How to Achieve Tax Compliance by the Wealthy: A Review of the Literature and Agenda for Policy

Tax compliance by the wealthy is relevant not only because their contributions are essential to maintain public budgets and social equality, but because their (non)compliance behaviour and the perceived (un)fairness of their contributions can fuel social unrest. In this paper, after giving a brief history of taxing the wealthy, we review the existing theoretical, empirical and policy literature on their tax compliance. We discuss how and why the wealthy differ from less affluent taxpayers because of specific interrelated political, social and psychological conditions. Understanding the psychological mechanisms that determine the tax compliance of the wealthy can provide policy insights on how to better integrate the wealthy in the tax system. Therefore, the present review is also a starting point for new policy approaches to increase tax compliance and tax morale among the wealthy.

and Bregman (2017) emphasize that the real problem of our time is tax avoidance by the rich who do not pay their fair share (see, e.g., recent 2019 World Economic Forum in Davos). Bregman, for example, emphasized the importance of taxes compared with the philanthropic schemes of the rich. 1 Tax compliance of the wealthy not only directly impacts a state's capacity to finance public goods, but it also influences the tax compliance of the general population and can be the cause of social and political turbulence (for historical examples, see Adams, 1993;Finer, 1999;Webber & Wildavsky, 1986). Recent examples are the "Occupy Wall Street" protest in the United States in 2011 or the "Mouvement des Gilets Jaunes" in France in 2018. Tax changes have become a divisive issue centered around fairness in which some politicians regard efforts to increase taxes as "class warfare," whereas others consider lower taxes for the wealthy as balancing the budget on "the backs of the poor" (Slemrod & Bakija, 2000, p. 50).
The wealthy's tax behavior is also socially important because they, by personifying society's measures of success, prompt other citizens to imitate their tax behavior (Fassin, 2005). This role model function, interpreted from an evolutionary perspective, is a strategy to improve survival chances by learning from those perceived as the best models, whose habits, styles, goals, and motivations are worth imitating (Henrich, 2015, p. 120). Thus, if accusations of tax fraud by sports stars, Chief Executive Officers, and politicians violate ordinary citizens' tax morale, these latter then start questioning the reasons for their own tax honesty. Massive fines for tax evasion rarely harm their fame and positive image, or even the role model function. For example, a fine of 18.8 million Euros imposed on Portugal's football superstar Cristiano Ronaldo did not diminish the cheers and adulation after a brilliant hat trick in the World Cup.
For their part, the wealthy do contribute substantially to the tax pool. As an example, the top 2.7% of the income bracket in the United States pays about 51.6% of total income taxes (Desilver, 2016), while in Germany, the top 5.6% contributes 43.25% (Bundeszentrale für Deutsche Bildung, 2013). Even taking into account the high portion of indirect taxes (e.g., value-added taxes) in total tax returns (between 30% and 55% in the EU; Carone, Schmidt, & Nicodème, 2007) paid mostly by the middle and lower classes, rich individuals' contributions are essential for financing public goods (OECD, 2008) such as infrastructure or health care. The problem is, as empirical data show (e.g., E. Hofmann, Voracek, Bock, & Kirchler, 2017b), that the motivation to engage in tax evasion and avoidance increases with wealth. Many wealthy individuals also support initiatives to reduce their tax contributions (e.g., Tea Party protests, see Martin, 2015) and promote alternatives to tax payments such as the philanthropic system (Giridharadas, 2019). Thus, understanding the political macro, social, meso and individual micro mechanisms that determine and psychologically motivate the wealthy to pay taxes are essential to maintain and increase state budgets and social cohesion.
Despite the importance of the subject, social science (and particularly psychological science) remains surprisingly silent on the topic. Most tax research focuses on compliance by average citizens, with only a limited number of studies explicitly comparing the compliance behavior of the wealthy with that of the middle or lower class. However, the wealthy are different from the average citizen in the sense that they not only have access to different political and legal possibilities, opportunities, and incentive structures, but also have different social environments and individual dispositions that are relevant for their tax behavior. The aim of the present review is to draw attention to these differences and their psychological origins and expressions, thereby highlighting the importance for more differential tax research and tailored tax policies.
The present article starts with an historical overview. This overview shows the importance of tax collection from the wealthy and demonstrates that the status quo is by no means unchangeable. We then move to contemporary tax research and give a definition of tax compliance before examining the empirical evidence, indicating that (on average) the wealthy are less tax compliant than less affluent taxpayers. Based on a review of the interrelated political (macro), social (meso), and individual (micro) factors, we discuss psychological causes, research gaps, and practical solutions concerning the lower tax compliance of the wealthy. Among other things, we show how the political and legal macro level allows the wealthy to "morally disconnect" from their own tax behavior and therefore from their impact on society. On a meso level, their ability to hire highly skilled tax practitioners transforms their tax decisions into a group decision with specific group dynamics allowing to "optimize" their tax behavior. Also, on the micro level, the wealthy differ from average taxpayers as wealth and status go together with specific personal values, which likely increase reactance to taxation. We argue that the entire range of these peculiarities calls for more tailored policy approaches, which (as our historical overview shows) can be built on good examples from the past. Finally, we discuss how the classical coercion-based and legitimacy-based instruments that are used to influence tax compliance can be applied to address the peculiarities of the wealthy. We claim that for each level-the macro, meso, and micro-a specific combination of hard coercive-based and soft trust-generating legitimacy-based measures is necessary to achieve tax compliance from the wealthy.
In this article, we highlight innovative social psychological research in addition to reviewing literature from a wide range of other academic disciplines (e.g., economics, sociology, political science, history) and practitioners (e.g., OECD, tax administrations). Applying this multidisciplinary approach allows us to generate new ideas that go beyond expensive and hard-to-implement tactics designed to foster tax compliance of the wealthy, such as international cooperation in closing legal loopholes given the armada of tax havens that offer tax "saving" schemes to the rich. Understanding the specific psychological differences between wealthy and average taxpayers and their causes is important to policymakers whose job is to increase compliance in these specific target groups. Such understanding is also relevant for researchers interested in cooperation, poverty reduction, inequality, and behavioral interventions in public management.

A Short History of Taxing the Wealthy
The history of taxes and the evolution of human societies are closely intertwined. In early societies, those with power created economic bottlenecks in trade routes so that they could collect payments from merchants in return for safe passage (Pennisi, 2012). These tax earnings were used to defend and further extend their rule (Pennisi, 2012) making stable finances the foundation of state power (Davies & Friedman, 1998;Webber & Wildavsky, 1986). For instance, from 5,000 BCE onward, the flourishing Egyptian culture had its own sophisticated tax system (Davies & Friedman, 1998) with the Rosetta Stone (inscribed around 200 BCE) as its most famous artifact, being a tax document granting exemptions to priests (Carlson, 2005) and reporting the reactions to a tax revolt (Adams, 1993). The Pharaoh regularly collected taxes from officials and ordinary citizens, with special levies as needed (e.g., for military campaigns; Ezzamel, 2002). To avoid the risk of scribes enriching themselves by cheating taxpayers, pharaohs increased the salary of their scribes (Adams, 1993). In addition, scribes were instructed to conduct tough enforcement strategies, but also to act kindly and were granted autonomy to reduce the tax for poor farmers (Adams, 1993, p. 8). Thus, even in ancient Egypt, the importance of a well-paid, and therefore trustworthy and competent tax administration was recognized as a key factor for successful tax collection.
Ancient Greece developed a tax system that depended on a sophisticated administration, but was also strongly based on social norms and social enforcement, in which wealthy citizens were expected to make voluntary contributions to various state projects (Reich, 2018). This so-called Liturgical system encouraged the rich to compete for honor and gratitude in a way that led to public improvements and beautiful buildings; that is, the liturgy. Public buildings, amusements, and, in particular, military equipment were purchased by rich citizens and donated to the city (Adams, 1993). An interesting feature of the liturgical system was the so-called "antidosis procedure": A wealthy person who was assigned a liturgical service could attempt to resist to perform the liturgy by nominating another, wealthier person who had not performed any significant liturgy recently. In cases where the two could not come to an agreement, a court would decide which of the two would carry out the liturgy (Reich, 2018).
With the Industrial Revolution, land and property were used to obtain credit, to invest and earn income based on profit (Seligman, 1913). The new developments allowed banking families like the German Fuggers to gain immense wealth. Wealth creation and contribution to state finances were intertwined. The bankers lent money to the Kings who often paid back their loans not with money, but by granting mining or other monopoly rights (Graulau, 2008). However, sometimes, the Kings defaulted; for example, around 1,600 CE, the Spanish King shirked his debts, thereby making the Fuggers pay for his wars. The Habsburg dynasty was particularly notorious for defaulting (e.g., five times just in the 19 th century, see Gasser & Müller, 2012). Thus, especially in times of war-and hence need-states took money away from the wealthy for state reasons.
In 1798, England needed to fund the Napoleonic Wars, which may have provided the motivational base to invent the first known income tax. This was a progressive system in which those with lower income paid less than those with higher income (Cooper, 1982). The new idea was probably influenced by the concurrent rise of the labor movement (Aidt & Jensen, 2009) and mirrored Smith's (1937, p. 777) similar proposal in his Wealth of Nations that "the subjects of every state ought to contribute towards the support of the government, as nearly as possible, in proportion to their abilities." Thus, although it was first implemented only temporarily as "national defense levies," the progressive income tax soon became the primary source of national finance in most countries (Steinmo, 2003).
In the early and mid-20 th century, progressive income taxation was extreme by today's standards, with national finances based on taxes from corporations and from fewer than 5% of the wealthiest citizens (Steinmo, 2013). In Spain, only 1,500 individuals paid taxes in 1933 (Alvaredo & Saez, 2009). In the United States, the Second World War increased national expenditures more than 12-fold (Steinmo, 2003), which meant that taxes on the easily identified rich were no longer sufficient, and the government began to include the identifiable income of an increased number of industry workers in the tax collection. To gain the acceptance of the workers, the marginal tax on the wealthy was pushed up to extreme levels, while the tax thresholds were substantially lowered (Steinmo, 2003), with a major propaganda campaign linking all income tax to the war effort (Jones, 1988). Hence, while only the richest in the United States paid income tax before 1930, by the end of World War II, 60% of income earners paid (Steinmo, 2003), and after the war, income tax was no longer an exclusive tax for the rich but a tax for the masses (Jones, 1988). Nevertheless, at this time, the rich were paying more than ever, with a 1957 U.S. federal individual income tax of 91% on incomes over USD400,000 (equivalent to around USD3,500,000 today; Slemrod, 2000) compared with the current rate of 39.6%.
After the Second World War, income taxes were maintained rather than being rolled back (Steinmo, 2003), with both politicians and economists positing that the state should manage the capitalist society through tax regulations. This political consensus ended in the early 1980s with representatives of neoliberal thought such as Friedrich von Hayek and Milton Friedman (Mirowski & Plehwe, 2015) questioning whether the complex tax code was really fair, and whether it helped accomplish the government's goals (Steinmo, 2003).
Then, in the mid-1980s, U.S. taxes were cut dramatically, resulting in an increase in public deficits. Further questions were raised about the state's ability to manage society, a trend that went hand in hand with disregard for a progressive income tax system (Steinmo, 2003). Because of increased income tax evasion (accelerated through the increased opportunities of a globalized financial market), consumption taxes-being hard to evade-became more popular (Graetz & Wilde, 1985), while at the same time, countries began competing to attract foreign capital through tax exemptions (Devereux, Griffith, Klemm, Thum, & Ottaviani, 2002). Across the globe, countries began following the U.S. tax reforms, lowering taxes while financing income tax reductions by increasing other taxes (e.g., VAT, FICA; Devereux et al., 2002;Steinmo, 2003). As a result, the average tax rate of 26.38% for the top 400 highest earning taxpayers in 1992 fell to 23.13% in 2014 (IRS, 2014); since Donald Trump's tax reforms began, this group now has lower effective tax rates than any other group in the United States (Saez & Zucman, 2019). The same trend is observable in Germany where the tax rates on capital companies declined from 53% in 2000 to 42% in 2005 (Hartmann, 2011). On the other hand, the tax burden of the masses increased with the rise in value-added taxes, so that in Germany today, around 50% (compared to 40% in 1990) of total tax monies come from value-added taxes, with only around one third from income tax (compared to 40% in 1980;Hartmann, 2011).
Overall, this history of taxation identifies several factors that facilitate the raising of tax revenue from the wealthy: (i) the ability of the state to identify wealth, thus taxable assets, (ii) a professional tax administration, (iii) budgetary necessity, often related to war, and (iv) political, social, and intellectual trends. If several of these factors come together as they did in the United States during the world wars, with the war against the Nazis generating patriotism and a common social drive, extremely high taxes can be collected from the wealthy (Steinmo, 2003). Thus, what seems to help is a "common purpose" under which the community can assemble and bond. Humans are "groupish" (Boyer, 2018) as within-group cooperation tends to favor success in intergroup competition (Henrich, 2015). Thereby, the statement of Nobel prize winner Joseph Stiglitz that "The climate crisis is our third world war" (Stiglitz, 2019) might be the adequate first step in creating such a new narrative and vision under which the global community, including the wealthy can unite, a narrative that is strong enough to even increase tax contributions of the top 1%.

What Is Tax Compliance?
In the present review, we use the term tax compliance in general terms such that high-versus low-tax compliance means that individuals pay more compared to less tax to the state. Tax compliance can be differentiated into a motivational and behavioral component (Kirchler, 2007). Tax compliance motivation is defined as the individual willingness to comply with the tax law. The literature typically differentiates (as we do in our research) between the sources of motivation: enforced, voluntary, and committed tax motivation represents the continuum between the two broad angles of extrinsic and intrinsic motivation (Feld & Frey, 2007;Gangl, Hofmann, & Kirchler, 2015;Ryan & Deci, 2000, Torgler, 2007. Enforced motivation means that someone only pays taxes because of the fear of audits and fines. Voluntary motivation means an individual gives in to the tax law and pays because it is easier than evasion. Committed motivation drives someone to pay taxes because of a felt moral duty (Gangl et al., 2015;Koessler, Torgler, Feld, & Frey, 2019) or due to emotional stress, probably related to anticipated guilt or shame (Blaufus, Bob, Otto, & Wolf, 2017;Dulleck et al., 2016).
Tax compliance behavior refers to the concrete behavioral compliance with specific tax laws (Gangl, Hartl, Hofmann, & Kirchler, 2019;Kirchler, Maciejovsky, & Schneider, 2003) such as honest and timely payment and tax filing (i.e., tax honesty), proper and transparent handling of documents (i.e., administrative compliance), registering as a taxpayer (i.e., tax filing), legal exploitation of the tax law (i.e., tax avoidance), paying less than the statutory tax (i.e., tax evasion), or criminally exploiting the tax law (i.e., tax fraud).
Although a coherent typology of tax compliance motivations or behaviors does not exist, many studies conclude that there is a positive relationship between tax motivations and tax behaviors such that an intrinsic motivation to be an honest taxpayer also should result in higher tax honesty or lower tax avoidance (Gangl et al., 2015;Torgler, 2007;Wenzel, 2005). However, compared to tax motivations, which can be assessed with specific questionnaires (e.g., Kirchler & Wahl, 2010;Torgler, 2016), the assessment of tax compliance behaviors is inherently difficult, operating as it does in the shadows. Therefore, real tax compliance behavior can only be estimated, even by the revenue bodies. In fact, due to complex tax laws, taxpayers themselves often do not know whether they are honest or dishonest (Kirchler, Niemirowski, & Wearning, 2006), with some believing themselves honest but actually evading taxes, while others report evading taxes but are, in fact, honest. To study tax compliance, therefore, a multimethod approach is used in which revenue data, survey data, experimental data, and qualitative methods are combined to understand self-reported motivations and behaviors.

The Wealthy Are Less Tax Compliant than Average Taxpayers
The empirical evidence is relatively clear, no matter whether tax compliance motivation or behavior is examined or what method is used: almost all studies report that the wealthy are (on average) less tax compliant than middle-class citizens. Research using World Values Survey data shows a negative relation between income and tax motivation (Doerrenberg & Peichel, 2013), although our own research indicates that the results are mixed when different regions are considered separately (for an overview, see Torgler, 2007). A meta-analysis of 334 survey studies also reports a negative relation between income and self-reported tax compliance (E. Hofmann et al., 2017b). Furthermore, E. Hofmann et al. (2017b) demonstrate that this negative relation is stronger in Eastern Europe and Central Asia than in other world regions. Most notably, the negative relation holds even when the data sets may not sufficiently capture particularly wealthy taxpayers (e.g., the top 5% of income earners).
Studies based on tax revenue data that control for opportunity to evade and for tax rates also find a negative relation between income and tax honesty (Crane & Nourzad, 1986;Feinstein, 1991;Pommerehne & Weck-Hannemann, 1996). Likewise, recent studies that match wealth records from Norway, Sweden, and Denmark with microdata leaked from two large offshore financial institutions (HSBC Switzerland and Mossack Fonseca) indicate that tax evasion increases sharply with wealth, with the top 0.01% of the wealth distribution (i.e., households with more than $40 million in net wealth) evading about 30% of their income and wealth tax versus 3% by taxpayers overall (Altstaeder, Johannesen, & Zucman, 2017). Another study using IRS data of around 55,000 taxpayers concludes that the relationship between income and tax compliance has a reverse U-shape (Cox, 1984), meaning that, on average, the most noncompliant taxpayers are those with either very high or very meager incomes, with middle-income taxpayers being the most compliant (Cox, 1984). On the other hand, reports from the United Kingdom claim that affluent taxpayers are more likely to submit their tax returns on time than other taxpayer groups (Tax Audit Office, 2016) In general, aggressive tax avoidance tends to be discussed in relation to companies, particularly with respect to large multinational corporations (e.g., Apple, Google, Starbucks, Amazon, Facebook) that engage in profit shifting from high-tax to low-tax jurisdictions (European Commission, 2016;Frijters, Gangl, & Torgler, 2019). Consequently, multinational enterprises pay relatively low amounts of tax (Crivelli, De Mooij, & Keen, 2016;Dharmapala, 2014), like the mere 10% tax on real profits paid by Google in the United Kingdom (Tax Justice Network, 2016). According to the OECD, an estimated USD100 to 240 billion are lost globally each year to the profit-shifting activities of multinationals (OECD, 2015a). Although OECD countries have begun implementing policies to reduce profit-shifting (BEPS), the success of these initiatives depends on each country's willingness to forego maximization of its own short-term financial benefit. Some suggest that a 10 percentage point reduction in a country's average effective tax rate would, in the long run, increase the stock of inward foreign direct investment by an average of over 30% (De Mooij & Ederveen, 2008).
Like large corporations, wealthy individuals are also widely involved in tax-avoidance activities, with top earners and football stars sometimes even moving to another country to avoid taxation (Kleven, Landais, & Saez, 2013;Kleven, Landais, Saez, & Schultz, 2014). However, most individuals may not only find such a step more difficult, but some star scientists (measured by the number of patents) and millionaires (Moretti & Wilson, 2017;Young & Varner, 2011) stay put simply because moving abroad also involves great costs; for example, costs involved in migration, living expenses, personal circumstances, and connections (Simula & Trannoy, 2010). Hence, rather than moving to another country, affluent individuals may prefer to avoid taxes by moving money to less-taxed assets (e.g., in the stock market, trusts, or real estate; Goolsbee, 2000) or by donating to charity (Peloza & Steel, 2005). On the other hand, as the so-called Paradise Papers show, wealthy individuals can also legally avoid taxes by moving money to offshore havens. Overall, however, in contrast to the tax loss from corporate tax avoidance, the loss from wealthy taxpayers stems from a combination of tax avoidance and tax evasion (Gravelle, 2009). For instance, the U.K. HRMC reports that their 2015 auditing and monitoring activities enabled the collection of £230 million from avoidance schemes, £140 million from disclosure of offshore facilities, and £80 from serious civil fraud, a clear indication that tax avoidance and tax evasion are equally important for tax loss among rich taxpayers (National Audit Office, 2016).
Finally, some wealthy also engage in criminal tax fraud, although they might frame it as a sort of clever tax avoidance. One such example are the cum-ex business models in which wealthy individuals (with the help of banks) claim unjustified tax refunds for investments, generating a total loss of 31.8 billion Euros for Germany (Ackermann et al., 2017). Thus, the simple business model of cum-ex investments is to withhold money from the public tax coffers by exploiting imperfections in the tax administration. A legal solution for the future would be to forbid any businesses whose profit is only generated by exploiting the tax law.
However, by considering the interrelated micro, meso, and macro conditions, governments can understand why the rich evade and avoid taxes, and therefore develop strategies to increase motivation to pay honestly. In the following, we present the political, social, and individual factors that are likely reasons for the wealthy's tax compliance and starting points for policy interventions.

Political and Economic Conditions That Shape Tax Compliance of the Wealthy
The macro context is shaped by international secrecy jurisdictions including tax havens, and generous national tax exceptions often related to philanthropic foundations. We argue that this political and legal environment not only offers the opportunity to evade and avoid taxes, but also creates a tax climate with an ideological setup that reduces the moral concerns of the wealthy when avoiding taxes.

Tax Havens and Secrecy Jurisdictions
Politicians, celebrities, billionaires, and sports stars, along with fraudsters and drug traffickers, use tax havens to hide assets in secret trusts (Weisbord, 2016). Such tax havens allow large-scale tax avoidance among the rich (Forsythe, 2017), with an estimated 80% of hidden assets escaping any taxation (European Commission, 2016). Wealthy taxpayer decisions to avoid and evade taxes might thus be related to the jurisdictive environment offered by their own or, more often, a foreign country. In fact, many nations, both large and small, are motivated to earn extra money at the expense of other countries by rather hypocritically motivating rich foreigners to commit tax evasion (Weisbord, 2016). Scholars such as Piketty (2014) suggest the automatic transmission of banking information as a solution, something he sees as a first step toward a global tax on capital.
The world's most important tax havens, according to the 2018 Financial Secrecy Index, are the following 10 countries (in descending order): Switzerland, USA, Cayman Islands (United Kingdom), Hong Kong (China), Singapore, Luxembourg, Germany, Taiwan, Dubai, and Guernsey (Islands in the English Channel). The vicious international competition (or "race to the bottom"; Sharman, 2006) to attract these funds is probably one reason that an estimated 8% of global financial wealth is placed untaxed in another country (Zucman, 2013). Although the developed nations (e.g., the EU) lose the largest absolute amount, developing countries (e.g., African nations) lose the largest fraction of their financial wealth to tax havens (European Commission, 2016;Zucman, 2013). Hence, in the context of these nations' evasion-friendly jurisdictions, large corporations and wealthy individuals might also avoid suffering moral compunction when shifting funds from one country to another. The existence of these jurisdictions offering financial secrecy allows moral disconnection and moral disengagement (Den Nieuwenboer & Weaver, 2019), thus generating the perception that the own tax avoidance behavior is legal and normal, something that does not violate moral values. To reduce these psychological evasion maneuvers, the state should formulate strict legal codes in order to clearly communicate expectations. Importantly, tax administrations should actively avoid euphemistic language sometimes used among tax practitioners, who talk about tax optimization, tax saving schemes, creative tax planning, or testing the limits to disguise the ethical implications of their actions (Tenbrunsel & Messick, 2004). Instead, tax administrations should clearly use the terms aggressive tax avoidance, illegal accounting practices, tax evasion, or tax fraud to show that such behaviors are socially disapproved.
The situation becomes even more complicated when lobbying organizations intentionally influence tax laws to include loopholes for the rich (McBarnet, 1992;Mirowski & Plehwe, 2015), an effect that may have increased greatly over recent years (Martin, 2015;Scheiber & Cohen, 2015). According to one analysis of the political elite and income tax in the United States since 1945, this shift in tax laws is marked by two phases (Hartmann, 2009). Between 1945 and 1980, two thirds of the political elite came from working class environments and tax rates for top earners were high; since 1981, however, almost 70% of the political elite have originated from the upper or upper middle class, while tax rates for top earners have declined (Hartmann, 2009). Although movements to cut taxes for the rich tend to be in the political majority (e.g., Trump and the Republicans in the United States, Conservatives in the United Kingdom or Germany; Hartmann, 2011;Martin, 2015), some wealthy individuals want to see their taxes increase. For example, 64 wealthy Germans belong to an association that lobbies for increases in such wealth taxes as the inheritance tax (www.appell-vermoegensabgabe.de), while in the United States, billionaire investment mogul Buffett (2011) publicly stated that those as wealthy as he should pay more taxes. In the same vein, in 2010, a group of 51 German millionaires and billionaires (the "Club of the Wealthy") unsuccessfully proposed to Angela Merkel that they should give up 10% of their income over a period of 10 years as a form of "rich tax." The two contradictory trends-that some wealthy people see tax rates as too low, whereas others see them as too high-mirror the mixed outcomes of research on the impact of tax rates on tax honesty. While some studies (Alm, 1999) argue that the estimated underreported income-tax rate elasticity is between 0.5 and 3.0, a large number of empirical studies find no support for tax rates as an explanation for the negative relation between income and tax compliance (Cox, 1984;Crane & Nourzad, 1986;Feinstein, 1991;Goolsbee, 2000;Pommerehne & Weck-Hannemann, 1996;Poterba, 1987). It may be that perceived fairness of the tax system (derived from knowledge of the own and others' true tax burden rather than the tax rate itself) explains a certain portion of lower tax compliance by the rich (Gangl, Kirchler, Lorenz, & Torgler, 2017;Lewis, 1978). The practical conclusion is that tax administrations should focus on increasing perceived fairness by informing taxpayers about their true tax rates, in addition to detailing expenditure of tax revenue.

Tax Exemptions for Philanthropic Foundations
The generous tax treatment of philanthropic foundations and charities may legitimize tax avoidance of the wealthy by reducing the moral obligation to contribute to society via taxes (Giridharadas, 2019;Reich, 2018). In many countries, but particularly in the United States, charity is heavily subsidized by the state with allowances for generous tax exemptions. For instance, in 2016, the United States faced $50 billion foregone tax revenue due to tax relief for charities (Reich, 2018). However, these tax exemptions are relatively recent-only since 1917 the United States has allowed tax deductions for donations to charity (due to interventions by Rockefeller; Reich, 2018). Now, initiatives such as the "giving pledge," in which billionaires such as Bill Gates promise to give away half of their fortune to charity, are increasingly received with a critical view (Giridharadas, 2019). These contributions are made outside democratic institutions and are often more likely to increase the power of the founder than they are to support more equal societies. Nonetheless, think tanks and lobbyist organizations legitimize charitable giving by disparaging collection of taxes, highlighting that the donors are engineers of success who created profitable businesses, and can (in the same manner) solve society's problems much better than the state (Giridharadas, 2019).
Moral licensing theory suggests that individuals who initially behaved in a moral way can later display behaviors that are immoral because they may believe that past good behavior frees them to do something bad in the future (Merritt, Efron, & Monin, 2010). Field experiments indicate that committing a moral act earlier in the day was associated with an above-average likelihood of a subsequent immoral act (W. Hofmann, Wisneski, Brandt, & Skitka, 2014). Other research on field data found that companies with a good reputation are more likely to engage in aggressive tax avoidance than firms with a poorer reputation (Bai, Lobo, & Zhao, 2017). For instance, the good reputation from charitable giving might function as a license that allows firms to accept reputational consequences of tax avoidance. Thereby, charitable giving also avoids moral dissonance between the own unethical tax behavior and the moral self-concept, which is seen as a major driver for moral behavior (Bastian, 2019). However, empirical research comparing tax systems with charity systems as well as research on the influence of charitable giving on tax compliance are scarce. The insights from the liturgy in Ancient Greece indicate that the contribution was linked to community spirit, public sentiment, duty, and public demands (e.g., for infrastructure such as bridges). As Athenian statesman Pericles (429 BC) stated in a famous funeral oration, "We regard wealth as something to be properly used rather than something to boast about. ( . . . ) Every one of us who survives should gladly toil on her [Athens's] behalf" (cited in Adams, 1993, p. 63). Sitaraman (2017, p. 302) emphasizes the misalignment between constitutional and economic structure in the United States due to economic and political inequalities, citing John Adams who in 1767 raised the concern that there is "so much Rascallity, so much Venality and Corruption, so much Avarice and Ambition, such a Rage for Profit and Commerce among all Ranks and Degrees of Men even in America, that I sometimes doubt whether there is public Virtue enough to support a Republic." 2 Thus, policy makers should consider how pragmatic management of large-scale philanthropy would function. Philanthropy should support democratic institutions instead of delegitimizing them by questioning their capability to solve societal problems. For instance, it is argued that foundations designed to operate beyond one's death should be limited or forbidden, that foundations should focus on testing new approaches, should be integrated into state policy and importantly, need approval of democratic institutions before implementation (Reich, 2018). In addition, tax administrations might learn from charity administrations on how tax contributions might be made more attractive for the wealthy, for instance, by granting honor and gratitude to large taxpayers (as in the Ancient Greece liturgical system). Most current tax systems focus on deferring tax evasion with shame and guilt rather than rewarding tax compliance with pride and honors; empirical studies on the shaming effect show that shaming increased the payments of individuals with small debts, but had no effect on individuals with larger debt amounts (Perez-Truglia & Troiano, 2018).

Special Enforcement Regimes
Increasingly, countries have specialized enforcement regimes for the wealthy. Ordinary taxpayers often complain that the rich and powerful do not pay taxes because they are not controlled or fined. Such beliefs are sometimes fostered by media reports such as a German government press release that the tax audit rate for the wealthy had declined from 1,838 cases in 2010 to 1,391 cases in 2014 (Deutscher Bundestag, 2017). On the other hand, in recent years, tax administrations in many countries (e.g., Australia, Spain, or Greece) have been more purposely targeting the rich. For instance, data on corporate audit probabilities from the German Ministry of Finance show that in 2015, 21.3% of large corporations were audited compared to 6.4% of medium corporations, 3.2% of small corporations, and 1.05% of tiny corporations (Bundesfinanzministerium, 2016). For the United States, detailed IRS (2016) data reveal that the 0.01% who earn USD10,000,000 or more had an audit probability of 18.79% on their 2015 tax return compared with an audit probability of between 0.41% and 0.80% for those with an income between USD25,000 and 200,000. Many countries also have specialized bodies that focus on auditing large corporations or even high net worth individuals (OECD, 2010(OECD, , 2015b, or that implement special measures to identify wealthy tax evaders, such as owners of luxury cars or villas and those who take many international flights (Casaburi & Troiano, 2016;. In the United States, the IRS has a whistle-blower program that pays informers who help to detect an evader (Davis-Nozemack & Webber, 2012). Germany also buys information on offshore tax havens from whistle-blowers to identify rich tax evaders (Reuters, 2014;Wittrock, 2012). Thus, in contrast to the public perception, large corporations and wealthy citizens may be more likely to be audited than average or poor citizens.
Nonetheless, the political, legal, and economic context of financial secrecy laws and national tax loopholes-including those for charities-not only give the wealthy many more opportunities to evade and avoid taxes than the average citizen, but might also create an ideological environment that legitimizes aggressive tax avoidance behavior.

Social Contextual Factors That Influence Tax Compliance of the Wealthy
The tax relevant social meso context of the wealthy is different from that of average taxpayers. First, compared to average taxpayers who often prepare their tax returns alone, most wealthy people have the help of tax advisors. Research claims that this professional support not only leads to tax returns that are less compliant, but it also creates social group dynamics that often accelerate opinions to the unethical extreme. Second, the social environment (and thus the social identity) of the wealthy is likely different from the social identity of average citizens, creating comparisons to other wealthy, and thinking along in-and out-group interests. We will now consider each of these concerns.

Tax Practitioners
Research classically models tax behavior as individual behavior. This assumes that it is an individual who decides whether or not to be honest on their tax return (Allingham & Sandmo, 1972), which might be true for many employed taxpayers and also small entrepreneurs, but it is generally not true for wealthy taxpayers. As outlined above, compared to ordinary taxpayers, wealthy taxpayers' tax decisions are made in groups involving tax practitioners and wealth managers. Those who assist wealthy taxpayers in exploiting national and international tax laws include professional tax specialists, banks, and international accountancy firms and wealth managers (Harrington, 2012(Harrington, , 2015Sikka & Hampton, 2005). Taxpayers who want to minimize their taxes and who are high risk-takers seek out tax agents who are adept at finding loopholes (Sakurai & Braithwaite, 2001). As one U.S. lawyer put it, "You have to understand, the smartest people in the country think 24/7 about the best tax saving schemes, which they seek to sell for millions of dollars to specific companies and rich individuals" (informal comment given to us at a conference). Thus, not only do the lower income classes have less ambiguous income than higher income classes (Klepper, Mazur, & Nagin, 1991; for instance, higher income classes may have revenues from entrepreneurial activities that are easier to conceal than income from employment), but "the poor evade [while] the rich avoid" with the assistance of their advisors (Slemrod, 2007). This stereotype of a wealthy individual who searches out and is convinced by the most aggressive tax advisors has some empirical support (Schisler, 1995). For instance, a study of 7,127 income tax returns found that those who used a tax preparer had lower tax liabilities, higher tax reductions, and lower prepayments than those who had no such help (Christian, Gupta, Weber, & Willis, 1994). Findings of another experimental study suggest that tax practitioners have lower moral reasoning about taxes than nonspecialists, a difference explained by tax advisors' specific professional environment (Doyle, Hughes, & Summers, 2013). Several other studies also report that the average level of noncompliance is higher for returns prepared by tax practitioners (Ayres, Betty, Jackson, & Hite, 1989;Erard, 1993), although there is also empirical evidence that tax practitioners are less aggressive than taxpayers (Schisler, 1995) or that taxpayers do not want tax practitioners to be aggressive (Hite & McGill, 1992). Nonetheless, tax professionals do tend to be more aggressive when audit and penalty risks are low (McGill, 1988). Overall, then, individuals conceal tax money through professional tax lawyers (Ackermann et al., 2017) and with the help of banks and agencies, such as Mossack Fonseca (the Panama Papers) or Appleby (the Paradise Papers; Weisbord, 2016).
In this context, an interesting sociological study was conducted by Harrington (2015) who spent about 8 years in participatory research and conducting in-depth interviews with wealth managers in tax havens all over the world. She reports that almost all of them saw themselves as misunderstood good guys. Their self-perceptions cast them as protectors of elderly clients from rapacious heirs, facilitators of development finance to emerging markets, and quasi-family members to wealthy parents seeking advice on how to prevent their children from being destroyed by idleness and easy access to drugs. Those wealth managers concerned about poverty urge their clients to donate to charity. Wealth managers allow the ultrarich personal freedom, mobility, and privacy-they keep them off the radar of regulatory authorities, which clearly indicates that policies to increase the tax compliance of the wealthy also need to consider the role of tax advisors and wealth managers.
Social psychological group research indicates that decisions made in small groups-even dyads-differ from individual decisions. For instance, in the context of economic games, and compared to individuals, groups make more rational and analytical (Kugler, Kausel, & Kocher, 2012;Luhan, Kocher, & Sutter, 2009) and more competitive and unethical decisions (Moore & Gino, 2013;Stawiski, Tindale, & Dykema-Engblade, 2009;Wildschut & Insko, 2007). Although comparatively rare, there is also research indicating that small groups do not always differ from individuals concerning competition and ethical behavior, or are even more altruistic than individuals. Power-to-take experiments (Bosman, Hennig-Schmidt, & van Winden, 2006) and lying experiments (Muehlheusser, Roider, & Wallmeier, 2015) did not find overall differences between groups and individuals. Another experiment observed that small groups in dictator games can also be more prosocial and altruistic than individuals (Cason & Mui, 1997).
Heterogeneity of results concerning ethical decision-making in groups thus might be the consequence of different dominant opinions that existed for the respective decision situations (Isenberg, 1986). It also shows, in contrast to some suggestions from the literature (Moore & Gino, 2013), that not all group decisions are more unethical than individual decisions, as groups mainly amplify existing joint preferences. This implies that tax authorities should use information on tax advisors and on past behavior to distinguish between dishonest and honest groups. For instance, this could be achieved by increasing the auditing intensity on wealthy taxpayers who are known to employ aggressive tax advisors, or on the wealthy who, based on screenings, demonstrate initial signs of dishonesty. In contrast, wealthy people who employ nonaggressive tax advisors or those who are known to have been honest in the past would not need to be subjected to such harsh audits.

Social Identity
For the wealthy, the perception of belonging to a particular group might be related to at least two social psychological processes that influence tax motivations. On the one hand, wealthy individuals might identify with and compare themselves to other wealthy individuals, and, on the other hand, they also might feel a social distance to less affluent individuals.
Enhanced identification and comparison with other wealthy individuals adjusts internalized social norms connected with the wealthy rather than to the ordinary individuals (Wenzel, 2005), motivating affiliation and compliance with images and perceptions of what rich people ought to do. The drive to be similar to the reference group applies not only to lifestyles but also to tax behaviors. If all wealthy friends move money to offshore tax havens, then the individual will also more likely do that. On the one hand, this is due to compliance with the in-group norms, but, on the other hand, it is also due to competition and comparison, because one does not want to fall behind in the financial race . Thus, tax evasion can be the result of fear of losing one's privileged position, or because taxes are a hindrance to upward mobility (Jetten, 2019).
Wealthy individuals' likely identification with other wealthy people is also relevant for the tax authorities, as wealthy taxpayers might be more focused on how the tax authorities treat other wealthy people and not the general population. Thus, harsh audits are more likely to be accepted if there is a perception that all wealthy people are subject to such audits, and thus, while the treatment might be harsh, it is also fair. A negative example would be the (publicly known) lower tax audit rates in the rich south compared to the poorer north of Germany (Balser, 2011). Currently, the southern German state of Bavaria employs 15% less than the recommended number of tax auditors. Thus, in Bavaria, the probability of having evasion detected is lower than in other German states (Balser, 2011). However, federal tax administrations should avoid the perception that different audit frequencies or procedures exist within one country, as this might generate a feeling of unfairness and enhance the motivation to move to another state to evade taxes.
Belonging to and group identification with the wealthy could also lead to felt social distance from nonwealthy individuals, or even to the perception of in-and out-groups (Cardenas, 2003;Kramer & Brewer, 1984;Taifel & Turner, 1986). Wealth increases real physical distance, with the wealthy tending to live in separated neighborhoods on large properties that do not allow much (spontaneous) contact with others, particularly with individuals from another social class. Such social distance reduces empathy and trust (Kraus, Tan, & Tannenbaum, 2013;Stellar, Manzo, Kraus, & Keltner, 2012) and increases the perceived difference between rich and poor, which leads to a decline in a felt shared identity (Poteete & Ostrom, 2004).
Classical social psychological experiments show that even minimal signs of group belongingness such as shared art preferences can lead to in-and out-group perceptions (Taifel, Billig, & Bundy, 1971). Wealth differences likely create much stronger signs of difference, which makes it possible that perceived out-group members are more disadvantaged, more often punished, and less rewarded than in-group members (Taifel et al., 1971;Vuong, Chan, & Torgler, 2018). In this vein, some wealthy individuals' agreement with the narrative that low taxes are good, or tax avoidance is okay, may also be the result of out-group derogation (Brewer, 1999;Brown-Iannuzzi, Lundberg, & McKee, 2017). The wealthy might argue that a low redistribution through taxes is justified; because in contrast to the poor, they are high achievers who use the money to create new firms and jobs (Jetten, 2019).
Practical interventions to reduce the perceptions of in-and out-groups could focus on increasing the likelihood of face-to-face contact between different social classes (Hewstone, 2015). Examples are the financing of social housing in rich neighborhoods, as well as providing excellent public schools, playgrounds, public parks, and sport facilities, or encouraging voluntary organizations such as the voluntary fire brigades.
In sum, as inequality increases, so does social distance, accompanied by a similar decline in shared identity among all citizens, the wealthy, and the poor. In turn, this affects empathy and trust and reduces cooperation in general and tax compliance in particular (Cardenas, 2003;Gangl et al., 2015;Kramer & Brewer, 1984).

Values
Values and attitudes vary with socioeconomic status (Brown-Iannuzzi et al., 2017). According to social-psychological research, the rich likely hold fewer egalitarian values than the average citizen (Piff, Kraus, Côté, Cheng, & Keltner, 2010) and prefer social hierarchy. For example, U.S. elites are less fair-minded than the general population and prefer efficiency to equality (Fisman, Jakiela, Kariv, & Markovits, 2015). Evidence from Germany indicates that acceptance for social inequality is also stronger among the rich who grew up in affluent families than among the rich who grew up in working-class families (Hartmann, 2013). The experience of relative poverty might increase empathy with the less affluent and the willingness to pay taxes to contribute to social services, which reduce financial hardship. Experimental research shows that income and wealth distribution seemingly create a social hierarchy that (especially) those on top may be motivated to sustain through less egalitarian values, and ultimately, less cooperative tax behavior (Magee & Galinsky, 2008;Oc, Bashshur, & Moor, 2015;Xie, Ho, Meier, & Zhou, 2017). For instance, one experiment showed that individuals are reluctant to redistribute money if the redistribution changes or reverses the relative ranking in a given hierarchy (Xie et al., 2017).
Research evidence that high socioeconomic status likely increases immoral behavior such as cheating, lying, or egoism (Cardel et al., 2016;Piff, Stancato, Coté, Mendoza-Denton, & Keltner, 2012) further suggests that wealth may be related to lower moral values. In fact, other experimental and survey-based studies indicate that the wealthy have less concern for others (Stellar et al., 2012) and more favorable attitudes toward greed . For instance, during the 2007 financial crisis, top-earning managers constantly increased their pay without delivering any additional benefit to shareholders (Haynes, Campbell, & Hitt, 2017). Nevertheless, it remains unclear whether or why the wealthier are greedier, given that income is positively related to financial satisfaction (Sahi, 2013). However, greed may not moderate the ethical behavior of either rich or poor (Balakrishnan, Palma, Patenaude, & Campbell, 2017): several empirical studies conclude that wealthy individuals are more prosocial and moral than less affluent individuals (Andreoni, Nikiforakis, & Stoop, 2017;Balakrishnan et al., 2017;Liebe, Naumann, & Tutic, 2017;Trautmann, van de Kuilen, & Zeckhauser, 2013), suggesting that there is no simple linear relationship between wealth and ethical behavior .
On the one hand, many of these studies demonstrating that the wealthy are more unethical than the poor only study relative wealth, and therefore may not permit conclusions on the super wealthy but only on people earning a bit more than the average. One suggestion is that moderately wealthy (but maybe also very wealthy) individuals' ethical behavior depends on felt entitlement and security concerning the own status and the perception that status group boundaries are permeable (Jetten, 2019;. Wealthier individuals behave more egoistically and harshly if they fear that they or their children could lose their status in the future (Scheepers, Ellemers, & Sintemaartensdijk, 2009). It could also be that they think that they need to be harsh to climb the social ladder even further . In contrast, if the wealthy feel secure in their position, they may also be more generous. From a practical perspective, such insights would suggest that tax authorities need to be aware that the wealthy have a sense of entitlement or deservingness and perceived need to protect both current and future wealth and status (Mols & Jetten, 2017, pp. 128-129). This can become particularly important in situations where unpredictability and instability increase (Jetten, Mols, & Healy, 2017). The goal of the tax administration and tax policy in general should not be to threaten their social position but rather to think of instruments that encourage intergroup cooperation, prosocial behavior, and empathy toward other groups while taking into account their "Achilles heel" or particular sensitivity around the fear of losing their status in future. This will also require a good understanding of how the perceived normative climate affects individual behavior (Sánchez-Rodríguez et al., 2019). In addition, as Bird (1995) points out, "[p]ractical tax policy is, and always will be, in large part an exercise in damage control" (p. 1041).
On the other hand, wealthy and less affluent individuals might not differ in their values but in their freedom to express them. That is, wealthy people are likely more able to show their true attitudes because they are not as dependent on others as poor individuals, who might face social pressures to suppress their true feelings (Na & Chan, 2016). Experimental evidence indicates that individuals who adhere to "tit-for-tat" rules (e.g., equivalent retaliation) become more self-interested when in power, while those who opt for a more communal strategy are more prosocial when in power (Chen, Lee-Chai, & Bargh, 2001). It therefore seems reasonable to suggest that, compared to people who feel poor, individuals who feel rich can express their values more readily and efficiently regardless of whether these are seen or perceived as good or bad by society and people around them. Thus, tax auditors can expect that wealthy clients communicate their values, interests, and attitudes. For instance, they may openly talk about their disregard or regard of the tax system, which, in turn, makes it easy to classify their true tax motivation and to target interventions accordingly. For instance, taxpayers who disregard the tax system need to be informed about the existence of professional auditing procedures while seemingly committed taxpayers should be thanked for their cooperation.

Cognitive Styles
A comparison between 130 German millionaires with a representative sample of the German population showed that the wealthy differ on some personality dimensions from average citizens (Leckelt et al., 2019). Wealthy people generally have higher scores on personality dimensions such as emotional stability, extraversion, openness, disagreeableness, and agentic narcissism, and were found to have a more internal locus of control. Individuals who perceived themselves to be in a high rather than low economic position enjoy an increased sense of personal freedom and control (Manstead, 2018). In other studies, the wealthy were reported to perceive themselves as having a greater ability to influence their own and others' social environment and to overcome external threats (Guinote, 2017;Keltner, Gruenfeld, & Anderson, 2003;Kraus, Piff, Mendoza-Denton, Rheinschmidt, & Keltner, 2012). Indeed, economic affluence is often equated with a feeling of power (Lammers, Stoker, & Stapel, 2010), which can lead to the illusion of control-sometimes over events that, in fact, cannot be controlled (Fast, Gruenfeld, Sivanathan, & Galinsky, 2009). This increased sense of control could make influential people more optimistic and augments their perceived self-sufficiency and self-esteem (Guinote, 2017). The rich are thus generally more willing to persuade others to adopt their goals (Guinote, 2017;Laurin et al., 2016), but also more likely to start an argument, make the first offer, and compromise less often in negotiations than less wealthy individuals (Fast et al., 2009;Kraus & Mendes, 2014). The poor, in contrast, tend to be more risk-averse and short-sighted (Haushofer & Fehr, 2014), making them more likely to control themselves and inhibit spontaneous responses than the rich (Na & Chan, 2016). Hence, while wealth may activate the behavioral approach system, poverty seemingly activates the behavioral inhibition system (Gray, 1990;Keltner et al., 2003;Lammers et al., 2010).
This heightened sense of freedom, self-esteem, and perceived control might increase the willingness to resist and oppose anything that restricts freedom (Brehm, 1966) in a response referred to as "reactance." Reactance can drive individuals to do forbidden acts or the exact opposite of what is requested simply to reestablish their personal sense of freedom, the magnitude of which governs the size of the corresponding reactance (Brehm, 1966). Given that taxes and tax authorities' attempts to increase tax cooperation can be perceived as a limitation on personal freedom, the rich find it harder to relinquish this freedom and easier to fight against such a loss. In fact, experimental research shows that coercive audits and fines increase taxpayer reactance more than less coercive attempts by the tax authorities (Gangl, Pfabigan, Lamm, Kirchler, & Hofmann, 2017). Thus, wealthier individuals faced with unfair treatment are more likely to fight for their rights (Kraus et al., 2013), while lower class individuals react with more self-conscious emotions, such as shame, guilt, or embarrassment. Additionally, the fact that the rich face less daily life risks than the poor might make them more willing to take extra risks (Guiso & Paiella, 2012;Haushofer & Fehr, 2014) of tax evasion, because compared to a poor person, a possible fine is not an existential threat. The outcome, as two field experiments (Castro & Scartascini, 2015;Slemrod, Blumenthal, & Christian, 2001) and another study (Tauchen, Witte, & Beron, 1993) indicate, that the rich, when faced with an increased audit probability, reduce their tax compliance, while the average citizen's tax compliances increases. These observations suggest that the rich may feel more reactance and a greater need to fight back when confronted by tax enforcement, which, in turn, leads to lower compliance. While more research is needed to analyze how the wealthy react to enforcement, from a practical point of view, it seems crucial to implement enforcement measures in a way that does not provoke resistance, thus increasing tax compliance and tax returns.

Fairness
Although conditional cooperation theory argues that the rich do not contribute more taxes because the poor cannot be expected to reciprocate (Cherry, Kroll, & Shogren, 2005;Frey & Torgler, 2007), another possible reason for the wealthy's lower tax compliance may be a perception that the tax system treats them unfairly (Fung & Au, 2014;Reuben & Riedl, 2013). Studies show that the rich are much more sensitive than less affluent individuals to violations of fairness, and more motivated to defend and restore it (Sawaoka, Hughes, & Ambady, 2015). They might believe that taxation is predominantly used to redistribute wealth-thus, no personal benefit is perceived at all. In addition, the perception that taxes reduce the capital base in addition to reducing capital growth might lead to perceived unfairness and fear of losing ones' privileged position. The opinion that "my money" is paid as taxes and then wasted by incompetent politicians and civil servants who are not held accountable for mismanagement might cause perceptions of unfairness. This belief might be more pronounced among the wealthy than average taxpayers because first, the wealthy (along with their companies) are often audited and thus held accountable; and second, they make tax contributions in magnitudes that allow a direct comparison to larger scale public projects that also helps to explain why liturgy in Ancient Greece was such a powerful mechanism to encourage the contribution of the rich.
Public good research suggests that wealthy individuals make lower contributions to the public good when their own economic activity is less dependent on the local commons (Cardenas, 2003;Martinsson, Villegas-Palacio, & Wollbrant, 2015). However, this research also indicates that wealthy individuals who believe that their contribution is critical to group success feel responsible and cooperate more (De Cremer & van Dijk, 2002). Yet, as our own research showed, the rich also seem to be less knowledgeable about their true tax rate-with the noncompliant rich having less tax knowledge than the compliant rich -impacting how they perceive fairness and reducing their tax compliance relative to poorer individuals (Lewis, 1978). This low perceived fairness may thus be related to a lack of knowledge about the wealthy's own relevance for the public good and/or their own exact tax contributions and direct benefits such as public infrastructure or security. A related, but not yet researched possibility is that a misperception of numbers might contribute to perceived unfairness, with individuals focusing on their total tax contribution rather than their relative tax contribution. Thus, the rich might see an absolute tax contribution of 5,000,000 Euro as more unfair than a middle-class taxpayer's absolute contribution of 5,000 Euro, although in relative terms, both might be contributing the same proportion of income. Nonetheless, from a practical standpoint, it is key to inform wealthy about their true tax contributions and benefits by reporting percentages, for instance, through information brochures. This increased knowledge about concrete relative costs and benefits may increase perceived fairness of the tax system.

Subjective Wealth Perceptions
Subjective wealth perceptions influence individuals' psychological processes and decisions more than true wealth (Brown-Iannuzzi et al., 2017). Individuals who believe that they are on the top of the wealth distribution disregard redistribution more than individuals who believe that they are on the bottom of the wealth distribution, regardless of the true position (Brown-Iannuzzi et al., 2017). For instance, millionaires who are reminded of their privileged status relative to most other citizens may act and comply differently than millionaires who compare themselves with billionaires and thus might feel "relatively poor." In laboratory experiments, subjective wealth is manipulated by endowing participants with different amounts of money or by allowing them to earn money. These experiments, such as those conducted by Durham, Manly, and Ritsema (2014), show a positive causal impact of subjective wealth on tax evasion and tax nonfiling by demonstrating that participants who earned more experimental income via a task (sorting numbers) evaded more than individuals who received less (Alm & McKee, 2006;Alm, Cherry, Jones, & McKee, 2010). In fact, data collected from both students and taxpayers show that those who earned the most during the experiment (by placing objects with the computer mouse) also evaded the most (Choo, Fonseca, & Myles, 2016). This effect even occurs within subjects: individuals evade more in tax or public good rounds (Bühren & Kundt, 2014;Grundmann & Graf Lambsdorff, 2017;Muehlbacher & Kirchler, 2009) in which they earn more than in rounds in which they earn less. Wealthy individuals evade more than the nonwealthy when money is endowed (Baldrey, 1987;Boylan & Sprinkle, 2001). Nonetheless, the effect seems to be stronger for earned money related to effort than endowed money related to a windfall gain (Kroll, Cherry, & Shogren, 2007). Thus, perceived income heterogeneity reduces tax compliance, particularly when individuals believe that the heterogeneity stems from actual effort rather than random allocation of windfall gains. To increase overall cooperation, tax administrations' communication efforts should counteract the perception that material well-being is based on individuals' own faults or efforts. In addition, the fact that tax money subsidizes services (e.g., childcare, care for the elderly) that are essential for the community could be used to increase the acceptance of redistribution through tax payments.
Tax experiments indicate that the subjective wealth effect on tax contributions also depends on visibility. In laboratory experiments, wealth heterogeneity must be visible and known to all participants to enhance positional concern or social comparisons that can crowd out, for instance, monetary contributions to public goods (Cardenas, 2003;Chan, Mestelman, Moir, & Muller, 1999), induce competition, reduce social or common identity, and decrease cooperation (Fung & Au, 2014;Nishi, Shirado, Rand, & Christakis, 2015). For example, if rich individuals know their own relative wealth position, it is possible that they consider their in-group to include other wealthy individuals, whereas the poor are perceived as members of the out-group. This in-group/out-group construction may reduce the willingness of the rich to cooperate with the poor (Fung & Au, 2014). In contrast, subjectively, wealthy individuals who perceive themselves to be alone among the poor tend not to reduce their cooperation (Reuben & Riedl, 2013). Thus, tax administrations' communication strategies should avoid the creation of a social class discourse in which a group of wealthy opposes a group of poor, as this likely enhances visibility of differences and reduces cooperation.
Tax attitudes are also impacted by subjective expectations about the own and children's future, anticipated gains or losses, or anxiety and worry about losing money (and thereby the associated status) . It has been suggested that the anticipated future wealth status is even more important than the current perceived status when it comes to cooperative behavior . For instance, the perception of permeable status group boundaries (upward mobility threat) and insecure relative status positions can fuel unrest among the wealthy, particularly during unstable economic conditions and among those who acquired wealth in the recent past (Jetten, 2019;Jetten et al., 2017;. Thus, anxiety about the future might motivate some to avoid paying taxes in order to secure their wealthy status for the future. However, until now there is no research that analyzes the subjective current and future wealth perceptions of the wealthy and their impact on tax behavior. Nonetheless, if tax authorities think that such fears are relevant for a specific taxpayer, the communication strategy could highlight the factual wealth difference between a millionaire and an average person, clarifying how taxes do not change relative status positions (some reassurance of the stability of wealthy taxpayers' status quo). For example, taxes financing social services ensure social peace and thereby secure system stability. Overall, however, research on subjective wealth suggests that policy makers might be well advised to emphasize the rhetoric that wealthy are just as much part of society as all other socioeconomic groups with a duty to contribute in a meaningful way and no more virtuous or "better" than other people so as to reduce egoistic status enhancing behavior.

Policy Interventions to Motivate Tax Compliance by the Rich
There is no silver bullet to manage tax compliance of the wealthy; rather different methods have to be applied in combination while considering the context. A general idea in the tax literature is that a carefully considered mix of coercive-based harsh and legitimacy-based soft measures is needed to ensure tax compliance. This basic idea is also likely true for wealthy taxpayers. However, given the discussed peculiarities of the wealthy compared to average taxpayers, some special adjustments might be necessary. In the following, we present research on the basic distinction between coercive-based and legitimacy-based instruments before discussing how this approach in general-and, in particular, for the macro, meso, and micro level-should be applied to wealthy taxpayers. The overall aim is to use a nuanced application of the "carrot and stick" approach to better integrate the wealthy into society, or at least to the community of citizens and taxpayers.
Theoretical models on tax compliance such as the slippery slope framework (Kirchler, Hoelzl, & Wahl, 2008) and its extension (i.e., eSSF, Gangl et al., 2015), the multifaceted approach (Alm & Torgler, 2011), or the responsive regulation theory (Braithwaite, 2003(Braithwaite, , 2007 assume that a combination of coercive-based and legitimacy-based interventions can efficiently impact tax motivations and behaviors. Coercive-based interventions are founded on various auditing tools (e.g., third-party information, personal audits) and punishments (e.g., fines, prison sentences) but also incentives (e.g., tax amnesties). Legitimacy-based interventions aim to convince citizens to comply voluntarily with tax rules through fair procedures, professional support, information provision, and a positive image of the tax administration . While both approaches are effective overall in generating higher tax payments, their psychological functioning is different (Kirchler et al., 2008). Coercive-based interventions lead to enforced compliance, and thus people pay because of the fear of audits (Kirchler et al., 2008). Legitimacybased interventions foster trust in the tax system and voluntary tax compliance, and thus, people accept their tax obligations without threatening audits and fines.
These theoretical assumptions receive empirical support from lab experiments. In these experiments, students or taxpayers are put into the role of a selfemployed person earning profit and paying taxes over several years. Importantly, participants receive remuneration, depending on audits and fines imposed by hypothetical tax authorities. We conducted a neurophysiological experiment in which students were asked to pay taxes, either in a country that relies on constant harsh controls and fines (i.e., coercive-based intervention) or on competent and helpful tax administrators (i.e., legitimacy-based intervention, . All participants made 40 tax-paying decisions in each condition, while ERPs (eventrelated signals on the cortex) were recorded. The analyses focused on signals between 150 and 500 ms after stimulus presentation (the page when the tax decision was to be made). Observed patterns of early signals indicated an enhanced response conflict and a higher arousal under the legitimacy-based than coercive-based intervention, while later signals suggested that the coercive-based intervention is related more to automatic processing than the legitimacy-based intervention. Using earlier economic decision-making studies as a reference, these results suggest that coercion-based interventions reduced the tax decision to a simple calculative problem, whereas legitimacy-based interventions maintained the complex moral and social dimension of tax compliance. Additional survey data indicated that the coercive (in contrast to legitimacy-based) intervention increased self-reported reactance and enforced motivation and decreased voluntary motivation.
Empirical evidence also shows that the combination of coercive-based and legitimacy-based interventions can reduce the negative effects of pure coercivebased interventions. Previous lab and online experiments framed participants as self-employed in a country in which tax authorities use harsh/lenient controls (coercive-based intervention) and employ well/poorly educated tax auditors (legitimacy-based intervention). Although no interaction effects were found, results suggested that coercive-based interventions lose some of their negative effects on trust and tax motivation when they are combined with legitimacy-based interventions (Gangl et al., 2013;Hofmann et al., 2014;E. Hofmann, Hartl, Gangl, Hartner-Tiefenthaler, & Kirchler, 2017a). However, together with the Austrian tax authorities, a field experiment was conducted in which the combination of coercion and legitimacy implemented through close supervision over the first year of young entrepreneurs' enterprise showed differential short-term effects. Entrepreneurs in the intervention paid more taxes but they were also late more often than entrepreneurs in the control group (Gangl, Torgler, Kirchler, & Hofmann, 2014).
In general, coercion-based and legitimacy-based interventions as classical administrative tools are also relevant for the tax decisions of the wealthy. However, some peculiarities need to be considered. Our review showed that on the micro level, the wealthy are more willing to promote and defend their view, seeking to restore fairness and remove restrictions on their freedom; this likely makes them more reactant to coercive interventions than average taxpayers who might more often give into coercion. This view is supported by field experiments that indicate how an audit threat increases compliance of average taxpayers but reduces compliance of wealthy taxpayers (Castro & Scartascini, 2015;Slemrod et al., 2001). In addition, wealthy taxpayers are less likely to give into coercion due to peculiarities on the macro and meso level that mean the wealthy are more easily able to exploit international secrecy jurisdictions with the help of skilled tax advisors. Thus, wealthy people who feel coerced by the tax administration are likely more reactant and also have more resources (compared to the average taxpayers) to escape this situation, making classical coercive attempts to increase their tax honesty less effective. Therefore, to reduce reactance and to ensure compliance of the wealthy, coercive-based interventions need to be cautiously combined with legitimacy-based interventions. The functionality of coercive-based interventions depends on perceived professionalism, fairness, and thus, legitimacy.
In detail, coercion-based and legitimacy-based interventions for each level (the macro, meso, and micro) have different priorities and different aims. On the macro level, coercive-based interventions are essential, because only compulsory legal environments and specialized revenue bodies-which can enforce compliance-can reduce the avoidance and evasion opportunities of the wealthy. However, accompanying legitimacy-based interventions through marketing campaigns are necessary to reduce the aversion to and increase the acceptance of coercive interventions. In addition, on the meso level, coercive-based interventions are needed to regulate tax advisors more strictly. However, legitimacy-based interventions should dominate on the meso level to leverage social dynamics and to create social norms of tax honesty. At the micro level, during the direct interaction between tax auditors and wealthy taxpayers, coercive-based interventions should be reduced to a minimum, whereas legitimacy-based interventions should be prioritized. Well-educated tax auditors should target coercive-based instruments only to known criminal tax fraudsters ("crooks"), thereby following a responsive regulation approach. Importantly, personal interaction should be characterized by appreciation and respect to increase perceived legitimacy of tax collection and trust in the tax system. Table 1 provides an overview of how, on each conceptual level, different psychological causes of tax compliance of the wealthy can be addressed by coercive-based and legitimacy-based interventions.
In the following, we present the different interventions in detail.

Compulsory Legal Environments
On the macro level, moral disengagement has to be addressed by coercive tax laws that clearly communicate that tax flight and aggressive tax avoidance are morally objectionable. Thus, countries with high-tax losses through tax flight need to establish a tax law with fewer loopholes and possibilities of escape to tax havens. For example, this could include the creation of national VAT taxes that (unlike profit taxes) cannot be shifted to another country, or it might involve establishment of legal virtual locations to facilitate taxation of profit in all countries in which a company operates virtually (Rohwetter, 2017). Aggressive tax avoidance can, for example, be addressed by a modern form of a tribute (Frijters, et al., 2019). Like some kings in the past, the state could decide (with the help of an independent agency) to calculate how much a company can afford to pay in tax without endangering the company itself-and then charge exactly this amount (e.g., based on market signals about the amount of surplus created in the relevant region by the companies taxed and therefore what they could pay). Such modern ways of compulsory acquisition are particularly an option if society needs to address extreme challenges such as war or environmental disasters.
Human history is full of examples in which states demanded tribute, for example, the tribute of grain in ancient Egypt and Rome.
Charities allow moral licensing as some might feel that giving to charity permits the avoidance of taxes. Thus, legal regulations need to counteract the spin that private charities are better (or as good) as the state in addressing societies' problems. The options are to regulate more strictly what counts as a deductible charity, and in general to reduce the tax exemptions for charities (Reich, 2018).
It is essential to focus on a strict legal environment. Without simultaneously securing public opinion and the voluntary compliance of the wealthy through additional legitimacy-based instruments, however, no legal remedies will suffice.

Specialized Revenue Bodies
Engaging in tax evasion and aggressive avoidance without consequences fosters moral disengagement and the belief that it is normal to evade taxes. The OECD (2015b) suggests that countries should establish special revenue bodies for both large corporations and rich individuals that can implement targeted auditing techniques (e.g., data mining or whistle-blower systems; Casaburi & Troiano, 2016;Wittrock, 2012). However, a 100% audit rate is unrealistic and high fines may not be an existential threat to the financial situation of the wealthy (e.g., one experiment found that a fine has to be 15 times the evaded amount to be effective; Friedland, Maital, & Rutenberg, 1978); it may even provoke their resistance and their willingness to sue (e.g., Slemrod et al., 2001). Nonetheless, suspending audits and fines offers no solution either, and tax amnesties, which allow rich taxpayers to repatriate their money from tax havens without being fined, also show no long-term positive effect (Alm & Beck, 1993;Toro, Story, Hartnett, Russell, & Van-Driessche, 2017). In fact, amnesties might even lower tax morale among honest taxpayers (Torgler, Schaltegger, & Schaffner, 2003). Thus, working with coercive-based interventions alone is not enough; for example, the OECD (2015b) suggests that it is essential that special revenue bodies combine coercive-based with legitimacy-based interventions.
A good example of how to combine coercion with legitimacy comes from the United Kingdom. There, a specialized revenue body exists that focuses on 6,500 individuals (0.02% of all taxpayers) who pay 1.3% of all income tax and 15% of all capital gains tax in a year (National Audit Office, 2016). To reduce tax avoidance schemes, this audit-focused unit employs 40 customer relationship managers, each responsible for about 160 taxpayers. According to the unit's data, it has been successful in substantially increasing tax compliance (National Audit Office, 2016), making this use of specialized tax officers implementing individualized customer treatments an idea worth further exploration. Such relationship managers could treat taxpayers differently depending on the business category and interaction history. They could ensure that wealthy taxpayers are sufficiently informed about their actual tax rates and tax rights, and emphasize how significant their contributions are for the community, while explaining what the money will finance, and how it profits the taxpayer (e.g., legal and social security; Chen et al., 2001;Guinote, 2017;Oc et al., 2015). Such a monitoring system may generate less reactance among the rich and increase their perception of the tax system as trustworthy and legitimate, thereby ultimately encouraging honest tax payment.
Some tax authorities have also opted to abandon many coercive measures. For example, the tax authorities in the Netherlands, the United States, Australia, and Austria have established new working and monitoring relationships with large corporations who were honest in the past (Colon & Swagerman, 2015;De Simone, Sansing, & Seidman, 2013;Torgler & Murphy, 2004;The Netherlands Tax and Customs Administration, 2010). In such trust-based relationships, the taxpayer agrees to be transparent about all tax data and tax strategies, while the tax administration promises to resolve all tax issues on time and abstain from auditing the taxpayer for prior years, thereby reducing uncertainty for a corporation. Nonetheless, although evaluations indicate that this system leads to reduced monitoring costs and faster issuance of final corporate tax statements (Elmecker et al., 2016), there is still no empirical evidence for its positive effect on tax payments itself, which makes this noncoercive approach less attractive.

Marketing Campaigns
Public marketing campaigns are especially suited to communicating the legitimacy of the tax system and thus to generate trust. However, the practical aim is to change public opinion such that tax evasion and aggressive tax avoidance is publicly disregarded as unethical. Such public campaigns need to back up (coercive) legal regulations and administrative procedures, promoting legitimacy and making moral disengagement and moral licensing more difficult. Campaigns could promote coercive measures as protection of the honest, important pillars of democratic societies, stability, and safety. To counteract the attempts of some lobbies to disregard government interventions (Reich, 2018), campaigns should show what is financed with the tax money, thereby creating a positive vision for the country (e.g., fighting climate change, security) and enhancing social belongingness and cooperation. One historical example of this was seen in the United States during the world wars. Our own experiments showed that priming patriotic feelings by explaining what the state had successfully accomplished in the past, or by exposing participants to typical landscape pictures of the country (e.g., of mountains or rivers; Gangl, Torgler, & Kirchler, 2016) could help to increase trust, felt social belongingness and cooperation (Macintyre, Chan, Schaffner, & Torgler, 2018;Torgler, 2004b, 2005a, Konrad & Qari, 2012. Importantly, in addition to involving different stakeholder groups, wealthy and famous individuals could be used as positive role models, thereby creating a social norm of tax honesty and pride in being a significant taxpayer. In India, the government has used celebrities to promote or change tax-paying behavior. For example, the 1997 tax amnesty campaign in India used sports and film stars to attract 350,000 delinquent taxpayers and generate 2.5 billion dollars in erstwhile lost revenue (Torgler & Schaltegger, 2005). Public campaigns may also motivate more whistleblowing among wealth managers.
Even though the positive effect of such campaigns on tax compliance seems plausible, empirical evidence is still scarce (Cyan, Koumpias, & Martinez-Vazquez, 2017), and thus, tax administrations should conduct field studies in this area. Overall, a good policy intervention to achieve lasting behavioral changes requires consideration of insights from social psychology (Mols, Haslam, Jetten, & Steffens, 2015).

Regulation of Tax Practitioners
On the meso level, tax decisions of the wealthy are often conducted in a group-setting involving tax practitioners. This tends to produce more exploitative tax returns than if taxpayers made their tax decisions alone (Christian et al., 1994;Doyle et al., 2013). One option for a coercive-based regulation would be to reduce or even exclude the involvement of tax practitioners through implementing direct relationships between tax authorities and taxpayers (e.g., based on automatic data transfers). One option for legitimacy-based regulation would be to interact only with tax practitioners who have earned certificates of trust by completing official training that addresses the interests of the community, and not only the individual. It would also be worthwhile discussing strategies to increase good ethics and practices that stipulate the type of professional code exemplified by medicine's Hippocratic oath (field evidence on promises, see Koessler et al., 2019). Such public training and professional codes could counteract the promotion of low-tax values in current private trainings. Thereby, social norms of tax honesty are built among the tax practitioners, which likely influence the tax decisions of the wealthy. Certificates could serve as both a gatekeeper and a criterion for promotion: only tax practitioners with a respective certificate would be allowed to submit tax returns in the name of a taxpayer.

Reputation Mechanisms
On the meso level, our literature review showed that social norms play a significant role. One option to influence the social norms of the wealthy is to combine coercive-based and legitimacy-based instruments in reputation measures by publicly disclosing a person as a negative or positive example. For instance, identifying tax evaders publicly (Casal & Mittone, 2016;Coricelli, Joffily, Montmarquette, & Villeval, 2010) may act as punishment and a deterrent from engaging in aggressive tax avoidance (Graham, Hanlon, Shevlin, & Shroff, 2013). Preliminary empirical evidence from the United States, however, does not confirm that a fear of lost reputation is effective. Moreover, the use of a tax haven seems to have no negative effect on stock prices or media reports (Gallemore, Maydew, & Thornock, 2014). Similarly, a field experiment in the United States that shamed the noncompliant via letters to neighbors increased tax compliance for small but not for large tax evaders (Perez-Truglia & Troiano, 2018). On the other hand, after Greece published a blacklist of over 4,000 citizens who owed tax money to the state (Aswestopoulos, 2012), it experienced a decline in the size of the shadow economy from 25.4% in 2010 to 22.0% in 2016 (Schneider, 2016). Shaming could also be applied to aggressive tax advisors or countries that serve as tax havens (Comte, 2017) by publishing blacklists (e.g., financialsecrecyindex.com). Admittedly, these nations' incentives to attract foreign investment through tax reduction strategies could reduce the efficacy of a shaming signal (Weisbord, 2016). However, in the future, it may be possible to link specific desirable benefits with a country's decision to not be a tax haven; for example, visa and market access, or access to international credit, which are more significant than the benefits from foreign tax evasion investment. Perhaps, shaming would be more effective if there were extensive media reports on the legitimacy of tax collection and the related harm of tax avoidance.
An alternative to shaming evaders is to reward honest taxpayers (Feld, Frey, & Torgler, 2006). An example is the "fair tax mark" assigned by a U.K. NGO to companies assessed as honest taxpayers (fairtaxmark.net). Recipients of this designation can leverage the label in marketing. If such a mark were accompanied by mass media reports, it could benefit honest companies, like the "fair trade" and "certified organic" labels. Another suggestion put forward by Feld et al. (2006) is that tax offices issue a certificate of "correct declaration and tax cooperation" showing the firm to be a good taxpayer. The resulting increase in a firm's reputation and image could attract more favorable conditions on the capital market and positive shareholder reactions through higher share prices. In general, it would seem worthwhile to reward honest taxpayers with special attention (e.g., being held up as honorable citizens and role models), and even small rewards (e.g., vouchers; Koessler et al., 2019). Another method of rewarding honest taxpayers would be to place their name on public buildings that were financed with tax money. The Liturgy system of ancient Greece surely can be an inspiration for new ways to leverage the social norms existent among the wealthy, using gratitude and honor to motivate tax payments. As the late Boulding (1992) pointed out, "[T]he dynamics which governs the creation, destruction, and distribution of various forms of pride and shame in society are very little understood, yet nothing perhaps is more crucial to the understanding of the overall dynamics of a particular society than the marked differences which exist among societies in this regard" (p. 93).

Increased Participation
On the meso level, participatory elements are key to increasing legitimacy. Participation and voice in the selection of tax rules foster perceived fairness, felt responsibility, identification with the state, and the community at large (Feld & Tyran, 2002) and, in turn, tax morale and tax compliance (Alm, 2019;Torgler, 2005b). Thus, both the rich and the less affluent citizens could be more involved in (i) the administrative processes, (ii) determination of tax rates, and (iii) the spending of tax money. In the administrative process, taxpayers could have a voice in setting deadlines and scheduling meetings. Concerning the determination of tax rates, citizens (like in Switzerland) could be asked which assets should be taxed more than others (Pommerehne & Weck-Hannemann, 1996). With respect to spending, taxpayers could be allowed to have a voice in the outlay of at least a certain proportion of their own tax money. Rich taxpayers may even become involved in the realization of particularly needed public goods; for example, by choosing from a list of such projects that they could finance and support personally. Once the project is underway, a committee could be formed with other citizens to bring it to fruition. Once again, ancient Greece's liturgy system shows how to encourage the rich by involving them in the tax collection and spending process in a way that leads to public improvements. The administration of charities could give further ideas on how to manage the tax system in a more participatory way allowing the relevant wealthy contributors more control over tax issues. Nonetheless, there are few extant field studies (Touchton & Wampler, 2019) concerning participation, and rarely do they focus on wealthy taxpayers.

Well-Trained Tax Auditors
On the micro level, the powerful situation of wealthy taxpayers has to be addressed by well-trained tax auditors to generate legitimacy, trust, and voluntary tax motivation. Thus, tax administration personnel need to not only be well versed in their legal and administrative skills but also on their social, emotional, and psychological skills. The Pharaohs in Ancient Egypt realized that those who collect the tax are key for a successful tax system (Adams, 1993). Tax auditors should be able to compete with their counterparts, i.e., tax practitioners of the wealthy, who are often better equipped with resources and are well trained and highly compensated. Thus, tax administrations need to offer competitive salaries, and also need to put the single tax auditors or auditing teams into an empowered and autonomous position, allowing them to directly negotiate with taxpayers by maintaining a high level of transparency to avoid the potential of corruption. Hence, tax auditors need to be highly skilled and trusted professionals who are given and are capable of taking a lot of responsibility. Some countries, particularly in the past, have employed tax auditors without a university degree. This was especially the case outside of big cities: tax auditors were not specialists, and had little knowledge of specific target groups or business sectors. In addition, the practice of giving the decision-making power to someone who is not present during an audit (e.g., to the supervisor of the auditor) reduces perceived competence of personnel directly interacting with the taxpayer. Accordingly, to ensure perceived legitimacy of those who interact with a taxpayer, an excellent and specialized education is necessary. International educational and exchange programs and international cooperation programs that create and exchange information and expertise such as the EU Fiscalis program 3 or sabbaticals at universities can also contribute to an increase in skills.

Responsive Regulation
On the micro level, it is important to recognize the peculiarities and diversity of wealthy taxpayers' attitudes, motivations, and cognitions in order to choose the most effective coercive-based and legitimacy-based strategy. Responsive regulation claims that taxpayers need to be treated differently depending on their tax motivation (Braithwaite, 2003(Braithwaite, , 2007. For instance, strict and harsh control measures should be targeted only to taxpayers motivated by enforcement, such as known fraudsters. Tax audits for new taxpayers should be used to educate and to demonstrate the professional way in which monitoring is conducted. Finally, harsh audits for committed motivated taxpayers should be suspended, and instead, these taxpayers should be respected and thanked (Gangl et al., 2015). Additionally, tax auditors should be aware of the special values, cognitions, and fairness perceptions of the wealthiest, in order to choose a convincing and trust-generating communication strategy. For instance, the wealthy who respond to gratitude should be made aware of gratitude measures, whereas wealthy who are afraid that they are the only ones who pay honestly should be made aware of auditing frequencies for fellow wealthy taxpayers. Thus, tax auditors should also be trained in how to psychologically diagnose a taxpayer in order to implement the right communication and enforcement strategy. One way of training these competencies is through role-plays that practice convincing arguments in response to the most common complaints of taxpayers. Additionally, escalation plans for coercive measures can be developed, giving tax auditors a road map of which situation (e.g., repeated postponed meeting) should be addressed with which measure (e.g., issuing a caution).

Appreciation and Respect
On the micro level, cognitive tendencies of the wealthy such as an increased willingness to defend own rights or to fight perceived unfairness likely increase their sensitivity to friendly or unfriendly interactions. Thus, appreciation and respect for those who are tax compliant represent very inexpensive legitimacybased instruments, which nonetheless are central to maintain, guarantee, or generate trust, particularly the basic trust in the system . In official documents and interaction, respectful treatment would include a thank you note in advance for future cooperation, past cooperation, and the tax contribution itself, which is a sign of appreciation. In addition, examples of what will be financed with the money could be shared. Personal interactions are important. In personal encounters, tax auditors should listen to the taxpayers and convey a feeling of genuinely caring for their situation, and for valuing their accomplishments and contributions to society. This empathic and respectful approach should, of course, be the rule for all taxpayers, not just the wealthy; however, it is suggested that it is especially needed for the wealthy. If unsatisfied by treatment from the tax authorities, the wealthy have more options to sue or to hire aggressive tax advisors.

Conclusions
Tax compliance by the wealthy is of paramount importance to a wellfunctioning society. The present review summarizes the existing evidence from various academic fields, from history, and from practitioners, delineating how and why the wealthy differ from average taxpayers. We do not yet fully understand the extent or the determinants of the wealthy's tax noncompliance. However, the tax decisions of the wealthy are based on specific interrelated political-economic, social, and individual differences with associated psychological consequences. We discuss how these peculiarities can be addressed by tax authorities, via tailored coercive-based and legitimacy-based instruments. Given the recent burgeoning of costly crises that are straining public budgets across the world (e.g., climate and demographic change, financial recessions, the refugee crisis), together with a loss of public confidence in governments' ability to establish a fair economic and political system, it seems both important and timely to intensify efforts into research and policies regarding wealthy taxpayers. Thereby, research and practice should go hand in hand. To foster an evidence-based focused tax administration, all new policy attempts should be accompanied by evaluation procedures that recognize contextual dependencies and examine differential effects. One way to test the usefulness of new potential instruments would be through randomized control trials conducted in close collaboration between researchers and the tax administrations (for a detailed discussion, see Torgler, 2016).
One main aim of this review was to shed light on the tax compliance of the wealthy, to show that they are different from average taxpayers, and to suggest that research and policy interventions in this regard can be successful. The historical overview at the beginning of this review should convince researchers and policy makers that there are many possible different ways to motivate the wealthy to contribute more taxes to the benefit of the society.
KATHARINA GANGL is a researcher at the Institute for Advanced Studies, Vienna. She has 10 years research experience in tax psychology, publishes her work in peer-reviewed journals and books and collaborates with tax administrations all over the world.
BENNO TORGLER is a Professor at the QUT Business School. He has published more than 200 scientific contributions in various social science fields and has pioneered the fields on social norms of tax compliance and tax morale.