Religiosity and Informal Economic Practices in Southeastern European Societies

Abstract

The dominant religions in Southeastern European countries (Serbia, Macedonia, Montenegro, Bosnia and Herzegovina, Kosovo, Albania), Orthodoxy, Islam, and Catholicism, contain social teachings, which include several norms that deal with certain forms of economic practices. These post-socialist societies develop various forms of informal practices, some of which are contrary to elements of religious social teachings and religious ethics. In the process of the revitalization of religiosity after the fall of socialism in this region, the question can be posed as to whether the attitude towards informality and the application of certain informal economic practices, which range from the illegitimate to the illegal (getting things “done” through informal connections, tax evasion, corruption), correlates to some extent with the level of religiosity and the type of religion. The results of the research show that there is a connection between belonging to a certain confession or religion, self-declared religiosity and level of religiosity, and approving of informal practices and engaging in them. At the state level, a specific dynamic was developed even when it came to approving of and engaging in informal practices depending on whether the members of certain confessions were a minority or a majority at the level of the observed country.

1. Introduction

On the territory of the countries that are the subject of this research, there are elements that can be considered common factors of the social context: A socialist past,1 persistent patriarchal culture, a current orientation towards the EU. What differentiates them and what has proved to be important in the post-socialist period, is, among other things, the religious belonging of citizens. During the conflicts and wars that took place between certain ex-Yugoslav republics or ethnic groups within these republics in the 1990s, religious belonging played an important role. In line with the general global trend (Berger 2008) yet instigated by the awakening of identities and conflicts in all of the observed post-socialist societies, a revitalization of religion has occurred (Gavrilović 2013; Zrinščak and Nikodem 2009; Abazović 2015; Cvitković 2004). The issue tackled by this paper is the extent to which the religiosity of the citizens in these societies affects their practices, or more precisely, the extent to which it affects those forms of economic behaviour that are opposed to the dogmatism of the religion to which they belong. The basis of this phenomenon can be found in the ethical teachings of Islam and Christianity (the dominant religions) that forbid certain forms of economic practices and the potential internalization of religious values by our respondents, i.e., believers belonging to these religions. The basic research question of this paper is: Does the attitude towards informality and the application of certain informal economic practices, which range from the illegitimate to the illegal (getting things “done” through informal connections, tax evasion, corruption), correlate to some extent with the level of religiosity and the type of religion confessed in certain countries?

This question is part of a wider problem of the relation between religion and moral judgement and behaviour (Stark 2001), which examines the almost undisputed assumption that religion facilitates the establishment and preservation of the moral order. Inglehart and Norris observe:

The predominant religious cultural traditions in any society, such as the legacy of Protestantism and Catholicism in Western Europe, are expected to leave a distinct imprint upon the contemporary moral beliefs and social attitudes that are widespread among the public in these nations. (…) The major faiths of the world express divergent teachings and doctrines on many moral values and normative beliefs, such as those surrounding the roles of women and men, the sanctity of life, and the importance of marriage and the family

(Norris and Inglehart 2004, p. 20).

More specifically, this paper deals with the correlation between religion and economic behaviour. There are a number of studies from Weber (2005) and his well-known and controversial hypothesis on the influence of Protestant ethics on the development of capitalism, through Iannaccone (1998), Rachel M. McCleary and Robert J. Barro, to date, which deal with the (non-)existence of the relationship between religion and various aspects of economic behaviour. For example, McCleary and Barro emphasize that “Religious beliefs affect the economy by fostering traits such as work ethic, honesty (and hence trust), thrift, charity, hospitality to strangers and so on” (McCleary and Barro 2006, p. 51). On a similar note, Landes (1998) attributes the failure of Spanish development in the 16th and 17th century to the culture of intolerance diffused by the Catholic Church, which forced some of the most skilful people out of the country. Finally, Stulz and Williamson (2001) attribute the low level of creditors’ protection present in Catholic countries to the anti-usury culture pervasive in the Catholic tradition (Guisoa et al. 2003, p. 226). Different religions are, thus, focused on specific behaviours to varying extents, including the economic aspects of social life in general or its different parts. In a cross-country study, both La Porta et al. (1997) and Inglehart (1999) found certain evidence for some of these relations. However, one should always exercise caution, as proposed by Iannaccone (1998), when conducting such research: “It is possible, of course, that religion’s statistical ‘effects’ are entirely spurious. One may readily posit the existence of underlying characteristics that shape both religious behaviour and any other behaviour. ‘Good’ kids may avoid drugs, stay in school, and go to church”. The study of this relation is multidimensional, and a number of elements have to be considered, such as the level and type of religiosity (Stark 2001).

According to Stark (2001), the results of research in Eastern Europe, in the post-communist countries in the Orthodox region, show a tendency towards a weaker link between religion and morality:

Within Eastern Europe, the more remote Orthodox conception of God will result in correlations that are weaker in Orthodox nations than in non-Orthodox nations. In combination with the effects of Communist repression, this will result in a lack of any significant correlations between God and morality in the Orthodox nations, except in Romania.

Similar findings are also present in research conducted in Croatia, even though Catholics are dominant in the country, showing that there is no strong connection between religiosity, morality, and forms of social behaviour in post-socialist countries. If religion cannot demonstrate its power to maintain the moral order in post-communist countries, does that mean that we are dealing with long-term effects of the communist order? Does the post-communist revitalization have no power to significantly change this effect? What is the purpose of religion in post-communist societies? “Particularities about Eastern Europe are confirmed also in this analysis. Religion plays a minor role with concerning different moral norms in comparison to Western European countries” (Zrinščak and Nikodem 2009, p. 5).

Previous research, conducted in Serbia, shows that there is a significant correlation between religiosity and the condemnation of the following forms of behaviour: Adultery, recreational drug use, lying for one’s own benefit, homosexuality, abortion, divorce, euthanasia, suicide, genetic experiments, and casual sex but there is no correlation with cheating in the economic sphere, receiving bribes, tax evasion, fraud and abuse of state welfare systems, not paying public transport fares (Gavrilović 2013).

Informal practices, although highly developed and omnipresent in social life (Ledeneva 1998, 2013), particularly in the period of transition under the conditions of adopting formal rules that are not close to the traditional matrix of Balkan societies (Chavance 2008), have rarely been the subject of sociological research in the observed region. Despite the comprehensive sociological and anthropological literature on these somewhat modish topics that relate to other parts of the world, there are only a few qualitative studies in this region (Cvejić 2016). The observed practices first developed before the period of transition and have been present ever since; however, this period has suited them well due to the existence of underdeveloped institutions, the current social moment and the unfavourable experience of the socialist period (Giordano and Hayoz 2013, pp. 11–12). The connection between religion and informal practices has not been studied in the region examined in this paper.

The religions that are dominant in these societies either perceive certain economic practices differently or they do not possess a differentiated perspective on this sphere of social life at all. What is of particular importance is which of the forms of behaviour are considered unallowable bearing in mind the observed attitudes. Previous sociological analyses speak of the following connections:

Judaism has the strongest negative impact on the willingness to cheat on taxes, followed by Protestantism (second), Catholicism and Hinduism (third), and Islam (fourth). The ranking changes somehow when it comes to accepting a bribe. The strongest negative effect is for Buddhist, with Protestants and Muslim next, and Catholics last. Protestants are the only religious group that favours incentives. This result vindicates Weber’s claims

(Guisoa et al. 2003, p. 228).

In this region, the dominant religions are Islam, Orthodoxy, and Catholicism; hence, the fundamental teachings of these religions that may affect the observed practices will be presented here briefly.

The economic teachings of Islam rest on sharia.

Islam and the Islamic economy based upon it aim to create a just, moral, ethic, and sustainable social system in which an individual and the society are mutually dependent and connected. Islam acknowledges private property, market economy, and encourages the accumulation of wealth and the striving of an individual to maximize their gain and satisfaction, however, exclusively on the basis of the principles of the permissible or the forbidden (Halal or Haram), discouraging lavishness, and prescribing the obligation of the rich to provide help to the poorer members of the community

(Čočić and Pešić 2012).

These tenets, which clearly delineate the possible actions of believers, can be used as behavioural guidelines in conditions of existing religiosity that significantly increased after the period of socialism. Catholicism insists on the moral aspect of the economic sphere, where taxes should contribute to the development of a community and solidarity. The general instructions that stem from the principles of Christian morality are present (Papsko veće za pravdu i mir 2016). Orthodoxy is without a clear discourse on the work ethic, with labour primarily serving the satisfaction of physical needs (food, clothes, senses), which is of secondary importance itself. Orthodox churches deal with other issues as primary ones, while economic issues are not in focus nor in the discourse of the representatives of religious communities (Savramis 1991).

The differences that exist in social teachings of the religions present in the region can affect the adoption and practice of certain behaviours, labelled as unsuitable in these teachings.

In light of the above mentioned sociological insights into the relation between religion and certain forms of economic behaviour, we will analyse the influence of belonging to a certain religion or confession,2 self-declared religiosity,3 and the level of religiosity4 as the factors of attitudes5 towards and engagement6 in specific informal economic practices (corruption, tax evasion, getting things “done” through informal connections).

The data presented in this research were collected within the project Closing the gap between formal and informal institutions in the Balkan, in a survey conducted during the period from May to June 2017. The target population used in this research consisted of the entire 18+ population of permanent residents of the target countries. The sample presents a national proportional, three-stage random representative stratified probability sample,7 where respondents were randomly chosen household members of 18+ years of age. The survey was conducted face-to-face as a computer-assisted personal interview (CAPI) on a representative national sample of six societies that were encompassed by our research—Bosnia and Herzegovina (1246 respondents), Serbia (1127 respondents), Macedonia (1015 respondents), Kosovo8 (930 respondents), Albania (919 respondents), and Montenegro (803 respondents).

The one-way and the two-way analyses of variance were used for the relations and effects analysis, with a Chi-square test performed for category variables, while in those cases where significant differences, i.e., significant main factors and/or interactions, were discovered, further appropriate tests were used—analysis of simple effects and adequate post-hoc tests (bearing in mind that the assumption of homogeneity of variance was compromised in the majority of cases and that the groups were not equal in size, the Games-Howell post-hoc test was used). The SPSS 23.0 software package was used for data processing. The obtained results are presented in graphs because we believe that they are thus shown in a more illustrative manner.

2. Results

2.1. Data on Religiosity in the Observed Region

First, we will present the religious panorama of the observed societies (Bosnia and Herzegovina, Serbia, Macedonia, Kosovo, Albania, Montenegro). In these countries, members of Orthodox Christianity and Islam are the dominant ones, with only 5% of Catholics. Figure 1 shows the distribution of confessions in the observed countries.

In Serbia, Macedonia, and Montenegro, Orthodox Christians are the most dominantly present, while in Bosna and Herzegovina, Albania and Kosovo, the majority of believers are Muslims. Taking into account the number and distribution of confessionality in the observed countries presented in the below graph, only the three most populous confessions—Orthodox, Muslim, and Catholic, will be included in the analyses below.

A specific phenomenon found in this region is the difference between confessionality and religiosity. Namely, due to the merging of ethnic and religious identities (Gavrilović et al. 2016), as well as the ethnic conflicts that took place in this region, respondents claim that they belong to a religious community (Orthodox, Muslim, or Catholic) but in some cases, they are not necessarily religious. This phenomenon of belonging without believing (Davie 2005) is not exclusively characteristic of the territory of Southeastern Europe, yet the causes here are specific because of identity struggles, as mentioned above (Gavrilović and Đorđević 2018). Figure 2 shows religious self-declaration in these countries.

Among those declared confessional members, there were some who claim not to be religious or even to be atheists. The number of those claiming to be atheists was the highest in Albania (and can be seen in Figure 1, over 80% of the Albanian population declared themselves Muslim). This percentage is significantly higher in comparison to all of the other observed countries since atheisation in Albania during the socialist period was the most extensive of all, and its consequences are still evident today. On the other hand, in Albania as a relatively ethnically homogenous country, there have been no open ethnic conflicts, unlike the case in the other countries, to initiate the renewal of religiosity as an indicator of ethnic differences. Albania also possesses the highest number of those who refuse to answer the question on religiosity—no less than 21.7%.9

Therefore, we used confession as a factor of religious self-declaration for further analyses (Figure 3). The lowest proportion of religious people can be found among the Orthodox population (around 25% of them consider themselves non-religious) but the fewest atheists as well (3%). Muslims are the most religious of all (76%), while the highest percentage of atheists can be found among Catholics (over 10%). The observed differences between confessions were also statistically significant at the level of χ² (5427,6) = 77.369, p < 0.001.

When examining confession and country of origin of respondents together, one can obtain new insights into confessionality and religiosity within certain societies (Figure 4).

The Chi-square independence test showed that there was a significant difference between religious self-declaration and the confession to which respondents belong in Serbia: χ² (1026,6) = 16.562, p = 0.011; Macedonia: χ² (984,4) = 26.571, p < 0.001; Albania: χ² (703,4) = 10.424, p = 0.03, while this was not the case in the other countries. In Serbia, the most religious were Muslims (almost 70%), while the least religious were Catholics (more than a half). In Macedonia, over 90% of Muslims declared themselves religious, while only 7% say that they were not, compared with almost 20% of Orthodox adherents in this country who declared themselves non-religious. In Serbia and Albania, the number of religious members of all confessions was lower than in Bosnia and Herzegovina and Macedonia. Even though the number of religious members was generally lower in the Orthodox population, in Kosovo, 90% of Orthodox adherents declared themselves to be religious. It was obvious that religiosity was affected by the position of religious communities within a country.

The nature of religiosity in the observed countries was further analysed by measuring the level of religious activities such as prayer and visiting temples. Figure 5 shows the level of religiosity and the differences that exist in the observed countries (in more detail see: Appendix A,

Table A1. Level of religiosity among countries.

Country	SS	df	MS	F	p	η2
Between Groups	4482.046	5	896.409	114.346	0.000	0.091
Within Groups	44,912.146	5729	7.839
Total	49,394.192	5734

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

).

The level of religiosity was also significantly correlated with the confession to which respondents belong (Figure 6).

A statistically significant difference was determined between the confessions, F(2,5615) = 64.482, p < 0.001, while additional post-hoc comparisons showed that both Muslims (m = 6.29, sd = 3.33, p < 0.001) and Catholics (m = 6.74, sd = 3.07, p < 0.001) were significantly more religious than Orthodox Christians (m = 5.48, sd = 2.37). There was no significant difference between Muslims and Catholics with regard to the level of religiosity. Orthodox Christians had the lowest level of religious practices, which had already been established in various studies (PRC 2017); however, another important finding was that it has been growing in those countries where Orthodox adherents are a minority (for more detail see Appendix A,

Table A2. Level of religiosity among confessions.

Confession	SS	df	MS	F	p	η2
Between Groups	1126.951	3	375.650	44.738	0.000	0.023
Within Groups	47,357.265	5640	8.397
Total	48,484.216	5643

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

).

However, what is specific for the region in question and in accordance with our previous statements, is the fact that these activities also possess an intra-state dynamic. Namely, the level of religious practices rises or falls not only on the basis of belonging to a religion or a confession but also in relation to the extent of these activities at the state level, as well as in relation to the position of a community—as a minority or a majority—and the relations between religious communities within a country. With this in mind, the level of these religious practices is the highest among the Catholic and Orthodox populations in Kosovo, followed by Muslims in Macedonia, where both of these communities form a minority (Figure 7).

The two-way analysis of variance of different groups was used to explore the influence of the country of respondents’ origin and the confession to which they belong on the level of religiosity. All influences were statistically significant at the level of p < 0.001. Furthermore, the influence of the interaction between the country of origin and the confession was statistically significant as well, F(12,5623) = 30.165, p < 0.001. In order to examine the influence of this interaction, we performed an analysis of simple effects (a one-way ANOVA for each country separately), which showed that belonging to a certain confession affects the level of religiosity in the majority of countries.

The analysis showed that there was a significant influence of belonging to a certain confession on the level of religiosity in Serbia, F(2,999) = 3.933, p < 0.020, however, the Games-Howell post-hoc test did not show any significant difference between the members of different confessions. The situation was similar when it came to Montenegro, where the analysis showed a statistically significant difference, F(2,762) = 3.027, p < 0.049, while the post-hoc test did not point to any specific difference between the groups. Contrary to this, the analysis showed that in Bosnia and Herzegovina there was a statistically significant difference in the level of religiosity between different confessions, F(2,1145) = 26.034, p < 0.001, and between all of the groups at that—Catholics (m = 7.51, sd = 2.87) were significantly more religious than both Muslim (m = 6.82, sd = 2.12) and Orthodox believers (m = 5.86, sd = 2.45), while Muslims were significantly more religious than Orthodox adherents. The analysis further showed that there was a statistically significant difference between the members of different confessions in Macedonia as well, F(2,983) = 262.455, p < 0.001, i.e., Muslims (m = 9.21, sd = 2.71) in Macedonia were significantly more religious than the Orthodox population (m = 5.15, sd = 2.99). The situation in Albania was the opposite, with the analysis showing that there was a statistically significant difference between the members of different confessions, F(2,879) = 17.408, p < 0.001, both between Muslims (m = 4.01, sd = 2.14) and Orthodox believers and Muslims and Catholics (m = 5.53, sd = 3.37), where Orthodox and Catholic believers were significantly more religious than Muslims. As far as Kosovo is concerned, the analysis showed that there was no statistically significant difference between the members of different confessions with regard to the level of religiosity (p < 0.066) (Appendix A,

Table A3. Country and confession on the level of religiosity.

Source	df	MS	F	p	η2
Country	5	211.970	30.159	0.000	0.026
Confession	2	147.087	20.928	0.000	0.007
Country * Confession	10	253.138	36.016	0.000	0.060
Error	5598	7.028
Total	5616
Corrected Total	5615

Note. df—degrees of freedom, MS—Mean squares.

). All the data corroborate the hypothesis that the level of religiosity of the members of certain religious communities is affected by their position and social context.

In Serbia, Macedonia, and Montenegro the most numerous were Orthodox believers, while in Bosnia and Herzegovina, Albania, and Kosovo the dominant group was Muslims. The members of the Catholic confession were most present in Bosnia and Herzegovina (even though it was a case of a smaller community than the Muslim and the Orthodox ones) and least present in Macedonia. Protestants were somewhat significantly present only in Serbia (1.6%).

The above analyses show that respondents from Albania were significantly less religious than respondents from all the other countries. Serbia was also characterised by a significantly lower level of religiosity compared with all the other countries, excluding Albania. The respondents from Kosovo were significantly more religious than the populations of the other countries. Bosnia and Herzegovina were characterised by a higher level of religiosity in comparison with Serbia, Montenegro, and Albania. Montenegrins were significantly more religious only when compared with Serbs and Albanians, while in relation to the other countries they were characterised by a significantly lower level of religiosity. When observing the influence of confession on the level of religiosity, both Muslims and Catholics were significantly more religious than Orthodox adherents. In Macedonia, Muslims were significantly more religious than Orthodox believers, while the situation was the opposite in Albania—Orthodox and Catholic believers were significantly more religious than Muslims. This result indicates the existence of “intrastate dynamics”, a generally lower or higher level of religiosity, as well as the influence of the position of a religious group (minority or majority) on the level of religiosity.

2.2. Religion and Attitudes toward Informal Economic Practices

Taking into consideration the nature of attitudes as relatively stable structures, we analysed the attitudes that (do not) justify engaging in informal practices, mentioned in the introductory methodological section of this paper and correlated them with the characteristics of religiosity (self-declared religiosity and level of religiosity) of our respondents. We believe that the existence of certain attitudes can be an indicator of potential behaviours in the observed societies.

2.2.1. Confession or Religion as a Factor of Justifying Informal Practices

In further analysis, we will determine the connection between one’s confessional belonging and religiosity, on the one hand, and attitudes toward informal economic practices, on the other. Belonging to a specific confession or religion (Orthodoxy, Catholicism, Islam)10 was analysed as a factor justifying these practices.

The first finding was that attitudes towards “using informal connections” to perform various activities were a very frequent practice in these transitional societies. Traditional societies are based on traditional support networks but also on new ones grounded in political networking. In the entire sample, a specific confession or religion played an important role in justifying such practices (Figure 8).

Generally, the degree of justification for such a practice was very low, yet the analysis of variance confirmed the existence of a statistically significant difference in the level at which the members of different confessions justify using connections and acquaintances for getting things done, F(2,5777) = 95,901, p < 0.001. A post-hoc test further showed that there was a significant difference in the scores of all three groups of respondents. Namely, Orthodox believers condoned this practice to the greatest extent (m = 2.44, sd = 1.31), followed by Catholics (m = 2.22, sd = 1.28), with the lowest degree of justification found in Muslims (m = 1.96, sd = 1.27). In Muslims, this phenomenon can be linked to the demand for “fair-play”, i.e., the behaviour that is “halal”, being important, where using one’s connections to achieve a disloyal advantage certainly does not belong to such behaviour (Appendix A,

Table A4. Using connections for getting things done among confessions.

	Confession	SS	df	MS	F	p	η2
Using connections for getting things done	Between Groups	322.157	2	161.078	95.901	0.000	0.032
	Within Groups	9703.193	5777	1.680
	Total	10,025.350	5779

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

).

The second observed attitude was regarding tax evasion (Figure 9).

The one-way analysis of variance indicated a statistically significant difference in the level at which the members of different confessions justify tax evasion, F(2,5755) = 7.051, p = 0.001. Additional comparisons using the Games-Howell test showed that Muslims (m = 1.48, sd = 0.94) justified the practice of evading taxes to a statistically much less significant degree both in relation to the Orthodox (m = 1.57, sd = 1.03) and the Catholic populations (m = 1.63, sd = 0.99) (Appendix A,

Table A5. Evading taxes among confessions.

	Confession	SS	df	MS	F	p	η2
Evading taxes	Between Groups	14.102	2	7.051	7.170	0.001	0.002
	Within Groups	5659.390	5755	0.983
	Total	5673.492	5757

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). This phenomenon can also be linked to the social teachings that prefer those actions that are “halal”.

Figure 10 shows the attitudes towards the approval of accepting a bribe with regard to a respondent’s confession.

The analysis of variance results showed that there was a statistically significant difference in the level at which respondents of different confessions justified accepting/giving a bribe, F(2,5781) = 8.511, p < 0.001. Additional comparisons performed by using the Games-Howell test showed that Catholics (m = 1.67, sd = 1.03) condoned this practice to a greater extent than both Orthodox believers (m = 1.46, sd = 0.91) and Muslims (m = 1.44, sd = 0.91). Between Orthodox and Muslim believers there was no significant difference (Appendix A,

Table A6. Accepting/giving a bribe among confessions.

	Confession	SS	df	MS	F	p	η2
Accepting/giving a bribe	Between Groups	14.396	2	7.198	8.511	0.000	0.003
	Within Groups	4889.097	5781	0.846
	Total	4903.493	5783

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). The fact that the number of Catholics in the sample was extremely small, and that the number of those who condoned this practice was even smaller, raises doubts about this finding, making the one that relates to there being no difference in the justification or condemnation of the use of bribes between Orthodox and Muslim believers more significant. All in all, there was almost a unanimous disapproval of this illegal practice among our respondents, regardless of their confessional belonging.

The findings show that confession was correlated with justifying illegal and illegitimate practices and that Muslims condoned these practices to a lesser extent than others.

Significant differences were found regarding the level of justifying the use of connections and acquaintances for getting things done between respondents of different confessions. Orthodox believers condoned this practice to the highest degree, followed by Catholics, while Muslims were the ones who approved of this practice to the lowest degree. Muslims also condoned the practice of tax evasion to a significantly lesser extent, both in relation to the Orthodox and the Catholic population. Catholics were the ones who condoned this practice to a much greater extent than both Orthodox adherents and Muslims. No significant difference was determined between Orthodox adherents and Muslims here.

Our findings corroborate the assumption that the social teachings of Islam on behaviour in the economic sphere contributes to such attitudes towards informal practices among its respondents, while the generalised attitudes of Catholics and especially the complete neglect of this sphere in the Orthodox population favours more liberal attitudes of members of these confessions towards practices that are categorised as illegitimate and illegal.

2.2.2. The Influence of Self-Declared Religiosity on Justifying Informal Practices

As already mentioned, in the countries that were the subject of this analysis, there exists a phenomenon of confessional identification that was not accompanied by an appropriate religiosity, thus, giving rise to a gap between those who declare themselves believers of one religion and those who speak of themselves as religious persons (belonging without believing, (Davie 2005)). This is why the data on self-declared religiosity and the level of religiosity were taken into account as a separate factor of the (dis)approval of informal practices.

The one-way analysis of variance indicated a statistically significant difference in the level of justifying the use of informal connections and acquaintances in order to “get a job done” between respondents of different self-declared religiosities, i.e., those that declared themselves religious, non-religious, or atheist, F(2,5419) = 20.475, p < 0.001 (Figure 11). Additional comparisons showed that non-religious persons (m = 2.45, sd = 1.31) justified this practice to a statistically significantly lower degree than religious persons (m = 2.16, sd = 1.31) and atheists (m = 2.23, sd = 1.29), while no significant difference was observed between religious persons and atheists (Appendix A,

Table A7. Using connections for getting things done among religious self-declaration.

	Religious Self-Declaration	SS	df	MS	F	p	η2
Using connections for getting things done	Between Groups	71.531	2	35.765	20.475	0.000	0.008
	Within Groups	9465.800	5419	1.747
	Total	9537.331	5421

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). Such a finding can, to some extent, be understood in terms of the established practice, in these societies, in which things get done outside institutions and through traditional connections (family, fellow citizens, members of the same confession) (Cvejić 2016); particularly if one bears in mind that religious people are usually more traditional than non-religious people, which is where one could look for an explanation for the obtained data. Believers in this region mainly fall under the type of traditional believers, with the number of those church-convicted being much smaller (Gavrilović 2017; Abazović 2015).

When it comes to justifying tax evasion and religious self-declaration (Figure 12), the analysis of variance did not show any statistically significant difference between religious, non-religious and atheist respondents F(2,5399) = 0.194, p = 823 (Appendix A,

Table A8. Evading taxes among religious self-declaration.

	Religious Self-Declaration	SS	df	MS	F	p
Evading taxes	Between Groups	0.383	2	0.192	0.194	0.823
	Within Groups	5325.554	5399	0.986
	Total	5325.937	5401

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

) and as can be seen in the graph, the degree of justification of this practice was very low among all respondents.

The next practice whose justification was observed in relation to the self-religious declaration was accepting/giving a bribe (Figure 13).

The one-way analysis of variance indicated a statistically significant difference in the level at which the respondents who self-declared differently on the issue of religiosity justified accepting/giving a bribe, F(2,5427) = 13.945, p < 0.001. Additional comparisons showed that religious persons (m = 1.42, sd = 0.88) approved of this practice to a statistically significantly lower degree compared with both non-religious persons (m = 1.53, sd = 0.97) and atheists (m = 1.64, sd = 1.03). No significant difference was observed between religious persons and atheists (Appendix A,

Table A9. Accepting/giving a bribe among religious self-declaration.

	Religious Self-Declaration	SS	df	MS	F	p	η2
Accepting/giving a bribe	Between Groups	23.172	2	11.586	13.945	0.000	0.005
	Within Groups	4508.922	5427	0.831
	Total	4532.094	5429

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

).

When examining the use of connections and acquaintances to perform certain activities, non-religious persons justified this practice to a significantly lower degree than religious persons. As far as tax evasion is concerned, there were no differences, while religious persons justified the practice of accepting or giving a bribe to a significantly lower degree.

2.2.3. The Influence of Level of Religiosity on Justifying Informal Economic Practices

The following segment will present the observed difference in approving of, i.e., justifying, informal practices in relation to the level of religiosity. Figure 14 shows the observed differences in relation to the approval of the practice of using connections and acquaintances to perform certain activities.

When it comes to the level of justifying the use of connections and acquaintances to get things done, the analysis of variance indicated a statistically significant difference, F(2,5630) = 8.808, p < 0.001, between the respondents characterised by different levels of religiosity, while the post-hoc tests showed that the respondents characterised by a high level of religiosity (m = 2.08, sd = 1.26) approved of this practice to a statistically significantly lower degree than the other two groups of respondents, both those characterised by a medium level of religiosity (m = 2.19, sd = 1.29) and those characterised by a low level of religiosity (m = 2.27, sd = 1.35) (Appendix A,

Table A10. Using connections to get things done among different levels of religiosity.

	Level of Religiosity	SS	df	MS	F	p	η2
Using connections to get things done	Between Groups	30.631	2	15.316	8.808	0.000	0.003
	Within Groups	9789.453	5630	1.739
	Total	9820.084	5632

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). This correlation is something that could be expected if one starts from the assumption that there is a link between religious practice and moral judgement.

When it comes to the level of religiosity and justifying tax evasion, the results are shown in Figure 15.

The analysis of variance indicated a statistically significant difference in the level at which respondents characterised by different levels of religiosity justified tax evasion, F(2,5613) = 5.430, p = 0.004. Additional comparisons using the Games-Howell test showed that respondents characterised by a low level of religiosity (m = 1.48, sd = 0.95) approved of this practice to a statistically significantly lower degree in relation to the other two groups of respondents, those characterised by a medium level of religiosity (m = 1.56, sd = 1.01) and those characterised by a high level of religiosity (m = 1.59, sd = 1.04) (Appendix A,

Table A11. Evading taxes among different levels of religiosity.

	Level of Religiosity	SS	df	MS	F	p	η2
Evading taxes	Between Groups	10.714	2	5.357	5.430	0.004	0.002
	Within Groups	5537.833	5613	0.987
	Total	5548.547	5615

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). This could, perhaps, be interpreted from the perspective of Stark’s (2001) findings that show that participation in rituals is not the only indicator of type of religiosity and that mere participation does not correlate with moral judgement but that the manner in which God is imagined and other aspects of religious awareness do (our variable was composed of two elements of religious practice).

When it comes to accepting/giving a bribe with regard to the level of religiosity, the results are shown in the following graph (Figure 16).

The analysis of variance showed that there was no statistically significant difference in the level at which respondents characterised by different levels of religiosity approved of accepting/giving a bribe, F(2,5636) = 960, p = 383 (Appendix A,

Table A12. Accepting/giving a bribe among different levels of religiosity.

	Level of Religiosity	SS	df	MS	F	p
Accepting/giving a bribe	Between Groups	1.618	2	0.809	0.960	0.383
	Within Groups	4751.145	5636	0.843
	Total	4752.763	5638

Note. df—degrees of freedom, MS—Mean squares.

). This practice was justified by an extremely low percentage of all respondents.

The results of the analyses in this segment show that those respondents characterised by a high level of religiosity approved of the practice of using connections to get things done to a statistically significantly lower degree compared to the other two groups of respondents, those characterised by a medium level of religiosity and those characterised by a low level of religiosity. Contrary to this, respondents characterised by a low level of religiosity approved of the practice of tax evasion to a statistically significantly lower degree in relation to the other two groups of respondents, those characterised by a medium level of religiosity and those characterised by a high level of religiosity. These findings could be in agreement with Stark’s (2001) findings that show that participation in rituals is not a good indicator of moral judgement or that social context negates the influence of religiosity in concrete cases.

2.2.4. The Influence of the Country and Confession on the Level of Justifying Informal Economic Practices

In accordance with our analyses of religiosity that showed that there exists a certain dynamic between belonging to a religion or confession and living in a certain country, further analyses will present the data on the influence of the country and confession on the level of accepting and justifying “the use of connections and acquaintances to get things done”. In line with the previous findings, once again it was confirmed that there is an intrastate dynamic of religiosity of the members of different confessions with regard to them being a minority or a majority (Figure 17).

The two-way analysis of variance of different groups was used to explore the influence of the country of respondents’ origin and the confession to which they belong on the level at which they justify the use of connections and acquaintances to get things done. All of the influences were statistically significant at the level of p < 0.001. Furthermore, the influence of the interaction between country of origin and confession proved to be statistically significant, F(10,5762) = 5.771, p < 0.001 (Appendix A,

Table A13. Country and confession on using connections for getting things done.

Source	df	MS	F	p	η2
Country	5	13.680	8.550	0.000	0.007
Confession	2	23.253	14.533	0.000	0.005
Country * Confession	10	9.234	5.771	0.000	0.010
Error	5762	1.600
Total	5780
Corrected Total	5779

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

).

To examine the influence of the interaction, an analysis of simple effects was performed (a one-way ANOVA for each country separately) and it showed that there was a significant influence of belonging to a certain confession on the level of justifying the use of connections and acquaintances to get things done in Kosovo, F(2,893) = 38.889, p < 0.001, i.e., the level of approving of this practice was significantly much higher among the Orthodox population (m = 2.80, sd = 1.01) compared with both Muslims (m = 1.45, sd = 0.98) and Catholics (m = 1.24, sd = 0.64). Similar results were obtained for Albania, F(2,887) = 4.872, p < 0.008. Namely, in Albania, the level of justifying this practice was also determined to be significantly higher for Orthodox believers (m = 2.56, sd = 1.53) than for Muslims (m = 2.12, sd = 1.36) and Catholics (m = 1.89, sd = 1.17). As far as the other countries are concerned, the analysis showed that there was no significant influence of belonging to a certain confession on the level of justifying the use of connections and acquaintances to get things done (Appendix A,

Table A14. Using connections for getting things done among confessions by country.

Country		SS	df	MS	F	p	η2
Serbia	Between Groups	6.523	2	3.262	1.864	0.155
	Within Groups	1822.832	1042	1.749
	Total	1829.355	1044
Bosnia and Herzegovina	Between Groups	7.137	2	3.569	2.162	0.116
	Within Groups	1932.945	1171	1.651
	Total	1940.083	1173
Macedonia	Between Groups	0.886	2	0.443	0.236	0.790
	Within Groups	1869.365	994	1.881
	Total	1870.251	996
Montenegro	Between Groups	1.869	2	0.935	0.664	0.515
	Within Groups	1090.214	775	1.407
	Total	1092.084	777
Kosovo	Between Groups	73.868	2	36.934	38.889	0.000	0.080
	Within Groups	848.114	893	0.950
	Total	921.982	895
Albania	Between Groups	18.185	2	9.092	4.872	0.008	0.011
	Within Groups	1655.533	887	1.866
	Total	1673.718	889

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). The Orthodox population is a minority in the societies where this connection is significant, and that is perhaps why they rely on social networks to achieve their own interests, which leads them to approve of such a practice.

The next measured influence was the influence of the country and confession on justifying “evading (not paying) taxes” (Figure 18).

The two-way analysis of variance of different groups was used to explore the influence of the country of respondents’ origin and the confession to which they belong on the level at which they justify tax evasion. The statistically significant major influence of the country of origin was determined, F(5,5740) = 8.756, p < 0.001. Furthermore, the influence of the interaction between the country of origin and the confession proved to be statistically significant, F(10,5762) = 5.771, p < 0.001 (Appendix A,

Table A15. Country and confession on evading taxes.

Source	df	MS	F	p	η2
Corrected Model	17	13.566	14.306	0.000	0.041
Intercept	1	1039.810	1096.573	0.000	0.160
Country	5	8.302	8.756	0.000	0.008
Confession	2	0.538	0.567	0.567	0.000
Country * Confession	10	4.547	4.796	0.000	0.008
Error	5740	0.948
Total	5758
Corrected Total	5757

Note. df—degrees of freedom, MS—Mean squares.

).

To examine the influence of the interaction, an analysis of simple effects was performed and it showed that there was a significant influence of belonging to a certain confession on the level at which respondents justify tax evasion in Bosnia and Herzegovina, F(2,1166) = 4.915, p < 0.009, and in Macedonia, F(2,986) = 16.171, p < 0.001. Additional comparisons using the Games-Howell test showed that in Bosnia and Herzegovina Orthodox believers (m = 1.7128, sd = 1.12307) justified and accepted tax evasion significantly more than Muslims (m = 1.5312, sd = 0.96254). In Macedonia, the situation was completely different. Namely, in this country, Muslims were the ones who justified tax evasion to a significantly higher degree in relation to the Orthodox population (m = 1.7219, sd = 1.01429). There were no significant differences noted in the other countries (Appendix A,

Table A16. Evading taxes among confessions by country.

Country		SS	df	MS	F	p	η2
Serbia	Between Groups	1.771	2	0.885	0.931	0.395
	Within Groups	990.117	1041	0.951
	Total	991.888	1043
Bosnia and Herzegovina	Between Groups	9.831	2	4.915	4.693	0.009	0.008
	Within Groups	1221.174	1166	1.047
	Total	1231.004	1168
Macedonia	Between Groups	32.342	2	16.171	18.363	0.000	0.036
	Within Groups	868.301	986	0.881
	Total	900.643	988
Montenegro	Between Groups	5.677	2	2.839	1.878	0.154
	Within Groups	1162.033	769	1.511
	Total	1167.710	771
Kosovo	Between Groups	2.838	2	1.419	2.047	0.130
	Within Groups	615.667	888	0.693
	Total	618.505	890
Albania	Between Groups	1.621	2	0.810	1.232	0.292
	Within Groups	585.580	890	0.658
	Total	587.200	892

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). We observed that these communities were minority communities in the said countries and that they were opposed to the state in some manner; thus, their attitude towards not paying taxes can be understood as a proxy for their attitude towards the country itself.

The analysis of the influence of the country and confession on justifying the act of accepting/giving a bribe is presented in Figure 19.

The two-way analysis of variance of different groups was used to explore the influence of the country of respondents’ origin and the confession to which they belong on the level at which they justify accepting/giving a bribe. The statistically significant major influence of the country of origin was determined, F(5,5766) = 5.229, p < 0.001. The influence of the interaction between country of origin and confession also proved to be statistically significant, F(10,5766) = 7.122, p < 0.001 (Appendix A,

Table A17. Country and confession on accepting/giving a bribe.

Source	df	MS	F	p	η2
Corrected Model	17	8.453	10.240	0.000	0.029
Intercept	1	971.762	1177.191	0.000	0.170
Country	5	4.316	5.229	0.000	0.005
Confession	2	0.457	0.554	0.575	0.000
Country * Confession	10	5.879	7.122	0.000	0.012
Error	5766	0.825
Total	5784
Corrected Total	5783

Note. df—degrees of freedom, MS—Mean squares.

).

The influence of the interaction was further examined by an analysis of simple effects, i.e., a one-way ANOVA for each country separately, which showed that there was a significant influence of belonging to a certain confession on the level of justifying the act of accepting/giving a bribe in Serbia, F(2,1047) = 3.373, p < 0.035, Bosnia and Herzegovina, F(2,1171) = 47.213, p < 0.001, Macedonia, F(2,996) = 26.058, p < 0.001, and Montenegro, F(2,771) = 4.419, p < 0.012. Additional comparisons showed that in Bosnia and Herzegovina Muslims (m = 1.45, sd = 0.89) justified accepting/giving a bribe to a significantly lower degree compared with Orthodox (m = 1.61, sd = 1.03) and Catholic believers (m = 1.76, sd = 1.12). In Macedonia, the situation was the complete opposite, i.e., Muslims (m = 1.74, sd = 1.15) were the ones who justify this practice significantly more than the Orthodox population (m = 1.29, sd = 0.79). In the other countries, additional comparisons did not yield any significant differences between members of different confessions (Appendix A,

Table A18. Accepting/giving a bribe by confession among countries.

Country		SS	df	MS	F	p	η2
Serbia	Between Groups	5.879	2	2.939	3.373	0.035	0.006
	Within Groups	912.313	1047	0.871
	Total	918.191	1049
Bosnia and Herzegovina	Between Groups	14.426	2	7.213	7.619	0.001	0.013
	Within Groups	1108.591	1171	0.947
	Total	1123.016	1173
Macedonia	Between Groups	44.093	2	22.046	26.058	0.000	0.050
	Within Groups	842.682	996	0.846
	Total	886.775	998
Montenegro	Between Groups	7.265	2	3.633	4.419	0.012	0.011
	Within Groups	633.816	771	0.822
	Total	641.081	773
Kosovo	Between Groups	0.016	2	0.008	0.014	0.986
	Within Groups	493.224	887	0.556
	Total	493.240	889
Albania	Between Groups	0.592	2	0.296	0.344	0.709
	Within Groups	769.163	894	0.860
	Total	769.755	896

Note. SS—Sum of squares, df—degrees of freedom, MS—Mean squares.

). It appeared that in this case, there was an intrastate dynamic at work as well, i.e., the influence on the position of members of a certain religion, their numbers, minority status, and their current attitude towards the state. There were no patterns when it came to justifying this practice in terms of only one confession.

The influence of the interaction between country of origin and confession proved to be statistically significant for justifying the use of connections as a way to get things done in Kosovo and Albania. In both these cases, Orthodox believers justified the use of connections to a greater extent. This could be a consequence of the higher tolerance among Orthodox believers towards this practice in general but also of the position of minority communities that rely on social networks.

We determined a statistically significant influence of the country of origin and confession on justifying tax evasion in Bosnia and Herzegovina in the Orthodox population and in Macedonia in the Muslim population. This sheds a somewhat new light on the generally lower level of justifying such behaviour among Muslims and corroborates the hypothesis that the position of a religious community determines their attitudes.

We also found a statistically significant influence of the interaction between country of origin and confession (Serbia, Bosnia and Herzegovina, Macedonia, Montenegro) regarding the approval of accepting/giving a bribe. In Bosnia and Herzegovina Muslims justified accepting/giving a bribe to a significantly lower degree than both Orthodox and Catholic believers. In Macedonia, the situation was the complete opposite, where Muslims were the ones who justify this practice to a significantly higher degree compared to the Orthodox population. It seems that, in this case, there was an influence of the position of a community in the observed society as well. In all other countries, additional comparisons did not lead to any significant differences between confessions.

3. The Influence of Religiosity on Respondents’ Informal Practices

Apart from the attitudes towards performing certain actions, i.e., informal economic practices, we also analysed the data on the personal participation of our respondents in informal practices and linked this phenomenon to religiosity. We are aware of the fact that respondents are not willing to admit that they engage in practices deemed not only immoral but also illegal, which to a certain extent affects the validity of the data and the results. Moreover, admitting to tax evasion or giving a bribe can easily lead to legal sanctions, which caused the level of “confessing” to such practices to be understandably much lower.

The percentages shown in Figure 20 point to a somewhat lower frequency of engaging in informal practices among Muslims (around 22%) and Catholics (around 21%) compared to Orthodox believers (almost 30%). The Chi-square independence test confirmed that this was a statistically significant difference, χ² (5895,2) = 36.511, p < 0.001.

The percentages shown in Figure 21 point to a somewhat higher frequency of engaging in informal practices by non-religious persons (28%), while atheists engage in them most frequently (around 35%). The Chi-square independence test showed that the observed difference between self-declared religiosity and engaging in informal practices was statistically significant at the level of χ² (5519,2) = 28.098, p < 0.001.

The Chi-square test showed that there was a statistically significant difference between the level of religiosity and engaging in informal practices, χ² (5735,2) = 12.265, p = 0.002. In this sense, those with an estimated lower religiosity engaged in informal practices more often (around 30%) than those with medium and high levels of religiosity (around 23%) (Figure 22).

As far as different countries are concerned (Figure 23), the Chi-square independence test showed that there was a significant difference between engaging in informal practices and the confession to which respondents belong in Macedonia, χ² (1008.2) = 9.421, p = 0.009, where Orthodox believers admitted to having engaged in informal practices in a higher percentage (around 30%) compared with Muslims (less than 20%), and in Montenegro, χ² (791,2) = 77.996, p < 0.001, where almost 40% of Orthodox respondents stated that they engaged in informal practices, followed by just over 30% of Catholics, with only around 20% of Muslims. There were major Muslim minority communities in these countries, and the results point to a lower level of engaging in informal economic practices. In the other countries, no significant differences were observed related to practising the use of connections, tax evasion and accepting/giving a bribe with regard to the confessions present in these countries.

The research data indicate a somewhat lower level of engagement in informal practices among Muslims (around 22%) and Catholics (around 21%) in comparison with Orthodox believers (almost 30%).

Non-religious people also engaged in informal practices to a slightly higher degree (28%), while atheists were the ones who were involved in such practices most frequently (around 35%). Those whose religiosity was estimated as being lower engaged in informal practices more often (around 30%) than those with medium and high religiosity (around 23%).

In Macedonia, the Orthodox population stated that they engaged in informal practices in a significantly higher percentage (around 30%) compared with Muslims (less than 20%). In Montenegro, almost 40% of Orthodox believers admitted to engaging in informal practices, followed by just over 30% of Catholics, and only 20% of Muslims.

At the level of engaging in informal practices, all of the examined relations indicated the existence of a connection between confessionality, religiosity, and involvement in informal practices.

4. Conclusions

This paper has examined religiosity in six Southeastern European countries (Bosnia and Herzegovina, Serbia, Albania, Kosovo, Montenegro, and Macedonia), as well as the relation between religiosity and informal economic practices. These analyses were performed on the basis of the well-known tenets of sociological but also economic thinking on the connection between, and the influence of, religion on various elements of human behaviour in the economic sphere. Confessionality and religiosity were treated as independent, while economic behaviour was treated as the dependent variable.

The research confirmed the already familiar notion that Muslims and Catholics generally display a higher level of religiosity compared with Orthodox adherents. The novelty introduced by the analysed data, obtained thanks to the comparative approach employed here, relates to the fact that it is not possible to establish consistency regarding the highest level of religiosity being present in one of the religions or Christian confessions across the entire region. It appears that there is a specific dynamic of the level of religiosity in a single country (this phenomenon has already been observed in previous research, (Gavrilović 2017)), which has been confirmed by this study as comprising the following elements:

(a)

The level of religiosity is generally higher at the level of a single country, in all confessions (Bosnia and Herzegovina, Macedonia), or lower (Albania);

(b)

our data show that religiosity is higher in minority groups within a country, which can be linked to the concept of “cultural defence” (Bruce 1999, 2002) and the feeling of being endangered, as well as the use of religion as an important identity marker in Balkan peoples who are connected by many elements—language, history, customs, and who suffer from the curse of the “narcissism of small differences”;

(c)

confessional orientation is present in almost the entire population, it does not correspond to one’s declaration of one’s own religiosity; namely, there are those who identify themselves confessionally but who claim to be non-religious, even atheists (the highest percentage is in Albania).

When it comes to the attitude towards observed informal economic practices that can be considered immoral or illegal, there is a lower level of approval of such practices by Muslims compared to Orthodox believers. These findings correlate with those already determined that also point to the fact that Orthodox adherents display a slightly greater permissiveness in relation to certain behaviours that are not explicitly part of church teachings or discourse (Stark 2001). On the other hand, Muslims show no approval of those moral practices deemed wrong by their religious teachings (Guisoa et al. 2003).

As far as self-declared religiosity and the level of religious practice as factors of accepting or rejecting informal economic practices are concerned, the findings partially correlate with expectations about people who consider themselves religious and substantiate that in practice being more disapproving of such practices. Namely, our findings here are also in broad agreement with those reached by Stark (2001) and Zrinščak and Nikodem (2009), who speak of the religious practice as not being the best indicator of moral behaviour, particularly in the sphere of the economy, with certain elements of religious awareness being more useful in such roles. The other important factor is related to whether we are dealing with a minority or a majority religious community. This fact, which is an important finding of our research, will, to a great extent, determine the behaviour of respondents; members of certain confessions. In the practices examined here, it is also important to note the relation between a religious community and the country in which they live. Changes in behaviour with regard to the position in a social community have been determined in the study on trust conducted by Guisoa et al. (2003).

Attitudes are a relatively good behaviour predictor; however, in addition to this, we asked our respondents about their personal experiences with informal practices. Despite the previously discussed difficulties, i.e., the fact that people are not willing to speak of their participation in illegal activities (tax evasion and accepting/giving a bribe), we still managed to reach the conclusion that at the level of engaging in informal practices, all of the observed relations indicate the existence of a connection between religiosity and engaging in informal practices. Even though a question is often raised on whether religiosity on the territory of the former Yugoslavia is intrinsic (Lavrič and Flere 2008), our research showed that we could say that the values proclaimed by religion are interiorized by respondents when it comes to their behaviour in social situations. These findings differ somewhat from the ones discussed by Siniša Zrinščak when claiming that this connection is relatively low in post-socialist societies: “Cheating on tax is rejected as immoral behaviour, but at the same time the majority believes that their compatriots cheat on tax in a large proportion and are inclined to do the same if they found it ‘personally constructive’.” Our data show that the approval of informal practices correlates with confessionality and religiosity; however, further explanations of possible behaviours should be sought using qualitative methods. In sociology, it is very difficult or almost impossible to determine a causal relationship, thus, even though the deterministic sequence goes from religion to informal practices due to the multiple mediations of this influence (through socialisation, for example) but also due to a number of other influences exerted in societies, we maintain our assertions on the determined correlation.