The aim of this article is to present two different roles of the state affecting individuals' religiosity. First, we provide evidence for the effectiveness of socialist regimes in influencing citizens' opinions by comparing religious beliefs among several generations of Eastern Europeans. Second, the article explores whether the democratization process in Eastern Europe led to a revival of religiosity by applying two strands of reasoning from the secularization framework: Berger's theory of plausibility structures (Berger 1969) and Norris and Inglehart's (2004) existential security hypothesis. The results show that due to an increased plausibility structure created by the democratic states a slight religious revival can be observed in several postcommunist countries.
Religion is a classical topic in the field of sociology and other disciplines. It has been hypothesized to be the main integrating force of societies, by uniting self-interested individuals into a moral community (Durkheim 1912). It is therefore not surprising that its role as a political instrument and a mean of exertion and legitimization of power has been evident throughout history (Martin 1978; Bruce 2003). Religion consequently continues to play an important role in the political domain (Berger 1999; Norris and Inglehart 2004).
The state oppression of Communist countries and the societal liberalization of these countries after the end of the Cold War offer two fascinating perspectives to illustrate the effect that the political domain can exert on individuals' religiosity. First, the state driven secularization of the socialist societies provides us with an example to study the expected decline of religious beliefs. Second, the following democratization raises the question of a “religious revival” (Froese 2004a; Greeley 1994) in Eastern Europe, which can be explained by a re-establishment of stronger ties between churches and the state, which is in line with Berger's theory of plausibility structures (Berger 1969), and thus classical secularization theory.
Using religious socialization and the consequential generational differences in believing as a proxy for state oppression, this research adds to the literature on religiosity in Eastern Europe by including generations and comparing these across time.1 Based on the intensity of forced secularization in a cross-national perspective, the article further speaks to the research community working on sociological theories of religion by testing rival models in the context of postcommunism.
The article is divided in two parts. In the first part, we establish whether socialism was effective in its attempt to secularize its societies through state enforced marginalization of religions. First the relevant literature on the relationship between the states and churches is discussed, before empirical evidence is presented to measure the effect of this process on individuals' religiosity. The second part is dedicated to the time after the end of these authoritarian regimes and the democratization period. Different sociological theories of religion that make different predictions about the development of religiosity are discussed and hypotheses regarding the expected dynamics in Central Eastern Europe are derived. The empirical analysis uses latent class analysis to determine the latent construct of religious beliefs among three different generations (pre-Cold War, Cold War and post-Cold War), and compare the development of belief in God since the collapse of state socialism over a period of 17 years in a sample of five East European societies.
State-Enforced Secularization in Socialist Societies
Based on the ideology of “scientific materialism” and disdain for religion as “opium of the masses,” socialist regimes imposed measures that were actively meant to undermine religion's traditional sources of socialization and transmission of religious beliefs.2 The ideological aim behind this process was a “political forced secularization” (Meulemann 2004:49; emphasis in original) of socialist societies with the creation of “pure atheists,” who do not believe in the existence of God any more. Research on religion during, but also after the collapse of Communism generally describe this state-controlled process to have been successful (c.f. Beeson 1982; Nielsen 1991; Tomka 1994). Borowik (1997:12-3), for example, noted that the “widely spread secularization was undoubtedly one of the most striking features in countries of the former Soviet block,” as “Communist Parties asserted complete control of religious affairs” (Stan 2009:90).
The state-church relationship during socialism was as a consequence characterized by a weak and anomic society, in which the role and power of religion was diminished (Tomka 1991:95). The means to successfully achieve this goal were numerous.3 Most importantly, churches were not able to play a role in public education and religious organizations were monitored or prohibited (Froese 2004b). Moreover, policies of state supplied childcare and increased female labor participation are deemed to have eroded traditional family structures that are seen as main instances of religious socialization (Myers 1996). As direct means, rebellious pastors were imprisoned and often tortured especially in the 1950s during Stalin's reign (Ramet 1987; Burgess 1997; Gautier 1997). Church properties were confiscated, and some places of worship were transformed into warehouses and restaurants (Ramet 1987, 1990; Michel 1992; Stan 2009).
But also after Stalin's death, the situation remained perilous for religious citizens. For example, it was difficult for children of churchgoing families to get one of the few places at the university or to find an adequate job (Burgess 1997). Ramat (1987:5) summarizes the consequences of this suppression as follows: “believers were treated as second-class citizens – excluded from membership in the party, from the officer corps in the military, and from upper-level positions in the government, industrial management, sociopolitical organizations, and the media.” This suppression was somewhat relaxed in the late 1970s and served as an important vehicle of religious and national identity in the wake of Communist opposition – especially in Poland and Hungary (Borowik 2002; Bruce 2003) – which helped to cause the peaceful downfall of socialism 10 years later (Swatos 1994; Minkenberg 1997; Pollack 2002).
The observed effects of this institutional suppression on individuals' religiosity (usually measured by belonging to a denomination) can be summarized as initially a relative stability in the early period of state socialism, which was followed by significant decline especially in the 1960s and 1970s. The perestroika of the 1980s led to an opening of socialist societies and consequently also helped religiosity to slowly blossom (Tomka 1991; Pollack et al. 1998). However, countries varied in their intensity and consequently their success in achieving the aspired secularization “(…) from the marginalization of religion in some countries, such as in the former Soviet Union, through almost total destruction in Albania, and to the strengthening of religion in Poland.” (Borowik 1997:13). Growing numbers of nonbelievers were observed especially in the Soviet Union, Czechoslovakia, East Germany, and Hungary, which were also characterized by “churches within socialism” (Froese 2001).4 In other countries such as Poland, the church remained a relatively strong societal force, as it was connected to national identity (Nowicka 1997) and the Communist regime failed to entirely collectivize agriculture and hence control rural areas (Tomka 2010). But even in a religious stronghold such as Poland, Roman Catholicism was generally portrayed as “an alien anti-Polish force” (Borowski 1985:394) and the Polish United Workers' Party also politically followed the program of “planned (directed or programmed) secularization” (Pomian-Srzednicki 1982:67-85).
In sum, despite country differences in the extent and success of church suppression, the ruling Communist Parties in all Eastern European countries at least attempted to force an ideological-driven secularization process. The question that arises from this discussion is whether the socialist regimes were successful in changing individuals' religious belief. In a first step, we attempt to answer this question by focusing on the time period right at the end of state socialism.
Socialist Socialization and Its Effects on Religiosity
Before 1945, “religion had been one of the main pillars of the societal order and the state, but under the Communist era it was persecuted and pushed to the private sphere” (Tomka 2010:1). As time passed, even the influence of religion over private life declined. Previous research so far has focused on overt religiosity to investigate the effect that this state oppression had on individuals' belief by comparing belonging to a church over time and across space. However, this might not tell the whole truth of state forced secularization, as “it is possible that in the case of many people the abandonment of religion was simply a behavior by which they conformed themselves to the official expectations” (Tomka 2010:3). The difficulty in measuring the extent of believing is the scarcity of survey data during socialism.
To investigate the degree to which the authoritarian regimes affected religious beliefs in socialist societies, generational differences serve as a proxy for forced secularization. Three different generations are distinguished: those socialized before, during and after the introduction of Communism in Eastern Europe. For this argument, socialization processes are assumed to strongly influence an individual's beliefs and values. The classical definition of socialization was established by Hyman (1959:25) as an individual's “learning of social patterns corresponding to his societal position as mediated through various agencies of society.” It is further assumed that what is learned earliest in life is most important, as it serves as an unchanging value basis (Alwin and Krosnick 1991).
In the context of state enforced secularization discussed above, we can formulate the following expectations. On the one hand, those individuals who were socialized during the Communist suppression of religion should be less religious also in later life, as these anti-church attitudes should have manifested themselves in their belief system through socialization processes during socialism. Müller (2009:80) ascribes a declined and continuing low level of religiosity to the fact that the “Communist rule interrupted the steady transmission of religious ideas from generation to generation,” as the here defined Cold War “generation received no religious upbringing” (Tomka 2010:2). On the other hand, the older generation that still grew up before the state oppression of churches already developed their religious beliefs before socialism was installed in their countries. As socialization research has demonstrated once beliefs have been established in young adulthood, these are less prone to change. The socialist impact on religiosity should hence affect the older generation less.
The challenge to test these expectations is the distinction of different generations. In most Christian denominations, typically the age of 14 years is particularly important for the religious self-identification (Myers 1996; Kelley and De Graaf 1997; Argue et al. 1999). Young adolescents need to “confirm” their belief in the traditional confirmation ceremonies, which also marks the coming of age. Therefore, the distinction of generations for this research is deducted from this age. As 1945 is usually considered as the starting point of the Cold War period in Europe, those born 1932 and later religiously came of age during the socialist regimes and are therefore classified as the Cold War generation. They are assumed to have been socialized during a time when church and religion was largely suppressed. The citizens born before 1932 on the other hand still experienced freedom of religion. They are therefore classified as the pre-Cold War cohort. The year 1990 marks the end of the socialist regimes. Hence, those born later than 1976 belong to the post-Cold War cohort.
In a first step, we calculate the difference between those reporting to be religious for the pre-Cold War cohort and those socialized during socialism (Cold War cohort) for 10 different Central and Eastern European countries. It is assumed that the bigger the difference between those two generations, the stronger the effect of the socialist regime on individuals' religious socialization. Using the first wave of the study “Consolidation of Democracy in Central and Eastern Europe 1990-2000,”5 it is possible to get a baseline of religious beliefs just after the demise of state socialism. The data were collected between 1990 and 1992 and yields a total of 10,213 observations.6 Respondents were asked to state the degree of their religiosity.7 Those indicating to be religious and follow the teachings of the church (1) or to be religious in their own way (2) were classified as believers.
Figure 1 plots the proportion of religious respondents for the two different generations – socialized before and during the Cold War. As can be clearly seen, the countries vary quite considerably in the extent the state doctrine of atheism affected their citizens. It appears that the generational difference is biggest in the Ukraine and the Czech Republic. The two generations hardly show any difference in religiosity in Slovenia and Poland, where overall levels of belief in God is also the highest.
The question now is whether the above-mentioned country differences in church repression can account for the generational differences in religious beliefs observed in Figure 1. For this, we use a classification of the state-church relation in 1980 developed by Madeley (2003:13) based on Barrett's (1982)Christian World Encylopaedia. Madeley distinguishes 10 different categories ranging from 1 “state exists solely to promote Christianity” (e.g., the Vatican) to 10 “state suppression of religion” (e.g., Albania). Western European countries range between 1 and 5, while the former socialist countries in Central and Eastern Europe are mostly classified as “state obstructs all churches” (8) or “active state hostility to religion” (9).
Figure 2 plots the score of each country on the Madeley classification against the generational differences in religiosity plotted in Figure 1. The 10 cases illustrated here clearly show a trend: the more repressive the state-church relation, the stronger the observed effects of religious beliefs, measured through generational differences. However, Figure 2 also reveals some interesting cases. First of all, Eastern Germany – one of the countries that included the Protestant church into the state, as a “church within socialism” – shows higher effects than anticipated. It is well above the expectation line. On the other hand, Polish citizens appear to have resisted the state influence of church suppression much better. On the contrary, Hungary, which was characterized by a comparable state-church relation than Poland, exhibits much larger differences between the generations socialized before and during the Cold War. State-forced secularization was strongest in the former Soviet Union, which appears to have been particularly successful. The Cold War generation in Russia and the Ukraine (as a former part of the Soviet Union) are significantly less religious than the older pre-Cold War cohort.
In sum, this section clearly confirmed that the state enforced secularization was successful beyond overt behavior such as church attendance or belonging to a specific denomination. Using generational differences in reported religiosity as a proxy, the empirical analysis reveals that those socialized under the socialist regime, are indeed less religious. Moreover, the extent of church suppression that was executed by the state can account for observed national differences.8 The next step is now to investigate the development of religious beliefs once the socialist regimes had collapsed and the role of religion and the churches often drastically changed. Data availabilities restrict, which countries can be analyzed over time. The remainder of the article therefore focuses only on Poland, East Germany, Hungary, Slovenia, and Russia. As Figures 1 and 2 show, these five cases are characterized by not only different levels of state suppression but also varying secularization effects. We therefore believe that a reduced set of cases is still useful in drawing conclusions about a general phenomenon, namely, the role of the state in the oppression and revival of individuals' religious beliefs.
Theories about the Development of Religiosity after 1990
There is a widespread agreement in the literature on religion in Eastern Europe that the process of forced secularization that took place during the Communist era differs widely from a “natural” process of secularization as it is mostly observed Western Europe (Tomka 1991; Need and Evans 2001; Froese 2004). The state's crucial role in the de-establishment of religion has been pointed out above. Despite this, we think that two particular components of the secularization paradigm still offer considerable explanatory power in explaining religious change after the end of state socialism in Central Eastern Europe. These are Berger's theory of plausibility structures (Berger 1969) and Norris and Inglehart's framework of “existential security” (Norris and Inglehart 2004). It seems that the alternative supply-side approach to religion (Stark and Iannaccone 1994; Finke 1997; Froese and Pfaff 2001, 2005; Froese 2004a, 2004b) has failed to explain the developments in Eastern Europe. In its current form the supply-side theory has lost its former appeal, because it has to explain too many anomalies in an ad-hoc fashion, which is discussed in more detail in the online supplementary material in Appendix I. We therefore concentrate on mechanisms that have been suggested in the field of secularization approaches.
The first component is Berger's theory of plausibility structures: societies create and maintain social conditions that are either benevolent or erosive to the maintenance of religious beliefs and their intergenerational transfer. Human actions and interactions are put in a religious context that creates a “sacred canopy” in which the most relevant social interactions gain plausibility and legitimization in a religious sense. Modernization undermines these plausibility structures through processes of structural and social differentiation (Berger and Luckmann 1966; Luckmann 1967; Berger 1969). The stronger religious beliefs are embedded in people's daily lives, the more plausibility they gain. The stronger religion is oppressed and persecuted, the more will religious plausibility be undermined. This reasoning is not deterministic. Antireligious propaganda and legislation were not able to eradicate religion in Eastern Europe because it provided “free spaces […] as refuges from surveillance and control” (Froese and Pfaff 2001:484).9
However, there are numerous examples in Eastern Europe, where the state re-established the standing of the church, or the churches themselves tried to impose a stronger influence on wider parts of the society (see Tomka 1995 on Hungary, Borowik 2002 on Poland, or Madeley 2003 on a more general account of institutional changes in Eastern Europe). As Titarenko points out in a commentary on Borowik (2002), “‘a turn to religion’ becomes a ‘return to the national roots’” (Titarenko 2008:251) and political leaders sometimes embrace the spread of conservative moral principles that support the idea of strong political power (Titarenko 2008:252). Thus, there is an intimate link between the (re-)construction of national identity and the support for established national churches.
The second strand of secularization reasoning that might explain religious change in Eastern Europe is Norris and Inglehart's “existential security” approach, which puts the socioeconomic development of societies at the centre of attention. The demand for religious value systems is determined by the levels of experienced “existential security” in the respondents' formative years or current life situation: “Individuals experiencing stress have the need for rigid, predictable rules” (Norris and Inglehart 2004:19). Individuals use religious values and belief systems as a mean to cope with fear and uncertainty.
Their chapter on Eastern Europe (Norris and Inglehart 2004:111-32) supports the idea that that secularization takes place when the general living conditions improve. Therefore, levels of religious beliefs in Eastern Europe can be explained by levels of socioeconomic development, whereby the state's antireligious policies might have exacerbated the process of secularization (Norris and Inglehart 2004:132). It also follows that younger birth cohorts should consistently be less religious than older birth cohorts, given they grow up under constantly improving living conditions. This relationship is confirmed in their analysis.
However, we are not entirely satisfied with their explanation, because it misses the role of the state that introduces new religious structures. Greeley (1994) does report an U-shaped pattern of religiosity over birth cohorts: those socialized before the Communist era as well as those after show the highest rates of religious beliefs and activity. The middle cohort suffered the strongest experience of antireligious policies and is therefore still the least religious. Although Norris and Inglehart attribute increases in religiosity in Eastern Europe to increased economic insecurities following from the massive restructuring of formerly planned economies, we argue that these increases might be attributable to proreligious state policies.
Based on the theories portrayed and the research findings outlined above, we formulate the following hypotheses. First, the fact that we can show the tendency that changes between pre-Cold War and Cold War cohorts have been strongest in countries where oppression policies have been stronger counts as a partial confirmation of Berger's theory of plausibility structures. However, we also maintained the idea that the increase of religious beliefs in Eastern Europe during the 1990s can be explained through a strengthening of religious institutions by political actors. Therefore, we formulate:
State-church hypothesis (H1.1): The stronger churches and the state are entangled, the higher should religious beliefs be.
However Norris and Inglehart's existential security hypothesis states that changes in economic development and experienced (in-)security are the main drive behind changes in religiosity. This should be true in Western Europe and Eastern Europe alike:
Existential security hypothesis (H2.1): The higher the economic development of a society, the lower the level of religious beliefs in Western and Eastern Europe.
But one could also argue that changes in religiosity are based on different mechanisms in Eastern Europe. While the process of secularization in Western Europe took place gradually, driven by changes in economic development, structural differentiation and a slow decrease of the role of religion, the process in Eastern Europe was politically forced and fairly abrupt. The institutional changes in state-church relations in Western Europe were fairly modest (Madeley 2003). Yet in the newly formed democracies of Eastern Europe, institutions only slowly consolidated, which gave political actors more degrees of freedom to establish religious structures advantageous to them or a newly formed “national ethos.” Thus, the changes after 1990 could be driven by political interventions rather than economic change. We hypothesize:
Differential change hypothesis West (H3.1): Changes in religiosity in Western Europe are driven by economic changes rather than institutional changes.
Differential change hypothesis East (H3.2): Changes in religiosity in Eastern Europe are driven by institutional changes rather than economic changes.
Testing Religiosity: Data and Method
Religious beliefs are difficult to measure (Kelley and De Graaf 1997; Hill and Pargament 2003). Research interested in religion and its effects often relies only on the question of church attendance, as done, for example, in Gautier (1997) or Need and Evans (2001). But if we want to get at the heart of religion – believing – it is difficult to achieve this by simple observed survey responses. This is particularly difficult if we ask citizens of former authoritarian regimes about their religious beliefs, as they may not be willing to reveal their religiosity. Two problems arise from this: first, measurement error, and, second, data measuring religious beliefs are often only categorical or, at most, ordered. To address these issues we utilize latent class analysis to measure “true” religiosity.
Latent class analysis (LCA) is used as a categorical data reduction method analogous to factor analysis. Its original approach was introduced by Lazarsfeld (1959).10 A main feature of LCA is the possibility to investigate the relationship among several categorical variables assuming local independence between these indicators. Imagine we have two items, an “individual's vision of God” (A) and “how close a respondent feels to God” (B), both of which are only proxies of what we are really interested in – true religiosity. The fact that a clear image of God and feeling close to God frequently occur together is assumed to be caused by the same type of belief. Hence, in a latent class model if we know whether somebody has an image of God, we have no additional predictive power to explain closeness to God. There is consequently no covariation between these two anymore. Any observed relationship between (A) and (B) is simply because of a common “underlying” unobserved trait – religiosity. The aim of LCA in this context is to classify respondents into “true” believers or “true” atheists. See Appendix II in the online supplementary material for a detailed account of this method and the models estimated.
Data are taken from the East and West European samples of the “Religion” module of the International Social Survey Program (ISSP), which was repeated in 1991, 1998 and 2008.11 The first time point of the study in 1991 represents the baseline of regime suppression, as it was surveyed right after the collapse of the Soviet Union. Unfortunately, only five former socialist countries were included in all three waves of the study, and, hence, this allows us to investigate the dynamics of individuals' religiosity over time after state socialism. The Western European countries thereby function as a reference point to which the former socialist countries are compared to. “The fact that Western European democracies have a long history of religious tolerance, human rights, and civic liberties during the twentieth century means that the comparison of these regions provides an exceptionally well-designed ‘natural experiment’” (Norris and Inglehart 2004:131) to test the influence of the state on individuals' religious beliefs.
The ISSP data are suitable for exploring religiosity because of its rich pool of questions on religious beliefs. Three different indicator variables are used to measure the latent construct of religiosity.12 These items tap the question of belief in God, which is assumed to be a personal and moreover private trait. The socialist systems might have affected overt religious behavior, such as churchgoing. But the interesting question is whether the authoritarian regimes altered the personal belief in God and whether this could be retrieved thereafter.
The three indicators were recoded into variables distinguishing between three different types of religiosity: (1. those absolutely not believing in God – atheists, (2. those being agnostic and (3. those that at least to some extent believe in God. The idea behind this approach is to quantify the true atheists and believers and compare these numbers across several East and West European countries across time. However, as some citizens are agnostic about their belief in God, it is important to allow for this by including an additional latent class. In sum, the model presented below is estimated using three latent classes.13
Results: Explaining Trends in Religiosity in Postsocialist Societies
This section is divided into two parts. First, we compare the proportion of religious citizens in Eastern and Western Europe across three generations from the downfall of socialism in 1991 to 2008 descriptively. Second, we are trying to explain this variation in religiosity across time and cohorts by taking into account modernization and the state-church relationship.
Trends of Religiosity in Postsocialist Europe
The first step is to investigate the overall trends in Eastern Europe and compare these to Western Europe. Figure 3 plots the proportion of believers across all cohorts over time, which is based on three items measuring belief in God. Most important for this research is the proportion or size of each latent class. To save space, only the aggregated “mixing” probabilities of respondents that are “truly” believing in the existence of God are presented. The LCA confirms the well-documented secularization trend in Western Europe. However, religiosity seems to slowly increase in the former socialist countries. Nevertheless, the included Eastern European countries vary quite considerably in their level and direction of trends in religiosity. The fraction of believers in some – mostly Catholic – countries in Eastern Europe remains at a high level. In Poland constantly around 95% were classified as true believers, despite their socialization experience. In comparative terms, East Germans are absolutely least religious.
Overall, Figure 3 is not entirely conclusive regarding the time trend of religiosity in former socialist countries. Eastern Catholic countries do not seem to become more or less religious, while believing in God is declining in East Germany – the only predominately Eastern Protestant country. Interesting is the development in Orthodox Russia, which observes a clear revival of religiosity. The mechanisms leading to these time trends are explored in the next section.
Turning to the generational comparison, Figure 4 plots the proportions of those believing in God for each cohort over three time points (represented by three bars, respectively, two for the post-Cold War cohort). First, it becomes apparent that the level of religiosity differs for the three cohorts in East and West Europe. The pre-Cold War cohort is the most religious. Over time, Western European countries exhibit a clear intra-generational and inter-generational decrease in religiosity. This supports the clear overall downward trend observed in Figure 3.
Even if the cross-country trend in religiosity over time in Eastern Europe was not very conclusive, the distinction of different generations yields a clearer picture. Among the pre-Cold War generation no trend is observable. They remain on a rather high level over the last 20 years. Interestingly, among the generation that grew up during socialism and was hence socialized in a church-hostile environment, a clear religious revival is visible. One could of course argue that this is a simple aging effect, as this generation also grew older. However, the same aging effect should hold for the Cold War generation in Western Europe, which shows decreasing religiosity.
Turning to the Eastern post-Cold War generation, we expected them to show the highest revival of religiosity over time, as they were not socialized into the secular socialist societies such as their parents and grandparents. The first impression when looking at Figure 3 might suggest that this proposition has to be rejected. However, if we compare this cohort with the Cold War generation in the West and East, the picture looks very different. In Western Europe the level of religiosity among the post-Cold War generation is, first of all, declining and, second, constantly about 15% lower than the previous cohort. In Eastern Europe, we observe a slight increase of belief in God between 1998 and 2008. Further, the difference between the two cohorts is very similar in both years despite very different ages: members of the Cold War generation being up to 68 years old (as they are born between 1932 and 1976) and the post-Cold War generation being maximum 21 years old in 1998. Considering that life-cycle effects play a role in religious beliefs, the post-Cold War cohort appears rather religious. It is however still striking that religiosity among the Cold War generation saw such a clear increase. The next section is presenting an approach to explain these trends in religion in postsocialist societies and compares these to Western Europe.
Explaining Trends in Religiosity
The aim of this section is to explain the trends in religiosity by taking into account two different aspects of modernization theory – the economic development as well as the change of the state-church relationship. First, it is expected that the levels of religious beliefs will decrease over time, as countries develop economically. Second, the modernization of a country is thought to lead to a decline in the plausibility structure of religion, which leads to a decline in the role of the church in public life.
We first explore the relationship between aggregated beliefs and economic development in more detail. Figure 3 plots the effect of economic development – measured by the gross-domestic product (GDP, in 1000s) – for each country and the proportion of those believing in God for three cohorts in East and West Europe. To account for the changes (rather than levels) it is necessary to center the variables by country. As the lower panel of the figure illustrates, positive economic development leads to decrease in religiosity in the consolidated societies of Western Europe. However, this typical modernization expectation is less clear for the newly transformed societies in Eastern Europe. Changing levels of GDP seem not to affect the religiosity of the pre-Cold War and post-Cold War cohort. But those socialized during socialism appear to become more religious once their countries economically develop.
To explore the trend in religiosity further, the next step is to look at the state-church relationship. Based on modernization theory and the role of functional differentiation, we formulated the expectation that the stronger churches and the state are entangled, the higher should religious beliefs be. For the measurement of Berger's concept of plausibility structures, we are using data from the Religion and the State Project (Fox 2004, 2006, 2008). The data were gathered in 2004 and quantifies the most important aspects regarding the separation of church and state in 175 countries during the period from 1990 to 2002. The codings are based on human rights reports, academic resources, as well as news media sources. We decided to use the variable “extent of religious legislation.” For various law domains the coders recorded whether governments legislate religion into law. The range of law domains is broad; examples of domains that have been coded are as follows: whether religious education in state schools exists and whether this is funded by the state; whether dietary or dress restrictions exist; whether the clergy is funded from public sources; whether minority religions are monitored by the state. The measurement used in this article is an added index of the number of different laws that exist in a country. A high number of religious laws indicate strong state-church relations and hence a good condition to create and uphold religious plausibility structures in a society. An overview of the law domains that are empirically found in the countries of our sample can be found in the online supplementary material in Appendix VI, which shows the development of religious legislation in each country over time. It further reports religious status scores that are based on the evaluation of the coders as of how the state-church relations in a country can be described overall.
First of all, it is interesting to note that the Eastern European countries are more active in introducing new legislations than Western Europe, which is of course a counter-reaction to state-forced secularization during state socialism. Figure 6 plots the changes of legislatives over time. The figure further illustrates again the changes in religiosity over time, as these two are assumed to correlate (both variables are centered by year). As expected, the state-church relationship, which is measured here using state legislations, is more or less constant in the established democracies in the West. These countries have an established relationship between church and state, which had been settled in the decades of democratic consolidation in the past. Only Norway, which was and still is characterized by a strong state-church link, reduced this close relation by abolishing religious education in school. However, Figure 6 suggests that the former socialist countries clearly tighten the state-church relationship, by introducing new religious legislations. Further detailed accounts of this, however, also reveal that considerable national differences exist. Especially, Russia, but also Slovenia, increased the role of the churches in the state. According to Madeley's (2003) classification (which was already used for Figure 2), the change in state-church relation was highest in Russia. It was still classified as 9 (“state hostility or prohibition”) in 1980, but was one of the most church integrating countries in 2000 reaching a score of 3, as the state is even subsidizing churches today. For comparison, the unified German state is following a policy of complete noninterference, which does not mean a huge increase in tighter state-church relations for citizens of former East Germany.5
If we disaggregate these distributions again for each country and plot religiosity and religious legislations for each cohort in each country over time, we get Figure 7. Obviously the variation in legislation is much bigger for the Eastern European countries. As modernization theory predicts, countries that introduce more religious legislations and hence tighten the state-church link obtain higher rates of believers. This positive relationship is most visible in Eastern Europe between the Cold War cohort and the youngest generation. The pre-Cold War cohort in Western Europe also exhibits a strong positive link. However, this relationship is driven by Norway, which reduced the state-church link between 1991 and 1998.
As a final test, we estimated a fixed-effects regression model using cohorts as quasi-panel units. Each cohort in every country represents a case, which we observe at three (or two) different points in time. We use this modeling strategy controlling for cohort specific effects. In contrast to standard pooled ordinary least square regression, this method allows us to account for the time ordering. Further, it is possible to control for the individual inter-generational unobserved heterogeneity, as the aggregated religiosity might be affected by time-invariant factors, which we cannot account for in our model. Such a fixed-effects estimation hence dissects the error term of the regression prediction into a time-constant within cohort specific component (ui) and a stochastic error across all cohorts and times (εi,t). This model can be expressed as:
where yi,t is the level of belief in God for each cohort at each specific time point, β1 and β2 measure the effects that the changing economic development and religious legislations have on religiosity. Table 1 reports the results of this model separate for Eastern and Western Europe. Because we used centered variables, the fitted lines in figures 5 and 7 represent the same effect, now measured statistically.
|Western Europe||Eastern Europe|
|GDP (in 1000s)||−.21*||.08|
|[−.41; −.02]||[−.39; .54]|
|[−7.07; 19.71]||[1.47; 12.06]|
|Western Europe||Eastern Europe|
|GDP (in 1000s)||−.21*||.08|
|[−.41; −.02]||[−.39; .54]|
|[−7.07; 19.71]||[1.47; 12.06]|
Note: Significance levels: * .05, **.01. Data: ISSP (1991, 1998 and 2008)
The results of the fixed effects model presented in Table 1 appear to be quite different for Eastern and Western European countries. Note that in the established democracies in the West, GDP negatively affects religiosity. The higher the economic development in these countries, the lower the level of religious belief. This finding is in line with classical modernization theory and Norris and Inglehart's existential security hypothesis (H2.1). It also supports our differential change hypothesis (H3.1). In former socialist countries, increasing levels of living standards, however, do not seem to matter for explaining variations in religious beliefs. On the other hand, religious legislations, which measure the strength between the state and church, appear to positively affect the level of belief in God in the East. This result is also in line with the theory of plausibility structures, as a tighter state-church relation, increases levels of religiosity (compare H1.1). Moreover hypothesis H3.2 is therefore confirmed.
Overall the model seems to do a pretty good job in explaining the variation in support among each cohort, especially in Eastern Europe. Thirty-six percent of the within cohort variance can be accounted for by the model. However, only 20 percent of the variation within cohorts over time can be explained by GDP and religious legislation in the West. The model explaining religiosity in the former socialist countries is also better in predicting the differences between cohorts (R2 = 0.186 compared to R2 = 0.015 in the West).
Summing up these results, the model presented in Table 1, but also illustrated in figures 5 and 7, confirms modernization theory, which predicts lower level of religiosity based on economic development in the West and predicts higher proportions of believers in the East, because of a tightening relationship between the state and the church.
Conclusion and Discussion
The aim of this article was to explore religious belief in former socialist countries, where religious freedom was heavily suppressed during the 40 years authoritarian regime. To put religiosity into context, we compared Eastern and Western European countries. A latent class analysis of belief in God confirmed the expected strong effect of religious oppression in authoritarian regimes on individuals' religious beliefs. The proportion of believers is much lower in the former socialist countries compared with established democracies, which guaranteed freedom of religion for a long time, which was further measured by comparing the level of religiosity among those socialized before and during the Cold War period.14
As time progressed after the collapse of state socialism and introduction of democracy and liberal rights in these countries, religion took different paths in different countries. Catholic countries remained on a relatively high level, but did not observe any increase in religiosity. Eastern Germany – the only mainly Protestant country in the former Soviet influence sphere – even experienced a slight decrease in belief in God, with more and more people being classified as atheists. On the other hand, Orthodox Russia truly observed a revival of religiosity in the aftermath of socialism. Among all generations, religion seems to become more important again. Overall, differences between generations socialized before, during or after the religious suppression is rather small. The Cold War and post-Cold War cohort show somewhat more movement than the older generation, which was still socialized before socialism.
In a next step, we used modernization theory to explain trends in religiosity over time. The results presented here are supportive of two different facets of the secularization paradigm, Berger's theory of plausibility structures and Norris and Inglehart's existential security hypothesis. Not only did we find that the state played a crucial role in the disestablishment of religion in Eastern Europe. It also is one of the driving forces of its re-establishment after 1990.
A first look at the effect of antireligious policies on differences in religiosity between pre-Cold War and Cold War cohorts in 10 Eastern European countries revealed that these tend to be stronger in countries with harsher regimes of religious oppression. Yet the state seems to have been fairly effective in the re-establishment of a religious plausibility structure as well. We find clear evidence that the religious revival in Eastern Europe is strongly linked to the implementation of policies that favored the established churches.
However, we also found evidence for the secularizing role of economic development and structural differentiation. The smooth over time decrease in the Western European countries in our sample is explained with increases in GDP rather than institutional changes. We attribute the differences between East and West to the fact that Western European democracies are already largely consolidated, whereas the transitioning East European democracies still had to struggle with a reinvention of their national identities, whereby religion played a crucial role (Titarenko 2008).
In the exemplary case of Russia, the political motivation of using the Russian Orthodox identity to legitimize political power is a viable explanation for the positive association between legislation and increase in religious beliefs. Not only was Boris Yeltsin himself blessed by the Patriarch of the Russian Orthodox Church when he was sworn in as president in 1991 (Bruce 2003), but also the 1997 Freedom of Conscience and Association Law granted orthodoxy a “special place in Russian history and in the development of the spirituality of the Russian nation” (Bruce 2003:53), thus establishing new plausibility structures for an ethnoreligious identity.
Regarding the future development of religiosity in Eastern Europe, we expect that the role of economic development and increasing existential security (Norris and Inglehart 2004) will become more important as an explanatory factor, thereby mirroring the process of gradual secularization in Western Europe. As democracies consolidate, the questions of national identity and the state-church relations will gradually be settled and the demand for religion gradually decline with further increases in living conditions.