate transparency and the regulatory environment. Stiglingh et al. [74] researched tax global transparency and its influence on compliance improvement in multinational hospitality companies. Kustono, Effendi, and Pratiwi [29] discussed political costs and financial distress as the most relevant factors driving tax strategies in the hospitality sector by presenting an amalgam of patterns in stable and pandemic periods. These were intensified in the events of the COVID-19 pandemic due to serious financial constraints and, ultimately, tourism enterprises attempting to reduce costs. The economic effect of the pandemic on Greek tourism fell within worldwide trends [13,14]. Abbas et al. [75] discussed how the COVID-19 pandemic severely shook the tourism and hospitality industry, underscoring the importance of corporate social responsibility and innovation in its recovery process. Collectively, these studies signal the interactive dynamics between regulatory frameworks, corporate characteristics, and strategic adaptations in tourism and hospitality.

In light of the above, it is evident that the phenomenon of tax avoidance is to a substantial extent detected in Greek tourism enterprises, which implement various methods to reduce the tax burden. In recent years, in tandem with technological advances, means of tax evasion by businesses in tourism have also evolved [66,67]. The current literature suggests that, in view of the elasticity of demand in the sector, tourism taxes can be counterproductive, with possible negative impacts on tourist arrivals and the greater economy [13,14,15,20,68]. This is further exacerbated by an unduly high general tax burden, combined with insufficient mechanisms for providing tax incentives and high corporate rates, further constraining Greek tourism enterprises from reaching their full growth potential [13,14,22]. Many enterprises, therefore, have to employ tax evasion techniques such as the understatement of revenues, manipulation of the POS system to connect with a foreign bank, and not issuing proper receipts for tourism services, especially package deal services [13,14,22].

Recent attempts have identified several determinants and causal factors for the emergence and consolidation of tax avoidance in tourism businesses, such as the method of distributing the size of the tax burden and the possibility of tax injustices and inconsistencies in the tax system [13,20,68]. Apart from that, public distrust and corruption in the tax authorities, along with low levels of tax education among entrepreneurs, further aggravate the problem, considering that the government manages the public revenues and there is a high degree of corruption in the agencies and authorities of the tax control mechanism. To increase the levels of tax compliance and to minimize tax avoidance, sanctions, penalties, and fines are imposed by the administrative authorities [14,20,26]. The nature of the specific sets of tax sanctions applied depends on the respective set of compliance models adopted and on the motivational sources for the individuals to comply with the legislation. Greek authorities have applied deterrence models based on high levels of punishment as a means of imposing compliance [14,15,20]. However, such strategies are not enough in order for there to be a sustainable tax culture of trust-building, transparency, and improvement in the quality that governmental services can provide, according to [13,14,76]. It should be a combination of deterrence and measures to ensure that the public has a high level of trust as a means of reducing cases of tax avoidance and improving compliance within the Greek tourism sector.

Across all of these themes, the literature review identified a lack of sector-specific studies concerning tax avoidance in tourism and, more precisely, for Greek SMEs. While there exist various studies on tax avoidance and compliance issues, many have failed to address these specific problems of tourism SMEs, such as seasonality issues, high cash flow, and the scarcity of resources. The precise role of CSR and digital transformation as moderators or mediators of tax avoidance behavior has also not been fully explored, especially for small-scale enterprises operating within the rapidly changing digital environment. This work contributes, therefore, to the literature by addressing these gaps and providing specific guidance that could be used by policymakers and managers in their efforts to control tax avoidance in tourism SMEs.

Based on the insights gathered from the literature review, we highlighted and examined several influential factors of tax avoidance intention among SMEs, including firm performance, perceived fairness, the risk of detection, tax knowledge, coercive power, and digital transformation, all of which have shown varying levels of impact on tax avoidance behavior. Thus, we formulated the following hypotheses:

H1. 

Firm performance (FP) has a positive influence on tax avoidance intention (TAI).

H2. 

Perceived fairness (PF) has a positive influence on tax avoidance intention (TAI).

H3. 

Risk of detection (RD) has a positive influence on tax avoidance intention (TAI).

H4. 

Tax knowledge (TK) has a positive influence on tax avoidance intention (TAI).

H5a. 

Digital transformation (DT) has a negative influence on tax avoidance intention (TAI).

H5b. 

Coercive power (CP) has a positive influence on tax avoidance intention (TAI).

H6a. 

Coercive power (CP) mediates the relationship between firm performance (FP) and tax avoidance intention (TAI).

H6b. 

Digital transformation (DT) mediates the relationship between firm performance (FP) and tax avoidance intention (TAI).

H7a. 

Coercive power (CP) mediates the relationship between perceived fairness (PF) and tax avoidance intention (TAI).

H7b. 

Digital transformation (DT) mediates the relationship between perceived fairness (PF) and tax avoidance intention (TAI).

H8a. 

Coercive power (CP) mediates the relationship between risk of detection (RD) and tax avoidance intention (TAI).

H8b. 

Digital transformation (DT) mediates the relationship between risk of detection (RD) and tax avoidance intention (TAI).

H9a. 

Coercive power (CP) mediates the relationship between tax knowledge (TK) and tax avoidance intention (TAI).

H9b. 

Digital transformation (DT) mediates the relationship between tax knowledge (TK) and tax avoidance intention (TAI).

H10a. 

Corporate social responsibility (CSR) moderates the relationship between coercive power (CP) and tax avoidance intention (TAI).

H10b. 

Corporate social responsibility (CSR) moderates the relationship between digital transformation (DT) and tax avoidance intention (TAI).

3. Research Methodology

3.1. Conceptional Model and Rationale

Tourism is among the main industries in the Greek economy, but especially with regard to employment and its relation to the country’s GDP [13,14]. For tourism businesses of small and medium sizes, challenges such as company size quite often come into play, and usually, there are limited resources to adapt with the regulatory changes. Among some of the most pressing challenges facing such ventures are issues related to tax compliance, as the entangling factor of tax regulation, coupled with financial constraint, which could promote tax avoidance behavior [9,13,26]. Understanding the drivers of tax avoidance intention is not only important to policymakers who strive to ensure fairness in taxation but also to SME managers who have to adhere to the regulations, manage, and maintain sustainability at the same time. This research investigates the intentions of SMEs in Greek tourism to avoid taxes in order to establish what dictates their decisions and inform policies that will balance enforcement with support for SMEs.

This study investigates a number of crucial determinants, including firm performance, tax knowledge, risk of detection, and perceived fairness. Firm performance is normally related to sustainability or growth capability, which might have a potential influence on the tax decision in a situation when the enterprise suffers from financial burdens [2,5,35,43]. Tax knowledge refers to the awareness and understanding of tax regulations among the managers of SMEs, which may significantly affect compliance or avoidance behavior. A lack of knowledge about taxes may lead to unintentional non-compliance or create avenues to intentionally avoid it [39,48]. Detection risk is the perceived likelihood of being caught in tax avoidance and thus could deter individuals [16,49]. In Greece, this perceived risk from detection is an important factor owing to increased scrutiny and enforcement by tax authorities. Perceived fairness is defined as how SME managers perceive the equity of the tax system, particularly about the horizontal distribution of taxes within economic sectors and social classes [21,44,48]. To that extent, if taxes were perceived to be unfair, then managers would be more likely to engage in avoidance.

This research additionally examines the mediation of digital transformation and coercive power [58,59,77,78]. In the context of the tax landscape, this has become increasingly relevant. Many countries have mandated organizations to transition to a digitized system for tax reporting [59,78]. Greece is incentivizing a move towards the digitalization of its tax administration through the connecting of companies to its newly introduced myDATA [79], a platform that enables uniformity in digital accounting and tax applications. However, the full extent of such changes implemented by tourism SMEs remains underexplored, as is how digital transformation influences their behavior regarding tax compliance and avoidance intentions [13,14]. Coercive power refers to the pressuring effect of regulatory authorities via penalties and enforcement [21,50]. The Greek government has been tightening its policies to reduce the evasion of taxes, especially in industries involving a lot of cash transactions, such as tourism. This study investigates the mediation role of digital transformation and coercive power between those independent variables and tax avoidance intention.

Finally, this research investigates the moderating role of CSR on tax avoidance intention. CSR is hereby defined as a voluntary commitment by a firm to ethical practices, social responsibility, and sustainability [30,44,51,64]. Even though CSR has traditionally been associated with large corporations, its applicability is increasingly gaining relevance among SMEs, especially in tourism, where the reputation and good relations a business builds with local communities often define its level of success. Thus, it presents an opportunity to examine whether fundamentally strong CSR practices by SMEs can reduce any intention to engage in tax avoidance, even in the presence of coercive pressures and digital demands [30,44,51,64].

In this study, the categorization of SMEs follows the European Union’s definition, also applied in Greece. According to the EU classification, microenterprises have less than 10 employees, small enterprises range from 10 to 49 and medium-sized enterprises range from 50 to 249 employees [80,81,82]. These categories are most relevant to this research as they set the basis for understanding the possible spectrum of compliance capacities and competencies that might be required for different types of SMEs in view of the use and adoption of digital tools. Firms were distinguished by workforce size so that each business type could be marked clearly and identically in light of similar research [11,33,47,62,63]. This will also allow the research to test whether smaller firms, with typically more restricted resources, have a greater propensity for tax avoidance intention relative to medium-sized firms. Workforce size has been of vital significance, as it represents one of the most important dimensions of the capacity of SMEs, which directly affects their digital readiness and the ability to implement CSR initiatives.

This proposed theoretical framework presents the different interactions and relationships of various factors that determine tax avoidance intention among tourism SMEs. While this includes well-established predictors such as firm performance, tax knowledge, risk of detection, and perceived fairness, digital transformation and coercive power are added as mediators. Moreover, the role of CSR will be examined as a moderating variable to represent its potential effect on these types of enterprises’ behavior regarding tax-related practices. Therefore, the proposed model, in conjunction with the selected analysis method of Structural Equation Modeling (SEM), allows for capturing both the direct and indirect influences on Greek tourism SMEs’ tax avoidance behavior and exploring casualities and interrelationships among latent and observed values. The conceptualized model is presented in Figure 1.

3.2. Data Collection and Sampling

To this end, this study employs a quantitative cross-sectional research design that investigates potential factors that influence tax avoidance intention among SMEs operating in the Greek tourism sector. A cross-sectional design was utilized as it enables the collection of data at one single point in time, hence the provision of an overview of the perceptions and behavioral factors of the SME managers related to tax avoidance [83,84,85]. The quantitative nature of the research ensures that the data collected are measurable and that the testing of hypotheses can be properly carried out, hence providing objective insight into the subject matter [83,84,85]. Since it is exploratory research regarding the relationship between more than one variable, a self-reporting instrument was chosen as an appropriate tool for data collection, allowing for a systematic investigation of the determinants of tax avoidance intention [86,87,88].

Data were collected from a sample of 534 SME managers operating in the Greek tourism sector. A non-probability convenience sampling technique was adopted for participant selection; while not random, this approach is efficient in a sector where managers are generally pressed for time [86,87,88]. Although this approach limits the generalization of findings to the wider population, it preserves the reliability and validity of the data as its assurances that the sample consists of individuals with varying levels of experience in managing finance and is specific to tax-related matters in organizations [89,90,91]. The main inclusion criterion was that participants engage in decision-making processes related to financial management, tax reporting, and compliance. Thus, it ensures that the respondents are knowledgeable enough and have the relevant experience in providing the researcher with dependable data concerning tax-related behaviors.

The questionnaire was developed and adapted based on already-established scales to fit the context of Greek tourism SMEs and included 20 items in total (Appendix A). This study’s instrument was divided into two broad sections: demographics and scale measurements. Scale measurements included sub-sets to acquire data on: the independent variables, firm performance, tax knowledge, risk of detection, and perceived fairness; the mediating variables related to digital transformation and coercive power; the moderating variable for CSR; and the dependent variable for tax avoidance intention. While the demographic data included related questions about managers/owners and SMEs, with the data being further analyzed to determine any significant differences based on the size of the SMEs, using the EU categorization of SMEs based on workforce size, and uncovering the differences between small- and medium-sized SMEs.

The questionnaire ranged from 1 to 5 on the Likert scale, where 1 indicated strong disagreement with the statement and 5 indicated strong agreement. In addition, all the constructs were measured in terms of the perceptions and behaviors of the respondents.

To check for clarity and relevance, a pilot test with a small number of tourism SME managers was carried out prior to its full deployment. Based on the comments, some minor adjustments were made that improved the specificity of the questions and overall instrument reliability for it to be culturally and thematically appropriate for the Greek tourism sector context. The questionnaires were distributed through online and in-person methods. Through e-mails and professional networks, the online questionnaire was sent to managers of SMEs across different regions in Greece, whereas the in-person interview was conducted in areas with a high concentration of tourism businesses. Following data collection, Structural Equation Modeling analysis was performed.

3.3. Measurement Scales

We adapted instruments from Farooq, Farooq, and Jasimuddin [64] to measure corporate social responsibility (CSR) using a 3-item scale. Items in the scale addressed a variety of aspects of CSR, including items relating to social responsibility, whereas other items focused on employee-related CSR. In addition, we included an item on charitable contributions and donations adapted from Farooq, Farooq, and Jasimuddin [64]. Firm performance was measured with an adapted 4-item scale from Abbas [92] that considered aspects such as financial growth, market share, and resource efficiency. We measured tax knowledge using a 4-item adapted scale based on Nyantakyi, Sarpong, Asiedu, Adjei Bimpeh, Kwasi Anenyah Ntoso, and Ofeibea Nunoo [38], which captured the awareness and understanding of the tax regulations by respondents. The risk of detection was based on an adapted 3-item instrument from Lima, Cunha, and Nascimento [12] and Kayaoglu and Wıllıams [8] that determined perceptions of supervision by the tax authority and the effectiveness of audits. For mediating variables, digital transformation was assessed with a 3-item scale developed by Slavković et al. [93], and coercive power was assessed using an adapted 4-item scale from Anisykurlillah, Sugiyat, and Mukhibad [50], which measured the extent of digital inclusion into enterprise processes and perceived regulatory pressure, respectively. Finally, tax avoidance intention was measured using a 3-item scale developed by Nyantakyi, Sarpong, Asiedu, Adjei Bimpeh, Kwasi Anenyah Ntoso, and Ofeibea Nunoo [38], which measured the intent of the respondents regarding their engaging in tax avoidance behaviors.

3.4. Sample Profile

This study used a sample made up of 534 SME managers in Greek tourism, as illustrated in

Table 1. Sample Profile.

		Frequency	Percentage
Gender	Male	275	51.5%
	Female	259	48.5%
Educational Level	High School	108	20.2%
	Bachelor’s Degree	244	45.7%
	Master’s Degree	122	22.8%
	PhD	60	11.2%
Firm Operation Years	Less than 5	168	31.5%
	5–10	221	41.4%
	More than 10	145	27.2%
Ownership Structure	Sole Proprietorship	138	25.8%
	Partnership	194	36.3%
	Family Business	202	37.8%
Total Employees	1–9	145	27.2%
	10–49	228	42.7%
	50–249	161	30.1%
SME Scale	Small	373	69.9%
	Medium	161	30.1%

. The gender distribution of male and female respondents was relatively balanced, with 51.5% male respondents and 48.5% female respondents. The largest group in terms of educational level was bachelor-degree holders, at 45.7%, followed by masters-degree holders, at 22.8%. High school graduates constituted up to 20.2% of the sample. A smaller portion of the sample held PhDs, at 11.2%. Specifically, the descriptive statistics regarding firm operation years indicated that 31.5% of the firms had been in operation for less than 5 years, 41.4% for between 5 and 10 years, while the remaining firms had been operating for more than 10 years. In addition, the firms belonged to the following ownership structures: sole proprietorships, at 25.8%; partnerships, at 36.3%; and family businesses, at 37.8%. While considering the number of employees, 27.2% of the firms had between 1–9 employees, 42.7% had between 10–49 employees, and 30.1% had between 50–249 employees. Lastly, the sample was predominantly composed of 69.9% small-sized enterprises, while 30.1% were medium-sized enterprises.

4. Data Analysis and Results

The analysis in the present study was conducted using Smart-PLS4 software, version 4.1.0.0, utilizing Structural Equation Modeling (SEM) as one of the main methodological approaches. In general, SEM is highly regarded as being able to perform effective variance-based analyses, especially in management and the social sciences, as noted by Nitzl et al. [94]. PLS-SEM was used based on the fact that it was originally designed to facilitate the analysis of causal modeling for maximum explained variance in dependent latent constructs. In addition, MGA was used to analyze any differences between the sub-groups, such as small-sized versus medium-sized SMEs, which is crucial for identifying whether certain relationships vary across different contexts. Such an approach allows for an analysis of heterogeneous effects that may not be captured through standard regression methodologies [95,96,97]. The methodological framework followed the criteria set by Wong [98] to accurately estimate the beta coefficients, standard errors, and measures of reliability. In tune with the reflective measurement model evaluation criteria, each of the indicators had to show their appropriate link with their respective latent construct and outer loadings in excess of 0.7 to satisfy the evaluative criteria.

4.1. Common Method Bias

To verify the validity and accuracy of the research, a methodical check was performed for common method bias following the established methodological framework by Podsakoff et al. [99]. More specifically, Harman’s single-factor test was performed to check whether CMB was present, particularly if one factor accounted for the majority of the variance in the model. The results of the unrotated Principal Factor Analysis indicated that the general factor explained 26.791% of the total variance, falling far below the critical threshold of 50%. Although common method bias was not an issue in this study, controlling for CMB is an important procedural step to ensure the validity of the relationships among the variables. It safeguards the findings from possible distortions or potential biases associated with the collection of data. Hence, the findings of this study can be interpreted more confidently, as the effect of CMB has been shown to not make much difference in the established relationships.

4.2. Measurement Model

The first stage of employing Partial Least Squares Structural Equation Modeling involves comprehensive measurement model testing. In this model, the constructs are quantified by reflective indicators. The assessment process includes an evaluation of crucial measurement metrics with composite reliability, indicator reliability, convergent validity, and discriminant validity, following the procedure specified by Hair et al. [100].

Indicator reliability is, according to Chin [101], a foundational step in the measurement model assessment that seeks to establish the proportion of variance in a variable attributable to the underlying construct it is purporting to measure. The magnitude of outer loadings primarily influences this assessment; these, according to Wong [98] and Chin [102], should be greater than 0.70. However, Vinzi et al. [103] observed that although the loadings above 0.7 are preferable, examples from social science studies have numerous outer loadings below this threshold value. Critical judgment on the deletion of the low-loading items has to be based on a decision related to the items’ contribution to composite reliability and convergent validity so as not to decide prematurely. Indicators with a loading between 0.40 and 0.70 could, according to Hair et al. [104], be deleted only if their exclusion significantly improves the composite reliability or the AVE of the respective construct.

Following the suggestions of Gefen and Straub [105], the optimization of the measurement model was achieved through the exclusion of two indicators, CP4 and TK4, as shown in

Table 2. Factor loading reliability and convergent validity.

Construct	Items	Factor Loadings	Cronbach’s Alpha	rho_A	CR	AVE
Coercive Power	CP1	0.910	0.887	0.888	0.93	0.815
	CP2	0.898
	CP3	0.901
Corporate Social Responsibility	CSR1	0.579	0.589	0.674	0.781	0.55
	CSR2	0.878
	CSR3	0.736
Digital Transformation	DT1	0.757	0.579	0.631	0.772	0.536
	DT2	0.833
	DT3	0.585
Firm Performance	FP1	0.776	0.795	0.796	0.867	0.619
	FP2	0.807
	FP3	0.791
	FP4	0.772
Perceived Fairness	PF1	0.832	0.766	0.787	0.863	0.678
	PF2	0.864
	PF3	0.771
Risk of Detection	RD1	0.865	0.903	0.902	0.939	0.838
	RD2	0.940
	RD3	0.939
Tax Avoidance Intention	TAI1	0.785	0.815	0.826	0.891	0.732
	TAI2	0.915
	TAI3	0.863
Tax Knowledge	TK1	0.929	0.859	0.871	0.915	0.782
	TK2	0.909
	TK3	0.811

, as they had a factor loading below 0.500.

In this research, reliability was assessed concerning Cronbach’s alpha, rho_A, and composite reliability as the most important measures. According to Wasko and Faraj [106], constructs CP, CSR, DT, FP, PF, RD, TK, and TAI are above the minimum threshold of 0.700, and the remaining constructs have a moderate to high degree of reliability, as mentioned in references [107,108,109]. The rho_A statistic, conceptually situated between the Cronbach’s alpha and composite reliability, in most cases also exceeded the 0.7 cut-off threshold, which was in line with the study by Sarstedt et al. [110], and thus supported the findings of Henseler et al. [111] regarding its reliability.

Convergent validity was acceptable as the average variance extracted (AVE) for most constructs exceeded the threshold of 0.50 recommended by Fornell and Larcker [112]. However, Fornell and Larcker [112] also suggested that constructs exhibit adequate convergent validity even when the AVE is less than 0.50, provided that composite reliability is greater than 0.60. Discriminant validity has been verified by comparing inter-construct correlations to the square root of the AVE, following the procedure by Fornell and Larcker [112], but also by the heterotrait-monotrait ratio, HTMT, as developed by Henseler, Hubona, and Ray [111]. As revealed in both

Table 3. HTMT ratio.

	CP	CSR	DT	FP	PF	RD	TAI	TK
CP
CSR	0.28
DT	0.427	0.363
FP	0.445	0.424	0.692
PF	0.074	0.062	0.131	0.1
RD	0.162	0.079	0.139	0.102	0.391
TAI	0.627	0.363	0.602	0.717	0.076	0.228
TK	0.449	0.276	0.641	0.688	0.034	0.036	0.571

and

Table 4. Fornell and Larcker criterion.

	CP	CSR	DT	FP	PF	RD	TAI	TK
CP	0.903
CSR	−0.205	0.741
DT	0.289	−0.231	0.732
FP	0.376	−0.306	0.513	0.787
PF	−0.063	−0.038	0.016	0.048	0.823
RD	0.145	−0.06	−0.02	0.087	0.321	0.915
TAI	0.533	−0.266	0.429	0.579	0.057	0.198	0.856
TK	0.396	−0.202	0.466	0.573	0.025	0.02	0.477	0.884

The bold values represent the square root of the AVE.

, all values were less than the strict threshold of 0.85, which means discriminant validity was established.

4.3. Structural Model

The testing of the structural model of the proposed research framework was carried out by evaluating the R² and Q² values, in addition to assessing the path coefficients’ significances [108]. Therefore, in this research, R² ranged between 0.384 for tax avoidance intention, 0.31 for digital transformation, and 0.22 for coercive power, confirming that they fell within the expected range between 0 and 1. Q² showed moderate to high predictive relevance of the model, with values of 0.322 for tax avoidance intention, 0.202 for coercive power, and 0.295 for digital transformation.

However, this model was further validated through hypothesis testing, which also confirmed the significance of the relationships among the constructs. The path coefficients were assessed through reliance on the bootstrapping method by following the recommendations from Sarstedt, Ringle, and Hair [110]. Additionally, mediation analysis was conducted using the approach by Preacher and Hayes [113] and Streukens and Leroi-Werelds’ [114] bias-corrected, one-tailed bootstrap approach, with a sample size of 10,000. The results of these analyses are presented in

Table 5. Hypotheses testing.

Hypothesis	Path	Coefficient (β)	SD	t-Value	p-Values	Results
H1	FP → TAI	0.323	0.041	7.834	0.000	Supported
H2	PF → TAI	0.017	0.033	0.508	0.306	Not Supported
H3	RD → TAI	0.118	0.036	3.230	0.001	Supported
H4	TK → TAI	0.100	0.041	2.454	0.007	Supported
H5a	DT → TAI	0.287	0.036	7.961	0.000	Supported
H5b	CP → TAI	0.430	0.036	11.799	0.000	Supported

.

The results of the hypothesis testing provided important insights into the relationships between the key variables. Hypothesis 1 (H1), which proposed a positive relationship between firm performance (FP) and tax avoidance intention (TAI), was supported by the data, with a significant path coefficient (β = 0.323, t(533) = 7.834, p < 0.001), indicating that firm performance has a strong influence on tax avoidance intention. However, Hypothesis 2 (H2), which posited a positive relationship between perceived fairness (PF) and tax avoidance intention (TAI), was not supported, as the path coefficient was non-significant (β = 0.017, t(533) = 0.508, p = 0.306), suggesting that perceived fairness does not have a meaningful impact on tax avoidance intention. For Hypothesis 3 (H3), the results showed that risk of detection (RD) significantly influences tax avoidance intention (TAI), with a positive path coefficient (β = 0.118, t(533) = 3.230, p = 0.001), supporting the hypothesis. Similarly, Hypothesis 4 (H4), which examined the relationship between tax knowledge (TK) and tax avoidance intention (TAI), was supported, with a significant positive path coefficient (β = 0.100, t(533) = 2.454, p = 0.007).

Additionally, Hypothesis 5a (H5a), which suggested that digital transformation (DT) would positively influence tax avoidance intention (TAI), was strongly supported by the data, as indicated by its significant path coefficient (β = 0.287, t(533) = 7.961, p < 0.001). Finally, Hypothesis 5b (H5b), which tested the effect of coercive power (CP) on tax avoidance intention (TAI), also showed strong support, with a significant path coefficient (β = 0.430, t(533) = 11.799, p < 0.001), confirming that coercive power has a substantial impact on tax avoidance intention.

4.3.1. Mediation Analysis

The direct effects of the independent variables on tax avoidance intention (TAI) were assessed first. The relationship between firm performance (FP) and TAI was significant, β = 0.323, t(533) = 7.834, p < 0.001, indicating a direct influence of FP on TAI. Perceived fairness (PF) did not show a significant direct effect, β = 0.017, t(533) = 0.508, p = 0.306, suggesting no direct relationship between PF and TAI. Risk of detection (RD) had a significant direct effect on TAI, β = 0.118, t(533) = 3.230, p = 0.001, as did tax knowledge (TK), β = 0.100, t(533) = 2.454, p = 0.007.

For the total effects, the path from FP to TAI remained significant, β = 0.197, t(533) = 6.993, p < 0.001, suggesting partial mediation by the mediators. The total effect of PF on TAI was negative and significant, β = −0.054, t(533) = 2.041, p = 0.021, indicating full mediation through the mediators. Similarly, RD exhibited a significant total effect on TAI, β = 0.053, t(533) = 2.171, p = 0.015, also reflecting partial mediation. TK maintained a significant total effect, β = 0.192, t(533) = 6.599, p < 0.001, confirming partial mediation by the mediators.

The specific indirect effects revealed several significant mediating relationships. The mediation of coercive power (CP) in the relationship between FP and TAI was supported, β = 0.090, t(533) = 3.854, p < 0.001 (H6a), suggesting partial mediation. The mediation of digital transformation (DT) in the same relationship was also supported, β = 0.107, t(533) = 5.133, p < 0.001 (H6b), confirming partial mediation. For PF, the mediation by CP was significant, β = −0.057, t(533) = 2.537, p = 0.006 (H7a), indicating full mediation, while the mediation by DT was not supported, β = 0.003, t(533) = 0.256, p = 0.399 (H7b), showing no mediation in this path.

Similarly, RD showed significant mediation through CP, β = 0.071, t(533) = 3.400, p < 0.001 (H8a), indicating partial mediation, while the mediation by DT was partially supported, β = −0.018, t(533) = 1.581, p = 0.057 (H8b), suggesting partial mediation. Finally, the mediation by CP for the relationship between TK and TAI was significant, β = 0.119, t(533) = 4.810, p < 0.001 (H9a), indicating partial mediation, and the mediation by DT was also significant, β = 0.073, t(533) = 4.556, p < 0.001 (H9b), again showing partial mediation. The results of the mediation analysis are illustrated in

Table 6. Mediation analysis.

Hypothesis	Direct Effects	Coeff. (β)	SD	t-Value	p-Values	Results	Mediation Type
	FP → TAI	0.323	0.041	7.834	0.000
	PF → TAI	0.017	0.033	0.508	0.306
	RD → TAI	0.118	0.036	3.230	0.001
	TK → TAI	0.100	0.041	2.454	0.007
	Total Effects	Coeff. (β)	SD	t-value	p-values
	FP → TAI	0.197	0.028	6.993	0.000
	PF → TAI	−0.054	0.026	2.041	0.021
	RD → TAI	0.053	0.024	2.171	0.015
	TK → TAI	0.192	0.029	6.599	0.000
	Specific Indirect Effects	Coeff. (β)	SD	t-value	p-values
H6a	FP → CP → TAI	0.090	0.023	3.854	0.000	Supp.	Partial
H6b	FP → DT → TAI	0.107	0.021	5.133	0.000	Supp.	Partial
H7a	PF → CP → TAI	−0.057	0.023	2.537	0.006	Supp.	Full
H7b	PF → DT → TAI	0.003	0.013	0.256	0.399	Not Supp.	No mediation
H8a	RD → CP → TAI	0.071	0.021	3.400	0.000	Supp.	Partial
H8b	RD → DT → TAI	−0.018	0.011	1.581	0.057	Partially Supp.	Competitive Partial
H9a	TK → CP → TAI	0.119	0.025	4.810	0.000	Supp.	Partial
H9b	TK → DT → TAI	0.073	0.016	4.556	0.000	Supp.	Partial

.

4.3.2. Moderation Analysis

The research evaluated whether corporate social responsibility (CSR) played a meaningful moderating role in the relationship that existed among digital transformation, coercive power, and tax avoidance intention (TAI). Prior to the inclusion of the moderating variable, the R² value for TAI was 0.367; hence, the 36.7% variation in TAI was explained by DT and CP. By adding the interaction term for CSR, R² increased to 38.4%, demonstrating a gain of 1.7% in the explained variance of the dependent variable TAI.

More precisely, the interaction term of CSR × CP on TAI was significant: β = −0.076, SE = 0.045, t(533) = 1.696, p = 0.045, thus supporting Hypothesis H10a. This indicates that CSR attenuates the positive relationship between coercive power and tax avoidance intention. However, the interaction term for CSR × digital transformation (DT) to TAI was insignificant: β = −0.009, SE = 0.040, t(533) = 0.221, p = 0.413, which indicates there is no moderating effect of CSR in the relationship between digital transformation and tax avoidance intention, and hence Hypothesis H10b is not supported. More interestingly, CSR itself exerted a negative direct impact on TAI, β = −0.117, SE = 0.038, t(533) = 3.117, p = 0.001, implying that the greater the level of CSR, the lower the intentions to engage in tax avoidance behaviors. The results are depicted in

Table 7. Moderation analysis.

Hypotheses	Relationship	Beta	SE	t-Value	p-Value	Results
	CP → TAI	0.430	0.036	11.799	0.000
	DT → TAI	0.287	0.036	7.961	0.000
	CSR → TAI	−0.117	0.038	3.117	0.001
H10a	Moderating Effect (CSR × CP → TAI)	−0.076	0.045	1.696	0.045	Supported
H10b	Moderating Effect (CSR × DT → TAI)	−0.009	0.040	0.221	0.413	Not Supported

.

Moreover, for an in-depth explanation of the significant moderating effect of CSR between coercive power and tax avoidance intention, H10a was further tested by performing a simple slope analysis to understand what magnitude and level of CSR will influence the strength of the relationship between coercive power and tax avoidance intention. In fact, Figure 2 distinctly shows the results, indicating the presence of a moderating effect. For low levels of CSR, the slope of CP and TAI is steeper (y = 1.012x + 1.599). This implies that, if the CSR practices are weak or underdeveloped, then, with the presence of coercive power, when the tax authorities apply a penalty, this leads to greater tax avoidance intention. For the high levels of CSR firms, on the contrary, the slope is notably less steep (y = 0.708x + 1.821), hence providing evidence that higher CSR practices weaken the influence of coercive power on tax avoidance intention.

This finding implies that CSR acts as a mitigating agent regarding the impact of coercive power on tax avoidance behavior. In other words, firms with higher CSR are less likely to engage in tax avoidance, even under coercive regulatory threat, as theoretically identified in the Discussion section.

4.3.3. Multi-Group Analysis (MGA)

The last part of this research involved conducting a Multi-Group Analysis (MGA) to explore the possible differences between different types of SMEs, which were separated into small-scale and medium-scale, and different ownership structures, such as partnerships, family businesses, and sole proprietorships, to assess how these groupings affect the key influential factors in tax avoidance behavior [95,96,97].

The MGA further revealed that the differences among several of the key paths were significant between small- and medium-sized enterprises. More specifically, the influence of coercive power on tax avoidance intention was significantly higher, Δβ = 0.23, p = 0.001, meaning coercive power has a higher effect on medium-sized enterprises’ tax avoidance intention. Similarly, the path from DT to TAI had a significant negative difference (Δβ = −0.212, p = 0.002), which suggested that the role of digital transformation in decreasing tax avoidance intention is greater for medium-sized enterprises than for small-sized ones; Δβ = −0.223, p = 0.025 finally showed that the strength of the relation was stronger in medium-sized enterprises regarding the influence that TK had on CP. The rest of the paths, such as the effects of CSR on TAI (Δβ = −0.006, p = 0.472), firm performance on CP (Δβ = 0.079, p = 0.234), and perceived fairness on CP (Δβ = −0.134, p = 0.132), did not reveal any significant differences.

Furthermore, THE MGA conducted to find the differences in path coefficients amongst family businesses, partnerships, and sole proprietorships indicated some interesting results. For example, for coercive power-tax avoidance intention: family business-sole proprietorship: Δβ = −0.159, p = 0.050; partnership-sole proprietorship: Δβ = − 0.134, p = 0.072. This infers that coercive power has a greater influence on the tax avoidance intention of sole proprietorships compared to other firm types. While for the effect of DT on TAI, there were significant differences between family businesses-partnership (Δβ = −0.198, p = 0.009) and partnership-sole proprietorship (Δβ = 0.252, p = 0.005), suggesting that DT plays an important role in partnerships. Moreover, the path from FP to DT was significantly different between partnerships and sole proprietorships, Δβ = 0.239, p = 0.030, where the impact of firm performance on digital transformation efforts was stronger for sole proprietorships. No significant differences were observed in the other paths. The significant differences are depicted in

Table 8. Significant MGA results with group comparisons.

Path	Group Comparison	Difference (Δβ)	p-Value
CP → TAI	Small vs. Medium	0.230	0.001
DT → TAI	Small vs. Medium	−0.212	0.002
TK → CP	Small vs. Medium	−0.223	0.025
CP → TAI	Family vs. Sole	−0.159	0.050
CP → TAI	Partnership vs. Sole	−0.134	0.072
DT → TAI	Family vs. Partnership	−0.198	0.009
DT → TAI	Partnership vs. Sole	0.252	0.005
FP → DT	Partnership vs. Sole	0.239	0.030

.

5. Discussion

The results of the direct effects are crucial in understanding what factors affect the intention of tax avoidance in SMEs within the tourism industry. Firm performance is found to significantly and positively affect TAI, thus confirming H1 and indicating that the higher the firm performance, the more inclined to engage in tax avoidance. This is consistent with the literature, which has argued that more successful firms are not only able but also have more incentive to exploit their resources in an attempt to explore tax avoidance strategies, especially when they regard the obligations of paying taxes as financial growth barriers [2,5,31,43]. This finding underscores that sensitive tax policies that would respond to the financial pressures on SMEs without compromising compliance are indeed called for. Contrary to expectations, perceived fairness did not significantly influence TAI (H2). While perceived fairness of the tax system has been argued in past studies to decrease tax avoidance, the nonsignificant result may indicate that perceptions of fairness alone cannot drive behavioral change in the context of SMEs [44,50]. That would imply that factors other than fairness perceptions, for instance, enforcement or organizational culture, are more decisive with respect to tax behavior, especially in the case of Greek tourism SMEs. Risk of detection (RD) has proved to have a positive and significant effect on TAI (H3), hence supporting the fact that the higher the level of detection by tax authorities, as perceived by firms, the more likely tax avoidance is. While this may seem counterintuitive, it can be justified by the fact that higher risks of detection push firms to consider even more strategic or disguised forms of tax evasion in those industries where the mechanisms of enforcement are perceived as inefficient or/and inconsistent [16,48,49]. The interaction between enforcement and compliance appears to be complex, as any increase in the perceived detection risk should therefore be associated with effective mechanisms of enforcement as a means of reduction of avoidance, a critical issue in the case of Greek tax authorities [20,26]. TK has also been positively influencing TAI, which supports H4, indicating that the more knowledge there is about tax regulations, the greater the likelihood of firms engaging in tax avoidance. This supports findings in that, while often assumed that greater tax knowledge can lead to greater compliance, on the other hand, it has also been found that tax-savvy firms might use their knowledge to find legal loopholes or reduce their tax liability through strategic means, a prominent case for the tourism sector [3,4,9,23,24,39,48]. It also calls for more specific educational support regarding knowledge but also ethical behavior in taxation.

Both DT and CP had positively significant impacts on TAI, thus confirming H5a and H5b, respectively. The strong influence of CP indicates that the regulatory pressure and penalties threatened by the authorities are among the key drivers of the behavior of tax avoidance. Those firms that perceive a higher degree of coercive power could use tax avoidance as a form of response to being put under closer scrutiny. Such findings confirm previous studies linking coercive regulatory environments with strategic avoidance behaviors, not only in industry-specific environments but extending to the case of the tourism sector [13,14,27,58,59,77]. On the contrary, the strong positive impact of DT has been somewhat surprising, considering the general perception was that digital transformation would result in more transparency and compliance. This may indicate that, while firms are adopting digital tools, they also strategically apply technology with the goal of reducing their tax burden, a relationship that might indeed be more nuanced than previously thought between digitalization and tax avoidance behavior [59,93].

The mediation analysis provided several important insights into the relationships between the independent variables and coercive power (CP) and digital transformation (DT) acting as mediators regarding tax avoidance intention (TAI). The analysis revealed that CP mediated a number of key predictors for tax avoidance intention. The mediating effect of CP, for example, was significant in the relationship between firm performance and TAI (H6a), suggesting that firms with stronger performance experience higher levels of coercive power, which in turn increases their intentions to avoid taxes. This finding validates the idea that the more successful the companies are, the greater the regulatory attention drawn to them, and hence the greater the coercive power, which in turn may affect their behavior of tax avoidance, with prominent evidence for the case of the Greek tourism sector [21,50]. Similarly, CP significantly mediated the relationship between tax knowledge (TK) and TAI (H9a), proposing that the higher the level of tax knowledge, the greater the perception of coercive power, which in turn eventually affects tax-avoidance behavior. This perhaps can be explained by the fact that there is more awareness of regulatory pressures and penalties for better tax knowledge managers; hence, they make more calculated decisions on engaging in tax avoidance [12,38,48]. This supports the argument that tax knowledge perhaps does not necessarily lead to better compliance but, in some cases, could even foster tax avoidance behavior by increasing awareness of loopholes and enforcement risks [21,50]. However, CP fully mediated the link between the PF and the TAI (H7a); that is, perceived injustice in the tax system activated perceptions of coercive power, which, in turn, enhanced tax avoidance intentions. This full mediation, wherein perceived unfairness in the tax system alone is not sufficient to prompt avoidance behavior, is the perception of coercive pressure that translates this perception into the act of avoidance behavior, as previously indicated in the Greek tourism sector [13,14,22]. These findings contribute to the broader literature on tax compliance by emphasizing the role of regulatory enforcement in translating perceived fairness or the lack thereof into behavior [12,38,48].

The findings showed that DT partially mediates the relationship of FP with TAI (H6b) and the relationship of TK with TAI (H9b). Firms with higher performance and greater tax knowledge were likely to adopt more digital tools, which then lowered tax avoidance intentions. This underlines how digital transformation ensures transparency and compliance, something also supported by findings in the literature that show how digitalization reduces manipulation in taxation [38,62,115]. However, when examining H7b, the mediation of DT in the relationship between perceived fairness and TAI was not supported. This shows that while perceptions of tax equitability influence behavior, they are not mediated by digital transformation. This might be the case, as fairness perceptions are more of a psychological or ethical issue that may not blend directly with technological adoption. Another key finding here is that this contradicts previous instances in literature pertaining to how digital tools may promote perceptions of fairness through the standardization of tax procedures and processes, as suggested by Slemrod and Yitzhaki [116]. Fairness perception and digital transformation also work independently in shaping tax avoidance intentions.

The results for H8b revealed partial competitive mediation and thus reflect a more sophisticated relationship between RD and DT with respect to TAI. Specifically, the direct influence of RD on TAI is positive, indicating that SME managers will be more likely to have greater intention to avoid taxes while perceiving a higher risk of detection by the tax authorities. This could be due to the fact that the pressure or threat of detection is greater, and hence managers give more serious consideration to avoidance strategies. On the other hand, the indirect effect via DT remained negative, suggesting that increasing digital tools and systems for tax reporting decreases the intention to avoid tax [38,61,62,115]. As firms continue to develop and become digitalized, the more their tax processes become transparent and accountable, thereby encouraging less tax avoidance. Competitive partial mediation means that RD could result in higher tax avoidance intention; however, the presence of DT acts as a mitigating factor due to DT’s constraining possibilities for evasion. This underlines how relevant digital transformation is to balance the forces of perceived detection risk and, correspondingly, the efficiency of policies supporting digitalization in tax reporting systems as a means of deterring tax avoidance behavior in SMEs [38,61,62].

This present study provides insights into the important moderating role of corporate social responsibility in explaining the relationships of coercive power and tax avoidance intention, as well as digital transformation. The significant moderation result (H10a) indicates that CSR weakens the positive relationship between coercive power and tax avoidance Intention [5,30,62]. However, the simple slope analysis of coercive power showed that, at low levels of CSR, coercive power is stronger in its relationship with tax avoidance intention, represented by a steep slope. The opposite occurs when firms highly engage in CSR; coercive power is weakly associated with tax avoidance intention. This would suggest that CSR may act as a safeguard, which reduces the likelihood of business organizations engaging in tax avoidance even when they face strong coercive regulatory pressures. Such findings support prior research on CSR, which can invoke ethical behaviors and lessen opportunistic action given by tax avoidance. In fact, previous research has asserted that CSR instills an attitude of commitment to ethical and socially responsible behavior within the firms themselves, as coercive power is being exerted by such external mechanisms but not exclusively studied in the Greek tourism context [5,30,62,63,64]. Along these lines, CSR perhaps could be thought of as a form of self-governance that would decrease whatever urge there may be to avoid paying taxes by companies in spite of regulatory scrutiny and sanctions. By contrast, the moderating effect of CSR on the relationship between digital transformation and tax avoidance intention is insignificant (H10b). This insignificance may imply that digital transformation itself promotes compliance with tax affairs, regardless of the firm’s CSR practices. This possible explanation may lie in the nature of digital transformation. Digital tools and systems, such as tax reporting platforms, promote greater transparency and accountability, as human errors are minimal, which may provide a timelier process for tax filing [62,63,64,65]. To the extent that such technology advancements reduce or eliminate room for such errors, technology can thus act as an inhibiting factor against tax avoidance, leaving little room for CSR to have any further impact on the relationship. In such a case, digital transformation could operate as the sole mechanism by which it may be enough to promote compliance with taxation irrespective of whether a firm is committed to CSR or not [62,63,64,65].

This finding supports prior studies that have identified technology’s supportive role in increasing tax compliance. While CSR plays a significant role in ensuring moral conduct, in our case, it had no significant moderation effect on the association between digital transformation and tax avoidance. It would appear that the nature of digital transformation itself plays a vital role in curtailing tax avoidance because of transparency and accountability issues, regardless of CSR business policies. This contrast between the two moderation effects provides constructive evidence on how coercive power and digital transformation interact with CSR in shaping tax avoidance behavior [30,63,92]. CSR moderates the relationship between coercive power and tax avoidance, and hence, this suggests complementary dynamics between ethical commitments and external regulatory pressures; no similar moderation effect has been evidenced as far as the impact of digital transformation is concerned [30,51,64].

Moreover, it was revealed that several determinants had considerable differences in their impact on TAI among different enterprise types and ownership structures. Significant differences were observed among small- and medium-sized enterprises, as well as family businesses, partnerships, and sole proprietorships. For medium-sized enterprises, CP impacts TAI more compared to small enterprises in the group of SMEs. This suggests that medium-scale enterprises are more sensitive to external regulatory drives with respect to tax avoidance [11,33,47]. Still, the DT factor had a more substantial negative effect on TAI in medium-sized enterprises, indicating that digital tools are more efficient in reducing tax avoidance intention in these businesses. This also established that TK had a stronger influence on the coercive power for medium-sized enterprises than small ones. This showed how higher levels of tax knowledge in medium-sized enterprises increased their perception of coercive power. As was evident from the ownership structure analysis, the influence of CP on TAI in sole proprietorships was significantly higher as opposed to family businesses and partnerships. This could be due to the fact that, with limited resources or governance structures, sole proprietorships are highly prone to coercive pressures. Significant differences were also noted in the impact of DT on TAI. DT had a stronger effect on partnership cases than on family business and sole proprietorship cases in relation to tax-avoidance intention [62,63,82]. Lastly, FP exerted a more positive influence on DT in the case of a sole proprietorship as opposed to that of a partnership, thereby implying that sole proprietorships have a greater reliance on a firm performance to drive digital adoption [11,33,47]. Therefore, the findings highlight the key distinctions between the sizes and ownership structures of firms in their response to the determinants of tax avoidance. Medium-scale enterprises and sole proprietorships are found to be more vulnerable to coercive power, while digital transformation has a greater influence on partnerships. This therefore implies that policy intervention and managerial strategies need to consider not only the firm size but also the ownership type if the intention is to reduce the incidence of tax avoidance behavior, a crucial element in the tourism sector [13,14,22].

6. Practical Implications for Policymakers and Managers

The practical implications of this research can be highly significant for both the policymakers and business managers who have to engage with the problem of tax avoidance among SMEs. In this case, for Greece’s tourism sector, incentivizing digital transformation by promoting grants or tax credits for specific purposes can provide more transparency and, therefore, encourage better tax compliance. On the other hand, Greek taxation authorities can downplay penal measures and instead engage in tax literacy programs so that SMEs can understand and work in consonance with complicated regulations [23,26,28]. For instance, it could be argued that CSR adoption by tourism SMEs will decrease negative impacts due to strict enforcement and increase their reputational assets. These, in turn, may allow for a more compliant and transparent business environment and, therefore, contribute to the sustainable development of the Greek tourism industry [13,14,22]. In general, the findings implicate that developing CSR policies could be crucial in helping reduce the level of tax avoidance within those industries where the problem is most significant, thus encouraging or requiring more active participation in CSR as a valuable tool for constraining tax evasion. From a manager’s point of view, integrating CSR into core business strategies will enhance ethical positioning and decrease negative impacts from the coercive power of the external regulatory authority [62,64,92]. That is to say, CSR can internalize ethical behaviors in firms and reduce the impact of coercive pressure on tax avoidance, while DT has its place as a complementary strategy that enhances compliance independently [62,64,92]. The fact that digitalization can make a vital contribution to decreasing tax avoidance further increases the case for technological adoption in tax reporting. Tax administrations could provide digital tools, grants, or tax incentives for promoting and encouraging SMEs, particularly regarding partnerships and medium-sized companies, to adopt digital platforms that provide better insight into their businesses and limit the scope for tax avoidance. Secondly, the findings of the relationship between coercive power and TK suggest that tax literacy programs should concentrate on ethics in tax decision-making [34,35,77]. Tax agencies should, therefore, develop programs to help businesses deal with complex tax legislation and prevent avoidance behaviors through promotional ethical practices. For example, sole proprietorships that are more vulnerable to coercive power may benefit through a simplification of tax procedures, advisory services, and performance bonuses that promote digitalization. A better performance by a firm, for instance, can spur digital adoption in the case of sole proprietorships; hence, the provision of targeted incentives to such high-performing firms can improve compliance even further [62,63,82]. Finally, simplifying the tax processes and advice supporting sole proprietorships and small enterprises would help them avoid certain problems in adhering to regulations concerning their taxes. Providing such enterprises with more accessible means of compliance can lighten the regulatory load while at the same time improving overall tax compliance. The approach can also be balanced by promoting CSR, incentivizing digital transformation, and improving tax literacy, contributing to overall reduced levels of avoidance. Given this trend, policymakers and managers should concentrate on enhancing business transparency, ethics, and compliance.

7. Conclusions, Limitations, and Future Directions

In conclusion, the novelty of the study’s research contribution lies in the identification of the tax avoidance intentions of SMEs in the Greek tourism sector. This study explores coercive power, digital transformation, tax knowledge, and CSR and, therefore, identifies the potentially complex dynamics of organizational behavior that lead organizations’ decision-making on tax non-compliance. These findings open up new directions for future research regarding the additional ways in which digitalization, regulatory pressures, and ethical concerns intertwine in the constitution of tax avoidance across a wide range of organizational contexts. Our findings also provide evidence that coercive power significantly influences the intentions of tax avoidance, hence hinting at the role of regulatory pressure in shaping SME behavior. The current wave of digital transformation, though promising increased transparency, may also be used strategically in order to minimize tax burdens. Other findings of the study suggest that tax knowledge could result in more sophisticated ways of tax planning, whereas CSR serves as a moderator and might reduce predispositions toward non-compliance. From a theoretical perspective, our study provides an understanding of how regulatory pressures and digital adaptation collectively influence organizational behavior with respect to tax compliance. Our efforts extend existing frameworks, including digital transformation, as a mediating factor, showing that technological integration can both promote compliance and enable avoidance [38,60,61]. By integrating CSR as a moderator, this study underlines the internal ethical practices that play an important role in shaping tax-related decision-making processes and adds depth to those theories that address the role of social responsibility in compliance behavior. These insights provide a better magnitude of understanding of what drives tax avoidance intentions and also enriches instances in the literature related to tax behavior in sectors marked by regulatory complexity.

This research is not without limitations. A non-probability convenience sampling technique and cross-sectional design were chosen due to the exploratory nature of the research. Despite that, this approach enabled efficient data collection; it inherently limits the ability to generalize the findings to all tourism SMEs in Greece or other regions and cannot track changes in behavior over time. To mitigate potential biases, we conducted common method bias tests to ensure the reliability of the collected data to provide greater confidence and robustness between the analyzed variables [84,85]. It would be of particular interest if future studies focused their efforts on extending the scope of the research to other industries, or even cross-country comparisons, to gain a deeper understanding of the sectoral and geographical effects on the formation of tax avoidance behavior [81,82]. The study relies on self-reported data through questionnaires, which may be affected by social desirability bias [99]. Future studies might consider using objective tax data or adding a longitudinal design to capture any change in tax behavior over time and reduce the biases associated with self-reporting. Furthermore, including additional relevant factors may provide a more detailed investigation of the intention and behavior of tax avoidance. For instance, data from tax accountants or the tax authorities may prove to be revealing regarding compliance behavior. Future research could also consider samples from other neighboring countries that have not been studied in this paper in order to investigate cross-cultural comparisons and widen the perspective of how contextual factors influence tax avoidance practices and the constructs’ adaptability to different regions and cultural settings [29,30]. More importantly, while this study implemented CSR as a moderator, there are several other internal and external pressures that are worthwhile for future studies to explore as moderators.

It should be noted that digital transformation adoption is not limited and extends beyond traditional business operations. Thus, it is beneficial in developing an efficient strategy that will enhance tax compliance with digital solutions, both in Greece and at an international level, for the monitoring of trends of emerging technologies, defining challenges regarding tax reporting, and collaborative work with tax professionals and SMEs to avoid tax evasion behaviors, something of great significance within the tourism industry.