The Determinants of Tax Aggressiveness in Family Firms: An Investigation of Italian Private Family Firms

A recent stream of research has focused on tax aggressiveness, the downward management of taxable income through tax planning activities, and has analyzed its antecedents and consequences, mainly on public companies. Only very few studies, however, have been carried out in the context of private family business and have investigated whether some family firms are more tax aggressive than others, considering some specific features of family firms, such as their distinctive agency conflicts and socioemotional wealth. In this paper, we investigate the antecedents of tax aggressiveness in a sample of private Italian family firms. Our research findings show that tax aggressiveness is positively associated with ownership concentration, the presence of independent members in the board, and the adoption of reporting mechanisms. Instead, we found a negative relation between tax aggressiveness and the use of both strategic planning and a combination of managerial control systems (both planning and reporting mechanisms). We did not find any relation between family CEO and tax aggressiveness. In summary, overall, our findings show that family involvement in ownership, an independent board. and managerialization (the use of managerial mechanisms) are relevant antecedents of tax aggressiveness in private family businesses.

The increased interest in tax aggressiveness can be explained, at least in part, by the increase in the number of cases of tax aggressiveness reported in the news and by the debate that these cases have triggered. From a stakeholder theory perspective [15], tax aggressiveness is ethically questionable because firms are expected to pay a fair amount of taxes, providing the funds for public services such as healthcare, education, and infrastructure, public services from which they benefit too, either directly or indirectly.
Tax aggressiveness, however, is problematic even when we adopt a shareholder theory perspective [16], because revelations of tax aggressive behaviors may damage a firm's reputation and affect its share price [12,13]. According to this perspective, then, the decision to engage in tax aggressive behaviors ultimately rests on an assessment of the relative expected costs and benefits associated, e.g., [17,18]. The main benefit of tax aggressiveness is obviously tax savings, which may subsequently be used to grow the firm, to distribute higher dividends to shareholders, and/or to increase managers' direct and With the aim of taking into consideration all the family firms' distinctive features that could influence tax aggressiveness behavior, as elements of heterogeneity, we analyze the aspects proposed by previous studies, such as the nature and involvement of the CEO and SEW [22], as well as specific agency conflicts that could emerge in a family firm as a consequence of the family involvement in ownership, governance, and management [17]. Furthermore, differently from previous studies, we also consider as elements of heterogeneity of family firms the adoption of managerialization tools, intended as management control systems. These latter include two main mechanisms: managerial reporting and strategic planning. While managerial reporting is devoted to measuring, monitoring, and evaluating a firm's performance, through the comparison between goals and results, strategic planning defines strategic goals and programs, determining the actions to achieve the goals and decisions on resource allocation. Specifically, firms with stronger or more effective managerial reporting systems are less likely to misstate or incorrectly report financial or taxable income, have reduced opportunities to engage in activities that are directed at benefiting executive management, and are less likely to engage in complex tax aggressive activities [24,25]. Moreover, the use of strategic planning allows firms to assess the costs and benefits of tax aggressiveness over a long-term horizon.
Therefore, aiming to fill the gaps in the literature, this study looks at private family firms displaying different characteristics that can help to ask the following question: Are some family firms more tax aggressive than others and why?
The research sample comprises 227 private Italian family firms. For these firms, secondary data were collected from many sources-including AIDA (a database of Italian companies providing mainly financial data) and the firms' websites. As data on governance and managerial mechanisms are not publicly available for private firms, we complemented this data with data collected through a questionnaire sent to a sample of private family firms.
We found that tax aggressiveness is positively related with ownership concentration, the presence of independent members in the board, and the adoption of reporting mechanisms. We found a negative relation between tax aggressiveness and the use of both strategic planning and a combination of managerial control systems (both planning and reporting mechanisms). Instead, we did not find any relation between family CEO and tax aggressive. With regard to control variables, stand-alone family firms with main operations located in different places than the headquarters are less tax aggressive than family firms that belong to groups and that have main operations concentrated in the same place. Moreover, tax aggressiveness is positively associated with profitability, firm size (in terms of total assets), and firm age.
Our findings contribute to research on family firms by analyzing tax aggressiveness within family firms and by providing evidence that some family firms are more tax aggressive than others. Our results also contribute to the tax aggressiveness literature by fleshing out the determinants of tax aggressive in family firms and, by doing that, responding to the call for more research on the impact of insider control and ownership structures on tax aggressiveness [26]. In particular, we outlined that family involvement in ownership, an independent board, and the use of managerial reporting are relevant antecedents of tax aggressiveness in family businesses. Our results also have relevant implications for attempts to curb tax avoidance in family firms. This paper is organized as follows. The next section presents the relevant literature on tax aggressiveness. The third section discusses the research hypotheses. The fourth section outlines the research design. In the fifth section, the empirical results are presented. Finally, the discussion of research findings and conclusions are provided.

Research on Tax Aggressiveness
Tax aggressiveness refers to any attempt by the firm to reduce income tax payments to the State. This topic has recently emerged as an important area of research within research on tax avoidance [17].
A firm's likelihood to engage in tax aggressive behaviors substantially rests on an assessment of the relative expected costs and benefits associated. Tax aggressiveness implies potentially both positive and negative outcomes [27]. From one side, for instance, aggressive tax strategies may contribute to increasing a firm's cash flows in favor of shareholders; on the other side, instead, these types of strategies can lead to taking risks in terms of tax risk management procedures, fines imposed by tax authorities, and reputation damages. Thus, tax aggressiveness can be seen as a mixed gamble, as it denotes a decision that implies the possibility of both gain and loss outcomes [27,28].
Although these studies undoubtedly shed new light on the antecedents of tax aggressiveness, they tend to focus on public firms and to neglect the potential role of insider control and ownership structure.
In fact, only few studies have been conducted about tax aggressive behaviors in private family firms [22], even though these organizations predominate in the world economy [30].
In their review of the empirical tax literature, [26] acknowledge this gap and call for more research on the impact of insider control and ownership structures on tax aggressiveness. Some scholars have started to answer this call by looking at tax aggressiveness in the context of family firms, which are characterized by a unique ownership structure in comparison to nonfamily businesses.

Tax Aggressiveness within Family Firms
Chen, S. et al. [17] compared US public family firms to US public nonfamily firms and found that public family firms are less tax aggressive than their nonfamily counterparts. Similarly, [22], relying on Finnish survey data, compared private family firms to private nonfamily firms and showed that private family firms are less tax aggressive than nonfamily firms. They both interpreted these results as indicating that family owners are more concerned with the potential penalty and reputation damage associated with tax aggressiveness than nonfamily firms and, therefore, tend to engage less in tax aggressive behaviors. In particular, [22] highlighted that the attention given by family businesses to noneconomic goals, such as firm and family's reputation and long-term survival of the firm, asks for a complementary adoption of both an agency and a SEW perspective in addressing tax aggressiveness in family firms.
Chen, S. et al. [17] proposed that both benefits and costs of tax aggressiveness are higher for family owners than for managers in nonfamily firms. However, their findings show that family firms exhibit lower tax aggressiveness, due to non-tax costs, arising from agency conflicts between dominant and minority shareholders and concerning potential price discount from nonfamily shareholders, the potential penalty imposed by the tax authorities, and the potential damage on firm reputation.
Mafrolla, E. et al. [23] focused on listed family firms and confirmed previous studies that highlighted less tax aggressiveness of family firms with respect to nonfamily counterparts. However, they found that family involvement has a non-linear impact on tax aggressiveness in family firms "as too much family involvement (which is otherwise beneficial) causes the detrimental outcome of higher tax aggressiveness" [23].
Steijvers, T. et all. [22] explored how tax aggressive activities vary within a heterogeneous group of private Finnish family firms. They suggest that SEW is a relevant determinant of tax aggressiveness in family business. SEW refers to the family's nonfinancial endowment in the firm, including family control and influence, emotional attachment and identification with the organization, family reputation and image, as well as the longterm success of the firm for future generations, [21]. Family firms are SEW loss averse [21]; as a consequence, they tend to be reluctant to adopt tax aggressiveness strategies that could damage their image and reputation because they intend to protect family values and reputation over generations [22]. Moreover, some scholars [31] underlined that family-induced diversity has an impact on whether and to what extent family firms prioritize SEW over financial wealth. Due to their concern for SEW, family firms focus on both financial and nonfinancial goals [32]. Then, in the specific family business context, tax decisions will be based both on potential financial gains and losses against potential SEW gains and losses, experiencing unique trade-offs [33].
Going further into family firms' behaviors, recent works suggest that the heterogeneity features of family firms influence the perception of risk to SEW and the pursuit of family interests [31,33,34] thus playing a critical role in tax aggressiveness strategies adoption [22,23,35]. Yet, a family firm's heterogeneity features may have important implications in making SEW a priority over financial goals when a family firm has to decide the strategies to be pursued.
According to [20], five main agency conflicts that arise from sources other than classic principal-agent issues, characterize family business: conflicts arising from asymmetric altruism [36,37]; conflicts of interest between family members in different roles [38][39][40]; conflicts of interest between family members and nonfamily members ; conflicts of interest between dominant (family) and minority (nonfamily) shareholders [45]; and conflicts of interest between owners and lenders [38].
Even though agency theory represents a relevant theoretical framework to analyze tax aggressiveness [46,47] and typical agency conflicts can be identified in family business, different and more complex than the classic conflict between owners and managers [20], previous studies did not go deeper into understanding how family businesses' specific agency conflicts may explain differences in tax aggressiveness, both between family and nonfamily businesses and among family firms.
For instance, [22] focused on the impact on tax aggressiveness of a high/low CEO ownership, thus focusing on the agency conflict between the CEO (family or professional nonfamily CEO) and the other shareholders. Chen, S. et al. [17] analyzing the impact of family ownership and control on tax aggressiveness in public family businesses, proposed that family firms distinguish from nonfamily businesses as they have lower classic ownermanager conflict, but bigger agency conflicts between dominant and minority shareholders.
Some recent studies on tax aggressiveness [23,27] take into consideration the heterogeneity of family firms, making a distinction based on family involvement in ownership, management, and governance. In our study, we simultaneously consider family involvement in ownership, governance, and management, adding also the dimension of managerial tools used by the company to address tax strategies and monitor related costs and benefits, as elements that could influence tax aggressiveness. However, although previous studies agree that nonfamily firms are more tax aggressiveness than family firms, what the main family businesses' characteristics that can explain a different behavior in terms of tax aggressiveness remains an open question that requires further investigation. In the next sections, we develop and test hypotheses about the antecedents of tax aggressiveness in family firms. With the aim of adding contributions to research, we propose a framework that explains why some family firms are reluctant to engage in aggressive tax strategies, while others show a greater inclination towards tax aggressiveness. We investigate how family firm heterogeneity drives variation in the tax aggressiveness of different types of family firms, considering how various sources of heterogeneity alter the perception of potential gains and losses to socioemotional and financial wealth.
In our study, we assume that heterogeneity is related to different types of agency conflicts, which depend on family involvement in ownership (FIO), management (FIM), and governance (FIG). Thus, we adopt as a main reference theoretical framework agency theory.
However, we complement it with a SEW perspective, as it considers other distinctive features of family business, as suggested by [22]. According to [17], we hypothesize that family owners in family businesses are the subjects that mainly determine the extent of tax aggressiveness.
We analyze the relationship between tax aggressiveness and FIO, considering the effect of concentrated family ownership, between tax aggressiveness and FIM, focusing on the role of family CEO, and between tax aggressiveness and FIG, taking into account the role of independent nonfamily members in the board. Finally, differently from previous studies, we also consider firm managerialization, that is, the adoption of managerial control systems, as a relevant characteristic that may have an impact on tax aggressiveness decisions in family business. In fact, these mechanisms represent an important corporate governance tool that enables the board of directors to better monitor and manage risks and to assess the impact of strategic choices on firm performance [48]. Our research hypotheses are presented in the next section.

Family Involvement in Ownership (FIO) and Tax Aggressiveness
Chen, S. et al. [17] proposed that the level of tax aggressiveness in family firms compared to nonfamily firms depends on the effect on benefits and costs arising from tax aggressiveness caused by differential features of the founding family of the firm (family owner) compared to those of managers in nonfamily firms. These authors showed that apparently, the tax aggressiveness level of family firms is less significant than in nonfamily firms and that this happens because family owners are willing to pay higher tax costs rather than tax penalties and face possibilities of the firm's reputation damage as consequence of auditing by tax officials. Chen, S. et al. [17] proposed that nonfamily firms have higher tax aggressiveness level than family firms probably because of a higher agency problem that occurs more in nonfamily firms than in family businesses. When the ownership and management are separated, inefficient processes of job contract and control occur. This inefficiency creates a chance for managers to perform opportunistic actions and bring out corporate governance problems [47].
Prior research also states that family firms may differ in the emphasis they place on SEW depending on the nature and the extent of family involvement [32].
In line with literature [27,31], we make a distinction between strong family-owned and weak family-owned firms to investigate how differences in family ownership influence tax aggressiveness.
Since family influence depends on the size of family ownership, we assume that a larger ownership stake confers more voting rights to the family, thus enhancing its ability to pursue the family agenda and protect SEW when making strategic decisions [35]. In contrast, lower levels of family ownership are associated with a lower ability of the family to impact strategic decisions [32] and decisions concerning tax aggressiveness.
A high level of family ownership determines a relevant emotional identification of the family in the organization and a greater engagement in the decision-making process. Furthermore, if the firm is family-named, the perceived firm identity is even stronger.
Accordingly, we assume that families with large ownership stakes tend to exert their influence to reduce the potential threats to firm's reputation and image [49].
However, family members are likely to base most of their wealth on the value of the company. Consequently, they are interested in maximizing the firm's financial performance, on the one hand, but they also tend to play an active role, influencing tax aggressiveness strategies in a way that mitigates potential threats to the family SEW, on the other hand.
If we consider smaller ownership stakes, with reduced voting rights, we detect a lower motivation and chance to influence strategic decisions on tax aggressiveness; thus, SEW preservation plays a secondary role.
This led to propose the following hypothesis: Family firms with more concentrated ownership are less tax aggressive than family firms with less concentrated ownership.

Family Involvement in Management (FIM): The Role of CEO and Tax Aggressiveness
Another characteristic that may affect a family firm's attitude toward tax aggressiveness is the involvement of family members in management. We consider family involvement in the CEO position, as the CEO represents the managerial role with the most relevant decision-making power. As CEOs who are members of the family tend to have a longer time horizon than CEOs who are not members of the family [32], they should be more concerned about the possible reputational and financial penalties associated with tax aggressiveness. SEW considerations are more likely to be a priority for family CEOs [21], who should be likely to support managerial decisions oriented towards the preservation of family interests, thus limiting her/his propensity to engage in aggressive tax activities. Therefore, a family CEO should balance the short-term benefits of tax aggressiveness (higher financial performance and higher dividends) with the long-term benefits of SEW preservation (long-term performance and reputation damage avoidance).
In addition, family CEOs may be less experienced and less knowledgeable than hired professional CEOs [50][51][52]; and thus, they may find it harder to manage taxable income.
When the CEO is not a family member, SEW considerations are less likely to be a priority because she/he will tend to be more strongly guided by short term financial and personal motivation. Moreover, non-family CEO remuneration is more likely to be short-term performance-based than that of family CEOs [53].
As the above arguments suggest that the potential SEW losses represent a priority issue for family firms in which the CEO is a member of the family compared to family firms in which the CEO is not a member of the family, therefore, we propose that:

Hypothesis 2.
Family firms in which the CEO is a family member are less tax aggressive than family firms in which the CEO is a non-family member.

Family Involvement in Governance (FIG): Board Composition and Tax Aggressiveness
In family firms, family members are usually involved not only in ownership (FIO) and in management (FIM), but also in governance (FIG).
The literature [20] outlines conflicts arising from asymmetric altruism, conflicts of interest between family members in different roles and between family and nonfamily members, conflicts between dominant (family) and minority (nonfamily) shareholders, and finally, conflicts of interest between owners and lenders.
From an agency theory perspective, the board of directors represents a key monitoring mechanism that is used to mitigate any residual loss to the firm's shareholders and thus control the agency problem [54]. The board of directors receives its authority for internal control and other decisions from the firm's shareholders. This makes the board the apex of decision control within the firm [54]. The board of directors has the authority to limit any residual loss arising from the agency problem because it has the ultimate responsibility within the agency framework to provide a relatively low-cost mechanism for replacing or re-ordering top management [55]. The board also has ultimate responsibility for all firm's strategic decisions in discharging duties owed to all stakeholders in society [2,56].
Family firms' boards of directors perform important monitoring and service roles [57]. Monitoring (control) focuses on managerial procedures, decision-making processes, operational activities, accounting practices, and compliance. Independent directors are a valuable feature of corporate governance [58,59]: they play a monitoring role as well as an advisory role [9], with the specific purpose of protecting minority shareholders and reducing all kinds of agency conflicts. Under this perspective, independent directors are required to properly prevent and manage agency conflicts. Thus, board monitoring is often associated with independent directors and their tendency to reduce the company's risk exposure [60].
The desirable outcomes of tax aggressiveness, such as higher dividends and cash flow, are largely reserved for the majority shareholders. Independent directors, considering their aim to protect minority shareholders, tend to be reluctant to pursue tax aggressive strategies that imply a large variety of risks and potential loss for all the shareholders.
According to these considerations, we propose that: Family firms with independent directors are less tax aggressive than family firms with only internal family directors.

Management Control Systems and Tax Aggressiveness
The adoption of agency cost control mechanisms, such as management control systems, tends to reduce agency costs and improve financial performance [54]. Management control systems comprise two main mechanisms: strategic planning, that is devoted to defining goals and programs, and managerial reporting that aims to measure, monitor, and evaluate firm performance. The main goal of management control systems is to formulate and monitor decisions throughout the organization and to guide employee behavior in a desirable direction to increase the chances of achieving the organization's objectives, including organizational performance [61].
Ref. [48] argue that a firm's managerial reporting system represents an important corporate governance tool. An effective managerial reporting system enables the board of directors to better monitor and manage risk. Previous research (e.g., [62][63][64][65][66][67] has considered the influence of effective monitoring systems on the likelihood of financial reporting and corporate fraud. Overall, the research findings indicate that firms with more effective monitoring of management are less likely to be involved in financial reporting and corporate fraud. According to [68], tax risks include the risk of paying less tax than is required under the tax legislation, and the reputational damage arising from such errors can result in additional costs. Consequently, the board has an important obligation to participate in tax-risk management so that the right balance is created between risk and opportunity in the firm [68][69][70]. In fact, tax represents an essential component of the risk management system and is included in the managerial reporting systems. As stated by [68] (p. 213), "the same rules apply for tax risks as apply for recognition and control of general business risks and establishing the control environment in general is normally the duty of the board." Specifically, firms with stronger or more effective managerial reporting systems are less likely to misstate or incorrectly report financial or taxable income, have reduced opportunities to engage in activities that are directed at benefiting executive management, and are less likely to engage in complex tax aggressive activities [24,25]. We thus expect firms in which the board can benefit from an effective managerial reporting system to be less inclined to participate in tax aggressiveness. Therefore, we propose: Hypothesis 4. Family firms with managerial reporting systems are less tax aggressive than family firms that do not have.
Strategic planning represents another critical component of management control mechanisms. According to [71], high performing family firms involve their board of directors in strategic planning more frequently than others. Since strategic planning attempts to systematize the processes that enable an organization to attain its goals and objectives, it allows firms to assess the benefits of strategic choices in term of performance.
When this approach is applied to tax strategies, it is possible to quantify, over a long-term horizon, the positive and negative outcomes of the strategies themselves.
On one side, in fact, aggressive tax strategies may increase the available cash flow and, on the other side, they may generate costs and potential losses.
With the adoption of strategic planning, family firms can be aware of the net economic/financial advantage to be exploited in the medium to long term.
The long-term orientation that is typical of family firms [72,73] shapes their risk preferences and influences their strategic investments, e.g., [74]. Family firms' long-term orientation results from the family's view of the firm as an asset to be passed on, rather than a source of wealth to be consumed during their lifetime [38].
Family firms are more likely to use extended time horizons when making decisions [38,73,75,76], which likely reduces their focus on short-term financial returns, i.e., myopic loss aversion [77], in favor of a focus on long term non-financial returns such as SEW preservation and firm reputation.
In the light of previous considerations, we propose that family firms that adopt strategic planning systems are less likely to be tax aggressive.

Hypothesis 5.
Family firms with strategic planning systems are less tax aggressive than family firms without.

Research Sample and Data Collection
We tested our research hypotheses, using a sample of Italian private family firms. As a research setting, Italy offers numerous advantages: a) about 85% of the firms in Italy are family firms [78], making this organizational form particularly relevant for the national economy and increasing the population of firms to draw from; b) in Italy, private firms are required to deposit their financial accounts in a public depository, making it possible to collect accounting data for these companies. This is not necessarily the case in other countries (e.g., the UK and the US).
The survey took place in 2015 (database data are referred to fiscal year 2013-2014). We collected our data using both primary (from questionnaires) and secondary (from the AIDA database) data. Through the questionnaire, we collected information on ownership, governance, strategy, and management control systems. Through a secondary source (AIDA), and in line with previous studies, e.g., [79][80][81], we identified firms' financial characteristics and performance. We randomly selected 15,000 firms out of the entire population of independent, incorporate Italian firms and sent a questionnaire to their CEOs. A total of 860 firms returned the questionnaire. Of these, 463 were classified as family firms. We define a family firm as one in which a family owns the majority of the equity [82][83][84][85][86], one or more family members sit on the board of directors [20,38,82], and one or more generations are involved in the firm's management [82,83,87,88].
The response rate of the survey conducted in the first wave of data collection (5.73%) was relatively low compared to typically reported response rates for surveys mailed to top executives, which range between 10% and 15%, e.g., [89,90]. This is, however, not surprising as prior research indicates that response rates are generally lower for private firms, e.g., [91], and particularly in Italy, e.g., [92,93]. From our dataset of Italian family businesses, we removed 106 family businesses that were listed.
After excluding 130 family firms with missing values from our variables (see description below), our final sample used in the analysis was N = 227.

Variables Definition
Dependent variable. Following prior research, we measured tax aggressiveness in three ways, e.g., [17]. The first measure we used was the effective tax rate (ETR), calculated as the ratio between total tax expense and pre-tax income. This measure reflects aggressive tax planning through permanent book-tax differences. The second measure we used was the cash effective tax rate (CETR), calculated as the ratio between cash taxes paid and pre-tax income. Unlike our first measure, this measure reflects both permanent and temporary book-tax differences. The third measure we used was the net cash effective tax rate (NCETR), calculated as the ratio between cash taxes paid and pre-tax income net of special items. Like our second measure, this measure reflects both permanent and temporary book-tax differences, but it also takes into account the existence of special items. For all three variables (ETR, CETR, and NCETR), the lower they are the higher tax aggressiveness is. All the values used to calculate these measures were taken at the end of fiscal year 2014. Following [17], each measure was set as missing when the denominator was zero or negative. Table 1, Panel A reports the correlations between these three measures. All measures were strongly and positively correlated with each other. In addition, the results of a factor analysis we performed on the three measures (Table 1, Panel B) show that all measures loaded on a single factor explaining around 82% of the total variance. Considering these results, and the fact that the net cash effective tax rate (NCETR) reflects both permanent and temporary book-tax differences, while also taking into account special items, we decided to run our subsequent analyses using NCETR as dependent variable. Independent variables. To test our first hypothesis, we measured ownership concentration (OWN_CON) as the logarithm of the percentage of shares owned to the same family. To test our second hypothesis, we built a dummy variable (FAM_CEO) that equals 0 if the CEO is not a family member and 1 if the CEO is a family member. To test our third hypothesis, we used a measure (EXT) that is a dummy variable that equals 0 if no independent directors sit in the board of directors and 1 if at least one independent director is member of the board. All independent variables were lagged one year (i.e., they refer to the fiscal year 2013). To test our fourth hypothesis, we built a dummy variable (REPORT) that takes the value of 1 if the firm adopts a managerial reporting system and 0 otherwise. Finally, to test our fifth hypothesis, we built the dummy variable (SPLANNING) that takes the value of 1 if the family firm adopts strategic planning mechanisms and 0 otherwise.
In particular, the variables REPORT and SPLANNING are referred to the presence (or absence) of these two specific management control tools. As far as the variable REPORT is concerned, we asked for the presence/absence of managerial reporting system, the mechanism that aims to measure, monitor, and evaluate firm performance, while, as far as SPLANNING is concerned, we considered the presence/absence of strategic planning, which is a mechanism devoted to define goals and programs and to allocate resources to strategic decision.
Control variables. In terms of controls, we controlled for the industry in which the firm operates by creating a variable (MAN) that takes the value of 1 if the firm is a manufacturer and 0 otherwise. We also controlled for whether the firm's main operations are located in the same place as the headquarters, because this may affect the extent to which the firm is embedded in-and willing to contribute to-its local community. This variable (EMB) takes the value of 1 if the firm's main operations are located in the same place as the headquarter and 0 otherwise. Further, we controlled for whether the firm is a stand-alone entity or is part of a group by creating a variable (GROUP) that takes the value of 1 if the company belongs to a group and 0 otherwise. Firms belonging to groups may, in fact, have different tax-related and reporting incentives than stand-alone firms [18]. We also controlled for the logarithm of the year in which the firm was founded (LOG_YEAR), as suggested by previous studies [27]. In addition, we controlled for the firm's profitability using the ROA at the end of the fiscal year, as firms with a higher profitability can be expected to have a higher taxation and so to be more tax aggressive. In fact, profitable firms are likely to reduce taxes relative to firms that are not so profitable or that record net operating losses [94]. Since tax aggressiveness is related to economies of scale and complexity [95], we controlled for firm size (represented by the total assets and employees). The logarithm of total assets, as size, may affect the expected benefits and the costs of tax aggressiveness [17]. As more complex organizations require more executive talent (which commands higher pay) and more opportunities to tax plan, we included the logarithm of the number of employees working in the firm to catch the dimension of such complexity [96]. All control variables were lagged one year.

Model Specification
We tested our hypotheses with OLS regressions. OLS regressions are appropriate as we were interested in cross-sectional variation in our observations and we did not expect our results to show path-dependency. Table 2 provides the descriptive statistics for and correlations between the variables in our study. All the correlation coefficients were low to moderate, with the exception of (a) the positive correlation between firm size and being a manufacturing firm (LOG-TA: r = 0.182, p-value ≤ 0.01; LOG-EMP: r = 0.257, p-value ≤ 0.01), being an old firm (LOG-TA: r = −0.349, p-value ≤ 0.01; LOG-EMP: r = −0.332, p-value ≤ 0.01), and belonging to a group (LOG-TA: r = 0.277, p-value ≤ 0.01; LOG-EMP: r = 0.272, p-value ≤ 0.01), (b) the negative correlation between family CEO and firm size (LOG-EMP: r = −0.173, p-value ≤ 0.01), (c) the positive correlation between board independent members and the firm's size (LOG-TA: r = 0.224, p-value ≤ 0.01; LOG-EMP: r = 0.202, p-value ≤ 0.01), (d) the positive correlation between reporting systems and the firm's size (LOG-TA: r = 0.241, p-value ≤ 0.01; LOG-EMP: r = 0.245, p-value ≤ 0.01), as well belonging to a group (r = 0.270, p-value ≤ 0.01), and (e) the positive correlation between strategic planning and belonging to a group (r = 0.200, p-value ≤ 0.01) as well reporting systems (r = 0.554, p-value ≤ 0.01). These correlations were not surprising and in the expected direction. Nonetheless, in order to mitigate possible concerns about multicollinearity, we decided to further investigate them. Our subsequent analyses indicated that multicollinearity problems are unlikely to affect our results, as all our V.I.F. were lower than 3.  Table 3 shows the results of our regression analysis. Model 1 includes only the control variables and shows that more profitable family firms are more tax aggressive than less profitable family firms (p-value ≤ 0.001), that larger family firms are more tax aggressive than smaller family firms if we consider the logarithm of total assets as a size measure (p-value ≤ 0.01), that stand-alone family firms are less tax aggressive than family firms that belong to groups (p-value ≤ 0.05), that firms with main operations in the same place as the headquarters are more tax aggressive than firms with activities dispersed in different places (p-value ≤ 0.05), and that older family firms are more tax aggressive than younger family firms (p-value ≤ 0.05). This may imply that first generation family businesses are less tax aggressive than family businesses with generations beyond the first. Finally, there was no significant relation between industry and tax aggressiveness.

Results
In Model 2, we added our independent variables, but considering the two main management control mechanisms separately: strategic planning, that allows to evaluate in advance the impacts of tax aggressiveness strategies, and reporting system, that focuses mainly on performance measurement and control. With regard to control variables, we mainly found the same results as Model 1, with the exception of firm age and belonging to a group, that were not yet significant. We also found that family firms with higher ownership concentration are more tax aggressive than family firms with a less concentrated ownership (p-value ≤ 0.01), that family firms with independent members in the board are more tax aggressiveness than family firms with only family board members (p-value ≤ 0.05), and that family firms which adopt strategic planning are less tax aggressiveness than family firms which do not use this managerial mechanism (p-value ≤ 0.05). Instead, we did not find any relation between tax aggressiveness and both family CEO and reporting system. These findings support our hypotheses #1 (Family firms with more concentrated ownership are more tax aggressive than family firms with less concentrated ownership) and #5 (Family firms with strategic planning systems are less tax aggressive than family firms without). Hypothesis #3 (Family firms with independent directors are less tax aggressive than family firms with only internal family directors) is not supported, while hypotheses #2 (Family firms in which the CEO is a family member are less tax aggressive than family firms in which the CEO is a non-family member) and #4 (Family firms with managerial reporting systems are less tax aggressive than family firms that have not) are not validated. As far as Model 3 is concerned, we introduced an interaction term that considered the combination of both strategic planning and managerial reporting systems in order to better understand the possible moderation effect of the adoption of a wide range of managerial mechanisms (i.e., managerialization) on the tax aggressiveness behavior of family firms. Research findings confirm those obtained in Model 2, but strategic planning was not yet a significant variable. Instead, we found a positive relation between tax aggressiveness and managerial reporting system and a negative one with the interaction of two managerial mechanisms (managerial reporting system and strategic planning system) (both relationships: p-value ≤ 0.05).
The moderation effect of the presence of strategic planning on the relationship with the presence of the managerial reporting system and net cash effective tax rate was examined. The result indicates that there is a significant moderation effect (Delta R 2 0.01). More specifically, family firms presenting managerial reporting systems have a negative effect on their net cash effective tax rate but that negative effect (−0.240) is removed/mitigated when they also show strategic planning systems, since the interaction term is positively significant (+0.245). The effect size of the moderation is relatively small (Delta R 2 0.01). As [97] argued, however, even a 1% increase in R 2 is not trivial. As such, the results of this study show the importance of considering the moderation effect when investigating the hypothesized relationships with managerial reporting and strategic planning systems. Therefore, Model 3 results confirm our hypotheses #1b (Family firms with more concentrated ownership are more tax aggressive than family firms with less concentrated ownership), while they do not confirm hypotheses #3 (Family firms with independent directors are less tax aggressive than family firms with only internal family directors) or #4 (Family firms with managerial reporting systems are less tax aggressive than family firms that do not have), while hypothesis #2 (Family firms in which the CEO is a family member are less tax aggressive than family firms in which the CEO is a non-family member) is not validated again and hypothesis #5 (Family firms with strategic planning systems are more tax aggressive than family firms that have not) becomes not validated too.

Additional Analyses
We conducted two additional analyses to check for the robustness of our model. First, we used an alternative measure of the independent variable of interest, NCETR ( Table 4). Instead of using the ratio between cash taxes paid and pre-tax income net of special items, we used the ratio between total tax expense and pre-tax income (ETR). Second, we used another alternative measure of the independent variable of interest, NCETR ( Table 5). Instead of using the ratio between cash taxes paid and pre-tax income net of special items, we used the ratio between cash taxes paid and pre-tax income (CETR). The two analyses confirm the relationships found in our model (see Table 3).

Discussion and Conclusions
In this paper, we contribute to research on tax aggressiveness by studying its antecedents in a context-that of private family firms-that has been largely neglected in prior research on tax aggressiveness.
We believe that looking at tax aggressiveness in private family firms is important both theoretically and empirically. Theoretically, as existing research tends to focus on public companies, a call for research that further investigates in different contexts the relationship between tax aggressiveness, on the one side, and insider control and ownership structures, on the other side [26] has been launched. In our opinion, it is worth also investigating private family businesses as they represent a large portion of the operating companies all over the world. Empirically, understanding what drives tax aggressive behaviors in family firms may help curb these behaviors.
We contribute to research on family firms by providing evidence that some family firms are more tax aggressive than others. In particular, we consider as main aspects that may have an impact on tax aggressiveness the kind of family involvement in ownership, management and governance, and managerialization (i.e., the adoption of managerial mechanisms such as reporting systems and strategic planning). FIO, FIM, and FIG are relevant aspects in the light of an agency perspective, while managerialization can be considered relevant as managerial tools may help to consciously plan and manage the impacts of tax aggressiveness and to detect main related costs and benefits. Moreover, managerial control system represents an important corporate governance tool [80].
Our research findings suggest that both high FIO, that is high ownership concentration, and less FIG, that is the involvement of independent members in the board, increase family businesses' propensity towards tax aggressiveness. In fact, we found that family firms with more concentrated ownership are more tax aggressive than family firms with less concentrated ownership. This evidence is not in line with the literature [31] which proposes that family members with a large ownership stakes are likely to protect SEW, showing a lower motivation to make tax aggressive decisions. Instead, our results suggest that in a private family firm, a high ownership concentration implies different types of agency conflicts (not only between dominant and minority shareholders) [20] that may explain the interest of the majority shareholder to engage in tax aggressiveness. Moreover, our findings confirm that in private family businesses, family owners are the actors who mainly determine the extent of tax aggressiveness [17]. Another explanation for these results may be the high level of taxation and tax avoidance and evasion in Italy, which may explain owners' choices towards tax aggressiveness.
Findings regarding the role of independent members on the board are in turn consistent with the context of private family firms. In fact, we found a statistically significant association between the presence of independent directors in the board and tax aggressiveness. Family firms with independent directors in the board are more tax aggressive than family firms with a family board. This evidence is not in line with the literature, which however focused on public family firms. For instance, [22] (p. 354) documented how the presence of independent members sitting on the board of directors moderates the relationship between having a family member CEO and tax aggressiveness, so that "the CEO ownership share no longer affects the tax aggressive behavior of the firm if the board includes an independent outside director". Our findings may be explained by a weak monitoring role of board's independent members in private family firms with respect to public ones. In fact, in private family firms board independent members are mainly selected from people trusted by the owners, such as certified public accountants, tax advisors, or consultants. They are asked to provide strategic advice rather than protecting minority shareholders. Finally, their number is usually quite low relative to the total number of board members (as in the case of our sample) and this situation does not allow them to have a significant impact on decisions.
As far as FIM, that is the presence of family CEO, is concerned, we did not find any significant relation with tax aggressiveness. This result runs against our expectations, as we were expecting firms in which the CEO is a family member to be less tax aggressive, as suggested by prior research [22] This result can be explained by the fact that in private family firms, the family CEO is often also the majority shareholder or, in any case, is aligned with the owners. Therefore, no conflict of interest between shareholders and managers emerges and thus the CEO does not play a different role with respect to the shareholders with regard to strategic choices and, therefore, also with regard to tax aggressiveness.
Coming to managerial mechanisms, we may say that they are related with tax aggressiveness even though the significance of this relationship is not very high and the relationship changes depending on what mechanism is analyzed and whether the mechanisms are considered in isolation or jointly. If we consider managerial reporting systems and strategic planning separately, only the first shows a positive and significant relation with tax aggressiveness. Our study, in fact, highlights that the presence of reporting systems is related to a more tax aggressive attitude from family firms. This result can be explained by the aim of reporting system itself, that is used mainly as short-term performance measurement and control tool that can monitor the effects of tax aggressiveness. Instead, the presence of strategic planning determines a negative effect on tax aggressiveness, as well as the combined presence of reporting and strategic planning. This may be due to the fact that a complete set of mechanisms can assure an effective management of tax aggressiveness impacts not only in the short term, thanks to the use of reporting system, but also in the long term, as a result of the adoption of strategic planning. In fact, our findings can be explained by the role of strategic planning as a managerial tool that allows companies to assess the long-term impact of tax aggressiveness. The use of strategic planning, or a comprehensive managerial control system that includes both reporting and strategic planning, can be an element that reduces tax aggressiveness in that a managerial firm is better able to assess and control the costs and benefits of tax aggressiveness and thus can make more informed decisions. In fact, private family businesses, through strategic planning and management control systems, can better evaluate in advance the unique trade-off between SEW and financial gains and losses [34].
Consequently, in the long term, they take more conscious decisions which may favor SEW preservation instead of financial returns.
Coming to our control variables, we found a positive association between belonging to a group and being a firm with main operations in a same place and tax aggressiveness. Moreover, our findings suggest that older as well profitable and larger private family businesses are more tax aggressive. However, firm age and belonging to a group are not significant when we consider all variables, in Models 2 and 3.
These findings are not consistent with previous studies which found that stand-alone firms use their individual accounts both for tax and for reporting purposes [18] and therefore have more incentives than firms belonging to groups to manage the taxable income reported in this document. The positive association between tax aggressiveness and having a firm's main operations located in the same place as the headquarters may be explained by the limited exposure to different national contexts that can foster policies of tax aggressiveness. It is noteworthy that a family firm embedded in its local community is expected to be more concerned with the impacts on its reputation and therefore should tend to preserve SEW rather than to increase short-term economic results. Our findings seem not to support this consideration. Moreover, as suggested by [27], younger firms are likely to be in the hands of the first generation and are less likely to be interested in tax aggressiveness. In fact, they are characterized by limited or no agency conflicts, both classic ones between owners and managers and others that characterize family firms. The majority shareholder and CEO therefore have less interest in tax aggressiveness. The positive association between tax aggressiveness and firm size is in line with previous studies which suggested that differently from large firms "very small, young private family firms may not have the experience to engage in tax aggressive behavior and are fully occupied with the core business and/or survival of the firm" [22] (p. 353). With regard to the statistically strong positive association between profitability-measured as ROA at the end of the fiscal year-and tax aggressiveness, this association is in line with prior research (e.g., [97]), which shows that profitable firms are more likely to reduce taxes and adopt tax aggressiveness policies.
Our findings are not consistent with those of [82] that, in private family firms, suggest a mitigation role of agency conflicts played by governance controls. Instead, taken together, our results show that concentrated ownership accompanied by independent governance and the use of reporting systems increase the likelihood of tax aggressiveness in private family firms. In contrast, the adoption of strategic planning or a complete set of management control mechanisms can reduce tax aggressiveness.
As a whole, our results suggest that private family firms represent a quite different context than public family firms. The literature proposes that compared to public family businesses, private family firms with a concentrated ownership are more interested in preserving the SEW in the medium-long term than in the short-term economic returns that tax aggressiveness could favor. On the other hand, the presence of independent members who use managerial tools to measure performance, such as the reporting system, represents an effective mechanism for monitoring and controlling agency conflicts between majority and minority shareholders and any other type, typical of family businesses.
In contrast, our findings seem to suggest that private family businesses with a concentrated ownership, even though they have independent members on the board and use reporting systems, are more willing to engage in tax aggressiveness, being more interested in short-term financial returns. As suggested by an agency perspective, rent extraction and short-term financial returns are considered more relevant than preserving SEW and firm reputation in the long term. This behavior can be found when the private family firm belongs to a group, has its main operations concentrated near the headquarters, or is a profitable firm, a large company, and an old enterprise where many generations are involved. This can identify a context where different types of agency conflicts may emerge among family members in different roles, arising from asymmetric altruism, but in particular between majority and minority shareholders [20]. Moreover, in private family firms, costs of tax aggressiveness are lower than in public ones and it is easier for majority shareholders to expropriate minority shareholders because minority shareholders cannot influence potential prices discount, as the firm is not a listed one. Furthermore, private family firms can be less accountable and transparent than public firms. Therefore, they can avoid minority shareholders' opposition and can be less scrutinized by tax authorities. However, these considerations are valid if a short-term perspective is considered, when mainly the benefits of tax aggressiveness are evaluated. On the other hand, if we consider a medium-long-term horizon, the trade-off between costs and benefits of tax aggressiveness is better assessed and, therefore, the preservation of SEW and corporate reputation seem to become more relevant, as witnessed by the negative impact of mechanisms such as strategic planning on tax aggressiveness.
From a theoretical perspective, our results confirm that the antecedents of tax aggressiveness in private family firms can be identified only combining the different theoretical frameworks of both agency theory and SEW.
This allows us to consider the main characteristics that are relevant in explaining different behaviors of private family firms in terms of tax aggressiveness: different kinds of agency conflicts, the attitude towards preserving the SEW, and firm managerialization. This last aspect has never been explicitly considered in previous studies, although it is important to ensure proper planning and management of the benefits and costs of tax aggressiveness. However, our findings highlight that agency theory may represent a relevant framework to explain tax aggressiveness in the short term, while SEW is particularly useful when a long-term horizon is considered.
From a practical perspective, our findings highlight that different benefits and costs of tax aggressiveness can be found in different types of family businesses, depending on both the degree of family involvement in ownership and governance, and the use of managerial mechanisms. In particular, our results show that effective governance requires not only a "really" independent board but also the use of managerial mechanisms that can support the board in its monitoring and control role.
Despite the contributions highlighted above, the present study has limitations that need to be acknowledged and that can also provide further opportunities for future research. First, a panel design would have been preferable. Yet, as some of the data we used in our analyses had to be hand-collected, this limited the possibility of relying on a longitudinal research design. The use of lagged data for our dependent variable, however, should alleviate-although admittedly not completely dissipate-concerns about reverse causality. Future studies could test the robustness of our findings using a panel data design, controlling for the stability of this variable. Second, our results may be culturally constrained, as our data were collected exclusively in one country (i.e., Italy). In particular, the Italian context is characterized by high taxation and avoidance and tax evasion compared to other countries and this may have affected our results. Future research may investigate whether our results hold in other geographic settings, in which family firms may or may not be the predominant organizational form and that are characterized by different tax systems and levels of tax avoidance and evasion.
Third, a variable, that has not been included in our model but that may be of interest for the research is the change in tax regulations over the years. Even if it is true that in the domain studied, Italy, the corporation tax regulation has been quite stable over years, this element can be included in further research considering a longitudinal analysis.
Fourth, future research should go deeper into different tax aggressiveness behaviors that private family firms may adopt with either a short-term or a long-term perspective. Moreover, the relationship between tax aggressiveness and different features of private family firms (different types of agency conflicts, SEW, FIO , FIG, and FIM) should be further investigated.
Moreover, it could be worth investigating tax aggressiveness and leverage considering the positive and significant effect on earnings management. Previous studies indicate that companies with low tax rates have an indication of high tax aggressiveness and this is possible because of the earnings management [98].
Earnings management can also be carried out in order to smooth out the variability of profits, change the market's perception of the risk of investing in company securities, and, in this way, reduce the cost of capital [99].
Finally, it is necessary to further study the role of managerialization in tax aggressiveness, analyzing in greater depth the role of various managerial mechanisms, both in isolation and in combination.