The state as shareholder and tax authority: Effects of state ownership on corporate tax avoidance

ABSTRACT

This study aims to analyze the relationship between state ownership (SO) and the tax avoidance practices of firms listed on the Brazilian capital market. The research addresses two gaps in the literature. The first concerns the absence of metrics able to capture the State's proportional influence within firms' ownership structures, moving beyond the use of binary variables. The second relates to the scarcity of integrated theoretical approaches that explain the relationship between SO and tax avoidance practices. Ownership structure is central to corporate governance and is especially important in emerging economies such as Brazil, where the State acts as both a policymaker and a major shareholder. This study contributes to the theoretical, methodological, and practical advancement of the tax avoidance literature by providing evidence that state ownership tends to mitigate aggressive tax practices, with implications for managers, investors, and policymakers. Drawing on firms listed on B3 S.A. - Brasil, Bolsa, Balcão over the period from 2010 to 2023, ordinary least squares and quantile regression models were estimated. The State's presence in firms' ownership structures is associated with less aggressive tax avoidance practices, promoting more conservative tax management aligned with the public interest. The study offers three main contributions: (i) a methodological innovation that models SO as a continuous component of the ownership structure; (ii) an improved measurement of tax behavior by proposing adjustments to the normalized differential effective tax rate, making it more suitable to the Brazilian context; and (iii) a novel integration of agency theory, transaction cost economics, and property rights theory into a unified framework that explains state influence on corporate tax behavior.

Keywords:
state ownership; tax avoidance; property rights theory; agency theory; transaction cost economics

RESUMO

O objetivo deste trabalho é analisar a relação entre a propriedade estatal (PE) e as práticas de tax avoidance das empresas listadas no mercado de capitais brasileiro. A pesquisa preenche duas lacunas. A primeira refere-se à ausência de métricas capazes de capturar a influência proporcional do Estado na estrutura de propriedade das empresas, superando o uso de variáveis binárias. A segunda diz respeito à escassez de abordagens teóricas integradas que expliquem a relação entre PE e práticas de tax avoidance. A estrutura de propriedade é central para a governança corporativa e assume especial importância em economias emergentes, como a do Brasil, nas quais o Estado atua como formulador de políticas públicas e acionista relevante. O estudo contribui para o avanço teórico, metodológico e prático da literatura sobre tax avoidance, oferecendo evidências de que a presença estatal tende a mitigar práticas tributárias agressivas, com implicações para gestores, investidores e formuladores de políticas públicas. Com base em empresas listadas na B3 S.A. - Brasil, Bolsa, Balcão no período de 2010 a 2023, aplicaram-se modelos de mínimos quadrados ordinários e regressão quantílica. A presença do Estado na estrutura de propriedade das empresas está associada a menos práticas agressivas de tax avoidance, promovendo uma gestão tributária mais conservadora e alinhada ao interesse público. O estudo oferece três contribuições principais: i) inovação metodológica ao modelar a PE como componente contínuo da estrutura acionária; ii) aprimoramento da mensuração do comportamento tributário ao propor ajustes na taxa efetiva de impostos diferencial normalizada, tornando-a mais aderente ao contexto brasileiro; e iii) integração inédita das teorias da agência, dos custos de transação e dos direitos de propriedade em um arcabouço unificado que explica a influência estatal sobre o comportamento tributário corporativo.

Palavras-chave:
propriedade estatal; tax avoidance; teoria dos direitos de propriedade; teoria da agência; teoria dos custos de transação

INTRODUCTION

Taxes are the main source of revenue for most countries and are essential for governments to correct market failures and provide the public goods and services that the population needs. However, tax avoidance practices have been recurrent in various societies since tax legislation was first enacted, constituting a problem because they result in revenue losses for both developed and developing economies (Uadiale et al., 2010).

This scenario has driven a significant expansion of the literature on the topic, with studies focusing on the various factors that influence firms' tax behavior (Fernández-Rodríguez et al., 2019; Hanlon & Heitzman, 2010).

Among the main determinants of tax avoidance practices are organizational characteristics, managerial incentives, corporate reputation, audit quality, corporate social responsibility, and, especially, the ownership structure of firms (Duhoon & Singh, 2023; França & Monte, 2020; Wang et al., 2020). This study, in particular, situates itself within the analysis of ownership structure as a key variable for understanding firms' tax behavior.

A firm's ownership structure is centered on its organizational core and can take various forms because of the different profiles arising from the dimensions of shareholder identity and ownership concentration. Shleifer and Vishny (1997) explain that this structure involves the balance (or imbalance) between control rights (voting) and cash-flow rights, shaping firms' governance and financial performance.

In emerging economies, economic relations tend to rely more on institutional connections than on purely market dynamics (Adhikari et al., 2006), which reinforces the State's role in formulating public policy and in the efficiency of tax collection. In this context, state ownership (SO) plays a relevant role in explaining firms' tax behavior, especially in countries with a strong government presence in the market.

This observation relates to the fact that the State is, indeed, the largest minority shareholder in many corporations, owing to its participation in cash flows through tax collection. This role directly influences corporate governance and is even more pronounced when the State holds a direct equity stake (Desai et al., 2007; Xia & Walker, 2015).

Bradshaw et al. (2019) also argue that the State's presence as the largest shareholder can alter a firm's ownership structure, since the taxes paid may be regarded as implicit dividends to the controlling shareholder, potentially constituting an expropriation of the other shareholders' wealth.

Given the relevance of the topic, several studies have investigated the relationship between SO and tax avoidance practices, with a particularly large volume of publications from China, owing to the singular characteristics of its economic system.

Among the Chinese studies on the topic, notable examples are Long et al. (2024), who analyzed the relationship between non-state shareholder coalitions and tax avoidance practices, and Huyghebaert et al. (2024), who investigated the influence of large shareholders in Chinese companies. These authors concluded that government participation as a majority shareholder is negatively associated with tax avoidance practices, highlighting the role of state control in mitigating aggressive tax practices.

Beyond China, other countries have contributed, albeit on a smaller scale, to research on the topic. In Italy, Mafrolla (2019) analyzed how state-owned enterprises handle tax management; in Sweden, Hilling et al. (2021) investigated whether state ownership affects the adoption of tax avoidance practices; in Spain, Fernández-Rodríguez et al. (2019) examined the effect of SO on effective tax rates; and, in Brazil, Medeiros et al. (2021) assessed the extent to which ownership concentration influences the level of tax avoidance in state-controlled firms.

Most studies, both international and domestic, that analyze the relationship between SO and tax avoidance practices have used dummy variables to distinguish state-owned from private firms. In these models, a value of 1 is assigned to state-controlled firms and 0 to private firms (Bradshaw et al., 2019; Long et al., 2024; Medeiros et al., 2021). This dummy-variable approach simplifies the analysis by treating SO as an exclusive control factor, disregarding the complexity of the coexistence of different types of shareholders within the ownership structure.

Thus, given the specificities of the Brazilian setting, the binary distinction between state-owned and private firms ends up ignoring relevant nuances regarding the intensity of government influence and its interaction with the incentives and behaviors of private shareholders.

For this reason, the present study proposes a methodological advance by treating SO as an integral and coexisting part of the ownership structure, taking into account its interaction with other private shareholders, and by proposing a continuous, integrated measure of SO that reflects the State's effective percentage stake in firms' ownership structures.

Thus, by treating state ownership as a graduated rather than merely binary variable, the study fills a gap in the literature and expands the capacity to analyze the marginal and proportional effects of this presence on corporate decisions, especially those related to tax behavior.

This methodological innovation has two main contributions. First, it improves the understanding of the corporate governance structure in emerging economies by allowing the State's influence to be examined at different intensities and in different contexts of coexistence with private shareholders. Second, the continuous measure used in this study introduces greater empirical precision into analyses of tax avoidance.

Furthermore, this study integrates, in a novel way, agency theory, transaction cost economics, and property rights theory as the theoretical basis for explaining the relationship between SO and tax avoidance practices, given that the synergy among these theories affords a deeper understanding of conflicts of interest, as well as of governance costs and the allocation of control rights over assets.

From this perspective, the State's presence as a shareholder may increase ill-defined property rights, intensifying agency conflicts and raising transaction costs because of the misalignment of objectives between the State and private shareholders: whereas the latter seek to maximize profits and may encourage tax avoidance practices, the State has no incentive to reduce the tax collection of firms in which it participates.

This scenario becomes even more relevant in the Brazilian capital market, which is characterized by high ownership concentration and significant state participation in strategic sectors, as shown by Santos and Machado (2024). Accordingly, motivated by these gaps in the literature and grounded in the aforementioned theories, the following research question is posed: what is the relationship between SO and firms' tax avoidance practices in the Brazilian capital market?

Grounded in property rights theory, agency theory, and transaction cost economics, the results indicate that, in the sample analyzed, the State's presence in the ownership structure inhibits tax avoidance practices, favoring more conservative tax management oriented toward the public interest. These findings have practical implications for managers, policymakers, and investors, showing that the State, as a shareholder, can play a relevant role in inducing more prudent tax behavior aligned with the collective good.

THEORETICAL FRAMEWORK

Corporate Tax Avoidance

The term corporate tax avoidance lacks a universal definition; it is assigned different meanings that vary according to the perspective of the person defining it. In this respect, many tax avoidance practices explicitly comply with the parameters of legality-that is, they do not deviate from what is expressly prohibited by law-and are also socially desirable, since the tax reduction resulting from such practices can foster economic growth and stimulate private investment, whether by contributing to the development of strategic sectors or by financing public projects (Hanlon & Heitzman, 2010; Lietz, 2015).

Regarding the definition of tax avoidance, Annuar et al. (2014) point out that the multiplicity of transactions encompassed by the term, related to practices intended to increase corporate profits, makes it challenging to establish a universal definition. Hence, the context in which the practice of tax avoidance is analyzed delimits its definition. Accordingly, given the scope of this study, the term tax avoidance is understood as a firm's ability to reduce its taxes in general, in line with the definitions proposed by Dyreng et al. (2008) and Hanlon and Heitzman (2010).

Still on the scope of the term tax avoidance, Lietz (2015) explains that this expression encompasses all actions taken by firms to reduce explicit taxes, without necessarily distinguishing between clearly legal, legally questionable, or potentially fraudulent practices. This broad definition is useful for research that seeks to measure the extent of corporate tax planning without delving into the specific nature of the strategies used.

In the context of studies on tax avoidance, Fernández-Rodríguez et al. (2019) highlight the significant growth of the literature in recent decades, with different empirical approaches and themes such as managerial incentives, corporate governance, and ownership structure as determinants of tax avoidance practices.

From this perspective, Colombo and Terra (2022) identify Brazil as one of the environments most conducive to tax avoidance, owing to the combination of a high tax burden and a complex tax system. In this context, several studies have sought to identify the factors that influence firms' tax behavior, particularly those that analyze the SO structure, generally represented by a dummy variable (1 = state-owned; 0 = private). The evidence indicates that government-controlled firms tend to adopt less aggressive tax positions (Martinez & Bis, 2020; Martinez et al., 2023; Medeiros et al., 2021).

2.2 Ownership Structure, State Ownership, and Tax Avoidance Practices

Ownership structure is one of the fundamental bases of corporate governance (Wang et al., 2021), as it relates to a firm's shareholding composition, defined by the proportion of shares held by specific shareholders relative to the total shares issued.

This configuration is influenced by ownership concentration and by the identity of the shareholders-factors that shape strategic preferences and corporate objectives, directly affecting the organization's guidelines, financial behavior, and operational efficiency (Okimura et al., 2007; Pedersen & Thomsen, 2003).

From this standpoint, different types of shareholders have distinct characteristics and objectives that influence strategic decisions and organizational performance. Velte (2024) identifies six categories of share ownership: institutional, state, family, foreign, managerial, and private. Each of these categories has particularities that, in certain cases, go beyond mere profit maximization. Considering the objective of this research and the relevance of SO in developing countries, the analysis focuses on this type of ownership, grouping the remaining shareholders, in a simplified manner, as private ownership and/or private agents.

Desai et al. (2007) argue that the State is, in fact, the largest minority shareholder in almost all corporations, given its involvement in corporate cash flows through tax collection, which confers on it a relevant role in corporate governance.

From this perspective, SO tends to orient its actions toward social and political objectives, going beyond the maximization of economic value and seeking to act in accordance with the public interest, correcting market failures and promoting social welfare (Long et al., 2024).

However, this orientation may lead to the use of corporate resources for political purposes, to the detriment of private shareholders. Thus, governments assume an ambiguous role in the shareholding structure: on one hand, as a revenue-collecting authority interested in maximizing tax revenues; on the other, as a controlling shareholder oriented toward reducing taxes and increasing firms' profits (Ouyang et al., 2020).

With respect to taxation, it should be noted that it represents the main source of government revenue and one of the most significant costs for firms, as it falls directly on profit and reduces their results. This impact encourages organizations to adopt tax avoidance strategies to minimize the tax burden and maximize profitability (Wang et al., 2020).

The literature indicates that ownership structure influences these practices, highlighting the role of SO, which may lead to distinct behaviors regarding tax avoidance (Wang et al., 2020). In this context, the State's presence as a controlling shareholder can either inhibit or foster such practices, depending on the institutional environment and corporate governance (Hilling et al., 2021; Long et al., 2024).

It is worth noting that, beyond the empirical evidence, different theoretical approaches-such as property rights theory, transaction cost economics, and agency theory-contribute to deepening the understanding of this phenomenon.

2.3 Integration of Agency Theory, Transaction Cost Economics, and Property Rights Theory

Property rights theory, agency theory, and transaction cost economics share theoretical foundations within the field of organizational economics. Although these three theories address corporate behavior from different perspectives, there is a synergy among them that allows for a broader and more consistent analysis of business decisions (Kim & Mahoney, 2005; Santos & Machado, 2024). Consequently, it is argued that this theoretical integration is equally applicable to the context of the adoption of tax avoidance practices.

Agency theory addresses the conflicts of interest generated by the separation between ownership and control, highlighting how economic incentives influence the behavior of managers and shareholders. Transaction cost theory, in turn, focuses on minimizing the costs involved in negotiating and executing contracts, especially in contexts marked by uncertainty, opportunism, and high asset specificity. Property rights theory, for its part, integrates elements of these approaches by emphasizing the importance of clearly defined and protected rights.

In this context, Barzel (1997) and Coase (1960) argue that these rights are not restricted to physical possession but also encompass the use, income generation, and transfer of assets, aspects that are fundamental to understanding corporate behavior. Thus, in ownership-structure configurations, these challenges are amplified. The identity of majority shareholders can mitigate or intensify such agency conflicts, generating transaction costs and ill-defined property rights.

In the case of SO, this depends on how social and economic objectives are balanced (Bradshaw et al., 2019; Long et al., 2024); thus, when political interests prevail over economic ones, they may give rise to less aggressive tax avoidance practices, whereas, in situations of pressure for financial performance, tax avoidance may be used as a strategy.

2.4 Research Hypothesis

Whereas private firms primarily seek to maximize profit, state-owned enterprises are guided by public-policy objectives aimed at social welfare and are legally obligated to follow the guidelines that justified their creation (Santos & Machado, 2024).

In this scenario, two strands stand out: on one hand, SO, guided by public interests and committed to collecting revenue for the government, tends to engage in fewer tax avoidance practices; on the other hand, private ownership, focused on profit maximization, shows a greater propensity for tax avoidance strategies, given that reducing taxes directly increases profits.

As for empirical research, findings on the relationship between SO and tax avoidance practices diverge. Many studies indicate that state participation is associated with less involvement in aggressive tax avoidance strategies (Bradshaw et al., 2019; Fernández-Rodríguez et al., 2019; Hilling et al., 2021; Huyghebaert et al., 2024; Long et al., 2024; Medeiros et al., 2021; Solikhah et al., 2024); however, this finding is not unanimous, as there are also studies whose results suggest that state participation may favor tax avoidance (Mahenthiran & Kasipilai, 2012; Tang, 2020; Velte, 2024; Ying et al., 2017).

Given the empirical divergences, there is no consensus on the relationship between SO and tax avoidance. Nonetheless, the integration of agency theory, transaction cost economics, and property rights theory suggests that the State, in exercising its property rights, can direct corporate actions toward public-policy objectives and social responsibility. It is therefore inferred that this influence tends to reduce aggressive tax behavior. On this rationale, the guiding hypothesis of this research is presented:

H1: SO is negatively associated with the tax avoidance practices of firms in the Brazilian capital market.

METHODOLOGICAL PROCEDURES

Sample

The sample comprises 139 firms listed on B3 S.A. - Brasil, Bolsa, Balcão (B3), the Brazilian stock exchange, analyzed over the period from 2010 to 2023 (Table 1). The selection considered the availability and consistency of the information, according to the following criteria: (a) disclosure of financial statements in the Economatica® database and of annual management reports on the website of B3 and/or the Comissão de Valores Mobiliários (CVM), Brazil's securities market regulator; and (b) exclusion of publicly traded financial institutions, owing to their regulatory and tax particularities.

Table 1 -
Filter of the firms comprising the sample

Balanced panel
Total active firms	382
(-) Financial institutions	30
(-) Firms with missing or insufficient data	213
Total exclusions	243
Final sample	139

• Source: Prepared by the authors.

Measurement of the Variables

3.2.1 Ownership structure

To assess ownership concentration from the perspective of conflicts among controlling shareholders, it is preferable to use a metric that reflects the resulting power index. This study adopts the methodology used by Abramov et al. (2017) and by Santos and Machado (2024). This approach considers that any stake exceeding 50% of the voting shares of a state-owned controlling firm is treated as 100%.

This methodology allows the state ownership stake to be measured accurately across different shareholding structures. In sum, the analysis of the ownership structure begins with the observed firm and ends when all links in the control chain belong to private investors, anonymous investors, or the State, considering only common shares (Abramov et al., 2017; Santos & Machado, 2024).

3.2.2 Normalized differential ETR as a measure of tax avoidance

The effective tax rate (ETR) is widely used as an indicator of the corporate tax burden. In addition, it is frequently used to measure firms' level of tax avoidance. This metric is defined as the ratio between the expenses for Corporate Income Tax (IRPJ) and Social Contribution on Net Income (CSLL) and pre-tax income (LAIR) (Dyreng et al., 2008; França & Monte, 2020).

Although widely used, the ETR has some limitations. A low ETR may reflect either the use of more aggressive tax avoidance strategies or legitimate tax incentives, such as tax exemptions, specific tax policies, or accounting choices. In the Brazilian context, the sector of activity or the economic activity plays a crucial role in determining corporate taxation (França & Monte, 2020). Thus, to adapt the analysis to the Brazilian setting, França and Monte (2020), drawing on the studies of Hanlon and Heitzman (2010), proposed the use of the differential ETR, which, in the Brazilian case, was adjusted by economic sector and calculated based on the following formula:

$$ETRdifferentiali,t=averagesectorETR-currentETR$$ (1)

In this equation, the average sector ETR corresponds to the average current ETR of firms in the same economic sector in year t. The current ETR, in turn, is calculated as the ratio between the current tax (the sum of IRPJ and CSLL, minus deferred tax) and the LAIR of firm i in year t.

$$currentETR=currenttax/LAIR$$ (2)

Subsequently, Bizerra et al. (2024) proposed the normalized differential ETR as an evolution of the differential ETR. This metric was developed to overcome the limitations of the differential ETR, which, by using absolute values, makes it difficult to compare firms of different sizes or economic sectors. The normalized differential ETR incorporates a proportional adjustment factor, allowing for more precise and consistent analyses.

The interpretation of the normalized differential ETR follows the same logic as the differential ETR, in which positive values indicate that the firm has an effective tax rate below the sector average, suggesting more aggressive tax avoidance practices; a value of zero denotes conformity with the sector standard; and negative values reveal an effective tax rate above the average, indicating less aggressive tax avoidance.

The formula for calculating the normalized differential ETR in this study presents a slight modification relative to the one presented by Bizerra et al. (2024), given that, instead of using the generally accepted accounting principles ETR, this study uses the current ETR (obtained from Eq. 2), as presented by França and Monte (2020). Thus, our measure is as follows:

$$normalizeddifferentialETR=1-(currentETR/averagesectorETR)$$ (3)

3.2.3 Control variables

The study used the control variables described in Table 2. In addition to these variables, the variable representing the presence of other large influential shareholders (OLSH) was included, defined as the participation of entities with significant influence over the firm (Santos & Machado, 2024). For methodological purposes, a shareholder holding at least 20% of the voting capital was considered influential, in accordance with Technical Pronouncement CPC 18 of the Comitê de Pronunciamentos Contábeis (2012).

Table 2 -
Dependent, independent, and control variables

Variable	Description	Proxy	Guidance	Basis
TAXAV	Tax avoidance	Normalized differential ETR	> 0: indicates an effective tax rate below the sector average; = 0: indicates an effective tax rate in line with the sector average; < 0: indicates an effective tax rate above the sector average.	Bizerra et al. (2024), Dyreng et al. (2008), Hanlon and Heitzman (2010), França and Monte (2020)
SO	State ownership	Percentage of voting capital	A stake exceeding 50% of the voting shares of a state-owned controlling firm is treated as 100%; below 50%, the percentage corresponding to the State's influence in the control chain is applied.	Abramov et al. (2017), Santos and Machado (2024)
SIZE	Size	Log of total assets	Larger firms have more resources for tax planning but are more subject to oversight.	França and Monte (2020), Gupta and Newberry (1997), Solikhah et al. (2024)
LEV	Leverage	Total debt/total assets	The deduction of interest reduces the taxable base, influencing tax behavior.	Dyreng et al. (2008), Fernández-Rodríguez et al. (2019)
INTANG	Intangibility	Intangible assets/total assets	Firms with more intangible assets can use amortization for tax purposes.	Bizerra et al. (2024), França and Monte (2020)
INVINT	Inventory intensity	Inventory/total assets	Inventories can affect taxable income, depending on the valuation method used.	Gupta and Newberry (1997), Long et al. (2024)
ROA	Financial performance	LAIR/total assets	More profitable firms may seek to reduce the tax burden to maximize returns.	Bizerra et al. (2024), Ouyang et al. (2020), Solikhah et al. (2024)
TLOSS	Tax loss	Dummy; equals 1 if the firm reported a negative LAIR in the previous year and 0 otherwise	Firms with a tax loss can accumulate tax credits, affecting the ETR.	França and Monte (2020), Long et al. (2024)
OLSH	Other large shareholders	Dummy; equals 1 when the entity has at least two large shareholders and 0 otherwise	Other large shareholders can influence the controlling shareholder's decisions	Abramov et al. (2017), Santos and Machado (2024)

• ETR = effective tax rate; LAIR = pre-tax income (earnings before income tax).
• Source: Prepared by the authors.

Econometric Procedures

3.3.1 Proposed econometric model

In accordance with the proposed objectives, the influence of SO on the level of tax avoidance is analyzed according to Eq. 4:

$$TAXAV=\beta 0+\beta 1SOi,t+\beta 2SIZEi,t+\beta 3LEVi,t+\beta 4INTANGi,t+\beta 5INVINTi,t+\beta 6ROAi,t+\beta 7TLOSSi,t+\beta 8OLSHi,t+\mu i,t$$ (4)

PRESENTATION AND ANALYSIS OF THE RESULTS

Descriptive Statistics of the Variables

Table 3 presents the descriptive statistics of the variables used in the study, encompassing the dependent variable and the independent and control variables. The sample comprises 1,945 observations, considering balanced panel data for the firms analyzed over the period from 2010 to 2023.

Table 3 -
General descriptive statistics of the dependent and independent variables

Variable	Observations	Mean	SD	Minimum	Maximum
Normalized differential ETR	1,945	0.150770	0.985897	-7.42392	1
SO	1,945	14.35886	27.76218	0	100
OLSH	1,945	0.825192	0.379899	0	1
SIZE	1,945	15.14706	1.969159	10.3242	20.7729
LEV	1,945	1.756002	28.86884	0.00011	999.2516
INTANG	1,945	0.099595	0.193217	-0.0292419	3.090428
INVINT	1,945	0.570839	8.263103	0.00012	209.7372
ROA	1,945	-0.523482	19.90916	-744.3299	107.35
TLOSS	1,945	0.275578	0.446920	0	1

• Note: All variables are presented in balanced panel format and are described inTable 2.
• SD = standard deviation; ETR = effective tax rate.
• Source: Prepared by the authors.

The results in Table 3 reveal interesting information about the profile of the firms analyzed and their tax avoidance practices. The positive mean of the normalized differential ETR (0.1508) suggests that, in general, the firms in the sample tend to have a tax burden below the average of their respective sectors, indicating the presence of tax planning strategies. However, the high standard deviation (0.9859) and the quite negative minimum value (-7.4239) signal wide variation in tax strategies, with some firms exhibiting more aggressive tax avoidance behavior.

As for SO, it recorded a mean of 14.36%, ranging between 0 and 100% participation. The variable addressing the presence of OLSH had a mean of 0.8252, showing that most firms have more than one large shareholder in their ownership structure, which can influence corporate governance by reducing agency conflicts and limiting opportunistic tax avoidance practices. Seeking a more in-depth analysis of the sample studied, firms with and without state participation were separated, and a difference-of-means test (t-test) was conducted to identify statistically significant differences between the two groups of firms, as shown in Table 4.

Table 4 -
Descriptive statistics of firms with and without SO

Variable	Firms with SO (n = 612)		Firms without SO (n = 1,333)		Mean differences
	Mean	SD	Mean	SD	Diff.	t-test
Normalized differential ETR	-0.0702353	1.171314	0.04102	1.160052	0.111255**	1.9603

• SD = standard deviation; ETR = effective tax rate; SO = state ownership.
• *, **, *** = significance at 10%, 5%, and 1%, respectively.
• Source: Prepared by the authors.

The results in Table 4 indicate significant differences between firms with and without state participation, especially regarding tax avoidance practices. The dependent variable had a negative mean (-0.0702) for state-owned firms and a positive mean (0.0410) for private firms, with a difference of 0.1113, statistically significant at 5% (p < 0.05). These findings suggest that firms with SO adopt less aggressive tax avoidance practices, corroborating previous studies (Bradshaw et al., 2019; Huyghebaert et al., 2024), according to which the State, having no incentive to reduce tax collection, tends to induce more conservative tax behavior.

Analysis of the Regression Model Results and Hypothesis Test

The research hypothesis was tested by estimating the empirical model presented in Eq. 4. First, a Pearson correlation (r) analysis was conducted to identify possible high correlations among the independent variables; no multicollinearity was found, which ensures the robustness of the model. Next, we proceeded to (i) estimate the model by robust pooled panel regression and (ii) apply quantile regression (QR) as a complementary approach to the ordinary least squares (OLS) method, notable for its greater efficiency and resistance to outliers, in addition to reducing estimation errors.

To identify the most appropriate model for estimation, the Chow test was first performed, and its results (p = 0.069) indicated the pooled OLS model as the most suitable option. Next, the Breusch-Pagan test (p = 0.077) was applied, which corroborated this choice, reinforcing the robustness of the statistical modeling. Thus, each firm was treated as a unit of analysis in each year.

The results of the pooled OLS panel regression, presented in Table 5, show a negative and statistically significant relationship between SO and firms' tax avoidance, represented by the normalized differential ETR. According to the results, hypothesis H1 cannot be rejected; it states that SO is negatively associated with firms' tax avoidance practices in the Brazilian capital market.

Table 5 -
Estimation of the pooled OLS regression model between the normalized differential ETR and SO

Explanatory variable	Coefficient	SE*	t-statistic	p-value
SO	-0.0100678***	0.0037575	-2.68	0.007
OLSH	0.2233302	0.3049589	0.73	0.464
LEV	-0.0045051	0.0047996	-0.94	0.348
INTANG	1.96148***	0.6693096	2.93	0.003
INVINT	0.022477	0.0185129	1.21	0.225
ROA	-0.0039635	0.0058706	-0.68	0.500
TLOSS	-6.192371***	0.2845356	-21.76	0.000
SIZE	0.1477058*	0.0787719	1.88	0.061
Constant	-1.00119	1.174241	-0.85	0.394

• Notes: Number of observations: 1,945; adjusted R²: 20.61%. The F test (F = 14.59; p < 0.01) confirms the overall significance of the model. The Shapiro-Francia (W′ = 0.848; p = 0.001) and Jarque-Bera (p = 0.020) tests reject the normality of the residuals, but the sample size (n = 1,945) allows this assumption to be relaxed by the central limit theorem. The Wooldridge test (p < 0.001) indicates autocorrelation, and the Breusch-Pagan test (p = 0.077) does not reject homoskedasticity at 5%. A robust variance-covariance matrix was used to correct for heteroskedasticity and autocorrelation. The VIF (variance inflation factor) values ranged between 1.05 and 4.89, all below the critical threshold of 10, indicating the absence of relevant collinearity. The variables are described inTable 2.
• SE = standard error; ETR = effective tax rate; OLS = ordinary least squares,
• *, **, *** = significance at 10%, 5%, and 1%, respectively.
• Source: Prepared by the authors.

From the perspective of agency theory, state participation acts as a mechanism for aligning interests, mitigating conflicts between managers and shareholders and discouraging tax avoidance practices. According to transaction cost economics, the State's presence raises the potential costs of aggressive tax conduct, making it less attractive. In turn, according to property rights theory, shareholding structures with state participation tend to adopt more conservative tax positions, reducing tax avoidance compared with private firms oriented exclusively toward profit maximization.

Taken together, the theories explain that the negative effect of SO on tax avoidance practices stems from three factors: a lower managerial incentive for such practices (agency theory), an increase in the costs associated with aggressive tax conduct (transaction cost economics), and the prioritization of public-policy objectives over profit maximization (property rights theory).

Overall, these results converge with the findings of Bradshaw et al. (2019), Hilling et al. (2021), Long et al. (2024), Medeiros et al. (2021), and Solikhah et al. (2024), which show that the level of tax avoidance is inversely related to SO. However, they diverge from the conclusions of Fernández-Rodríguez et al. (2019), Mahenthiran and Kasipilai (2012), and Ying et al. (2017), who identified a positive relationship between SO and tax avoidance practices.

Among the control variables, some had a significant impact on tax avoidance. Intangible assets (INTANG) had a positive coefficient and significance at 1%, indicating that firms with a larger proportion of such assets, such as patents and trademarks, tend to adopt more aggressive tax strategies, given the flexibility in allocating revenues and expenses. Tax loss (TLOSS) had a negative and statistically significant relationship, suggesting that firms with losses in the previous year reduce or eliminate tax avoidance practices owing to the offsetting of accumulated losses.

Firm size (SIZE) also had a positive relationship, with significance at 10% (p = 0.061), signaling that larger firms have more resources and expertise for tax planning. In turn, the variables OLSH, leverage (LEV), inventory intensity (INVINT), and profitability (ROA) were not statistically significant.

To better analyze the relationship between SO and tax avoidance, QR was used, as presented in Table 6.

Table 6 -
Estimation of the QR model between the normalized differential ETR and SO

Variables	Q25	Q50	Q75
Constant	-0.1607807 (0.2182594)	-0.0701237 (0.1767972)	0.9984037** (0.0611083)
SO	-0.00944 (0.008541)	-0.038912*** (0.01085)	-0.02725** (0.011657)
OLSH	-0.2050644*** (0.057533)	-0.1525258** (0.0726676)	0.005291 (0.061408)
LEV	0.0010107 (0.0167574)	0.0013921 (0.0256016)	-0.002925 (0.0025409)
INTANG	-0.391376** (0.1824457)	-0.6359113*** (0.2017707)	-0.0140018 (0.1265034)
INVINT	0.0097426 (0.1650972)	0.0083929 (0.1027134)	0.000214 (0.0043015)
ROA	0.0026524 (0.1578956)	0.0035119 (0.121198)	0.0021869 (0.0030662)
TLOSS	-0.1403712* (0.0770072)	0.063539 (0.0634685)	0.005739 (0.008949)
SIZE	-0.0236379 (0.0150951)	0.035524*** (0.013334)	0.000581 (0.01001)
Pseudo R2	0.116	0.123	0.035

• Notes: Wald test: SO (p = 0.043), OLSH (p = 0.047), LEV (p = 0.9968), INTANG (p = 0.0103), INVINT (p = 0.9954), ROA (p = 0.899), TLOSS (p = 0.0333), and SIZE (p = 0.006). The variables are described inTable 2.
• SE = standard error; ETR = effective tax rate; Q = quantiles; QR = quantile regression.
• *, **, *** = significance at 10%, 5%, and 1%, respectively.
• Source: Prepared by the authors.

The QR results in Table 6 are, in general, consistent with those obtained in the OLS regression. Thus, regarding the relationship between SO and firms' tax avoidance, the results show a negative and statistically significant relationship in the Q50 and Q75 quantiles, suggesting that, in firms with median or aggressive tax avoidance practices, state participation tends to discourage such practices.

As for the Q25 quantile, which showed a negative and non-significant relationship, the results indicate that, in firms with a lower level of tax avoidance, state participation does not exert a relevant influence, suggesting that, by holding a tolerable level of tax avoidance, such firms operate within a limit considered acceptable for tax planning.

The influence of OLSH is significant and negative in the lower quantiles (Q25 and Q50), indicating that the presence of large shareholders acts as a control factor, reducing aggressive tax practices, and is less relevant in the more aggressive firms (Q75). The INTANG variable was statistically significant in the Q25 and Q50 quantiles, being negatively associated with tax avoidance in firms with a lower propensity for this practice.

SIZE was significant only in the Q50 quantile. Tax loss (TLOSS), in turn, was significant only in the Q25 quantile, with a negative coefficient significant at 10%, suggesting that, among firms with less aggressive tax avoidance practices, TLOSS further reduces the use of these strategies. Finally, LEV, INVINT, and ROA were not statistically significant.

Overall, the complementary and comparative analysis between the pooled OLS regression and QR demonstrates that SO has a consistent and negative effect on tax avoidance, which is stronger among firms with moderate and aggressive practices.

Interaction Tests

To deepen the analysis, interactions were considered between SO and the variables that proved statistically significant in their relationship with tax avoidance: SIZE, INTANG, and TLOSS.

The interaction between SO and SIZE (SOxSIZE) seeks to verify whether the effect of SO on tax avoidance is more or less pronounced in larger firms, which are subject to greater public and regulatory scrutiny. In the case of the interaction between SO and TLOSS (SOxTLOSS), the objective is to assess whether TLOSS represents a tax incentive or a condition of vulnerability. Finally, the interaction between SO and INTANG (SOxINTANG) aims to identify whether the state presence acts as a factor that restricts the use of more sophisticated tax planning strategies, typical of firms with a higher proportion of intangible assets. The specification of the regression model is presented below:

$$TAXAV=\beta 0+\beta 1SOi,t+\beta 2SIZEi,t+\beta 3LEVi,t+\beta 4INTANGi,t+\beta 5INVINTi,t+\beta 6ROAi,t+\beta 7TLOSSi,t+\beta 8OLSHi,t+\beta 9(SOi,txSIZEi,t)+\beta 10(SOi,txINTANGi,t)+\beta 11(SOi,txTLOSSi,t)+\mu i,t$$ (5)

Following the proposed methodology and in accordance with the results that indicated pooled OLS modeling, each firm was treated as a unit of analysis in each year. To reduce multicollinearity when including the interaction terms, the variables were previously centered. It should be noted that this centering does not alter the coefficient or the significance level of the interaction term, affecting only the coefficients of the main effects (Enders & Tofighi, 2007).

The results in Table 7 indicate that the SOxTLOSS interaction shows a positive association and significance at 5% in the OLS regression, suggesting that firms with state participation tend to adopt greater tax flexibility in the face of accumulated losses, recognizing such strategies as legitimate mechanisms for financial recovery. In contrast, SOxSIZE and SOxINTANG were not statistically significant in the OLS model.

Table 7 -
Estimation of the pooled OLS and QR regression model between the normalized differential ETR and SO to test the interactions among the variables SIZE, TLOSS, and INTANG

		SOxTLOSS		SOxSIZE		SOxINTANG
Model	Q	Coefficient	p-value	Coefficient	p-value	Coefficient	p-value
OLS	Mean	0.020756**	0.013	0.001469	0.549	0.009026	0.609
QR	0.25	-0.004850***	0.000	-0.000855***	0.000	-0.003975	0.289
QR	0.50	-0.001841***	0.001	-0.000237	0.284	-0.004330*	0.068
QR	0.75	-0.000100	0.916	-9.750000	0.900	-0.000881	0.816

• Notes: Number of observations: 1,945; adjusted R²: 22.32%. The F test (F = 61.11; p < 0.01) confirms the overall significance of the model. The Shapiro-Francia (W′ = 0.881; p < 0.001) and Jarque-Bera (p < 0.001) tests reject the hypothesis of normality of the residuals; however, given the sample size (n = 1,945), the central limit theorem justifies relaxing this assumption. The Wooldridge test (p < 0.001) indicates autocorrelation of the residuals, while the Breusch-Pagan test (p < 0.001) rejects the hypothesis of homoskedasticity. To ensure robustness, a robust variance-covariance matrix was used, correcting for heteroskedasticity and autocorrelation. The VIF (variance inflation factor) tests showed values between 1.11 and 1.92 for the variables analyzed, all below the critical threshold of 10, indicating the absence of relevant collinearity. The Wald test for the interaction variables showed p = 0.0045 (SOxSIZE), p = 0.0001 (SOxTLOSS), and p = 0.14 (SOxINTANG). The variables are described inTable 2.
• ETR = effective tax rate; OLS = ordinary least squares; Q = quantile; QR = quantile regression.
• *, **, *** = significance at 10%, 5%, and 1%, respectively.
• Source: Prepared by the authors.

In the QR, the results were heterogeneous. The SOxTLOSS interaction was significant in the Q25 and Q50 quantiles, with a negative relationship, whereas, in the upper quantile (Q75), the effect was not significant. This indicates that the state presence mitigates the impact of losses on tax behavior predominantly at the lower and median levels of tax avoidance. The SOxSIZE interaction, in turn, had a negative effect and was significant only in Q25. Finally, the SOxINTANG interaction maintained a negative coefficient across the entire conditional distribution of tax avoidance, as expected, but was significant in the median quantile (Q50).

Robustness Analysis

Although the pooled OLS regression model proved statistically significant and consistent, the possibility of endogeneity is acknowledged, arising mainly from reverse causality between SO and the level of tax avoidance. Thus, as an additional robustness test and check for endogeneity, the explanatory variable was lagged by one, two, and three periods (Clemens et al., 2011).

The results in Table 8 show the stability of the signs and significance for the SO variable in the lagged models. The coefficient remained negative and statistically significant, with a magnitude similar to that observed in Table 5, indicating that the effect of SO on tax avoidance persists even when the values of the variable are lagged by up to 3 years. This consistency suggests that the observed relationship does not stem from reverse causality, which lends robustness to the study.

Table 8 -
Estimation of the pooled OLS regression model between the normalized differential ETR and SO lagged by 1, 2, and 3 years

Explanatory variable	Model 1	Model 2	Model 3
State ownership	-0.0110718*** (0.0041059)	-0.0107844** (0.0042611)	-0.0092938** (0.004527)
Interaction - SOxTLOSS	0.0010376 (0.0097474)	0.0116748 (0.009707)	0.0026975 (0.0113885)
Interaction - SOxSIZE	0.0019205 (0.0028887)	0.002999 (0.0030425)	0.00269 (0.0032915)
Interaction - SOxINTANG	0.0181496 (0.0198676)	0.0023923 (0.019260)	-0.0081393 (0.0177394)
Constant	-3.004209** (1.271244)	-1.840952 (1.327955)	-2.293655 (1.514043)

• Notes: Model 1 - 1-year lag (R² 0.09, Prob F 0.000); Model 2 - 2-year lag (R² 0.088, Prob F 0.000); Model 3 - 3-year lag (R² 0.063, Prob F 0.000). The variables are described inTable 2.
• ETR = effective tax rate; OLS = ordinary least squares.
• *, **, *** = significance at 10%, 5%, and 1%, respectively.
• Source: Prepared by the authors.

Furthermore, the stability of the control variables' coefficients and the maintenance of the models' overall significance (Prob F < 0.01) indicate that the lags did not substantially alter the explanatory structure of the model, suggesting that there is no evidence of endogeneity capable of invalidating the inferences obtained in the main model.

FINAL CONSIDERATIONS

This study aimed to analyze the impact of SO as an integral part of the ownership structure of firms in the Brazilian capital market, investigating its relationship with tax avoidance practices. To measure firms' tax behavior, the normalized differential ETR was used, an adjusted metric that more precisely measures the adoption of aggressive tax planning strategies, making it possible to identify relative deviations in the tax burden and representing an advance over traditional metrics based on absolute ETR.

The results show that SO has a negative and statistically significant coefficient, indicating that the State's presence in the shareholding structure acts as an inhibiting factor in tax avoidance practices. State participation, therefore, tends to discourage aggressive tax strategies, favoring more conservative tax management aligned with the public interest. The QR confirms the complexity of this relationship by demonstrating that the effect of SO varies according to the level of tax avoidance: in firms with median or more aggressive tax behavior, the state presence plays a moderating and restrictive role.

Thus, these results can be explained by alignment with public objectives (agency theory), since the State, as a shareholder participating in a firm's ownership structure, has no direct incentive to reduce tax collection, because tax revenue finances public policies and essential services. They are also grounded in the ownership structure oriented by public-policy objectives (property rights theory), given that, under the influence of SO, firms tend to adopt more conservative tax conduct.

França and Monte (2020) point out that firms more committed to social responsibility tend to show less involvement in tax avoidance practices, owing to the pursuit of legal compliance and the preservation of reputation. Along these lines, transaction cost economics suggests that the state presence in the shareholding structure raises the costs associated with tax avoidance, owing to reputational risk and greater regulatory requirements.

As for its methodology, this study stands out by advancing beyond the usual dichotomous approach to measuring SO, treating it as a continuous variable, and by combining OLS and quantile regression models, which strengthens the robustness of the analyses and broadens the understanding of the marginal and conditional effects of the state presence.

The results contribute to the theoretical and empirical field by showing that state participation in the corporate ownership structure influences tax behavior in a nonlinear manner, acting as a partial disciplining mechanism for tax avoidance practices. In addition, they provide input for policymakers and regulators regarding the potential fiscal and tax effects of the State's presence in the shareholding structure of firms.

It should be noted, however, that some limitations must be considered, such as the indices and proxies adopted and the theoretical strands used, which may restrict other analytical possibilities. For future research, we recommend adopting more dynamic methodologies for measuring SO, capable of capturing the interaction among different types of shareholders, as well as replicating the model in other emerging contexts in which the State's presence in the capital market is relevant.

REFERENCES

• Abramov, A., Radygin, A., & Chernova, M. (2017). State-owned enterprises in the Russian market: Ownership structure and their role in the economy. Russian Journal of Economics, 3(1), 1-23. https://doi.org/10.1016/j.ruje.2017.02.001
  » https://doi.org/10.1016/j.ruje.2017.02.001
• Adhikari, A., Derashid, C., & Zhang, H. (2006). Public policy, political connections, and effective tax rates: Longitudinal data from Malaysia. Journal of Accounting and Public Policy, 25(6), 574-595. https://doi.org/10.1016/j.jaccpubpol.2006.07.001
  » https://doi.org/10.1016/j.jaccpubpol.2006.07.001
• Annuar, H. A., Salihu, I. A., & Obid, S. N. S. (2014). Corporate ownership, governance and tax avoidance: An interactive effects. Procedia - Social and Behavioral Sciences, 164, 150-160. https://doi.org/10.1016/j.sbspro.2014.11.063
  » https://doi.org/10.1016/j.sbspro.2014.11.063
• Barzel, Y. (1997). Economic analysis of property rights (2nd ed.). Cambridge University Press.
• Bizerra, H. K. A. I., Abrahim, Y. F., França, R. D., & Lucena, W. G. L. (2024). O fator moderador da reputação corporativa sobre a RSC na agressividade fiscal. Revista de Educação e Pesquisa em Contabilidade, 18(3), 407-426. https://doi.org/10.17524/repec.v18i3.3444
  » https://doi.org/10.17524/repec.v18i3.3444
• Bradshaw, M., Liao, G., & Ma, M. (2019). Agency costs and tax planning when the government is a major shareholder. Journal of Accounting and Economics, 67(2-3), 255-277. https://doi.org/10.1016/j.jacceco.2018.10.002
  » https://doi.org/10.1016/j.jacceco.2018.10.002
• Clemens, M. A., Radelet, S., Bhavnani, R. R., & Bazzi, S. (2011). Counting chickens when they hatch: Timing and the effects of aid on growth. The Economic Journal, 122(561), 590-617. https://doi.org/10.1111/j.1468-0297.2011.02482.x
  » https://doi.org/10.1111/j.1468-0297.2011.02482.x
• Coase, R. H. (1960). The problem of social cost. Journal of Law & Economics, 3, 1-44.
• Colombo, J. A., & Terra, P. R. S. (2022). Interest on equity versus dividends: The role of shareholder identity in corporate tax avoidance. Revista Brasileira de Gestão de Negócios, 24, 175-205. https://doi.org/10.7819/rbgn.v24i1.4155
  » https://doi.org/10.7819/rbgn.v24i1.4155
• Comitê de Pronunciamentos Contábeis. (2012). Pronunciamento Técnico CPC 18 (R2) - Investimento em coligada, em controlada e em empreendimento controlado em conjunto http://www.portaldecontabilidade.com.br/legislacao/CPC_18_(R2).pdf
  » http://www.portaldecontabilidade.com.br/legislacao/CPC_18_(R2).pdf
• Desai, M. A., Dyck, A., & Zingales, L. (2007). Theft and taxes. Journal of Financial Economics, 84(3), 591-623. https://doi.org/10.1016/j.jfineco.2006.05.005
  » https://doi.org/10.1016/j.jfineco.2006.05.005
• Duhoon, A., & Singh, M. (2023). Corporate tax avoidance: A systematic literature review and future research directions. LBS Journal of Management & Research, 21(2), 197-217. https://doi.org/10.1108/LBSJMR-12-2022-008
  » https://doi.org/10.1108/LBSJMR-12-2022-008
• Dyreng, S. D., Hanlon, M., & Maydew, E. L. (2008). Long-run corporate tax avoidance. The Accounting Review, 83(1), 61-82. https://doi.org/10.2308/accr.2008.83.1.61
  » https://doi.org/10.2308/accr.2008.83.1.61
• Enders, C. K., & Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods, 12(2), 121-138. https://doi.org/10.1037/1082-989X.12.2.121
  » https://doi.org/10.1037/1082-989X.12.2.121
• Fernández-Rodríguez, E., García-Fernández, R., & Martínez-Arias, A. (2019). Influence of ownership structure on the determinants of effective tax rates of Spanish companies. Sustainability, 11(5), 1441. https://doi.org/10.3390/su11051441
  » https://doi.org/10.3390/su11051441
• França, R. D., & Monte, P. A. (2020). Efeitos da reputação corporativa na tax avoidance de empresas brasileiras de capital aberto. Revista Universo Contábil, 15(4), 109-126. https://doi.org/10.4270/ruc.2019430
  » https://doi.org/10.4270/ruc.2019430
• Gupta, S., & Newberry, K. (1997). Determinants of the variability in corporate effective tax rates: Evidence from longitudinal data. Journal of Accounting and Public Policy, 16(1), 1-34. https://doi.org/10.1016/s0278-4254(96)00055-5
  » https://doi.org/10.1016/s0278-4254(96)00055-5
• Hanlon, M., & Heitzman, S. (2010). A review of tax research. Journal of Accounting and Economics, 50(2-3), 127-178. https://doi.org/10.1016/j.jacceco.2010.09.002
  » https://doi.org/10.1016/j.jacceco.2010.09.002
• Hilling, A., Lundtofte, F., Sandell, N., Sonnerfeldt, A., & Vilhelmsson, A. (2021). Tax avoidance and state ownership - The case of Sweden. Economics Letters, 208, 110063. https://doi.org/10.1016/j.econlet.2021.110063
  » https://doi.org/10.1016/j.econlet.2021.110063
• Huyghebaert, N., Kang, S., Wang, L., & Wu, W. (2024). Multiple large shareholders, identity, and corporate tax avoidance. Management and Organization Review, 20(2), 235-264. https://doi.org/10.1017/mor.2024.11
  » https://doi.org/10.1017/mor.2024.11
• Kim, J., & Mahoney, J. T. (2005). Property rights theory, transaction costs theory, and agency theory: An organizational economics approach to strategic management. Managerial and Decision Economics, 26, 223-242. https://doi.org/10.1002/mde.1218
  » https://doi.org/10.1002/mde.1218
• Lietz, G. M. (2015). Tax avoidance vs. tax aggressiveness: A unifying conceptual framework [Working Paper]. Social Science Research Network. https://doi.org/10.2139/ssrn.2363828
  » https://doi.org/10.2139/ssrn.2363828
• Long, W., Wu, H., Li, L., Ying, S. X., & Li, S. (2024). Mixed-ownership structure, non-state-blockholder coalition, and tax avoidance. International Review of Financial Analysis, 91, 102988. https://doi.org/10.1016/j.irfa.2023.102988
  » https://doi.org/10.1016/j.irfa.2023.102988
• Mafrolla, E. (2019). Tax avoidance in government-owned firms: Evidence from Italy. Public Money & Management, 39(3), 186-192. https://doi.org/10.1080/09540962.2018.1516955
  » https://doi.org/10.1080/09540962.2018.1516955
• Mahenthiran, S., & Kasipillai, J. (2012). Influence of ownership structure and corporate governance on effective tax rates and tax planning: Malaysian evidence. Australian Tax Forum, 27(4), 941-969. https://dx.doi.org/10.2139/ssrn.2384718
  » https://dx.doi.org/10.2139/ssrn.2384718
• Martinez, A. L., & Bis, C. (2020). Foreign capital participation and tax aggressiveness in Brazilian companies. Management and Business Research Quarterly, 15, 43-53. https://doi.org/10.32038/mbrq.2020.15.04
  » https://doi.org/10.32038/mbrq.2020.15.04
• Martinez, A., Rodrigues, J. C., Dias, J. M., Filho, & Nakao, S. H. (2023). Agressividade fiscal e financeirização corporativa no Brasil. Revista Contabilidade, Gestão e Governança, 26(3), 378-412. http://dx.doi.org/10.51341/cgg.v26i3.3130
  » http://dx.doi.org/10.51341/cgg.v26i3.3130
• Medeiros, J. T., Soares, R. A., De Luca, M. M., & Teodósio, I. R. M. (2021). Concentração acionária e tax avoidance: E quando o Estado é o maior acionista? Revista Universo Contábil, 17(2), 61-76. https://doi.org/10.4270/ruc.2021213
  » https://doi.org/10.4270/ruc.2021213
• Okimura, R. T., Silveira, A. M., & Rocha, K. C. (2007). Estrutura de propriedade e desempenho corporativo no Brasil. RAC-Eletrônica, 1(1), 119-135. https://repositorio.usp.br/item/001597216
  » https://repositorio.usp.br/item/001597216
• Ouyang, C., Xiong, H., & Huang, Z. (2020). Dual roles of government in corporate tax avoidance: Evidence from China. Pacific-Basin Finance Journal, 62, 101355. https://doi.org/10.1016/j.pacfin.2020.101355
  » https://doi.org/10.1016/j.pacfin.2020.101355
• Pedersen, T., & Thomsen, S. (2003). Ownership structure and value of the largest European firms: The importance of owner identity. Journal of Management and Governance, 7(1), 27-55. https://doi.org/10.1023/A:1022480016567
  » https://doi.org/10.1023/A:1022480016567
• Santos, L. C., & Machado, M. R. (2024). A influência da propriedade estatal no desempenho das empresas no mercado de capitais brasileiro. Revista Universo Contábil, 19, 1-12. https://doi.org/10.4270/ruc.2023112
  » https://doi.org/10.4270/ruc.2023112
• Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. The Journal of Finance, 52(2), 737-783. https://doi.org/10.1111/j.1540-6261.1997.tb04820.x
  » https://doi.org/10.1111/j.1540-6261.1997.tb04820.x
• Solikhah, B., Chen, C.-L., Weng, P.-Y., & Al-Faryan, M. A. S. (2024). Related party transactions and tax avoidance: Does government ownership play a role? Corporate Governance: The International Journal of Business in Society, 24(1), 1-23. https://doi.org/10.1108/CG-01-2024-0003
  » https://doi.org/10.1108/CG-01-2024-0003
• Tang, T. Y. H. (2020). A review of tax avoidance in China. China Journal of Accounting Research, 13(4), 327-338. https://doi.org/10.1016/j.cjar.2020.10.001
  » https://doi.org/10.1016/j.cjar.2020.10.001
• Uadiale, O. M., Fagbemi, T. O., & Ogunleye, J. O. (2010). An empirical study of the relationship between culture and personal income tax evasion in Nigeria. European Journal of Economics, Finance and Administrative Sciences, 20, 116-126. https://bura.brunel.ac.uk/bitstream/2438/18670/1/FullText.pdf
  » https://bura.brunel.ac.uk/bitstream/2438/18670/1/FullText.pdf
• Velte, P. (2024). Ownership structure and corporate tax avoidance: A structured literature review. Journal of Applied Accounting Research, 25(3), 696-731. https://doi.org/10.1108/JAAR-10-2022-0259
  » https://doi.org/10.1108/JAAR-10-2022-0259
• Wang, W., Wang, H., & Wu, J. (2021). Mixed ownership reform and corporate tax avoidance: Evidence of Chinese listed firms. Pacific-Basin Finance Journal, 69, 101648. https://doi.org/10.1016/j.pacfin.2021.101648
  » https://doi.org/10.1016/j.pacfin.2021.101648
• Wang, F., Xu, S., Sun, J., & Cullinan, C. P. (2020). Corporate tax avoidance: A literature review and research agenda. Journal of Economic Surveys, 34, 793-811. https://doi.org/10.1111/joes.12347
  » https://doi.org/10.1111/joes.12347
• Xia, F., & Walker, G. (2015). How much does owner type matter for firm performance? Manufacturing firms in China 1998-2007. Strategic Management Journal, 36, 576-585. https://doi.org/10.1002/smj.2233
  » https://doi.org/10.1002/smj.2233
• Ying, T., Wright, B., & Huang, W. (2017). Ownership structure and tax aggressiveness of Chinese listed companies. International Journal of Accounting & Information Management, 25(3), 313-332. https://doi.org/10.1108/IJAIM-07-2016-0070
  » https://doi.org/10.1108/IJAIM-07-2016-0070

This is a bilingual text. This article was originally written in Portuguese, published under the DOI https://doi.org/10.1590/1808-057x20262513.pt

author-notes-fn-presented-at

Study presented at the 25º USP International Conference on Accounting, São Paulo, SP, Brazil, July 2025, and selected for submission to the Accounting & Finance Review through the conference's fast-track process.

DATA AVAILABILITY STATEMENT

The entire dataset supporting the results of this study can be made available upon request to the authors.

GENERATIVE AI DISCLOSURE

The authors declare that generative artificial intelligence was used in the following stages of the production of this manuscript: - Text refinement: Gemini, for APA Style compliance, grammatical review, and improved textual fluency. The authors declare that, regardless of the use of the tools mentioned above, all generated content was supervised, verified, and critically validated by humans. The authors assume full and exclusive responsibility for the accuracy of the data, integrity of mathematical/statistical formulas, originality of the text, and the conclusions presented in the published article.

Email:

[e-mail]

Email:

[e-mail]

Email:

[e-mail]

Academic Editor-in-Chief:

Andson Braga de Aguiar

Associate Editor:

Paulo Roberto da Cunha

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

Publication Dates

• Publication in this collection
  05 Oct 2026
• Date of issue
  2026

History

• Received
  02 Dec 2025
• Reviewed
  22 Dec 2025
• Accepted
  12 Feb 2026