1. TAX COMPLIANCE GAMES METHODS AND EVIDENCE FROM BEHAVIOURAL DYNAMICS IN GREECE V. Vlachos, A. Bitzenis, P. Kontakos University of Macedonia, Thessaloniki, Greece Thursday, September 18, 14:00-15:30, Univ. of Verona, Palazzo di Economia Aula B Session 4, Entrepreneurship and Innovation (11) 8th Annual Conference of the Euromed Academy of Business Verona, Italy, September 16-18, 2015

2. 2 Note: The current paper is presented under the auspices of the THALES Research Programme. THALES Programme has been co- financed by the European Union (European Social Fund - ESF) and Greek national funds through the Operational Program “Education and Lifelong Learning” of the National Strategic Reference Framework (NSRF).

3. Foreword to our Project Our aim is to research and measure all various aspects of shadow economy in Greece, including corruption, tax evasion, tax avoidance, social contribution avoidance, undeclared and illegal work, self-consumption, tax morale level, tax compliance level, illegal and criminal acts (black or underground economy, money- laundering, human and drug trafficking, briberies). It covers all economic agents in Greece, such as citizens and corporations (e.g. public servants and private individuals, companies and all professional categories, etc.) The research is also performed at sector levels, e.g. to identify the extent of tax evasion and corruption in the trading of oil in Greece. 3

4. Foreword to our Project (cont.) It does not aim to the precise percentage regarding the measurement of Greek shadow economy but aims to the qualitative analysis of questionnaire results and the comprehension of the problem. The implementation of our interviews, scientific games and economic experiments (tax compliance games) involves at least 2,000 individuals and business owners (in majority small businesses). The project aspires to achieve numerous objectives, among which the development of a relevant theoretical background, and perform cross-country comparisons at regional level, but also with country groups with advanced taxation systems. 4

5. Foreword to our Project (cont.) In Greece, during the period 2008-2014, both unemployment increased from 8% to 27% and the GDP per capita decreased almost 25%. THALES results show that in contrast to the reduction of the shadow economy as a percentage of GDP from 28% before the crisis at 24% in 2013, we estimate that at the end of our full research analysis, probably, this percentage will have exceeded 30% in 2014/2015. 5

6. Agenda 1. Aims of the paper 2. Introduction 3. An Exploration of Tax Compliance Games 4. Preliminary Evidence from our Questionnaire Survey 5. Conclusion 6

7. 1. Aims of the paper Discuss the alternative methods adopted by researches based on tax compliance games. Introduce a behavioral approach to game theory and tax compliance games. Present the preliminary results from our structured interviews aiming to identify the tax behavioural dynamics of individuals as well as Greek SMEs. Our final aim is to develop a policy mixture that would significantly enhance tax morale and tax compliance and thus, affect positively tax revenues and at the same time improve the country's attractiveness for inward FDI. 7

8. 2. Introduction The growing body of literature on tax compliance indicates that economic, social & psychological variables have to be jointly considered. The researchers performing tax compliance games adopt a behavioral approach to game theory and use data from group experiments. The tax compliance games reviewed here concerns such (usually agent-based) models, which are either tested by data from group experiments or by simulations. Our discussion in this paper is based on studies that have been published from 2010 onwards in journals catalogued in internationally recognized lists. 8

9. 3. An Exploration of Tax Compliance Games This section constitutes the main body of the paper by discussing the methodology and methods of tax compliance games. Despite the numerous concerns about the predictive ability of game theory, the approach is still considered useful for approximating the behavior of players engaged (Rubinstein, 2007) –The next subsection pinpoints the main issues in tax compliance research (definition, context, methods, etc.) through past literature surveys. –A brief discussion of multivariate methods takes place in the second subsection, for the consideration of decisive factors that may not be included in game theory models. –The last subsection discusses the methods of tax compliance games through a review of empirical studies (experiments or simulations). 9

10. 3.1 Issues in tax compliance research highlighted by literature surveys The importance of non-financial motives and their modeling, per the latest literature survey is discussed (Pickhardt and Prinz, 2014). Andreoni et al. (1998) stated that the problem of tax compliance is an issue of public finance, as much as it concerns organizational design, law enforcement, ethics and labour supply. Torgler (2010) focuses on the undeveloped aspects of Andreoni et al. (1998), by addressing the effect of cultural or social norms. Alm (2014) reviews the social factors described by social norms, morality, etc. and argues that these group motivations are more important than the respective individual. Kirchler and Wahl (2010) identify the major issues with regard to the methods adopted by relevant studies (econometric, field experiments and laboratory experiments) 10

11. 3.2 Multivariate approaches to tax compliance Multivariate analysis of studies can explore the significance of a great number of independent variables. Wahl et al. (2010) perform two laboratory experiments, based on data from 120 students and 127 self-employed individuals, for the exploration of the influence that trust in authorities and their enforcement power exert on tax compliance. It is the most cited paper (i.e. over 40 citations in Google scholar since 2010). Their findings indicate that both high trust and high power lead to increased tax payments. They also find that age significantly influences strategic taxpaying behavior. The reason for not including this work in the next section, is because they based the most important part of their analysis on data gathered from an online questionnaire and not from an experiment (i.e. a procedure through which responses where reported). 11

12. 3.3 Tax compliance games Our work has considered the work performed by a significant number of prominent researchers in the field. Indicatively we can refer to the articles of Alm et al. (2010); Alm et al. (2014); Bloomquist (2010); Kastlunger et al. (2011); Muehlbacher et al. (2012); Andrei et al. (2014); Hashimzade et al. (2014); Bazart and Bonein (2014); Bruttel and Friehe (2014); Durham et al. (2014). Most of the aforementioned reserchers rely on the participation of students and university staff, and small business owners, and combine ideas from behavioral economics and social networks in the conducting of the laboratory experiments and tax compliance games. 12

13. 4. Preliminary Results from our Questionnaire Survey The preliminary evidence presented is based on our survey conducted during July 1st to Dec. 31st, 2014, with the participation with the participation of individuals as well as of micro and small business owners (i.e. owners of SMEs). Greece is a country with low level of public services hence tends to have low-level governance (including over-regulation and corruption) and that in turn tends to enhance poor tax morale and tax compliance. We realized that the most wealthy individuals are the most likely to avoid paying taxes. Top tax evaders are professionals, with politicians protecting their own occupations. Firms and the self-employed tend to participate mostly in tax evasion, compared to salary employment. 13

14. 4. Preliminary Results from our Questionnaire Survey (cont.) Tax evasion in Greece consists more a social norm, connected to the belief that taxes are wasted. In principle, tax evasion can be largely rooted to the historical mistrust between the Greek state and its citizens. The complexity of tax law and the tax system are found to have negatively impacted the tax compliance behavior, whereas importance was assigned by the small business owners that participated in the survey to the actual (high) level of tax rate which tends to encourage corruptive behaviors. The probability of tax audits, appears however to have altered in certain cases their attitudes towards declaring, instead of concealing, their real income. 14

15. 4. Preliminary Results from our Questionnaire Survey (cont.) As possible solutions to contain the size of tax evasion, according to feedback received from participants in the experiments, we might propose the following: (a) The court system must act instantly. Tax evasion cases should be legally treated within a month after presented evidence. (b) A financial reward scheme should be established to incentivize disclosure of tax evasion (c) A publicity campaign should be undertaken to encourage people to disclose tax evasion (even through SMS - mobile devices, internet websites). (d) Punishments do not have to be severe but at the same time if someone repeats the same behavior (for example, between three to five times) then the punishment would be severe. (e) Introduction of a flat rate of tax not below 10%, and not above 15% (for individuals and for companies). 15

16. 5. Conclusion Our discussion about the methods of tax compliance games indicates that the studies based on the particular method cannot explore all the determinants of tax compliance concurrently. A friendlier tax service fosters tax compliance, while uncertainty and inequity foster tax noncompliance. Networks and social norms play a major role in tax compliance. Overall, attention has to be brought to the group of individuals on which the experiment takes place. Interestingly enough, all studies discussed in the previous section except from Bloomquist (2010), are based on experimental data from student groups. Finally, summarized preliminary evidence and proposals to tackle tax evasion was presented from our structured interviews with the participation of individuals as well as SME owners. 16

17. Appendix: Our Research Aim & Methodology The implementation of our scientific games / economic experiments (tax compliance games) will involve at least 2,000 individuals and business owners (in majority small businesses, as explained in the next section), which will be identified via random or other statistically sound sampling methodologies. At least 20 different occupational categories and income groups will participate in the execution of the experiment, in order to determine their behavioral aspects related to tax compliance and attitude arising from the impact of the economic crisis, high unemployment, undeclared work, informal economy, corruption, tax implications, etc. The experiment is performed within an environment where the decisions or actions of participants are based on monetary incentives (and not vouchers) in which participants are exposed to frameworks resembling real-life taxpaying situations. 17

18. Appendix: Our Research Aim & Methodology (cont.) The real cash earnings / incomes cannot be observed by other participants or the experiment coordinator or other researchers (no impacts from the breach of rules will apply). Various levels of tax rates will be involved during the game in order to identify the respective levels of tax evasion incentives. Currently, the phase of the Project related to the implementation of a pilot survey involving 500 participants has been concluded and the preliminary evidence is discussed in the current paper. The methodology we have followed was based on the aforementioned performed literature review and is described in the next paragraphs. 18

19. Appendix: Our Research Aim & Methodology (cont.) a. Experiments They are used widely in the last 30-y, also by the economic science. Although studies have conducted laboratory experiments to estimate the extent of tax compliance & the degree of responsiveness in key economic variables, only few of them have advanced to the examination of other non-economic factors. The aim of this study is exactly to identify the social, psychological, political & cultural factors that influence the decision of individuals to comply with their tax obligations and the influencing of each factor in their final choices. Main drawback of this method is that the high administrative cost does not leave room for wide participation, which leads us to select and check a few parameters, which are considered more important in the final selection of the taxpayer. 19

20. Appendix: Our Research Aim & Methodology (cont.) b. Objectives of the experiment For the successful performance of the experiment should be, firstly, identified the targets, the criteria for selecting the individuals, the way of presentation and the rules of the game. Equally important parameter is the timing and whether it is static or dynamic. The main objective is to see whether the concepts of reward, ethical barriers and reciprocity positively influence the tax behavior of individuals. Our experiment is a synthesis of previous experiments adapted to the Greek reality, mainly from laboratory experiments of Alm et al. (1992), Torgler (2003) and Bosco and Mittone (1997). The common denominator of all these studies is their emphasis on the social & psychological factors affecting tax compliance & the introduction of variables involving ethical dilemmas of the taxpayer and that they render special role in the value of reciprocity. 20

21. Appendix: Our Research Aim & Methodology (cont.) c. Rules and Procedure of the Game We have first attempted that the game variables approximate as much as possible the values that taxpayers are facing in reality. In particular, the tax rate should neither be very high so that participants do not feel possessed by injustice in taxation of income, nor too low, thus resulting in the lack of any incentive for tax evasion. This study proposes a tax rate equal to 30%. Indicatively are mentioned the variables that can be changed if deemed that they do not fully reflect the Greek reality, such as: the tax rate (t = 0.30), the probability of a tax audit (p = 0.10) and the tax fine or penalty (equal to two times the undeclared income). it is suggested the involvement of students in the experiment, despite some potential bias in the sample (due to young age inexperience or higher than average education levels). 21

22. Appendix: Our Research Aim & Methodology (cont.) The group of students might be a good sample because the teachers will check their tax behavior, exerting a form of authority over students - even limited, which can be compared with the power of the tax authorities. Also some groups will be exposed in ethical dilemmas in order to determine whether the moral constraints affecting their tax consciousness. Another challenge is whether all the rules of the game will be clearly explained or only the most important. Taxpayers are known as not being fully aware of the tax laws, especially when these are highly complex. Our view is that a simple and clear tax legislative framework can positively affect the tax consciousness of people. 22

23. Appendix: Our Research Aim & Methodology (cont.) Because our aim is to understand whether and how the person is affected by moral constraints in the final decision to evade taxes or to tax comply, in order to arrive at clear conclusions we have preferred to conduct the game with the introduction of a small number of clear rules. Further, the influence of the complexity of the legislation in the tax moral citizens will be examined through the questionnaire survey. Those taking part in the experiment are randomly divided in five groups. At the start of the game players receive an amount of money as compensation for their participation in the game (e.g. 60 Euro), out of which they will be requested to reimburse 30% in the form of a tax (i.e. 18 Euro). 23

24. Appendix: Our Research Aim & Methodology (cont.) Players choose the amount of tax that will pay, knowing that their obligation is to pay 30% (or 18 Euro). The payment is done in such a way as to ensure their anonymity (except the players in Group E). Checks are carried out by lottery in the presence of all players. The results of the tests are published unless the checks are carried out in group E. Clear instructions are provided to the players regarding the tax payment, the tax audits, the payment of fine or penalty in the case of a tax audit, and finally the payment of the remaining balance in the form of remuneration for their participation in the research project. Players are notified in writing shortly before carrying out the game for the conduct and the rules. 24