Religion, Culture, and Tax Evasion: Evidence from the Czech Republic

Our paper analyzes the impact of culture and religion on tax evasions in the Czech Republic, which represents one of the most atheistic countries in Europe, and a very interesting example of attitudes to the church and religion, as well as the influence of religion on the social and economic aspects of everyday life. Our results suggest that, in the Czech Republic, religion plays the role of tax compliance, but only through a positive effect of visiting the church. National pride supports tax morality while trust in government institutions and attitudes towards government are not associated with tax compliance. These results suggest that the Czech Republic is no different from other countries regarding the relationship between religion and tax compliance. Moreover, the role of government as the authority for improving tax compliance is different from what is observed in other countries.


Introduction
The question about which role should be played by the state in economic life is one of the most important ones that economic science has attempted to answer during the last century and at the beginning of the current one.Liberal economists, such as the Nobel Prize winner (1976) Milton Friedman [1], claim that the role of the state should be minimized and that most of the economic processes should be given a chance for self-regulation.According to them, the level of taxation, being one of the most important sources of income of the state, should be minimized.On the other hand, proponents of Welfare State Theory support the idea of the state as a provider of "government protected minimal standards of income, nutrition, health, housing and education provided for each citizen" [2].In order to provide goods and services for its citizens, the state should collect considerable tax revenue.What is frequently forgotten in this kind of discussions, is that the role of the state may be to a large extend determined by the culture of the country in question and it may be not possible to find one universal solution for the optimal amount of taxation and optimal size of state due to culture differences.
The levels of taxation in various countries differ.For example, despite very similar GDP per capita values in Norway and the USA, or in Germany, Denmark, Sweden, and the United Kingdom, there is a huge difference in tax revenues expressed as the percentage to GDP (Table 1).Source: [3].Copyright obtained.
The aim of our paper is to analyze the links between the level of culture, taxation (and tax revenues of the government measured as a percentage of GDP), as well as the role and the place of the tax morale (and the issue of cheating) in this process.
This paper is organized as follows: Section 2 describes empirical effects of culture on economics.Section 3 elaborates on the tax morale and the civic duty to pay taxes and introduces the tax morale empirical econometric model for the Czech Republic, which explains the links between various cultural dimensions and the tax morale.Section 4 concludes the paper with comments on the main results and their implications.

Effects of Culture and Religion on Economics
In economic science, the sub-field that was dealing with religion, cultural values and their empirical effects experienced a boost at the end of the 20th century and the beginning of the current century.Among the many authors who attempted to systematically describe systems of cultural values, four of the most relevant ones deserve mentioning [4][5][6][7][8][9].All of these scholars succeeded, both in providing a systematic concept of human values applicable in different countries, and in computing indicators, which allow the comparing of cultural values.To the best of the authors' knowledge, nowadays, most economists rely upon the indicators computed by Geert Hofstede et al. [10], who were the first to introduce the concept of dimensions of national cultures and compute usable quantification of such.However, the most important question is whether Hofstede's concept of cultural dimensions, including its quantification, should be used in scientific research in 21st century given the fact that most of the indices were computed in the 1960s and the 1970s.Generally, Kirkman et al. [10] agree with Smith and Bond [11], that other authors who studied concepts similar to Hofstede's (including [5,6,9,12] as examples), supported Hofstede's conclusion rather than contradicted them.Further on, the review of Kirkman et al. [10] showed that research, using indices of Geert Hofstede, successfully managed to differentiate countries according to the cultural values and, thus, to increase variance.In addition, Kirkman et al. [10] claimed that most of the country differences predicted by Geert Hofstede were supported.
The research literature on the effect of cultural values on economic, social and psychological processes grows exponentially.More recent studies based on Hofstede's cultural indices can be divided to two categories.The first category continues to look for the effects on cultural values on economic, social, psychological and other processes either through simple correlations or via more complicated procedures partially described above.The second category of papers discusses the whole concept of cultural dimensions and correctness and relevance of Hofstede's estimates.
This study does not attempt to present overall review of literature published after the year 2002 (this would be the task for a separate paper), but some examples are mentioned below.
For instance, Scholtens et al. [13] studied the influence of cultural values on the international differences in business ethics.The authors analyzed ethical policies of almost 2700 firms in 24 countries and related them to the Hofstede's cultural indices.These ethical policies include human rights policy, policy against bribery and corruption, and the overall comprehensiveness, implementation and communication of firm's codes of ethics.The authors found that there is a positive relation of individualism and uncertainty avoidance dimension to the ethical policies and negative relation of masculinity and power distance.
Moreover, Chui and Kwok [14] studied the way that national cultures influence consumption of life insurance on a simple of 41 countries (data collected from 1976 until 2001).The findings suggest that power distance and masculinity/femininity dimension are negatively related to life insurance consumption, whereas individualism is a positive impact.
Furthermore, Nadler and Zemanek [15] recomputed the effects of culture on economic development in 31 countries in order to update similar types or research published in the 1980s and the 1990s.The authors studied the effect of cultural values on per capita gross domestic product, literacy rates, the negative population growth rate, and life expectancy development.Their study also partially supported Hofstede's findings.
In addition, Singh [16] studied the relationship between national culture and adoption of new products, ideas, or behavior in order to suggest a framework for distinguishing between innovative and imitative behavior.His results suggest that cultures where power distance is small, uncertainty avoidance is weak, and cultural values suggest masculinity are more prone to innovative behavior in their consumption patterns.The other result suggests that consumers coming from a large power distance, strong uncertainty avoidance and/or feminine culture are more susceptible towards normative influences, while those from collectivist countries tend to relay on interpersonal communication.
There is an ongoing discussion in the research literature about whether the dimensions of Geert Hofstede are still usable, given the fact, that most of the original data were collected in the 1960s and the 1970s (some countries, especially but not only the countries of transition, were added later).
Drogendijk and Slangem [17] studied the effects of five cultural distance measures on the choice of multinational enterprises between expanding abroad through greenfield or acquisition.The results suggest that high scores on all cultural distance measures significantly increase the likelihood that multinational enterprises choose greenfields and that the explanatory power of Hofstede's measures is comparable to explanatory power of Schwartz measures, while the explanatory power of managerial perceptions is somewhat lower.The authors concluded that it was too early to consider Hofstede's dimensions to be outdated and Schwartz's framework to be superior.
Soares et al. [18] examined different approaches as for conceptualization of cultural values for marketing research.The paper compares the five dimensional model of culture designed by Hofstede with eight other models, which attempt to operationalize cultural values and presents arguments for Hofstede's approach.Most importantly the authors argue that Hofstede approach is "simple, practical and usable shortcut to the integration of culture into studies" ( [18], p. 283).
Cultural indices represent a valid tool for international differentiation of countries according to the relevant cultural values.Even though some indices were computed at the end of the last century, their explanatory power seems to be comparable to the explanatory power of other similar tools computed more recently, while the simplicity and practical applicability exceeds the alternatives.Using cultural indices in economic research bring new meaningful insights into economic processes.
According to Torgler [19], who analyzed the main determinants of tax morale in Canada and employed game theoretic literature which analyzes tax evasion in all the types cooperative and non-cooperative games, in the real world people pay more taxes than traditional economic approaches would predict, and therefore it is necessary to analyze other than economic determinants of the willingness to pay or evade taxes [19].The author presents a set of four macroeconomic models which differ by control variables and are based on the data of World Value Study for Canada.
Furthermore, Torgler and Schneider [20] studied the attitudes towards paying taxing in multicultural European countries, represented by Belgium, Switzerland and Spain, and found that direct democracy, citizens' involvement and religiosity contribute to raising the tax morale.Additionally, Schneider finds that the willingness to pay taxes is also predetermined by the size of the shadow economy, which according to his estimations, makes from 20 to 28 percent for the countries of Central and Eastern Europe.
According to Hanousek and Palda [21], tax evasion is one of the central problems facing the governments of transition countries.Quite often, the system of taxation in those countries target narrow groups of wealthy people that are capable of endowing the state budget, which leads to inefficient system of taxation and low tax morale.Transition economies often lack quality government services that can be found in their Western counterparts.Moreover, massive spread of international travel, which became possible after the fall of the Iron Curtain, and the possibility to compare the level of public services at home and abroad, further reduce tax morale in transition countries.With regard to this, Feld and Tyran [22] find the link between the willingness-to-pay taxes and the quality of government services in transition countries.
Surely, the level of tax morale and the perception of civic duty that result in the willingness-to-pay taxes or engage in corruption, differ from country to country.Cabelkova and Hanousek [23] find, in the example of data from the Ukraine, that the level of corruption and trust in the rule of law and the state might be considerably lower in some transition economies than others.Further, Hanousek and Palda [24] analyze the tax morale and level of tax evasion in the Czech Republic and come to the conclusion an evasion Kuznets curve (meaning that the tax morale rises with the growing wealth of the citizens) exists, and that tax revenues are likely to increase with the increasing quality of public services.
Gupta and McGee [25] studied the effect of religion on tax evasion perceptions in Australasia.For an Australian study, Buddhists were significantly less opposed to tax evasion than Roman Catholics, Protestants, or Orthodox Christians.Roman Catholics were significantly less opposed to tax evasion than Protestants.Differences in mean scores for other religion comparisons were not statistically significant.
Ross and McGee [26] studied the effect of religion and socio-demographic variables on tax evasion in Malaysia.They found that Protestants were most opposed to tax evasion, followed by Roman Catholics, Muslims, Hindus, and Buddhists.In addition, those who were in the middle of the "God is important" statement spectrum were least opposed to tax evasion.
Ross and McGee [27] studied attitudes toward tax evasion in Switzerland.They found that respondents in the "Other" category, which presumably included atheists and agnostics, were most opposed to tax evasion, followed by Protestants, Roman Catholics, and Muslims.Those who attended religious services more frequently were also opposed to tax evasion than those who attended less frequently.
We continue with this line of research and estimate the effects of religion on the willingness to pay taxes in the respondents in the Czech Republic.

Empirical Evidence: A Tax Morale Model
Similarly to Torgler [19], this paper analyzes the determinants of tax morale using World Value Study (2008) data.Contrary to mentioned study, it employs the data for only the Czech Republic and uses more than one indicator of tax morale in order to test for the robustness of our findings with respect to the chosen indicator.Namely, we work not only with subjective attitudes towards cheating on tax in general, but also with the subjective opinions on paying cash to avoid taxes and claiming state benefits when not eligible.
We chose the case of the Czech Republic due to the specific situation with regard to the political and economic situation amongst all the post-Communist countries of Central and Eastern Europe.Although the Czech Republic is probably the most successful case of economic transformation and (alongside with Poland) revealed the highest economic growth (prior to the economic and financial crisis of the 2008), the trust in the government and in the state are quite low.In 2011, the prestigious British journal "The Economist" called the Czech political system a "rotting partitocracy".According to the journal, political parties in the country hijacked the democratic process and their leaders became arrogant and corrupt [28].This situation might be described as a "state capture"-an excessive exploitation of public resources by the ruling elite [29].An independent study conducted by researchers from the Charles University in Prague introduced the so-called "zIndex" of public procurement.The index shows that 67% of the €13.7 billion spent in the Czech Republic between 2006 and 2010 is not tracked in the government's official procurement database.About 14% of all tenders during that period (worth some €2 billion) only had one bidder, and none meet the criteria of the Organization for Economic Co-operation and Development (OECD) and the Regional Development Ministry [30].
In order to identify the determinants of the tax morale in the Czech Republic, we constructed three ordered Logit models.All of the models include same sets of independent variables and differ by the estimates of tax morale variable.We used three estimates for tax morale, all coming from the World Values Survey (WVS) conducted in 2008.In formal terms, our empirical model takes the following form: where β are coefficients to be estimated.The right hand side of the equation is in the form of a logistic cumulative distribution function.The dependent variable Y in probability model is binary and indicates whether the subject would be cheating on taxes if he or she had the chance, whether she or he would be paying cash to avoid taxes, and, finally, whether the respondent would claim government benefits to which he or she were not entitled.
We use data collected in the World Values Survey [31].It is based on face-to-face interviews, performed at respondents' homes and in their national language.The sample includes at least 1000 individuals.
Our dependent variable is tax morale.We consider three indicators for this dimension.The first indicator is based on the general question to assess the level of tax morale, as described in the research literature.It is the opinion on cheating on tax, and it refers to the question from the World Values Survey: "Please tell me for each of the following statements whether you think it can always be justified, never justified, or something in between: Cheating on tax if you have the chance".While some studies, such as Torgler [19], consider only this indicator, we utilize two additional indicators.They provide robustness checks and reveal additional information on tax morale and civic duty.
The second indicator considers the opinion on paying cash to avoid taxes.It also reveals information on tax morale and refers to tax evasion in all possible aspects.The third indicator is the opinion on claiming government benefits to which one is not entitled.This does not directly reveal information on tax morale, but does explain the attitude towards civic duty.
The original coding was on ten-point scale with 1 standing for never justifiable and 10 being always justifiable.The dependent variables (Cheating on Tax and Claiming State Benefits) were developed by recoding the ten point scale to a three point scale.Value one is an aggregation of first three points, value two is the aggregation of the second three points, and value three is the aggregation of the last four points.The dependent variable, Paying Cash to Avoid Taxes, was developed by recoding the original ten point scale to a five point scale, where value one is an aggregation of first two points, value two is an aggregation of the second two points, value three is an aggregation of the third two points, value four is an aggregation of the fourth two points, and value five as an aggregation of the last two points.
Apart from the necessary control socio-demographic variables, such as age, sex, occupation, income, and marital status, we include variables, which are to indicate the level of identification with the country as a group of people (national pride variable), due to the fact that pride to be a group member might encourage cooperative behavior within a group.The question of whether paying taxes might be viewed as cooperation in all the cultural environments remains.For example, in the case where citizens do not view the government to be a reliable agent regarding the reasonable use of the taxes collected, it might be viewed by the population that not paying taxes represents cooperative behavior.For this reason we include two variables, which indicate confidence in the government and subjective attitude towards the government, namely, whether the government is viewed to be "good" or "bad".
One variables, very often mentioned in the literature, with respect to any type of morale is religiosity (see, for example, Margolis [32]).One might have certain doubts about what religiosity is and how to measure it.Similarly, as there are many definitions of religiosity, there might be several indicators for it.
The most straightforward might be whether a respondent believes in God.On the other hand, one can claim that belief in God might not as important for a person so as to influence his or her behavior.For this reason we include both the variable indicating the belief and the variable indicating how important God is in one's life.Another critic of these indicators might be the idea that what is important in religiosity is not only in what one believes in but what one does.For this reason we also included the variable mapping how often one visits religious services.This variable is also important in that one may suggest that morality of the respondent is influenced to a large extent by the community that one belongs to and might be reinforced by the information one gets from the priest.
The results of all three models' estimations are presented in Table 2.In all three models, similar to Torgler [19], McGee, [33], Gupta, and McGee [25], and McGee and Bose [34], older respondents have a higher tax morale than younger people and males report lower tax morale than females.
The Logit model that is employed in this paper, mostly the signs of the estimated coefficients, are important, but the magnitude also reveals some information.Here, positive signs indicate that ceteris paribus, the odds of remittance occurrence, are rising with an increase in the explanatory variable.If one is interested in the magnitude, taking the value of exp of the estimate yields the odds ratio, since coefficient actually are log odds ratios.
In accordance to Torgler [19], in our models, education seems to have no significant effect on tax morale.Looking at marital status, in two out of three models, only divorced or separated, out of three possible categories, have lower tax morale than never married.This difference goes in opposite direction than in Torgler [19], who found, that, in Canada, separated people have higher tax morale than singles.Similar to Torgler [19], higher income has a significant negative effect on tax morale.
Looking at employment and occupation, employment status does not seem to have a significant effect on tax morale, whilst, out of eight occupational statuses, only higher and lower controllers, together with manual supervisors, have a significantly higher tax morale compared to farm labor.Probably the most interesting effects on tax morale are given by the variables of attitude to government and national pride.Contrary to Torgler [19], confidence in government and attitude toward government seem to have no significant effect on tax morale (in Torgler [19], trust in government had a highly significant positive effect on tax morale), whilst national pride (how proud you are to be a country's citizen), has a significant positive effect on tax morale (in Torgler, 2003, the effect was also highly significant and positive).Internationally, this finding is interesting from the perspective of how tax morale can be influenced by government interventions.
In the case of Canada, according to Torgler [19], higher national pride and trust in government have a positive effect on tax morale, therefore, government actions might have an effect on tax morale.Similarly, as we found, in the Czech Republic, the effect of the national pride variable on tax morale is positive, however, nonexistent in case of the trust to the government (as our model and data suggest).
Religious beliefs are another issue, which are frequently associated with morality in general and tax morale in particular.According to Torgler [19], religiosity has a positive effect on tax morale in Canada.Church attendance in his model of tax morality had a highly significant and positive effect on tax morale, whilst the particular religious confession of the respondent was not statistically important.Thus, the author concluded that religious confession does not influence tax morale, but religiosity, in of itself, represented in his model by church attendance variable, imposes those behavioral constraints, increasing tax morale.Contrary to his findings, our own suggest that, in the Czech Republic, neither belief in God nor the importance of God in one's life has a statistically significant effect on tax morale.On the other hand, church attendance has a highly significant and positive effect on tax morale in all the three models.These results are similar those obtained to Ross and McGee [27] and are the reverse to those obtained by McGee [33].

Conclusions
We analyzed links between cultural values and taxation in order to find factors which influence tax morale in the Czech Republic.Using the data from the World Value Study, we compared the factors identified for the Czech Republic with those reported in similar research for other countries.
Strong evidence was found that similarly as elsewhere (in particular, in Canada) pride to be a country's citizen and church attendance have significant positive effects on tax morale.On the other hand, confidence in government and attitude toward government were not statistically important, which may, under certain conditions, imply that actions of government do not have an effect on tax morale.
These results are especially interesting due to the fact that the Czech Republic is one of the most atheistic countries in Europe, which was caused by its turn to Protestantism in the 16th century, the Hussite Wars and facing the Crusaders' armies from all over Europe, and finally the restoration of Catholicism by the Habsburgs after the defeat of the Czech Protestant forces at Bílá Hora in 1620.As a result, the Czechs remain one of the most religion-neutral nations in the world (with the exceptions of Catholic strongholds in Southern Moravia).At the same time, the issue of national pride and "belonging" to the country remains very strong in the country (which is typical for all small nations) and typically gets stronger during football and ice hockey matches that involve the national team.However, when it comes to the attitude towards government or the belief in government, the Czech

Table 1 .
Tax revenues (% GDP, 2010) and GDP per capita in selected countries.

Table 2 .
Determinants of tax morale in the Czech Republic, ordered Logit estimates.