The Religiously Unaffiliated in Norway
- Side: 61-81
- DOI: https://doi.org/10.18261/issn.1890-7008-2017-01-04
- Publisert på Idunn: 2017-05-15
- Publisert: 2017-05-15
During the last 20 years, the number of people who are religiously unaffiliated in Norway has increased considerably, from four percent in 1991 to 14 percent in 2012. Using Norwegian survey data collected in 2012, I examine two aspects of the religiously unaffiliated: i) their social characteristics compared to the religiously affiliated and members of the Humanist Association, and ii) the predictors of unaffiliation using logistic regression. The results show that the unaffiliated are younger and include a larger proportion of individuals with a higher education than the religiously affiliated. Their social characteristics are often placed in between the affiliated and the humanists, suggesting a multifaceted social category. Certain factors, such as having had an unreligious upbringing and having friends with different religions, are highly predictive of unaffiliation. The most influential factor is rejection of a God or a higher power. Surprisingly, gender is insignificant in predicting religious unaffiliation in Norway. Possible interpretations are discussed.Keywords: religious unaffiliation, “the nones, ” Norway, secularization
In many Western countries, there has been a significant increase in the proportion of the population not affiliated with any religious organization (Baker and Smith 2009; Cheyne 2010; Shermer 2013; Singleton 2015; Winston 2012). This growing element in the religious landscape has popularly been referred to as “the nones”1In this paper, the term “religiously unaffiliated” will be utilized instead of “the nones” as it has less normative connotations. (Lugo 2012). The growth of the religiously unaffiliated has often been connected to a decline of religious beliefs and practices, and is consequently attributed to a wider, ongoing process of secularization (Lugo 2012; Singleton 2015). However, personal belief and membership in religious organizations do not necessarily coincide. The unaffiliated may be believers without belonging (Davie 1990; Lee 2015); thus, the assumption that the religiously unaffiliated are secular can be questioned.
In this paper, I study the unaffiliated in Norway. The interest in studying this phenomenon in Norway is that, alongside the other Scandinavian countries, Norway is sometimes addressed as the world’s most secular (see: Davie 2002; Norris and Inglehart 2011; Bruce 2011). Still, by January 1st 2012, about 75 percent of the Norwegian citizens were members of the Church of Norway (The Evangelical Lutheran Church of Norway – the state church up until 2012), but less than ten percent attended church services on a monthly basis (Botvar and Urstad 2011). To sum up, there is a high degree of membership, but a relatively low degree of religious activity and personal faith. What adds to the complexity in the Norwegian case is an existing, well-established secular alternative to religious organizations – the Humanist Association. Regardless of an institutional secular alternative, we are seeing a marked increase in people who are not members of any religious organization. The question of interest is: who are the religiously unaffiliated in Norway?
This paper investigates the religiously unaffiliated in Norway by answering two research questions: i) what are the social characteristics of the religiously unaffiliated in Norway, and how do they differ from the religiously affiliated and the secular humanists? ii) What social factors are central in understanding unaffiliation?
The Norwegian Case
One of the factors making the Norwegian case interesting is the government funding of religious organizations. In Norway, the state provides financial support for all registered religious and faith-based organizations based on the number of members – a financial arrangement that is quite unique to Norway. Since 1969 all registered religious and faith-based organizations have been funded by the same principle (Hylland 2006). Interestingly, this includes the Norwegian Humanist Association and other registered worldview organizations that are not religious in themselves, but have a philosophy of life or set of values not rooted in (ir)religious convictions – as does secular humanism. The reason for a membership-based funding system is not only to ensure freedom of religion, but to provide the means for doing so (Schmidt 2015). To qualify for funding, the organization must have a minimum of 500 members and register with the Ministry of Church Affairs.2https://lovdata.no/dokument/SF/forskrift/1988-12-01-996 – KAPITTEL_1. In 2013, the annual amount received per member was equivalent to NOK 422 (which is approximately €50). A restriction is that the same individual cannot be a member of more than one religious organization. If a person is a member of two (or more) religious organizations, the organization(s) will not receive funding for that individual. With the financial agreement in mind, the religiously unaffiliated become interesting to study since there are established secular alternatives.
The Church of Norway still holds a dominant position in Norway, but is losing members every year. In 1991, over 90 percent of the population were members of the Church of Norway, but there has been a marked decline in membership since then.
Figure 1 clearly illustrates how the Church of Norway’s dominating role is weakening quite rapidly. Indeed, the negative trend is strengthening over time. From 1991 to 2015, the proportion of citizens that were members of the Church of Norway declined from approximately 90 to 73 percent (see Botvar and Schmidt 2010; Botvar and Urstad 2011, 2012; Furseth 2015; Taule 2014, 2016; Urstad 2010). Some of the changes in the religious landscape may be attributed to the fact that Norway is becoming more pluralized or religiously complex (Furseth 2015). There is a small but steady increase in members of other religious organizations outside the Church of Norway, which can be attributed mainly to immigration. In 2015, about nine percent of the population (473,277 individuals) were members of these organizations. Of these nine percent, more than half (57 percent; 296,521 individuals) were members of Christian organizations. Islamic organizations are the second largest category of religious organizations, with 25 percent of the total membership (141,027 individuals) (Taule, 2014, 15). The third largest category of organization is the Norwegian Secular Humanist Association. As already noted, the humanists do not define themselves as a religious organization, but as a life stance organization (Livssynsorganisasjon). From figure 1, we see a rather stable number of members from 1991 to 2015. In 2015, there were 85,700 registered members of the Norwegian Secular Humanist Association, equivalent to slightly less than two percent of the population (Taule 2016).
The decline in membership of the Church of Norway is not proportional to the increase in membership of religious organizations outside the majority church. Even though there has been a growth of religious organizations outside the Church of Norway, this does not indicate that it is at the expense of the Church of Norway. This implies that the Church of Norway is losing ground to the religiously unaffiliated: the unaffiliated more than tripled in size from 1991 to 2012, encompassing 14 percent of the total population in 2012. The growth of the religiously unaffiliated is an ongoing development in the Norwegian religious landscape that is important to address. Despite financial incentives from the government, based on the principle of equal treatment – that includes secular alternatives – religious unaffiliation is increasing.
Previous Research on Religious Unaffiliation
The phenomenon of “religious unaffiliation” is a fairly new area of research as observations of this category grow in different countries around the world – although it is much more frequent in the West (Zuckerman, Galen and Pasquale 2016). Regardless of being a “new” phenomenon, it has received attention from researchers in the last few years. Within the Norwegian context, there is little research on unaffiliation. Most research has focused on explaining the downturn in the membership in the Church of Norway or why the level of membership is still relatively high (Botvar and Schmidt 2010; Breistein and Høeg 2012). The research in this section is therefore informed mostly by studies in the USA or Western Europe. Within the studies on religious unaffiliation, two main directions of theoretical explanations can be identified.
The first direction places religious unaffiliation within a “traditional” understanding of secularization, by arguing that the phenomenon is a consequence of individuals turning away from religion. Being mainly empirical studies, the theories look at social factors to explain this sudden (and to some, unexpected) increase in unaffiliation. However, the assumption that unaffiliation is a consequence of an increase of secular individuals is contested. Other contributions argue that this phenomenon should not be understood within the paradigm of secularization, but instead how individuals are leaving religious institutions, not religion all together.
Unaffiliation as consequence of secularization
In the United States alone, the proportion of the religious unaffiliated has increased considerably since 1990s. In 1990 the estimate of religiously unaffiliated was below 10 percent, while by the beginning of 2016, this had almost tripled in size to between 25 to 30 percent (Zuckerman, Galen and Pasquale 2016, 4). This growth makes the phenomenon one of the most evident tendencies in American religion.
Within studies on American unaffiliation, certain traits have been identified regarding this increase in unaffiliation. The “typical” American unaffiliated is portrayed as a young adult, white male, with a higher education, who is not married and has no children (Merino 2012; Lugo 2012; Baker and Smith 2009). Moreover, the individual is likely to be unreligious, with unreligious parents; and in the cases where the respondents are married, their spouses are usually unreligious as well (Baker and Smith 2009, 1257). These key findings establish that primary socialization and the respondents’ immediate social environment explain more than socio-demographic factors (such as age, residence and education) (Baker and Smith 2009; Lugo 2012). The gender aspect is also important to highlight as men are twice as likely to be religiously unaffiliated as women (Baker and Smith 2009). In other words: Gender and the individual’s immediate social environment are highly predictive for an individual’s chance of being unaffiliated.
In an article, Stephen Merino argues that a generational shift is important in understanding the increase in unaffiliation. Socialization and social experiences in an increasingly secular environment increase the probability of not having a religion (Merino 2012, 1).
This result has resonance with European studies as well. In an article by David Voas and Ingrid Storm (2012), the authors show that socialization and the social environment are important factors when studying religion. Another article based in the British context examines religious transmissions in depth (Voas and McAndrew 2012). Although they do not address unaffiliation directly, the authors conclude that church attendance and the respondents’ different religions correlate with parents’ beliefs and practices. Indeed, if both parents are religious, their child has an approximately 90 percent higher chance of adopting a religious belief than if none of the parents are believers (Voas and McAndrew 2012, 36). However, the picture is complex. The authors also conclude that generation, religious socialization and cultural environment are all important factors for explaining religious reproduction and change.
These results confirm the assumptions that religious unaffiliation is a consequence of secularization, and this increase indeed culminates the development. In the report from Pew Research Centre, “Nones”on the rise: One in five have no religious affiliation (Lugo 2012), the author presents four theories to explain the increase of the nones: i) “Political backlash”: the young turn away from organized religion due to its entanglement with (conservative) politics; ii) “Delays in marriage”: due to new marriage and cohabitation patterns among young people, membership is declining as other alternatives to marriage partnership are becoming more common; iii) “A broad social disengagement”: young Americans are less likely to be involved in social or communal activities in general, and iv); secularization.
Declining institutional religion, not secularization
The assumption that religious unaffiliation is a consequence of secularization has been challenged. In the book Recognizing the Non-religious the sociologist Lois Lee (2015) argues that non-religion is an abandonment of institutional religion, not necessarily religion altogether (Lee 2015). Lee creates a typology of different types of the non-religious, specifying that there is a need for a differentiation to understand this phenomenon. Lee objects to the notion that religious self-identification as non-religious is sufficient for categorizing an individual as secular. For instance, identifying as non-religious may be an effect of not finding the alternative he or she prefers in the questionnaire. In an article by Abby Day and Lois Lee (2014), the authors also emphasise caution in using self-identification for categorizing individual religiosity, especially “non-religion.” Leaving traditional religion for personal spiritualties might get sidelined when looking solely at self-identification. Indeed “[…] those who are not traditionally religious can in fact reassert the value of ʻreligion’ questions in social research, so long as those questions are constructed and analyzed with care” (Day and Lee 2014, 353).
The literature reviewed above places itself in two directions. One of them may address unaffiliation as a consequence of secularization, identifying several factors that try to explain religious unaffiliation. These include demographic factors and factors related to the individual’s social environment and gender. The other direction focuses on the limitations of categorizing the unaffiliated as secular, claiming that persons within this category may be religious, just not in a traditional way.
As most of the literature is drawn from outside the Norwegian context, it is worthwhile to replicate these findings using Norwegian data. A detailed discussion of the variables will be given in the section covering variable definitions. At this point it is important to note a particular difference from the American studies to this study on Norwegian unaffiliation. The difference is related to how one defines or specifies unaffiliation. In several of the international studies the focus is on religious self-identification, not membership per se. My study targets individuals who stand outside any religious organization, no matter how religious they feel or which religion they identify themselves with. The dimension of belief will therefore be added in the analysis to accommodate this aspect of unaffiliation.
Data and Methods
The empirical data are drawn from a national internet-administered Norwegian survey on beliefs and faiths (hereafter referred to as NBF)3A part of the data used in this article is acquired from Tros- og livssynsundersøkelsen 2012 (Norwegian Survey on Beliefs and Faiths 2012), conducted and financed by KIFO, Institute for Church, Religion, and Worldview Research, represented by Ida Marie Høeg, Ulla Schmidt, and Sunniva Holberg. KIFO is not responsible for the analysis or the interpretation of data presented in this article. – Tros- og Livssynsundersøkelsen – from 2012. The survey was developed and funded by KIFO, the Norwegian Institute for Church, Religion, and Worldview Research (KIFO – Institutt for Kirke-, religions- og livssynsforskning). The survey was administered and carried out by Norstat, a provider of data collection services. Norstat’s Norwegian panel consists of 83,500 individuals, encompassing a representative selection of the Norwegian population. (For a description of the panels, representation and recruitment policy, see Norstat 2015.) During the sampling of the NBF survey, a total of 16,745 invitations were distributed, resulting in answers from 4,005 respondents (Høeg and Gresaker 2015, 15–6). The main challenge when using the NBF data is the response rate, which in turn is linked to the question of validity. In this case, the response rate is less than 30 percent (for overview and details of NBF response rates, see Appendix Table 1). This may affect the external validity of the analysis, as the low response rate may affect the representativeness of the data.
When checking the representativeness of the data, there are areas where the data differ from the Norwegian population, and this needs to be addressed. The proportion of unaffiliated in the NBF data is 12.5 percent, which is a bit lower than the proportion of the population, which according to register data from Statistics Norway (SSB) is about 14 percent (Taule 2014) (For an overview of distribution of belief/faith communities in the NBF data, see Appendix Table 2). Moreover, the proportion of persons in the age categories “below 25” and “above 76” is less than what we find in the population as a whole. In addition, there is a higher proportion of persons in the 66–76 range (see Appendix Table 3). For the regional variables, the survey represents the population quite well; the relative differences between the population and the survey are within a margin of 1.2 percent. The central area where the NBF data differ from the population is related to the level of education: the proportion of persons with higher education is significantly higher in the NBF data. This provides some uncertainties regarding the education variable within the analysis. In other areas, however, the data seem to give a good representation of the population as a whole, thus making the data relevant for this purpose (See Appendix Tables 2, 3 and 4).
The number of respondents (n) defining themselves as religiously unaffiliated is 500. One important issue to address is how to treat members of the Norwegian Secular Humanists Association. Although they are not members of a religious organization itself, they are indeed members of a worldview organization. Thus, it is possible to include their membership among the affiliated, but also among the religiously unaffiliated. In the Norwegian context, they represent a small proportion of the population, just below two percent according to SSB. On January 1st 2012, this is equivalent to 83,100 registered members (Taule 2012, 46). In the selection represented by the NBF data, they represent about six percent, corresponding to 221 respondents. The manner in which they are categorized affects the results, especially if they are combined with the unaffiliated, as the number of unaffiliated persons would then be much larger (N=771) (See Appendix Table 5). The focus of this paper is the unaffiliated: persons who are not members of any organization, whether it is a religious organization or a worldview organization. Furthermore, to assume that the unaffiliated are secular may be a fallacy. Persons who are members of the Norwegian Secular Humanists have actively joined a secular life stance organization. This challenge is accommodated in two ways: i) For the descriptive statistics, I have isolated the secular humanists in a separate column. This way, the results show how the unaffiliated differ from both the religiously affiliated and the humanists. By addressing them separately, we also gain knowledge into the social differences of the three groups – the religiously affiliated, the unaffiliated and the Humanist Association; ii) In the logistic regression analyses, I have excluded members of the Humanist Association, to avoid inference. The unaffiliated will then be analyzed in relation to the religiously affiliated.
Choice and definitions of variables
The analyses are carried out in two steps. First, I describe how the unaffiliated and the affiliated (as well as the secular humanists and the full sample) are distributed across all analysis variables. The social characteristics provide the basis for understanding the unaffiliated and how the religiously unaffiliated differ from the religiously affiliated and the secular humanists. The unaffiliated may be overrepresented in certain categories, but there is likely a high overlap between some of these categories. In order to assess the controlled, partial correlations of religious unaffiliation and the independent variables, I estimate a series of logistic regression models in a second step. In short, step two of the analysis identifies both the significant influence of the independent variables to predict unaffiliation, and the strength and direction of the predictors. This way I can identify which of the variables have the least and strongest influence on unaffiliation, and which variables predict unaffiliation the most.
The variables included in the analyses below are motivated by previous research on the unaffiliated. In addition, other variables are included that have proven to be relevant in research on religion in Norway.
Gender is a relevant factor to consider. There is a strong gender gap in religiosity, and thus likely also in religious affiliation (see Furseth 2004, 2006; Furseth and Repstad 2006; Lövheim 2013; Voas, McAndrew, and Storm 2013; Voicu 2009).
Age is also an interesting factor for further investigation in relation to unaffiliation in Norway. Although the age composition in Norway has not changed dramatically over the course of the past few years, age might still be a contributing factor when it comes to religious affiliation. When studying religion and religious beliefs, age and generations must be taken into account. Previous research has shown that age is relevant in regard to religion and should be included to see whether it explains variances within categories (Botvar and Schmidt 2010; Furseth 2006; Voas and Storm 2012). In the bivariate analysis, age is coded into seven groups, whilst, in the logistic regression, the variable is continuous (18–79 years).
Education is an important factor to include. In 2008, less than 30 percent of the population had lower secondary school as their highest completed level of education. By contrast, more than one third of the population had completed some form of higher education, at a university or university college (SSB 2016b). Recent research has shown that persons with higher education are less likely to be religious in Norway (Botvar and Schmidt 2010; Urstad 2010). This does not mean that it automatically applies to the religiously unaffiliated, but will be examined in the subsequent analysis. The education variable contains the respondents’ highest completed level of education, and consists of four values: Lower secondary school (grunnskolen), upper secondary school (videregående), university short (1–3 years) and university long (four years or more).
Although income alone has not proven to be an important factor for studying religion in Norway, it has been shown to have an impact in other countries. Individuals with low personal income are in some cases overrepresented in regard to personal faith, but the differences are not always substantial (Urstad 2010). In North America, however, people with high income are somewhat underrepresented in regard to religiosity and religious affiliation (Lugo 2012). In the bivariate analysis, income is coded into eight values. In the logistic regression it is continuous.
City size is added. Although correlation between the degree of urbanization and religiosity in Norway is not always clear, many cases have shown correlations between religious traditionalism and degree of urbanization, especially related to the capital city (Botvar 2009). Individuals living in rural areas are more prone to participate in religious rites of passage than persons living in urban parts of Norway (Høeg and Gresaker 2015, 30–3). Pål Ketil Botvar (2014) finds that young people in larger cities are skeptical to religion in the public sphere. Thus, it is possible that city size may be influential when addressing unaffiliation. The variable has four values: Oslo (the capital), cities with more than 50,000 inhabitants, towns with between 5,000 and 50,000 inhabitants, and towns/places with 5,000 inhabitants or less. In the logistic regression, Oslo is the reference category.
I’ve also added voting behavior to the analysis. Lugo (2009) finds that the unaffiliated in the USA are more prone to be left-of-center on the political spectrum, and this may also be the case in Norway. In the logistic regression, the parties have been divided into three: politically left (the Red Party (communist party), the Socialist Left Party and the Labor Party), politically center (the Center Party, the Liberal Party and the Christian Democrats) and politically right (the Conservative Party and the Progress Party).
Marital status is significant in recent studies of unaffiliation, which makes it likely that this is the case in this study as well. The variable has five values. In the logistic regression, the reference group is married/registered partner. Having one or more children is also added to the analysis.
Membership of a Non-Governmental Organization (NGO) is added to the analysis, on the basis that there might be a connection between religious affiliation and civil engagement (Putnam 2001). For instance, correlations have been proven between participation in voluntary work and church-going (Herbert 2003, 91). Indeed, religious involvement has sometimes been linked as a predictor of voluntary work, as faith-based organizations can be seen in relation to social capital (Putnam 2001). In this respect, it is of interest to see whether there are correlations between membership of NGOs and affiliation with a religious or faith-based organization. The variable is dichotomous: persons who are members of one or more NGO’s and those who are not.
Religious socialization is an important aspect to include. In the analysis, this involves two variables. The first one is whether the respondents report that religion was important in their upbringing. The second one is having friends that have a different religious life stance than themselves. Both variables are dichotomous.
Finally, the variable capturing the respondent’s personal belief is interesting to examine. This will also accommodate Grace Davie’s notion of believing without belonging (Davie 1990); it is not necessary to be a member of an organization in order to have a religious faith. In addition, Singleton (2015) finds that the unaffiliated in Australia are more prone to be secular. In the analysis, belief in God and belief in a higher power of some sort is included. The reason for this is to compare the unaffiliated to the affiliated with respect to personal faith.
|Lower secondary school||6.5||3.6||4.3||6.0|
|Upper Secondary School||34.4||29.3||20.5||33.0||***|
|Over 800,000 NOK||2.8||2.9||5.2||3.0|
|More than 50,000||25.3||28.4||27.9||25.8|
|Less than 5,000||17.5||16.6||14.9||31.3||*|
*** = p > 0.001 – ** = p, 0.010 – * = 0.050 Asterisks mark that the unaffiliated differs significantly from the affiliated or the humanists.
|One or more children|
|Religious important in upbringing||0.0||0.0||0.0|
|Friends other religions/life view|
|Belief in God or higher power of some kind|
|Former members of The Church of Norway|
*** = p > 0.001 – ** = p, 0.010 – * = 0.050 Asterisks mark that the unaffiliated differs significantly from the affiliated or the humanists.
Results step 1, bivariate analysis
The results from Table 1 show no statistically significant gender differences between the affiliation categories; the distribution is approximately fifty/fifty. Looking at the age distribution, there are some clear distinctions. The unaffiliated are on average younger than both the affiliated and the humanists. The mode for the unaffiliated is the age category 25–35, with close to 40 percent being 35 years and younger. For the humanists, this is under 25 percent and for the affiliated, it is 30 percent. The age difference is also reflected in the mean, where the unaffiliated have the lowest score. Regarding education, there are also differences. Although the number of persons with higher education is high, both the unaffiliated and the humanists have a higher proportion of persons with a long university degree than the affiliated. Indeed, over 60 percent of the unaffiliated and the humanists have at least one year at a university. As for income, the differences are not substantial for the unaffiliated and the affiliated. The humanists have a larger proportion of persons making more than NOK400,000 than the affiliated and the unaffiliated. Looking at city size, the proportion of unaffiliated living in Oslo is relatively larger than the affiliated. There are also more affiliated persons in the less populated categories. Politically, the unaffiliated are more left-oriented than the affiliated. They have a significantly higher proportion of people declaring that they will vote for the Red Party (Rødt), Labor Party (AP) and Socialist Left Party (SV). The affiliated have a higher proportion of conservatives than the others, whilst the humanists have an even higher proportion on the political left side than the unaffiliated. In regard to marital status, the unaffiliated stand out in that over one in four are unmarried, while one third are married. For the humanists and the affiliated, close to 50 percent are married. Moreover, about 70 percent of the humanists and the affiliated have one or more children, whilst this goes for half of the unaffiliated. The humanists are the category where the most persons report being a member of an NGO, with two thirds of the persons. This is significantly more than the unaffiliated and affiliated, where half report the same. Four out of ten of the affiliated report that religion was important in their upbringing, while for the humanists and the unaffiliated, this is not the case: Less than 10 percent of the humanists and 17 percent of the unaffiliated report that religion was important in their upbringing. When looking at friendship, the majority of all categories have friends with other religions or life views, but the differences are significant: 80 percent of the unaffiliated and 96 percent of the humanists report having friends with other religious and worldviews, while 65 percent of the affiliated report the same. The final variable, relating to personal belief, shows significant differences. Close to 70 percent of the affiliated report that they believe in God or a higher power of some sort. Looking at the humanists, less than 10 percent report the same. As for the unaffiliated, about one in five say they believe in God or a higher power of some sort.
|Less than 5,000||-0.304||0.184||0.099||0.738||0.377–1.099|
|One or more children||-0.333||0.166||0.045||0.717*||0.391–1.043|
|Religion important in upbringing||-0.676||0.159||0.000||0.508***||0.197–0.820|
|Friends other religions||0.932||0.147||0.000||2.538***||2.250–2.826|
|Interaction gender – unreligious||.057||.245||.815||1.059||0.579–1.539|
|Nagelkerke R Square||.182|
|Hosmer & Lemeshow||.978 df= 8 Chi-square = 2.079|
*** p ≤ 0.001, ** p ≤ 0.01, * p ≤ 0.05 Probability or proportion for intercept (male, 18 years old, lower secondary school, making less than NOK 100,000, from Oslo, politically center, not member of any NGOs, married, no children, religion was not important in upbringing, does not have friends with other beliefs, and believes in God or a higher power of some sort) = 0.023 ≈ 0.2 percent. The Nagelkerke pseudo Square indicates that less than 20 percent of the variance in the dependent variable is explained by the predictor variables. Hosmer & Lemeshow (.063 > .050) showing that the model is adequate for its purpose.
Results step 2, logistic regression
Table 2 displays the results from the logistic regression predicting which variable has the highest odds ratio for unaffiliation. In the model, gender, age, income, membership of NGOs and marital status show no significant results in predicting unaffiliation. The results do show that each level of completed education increases the odds of being unaffiliated. The reference group for city size is Oslo, indicating that persons living in a town or city with a population of 5–50,000 inhabitants have a lesser probability of being unaffiliated than persons living in the capital. The other categories for city size are not significantly different from Oslo. Of the variables capturing the voting behavior, persons who are left-of-center and persons who are politically undecided have a higher probability of being unaffiliated. None of the marital statuses give significant results for persons who are married. Persons who have one or more children, however, are less likely to be unaffiliated than persons without children. Furthermore, the results also show that persons who report having a religious upbringing are less likely to be unaffiliated than persons who report religion as being important in their upbringing. Moreover, a highly predictive factor is having friends with other religions or world views different than themselves. They are two and a half times more likely to be unaffiliated than persons who are surrounded by more religiously homogenous surroundings. Persons who do not believe in any God or a higher power of some kind, are four times more likely to be unaffiliated than persons who are believers. Indeed, this is the highest predictor in the model. Finally, to address the possibility that unbelief and gender is correlated, an interaction variable is added to the model. The result is insignificant, meaning that gender and unbelief are not interacting when predicting unaffiliation in Norway.
In this article, I have analyzed the religiously unaffiliated in Norway. Using two sets of analysis, I have presented both the social characteristics of the unaffiliated and the importance of social background for this growing category within the religious landscape.
What I find is that the unaffiliated in Norway, in some areas, show similar traits as those described in the research from the USA; this despite Norway being different contextually and historically. The first of these similarities is that religious unaffiliation is growing rapidly. Over a short period of time, unaffiliation has gone from a relatively small proportion of the population to a substantial size. A second similarity is related to education. Persons with higher education are overrepresented among the unaffiliated in Norway. Although the types of higher education are not differentiated, the effect alone is clear, even when controlling for other variables. The third similarity concerns politics. Similar to the unaffiliated in USA, the unaffiliated in Norway are more prone to be politically left. A fourth likeness is that most of the unaffiliated are young, unmarried and have no children. Another way of looking at it is that there is a demographic change. The younger generations are replacing the older, more traditional religious generation. There are also notable differences. One of them is that the religiously unaffiliated are no less likely to be involved in voluntary organizations than the affiliated. The proportions are the same. Indeed, it is the members of the Humanist Association that have the highest proportion of volunteers. Another is that there are no statistically significant gender differences among the religiously unaffiliated. This is somewhat surprising, as a difference in gender and religiosity is notable also in Norway.
The first part of the analysis aimed to answer the following research question: what are the social characteristics of the religiously unaffiliated in Norway and how do they differ from the religiously affiliated and the secular humanists? The bivariate analysis showed that the religiously unaffiliated is typically a person who has high education, lives in Oslo or urban settlements, is politically left, unmarried, has no children, did not have a religious upbringing, has friends with different religious views, and is personally unreligious. The distribution of the research variables shows that the unaffiliated do differ from both the religiously affiliated and the organized secular humanists in several areas. In education, politics, marital status, having one or more children, not being raised religiously, having friends with pluralistic worldviews and believing in God or a higher power, the unaffiliated place themselves somewhere between the religiously affiliated and the secular humanists. In a way, this emphasizes the point of distinguishing the secular humanists from the unaffiliated. They are indeed different. For instance, there is a much larger proportion of unaffiliated who have not previously been members of the Church of Norway. Another way of looking at it: it is likely that a proportion of the religiously unaffiliated are religiously “passive.” They have not been baptized or previously registered in any religious organization; their status of being religiously unaffiliated is not necessarily by active choice. Hence, religious unaffiliation may be related to irreligious socialization (Merino 2012). Several social factors may interact so as to strengthen each other. Not having a religious background and not being a member of a religious organization may naturally coincide. Contrary to being religiously passive, for some, being members of the Humanist Association is not only an expression of their secular beliefs, but also a political stance regarding state-church issues (Repstad 1999). An interesting observation is that the religiously unaffiliated are in some aspects placed in between the religiously affiliated and the secular humanists; they are in many respects religious in-betweeners. Looking at the variables covering personal belief, former membership, political preference, education, and place of residence, the unaffiliated fall in between the Humanist Association and the religiously affiliated. This may reflect both that many of the unaffiliated are passive and that there is a large variety within this category. Being younger, and thus less likely to have started a family, they are also, in a sense, in between phases of life.
My second research question – what social factors are central in understanding unaffiliation was further accommodated by using logistic regression. What I find is that personal belief and the influence of friends are the two highest predictors. I also find that having an unreligious upbringing is correlated with being unaffiliated, in addition to higher education, being politically left-oriented and not having children. In the continued discussion, I wish to focus on three areas that may help to understand how these social factors influence unaffiliation.
1) The first of these areas is related to how unaffiliation may be understood as a consequence of individual secularization. I base this on the fact that the religiously unaffiliated is the fastest growing category within the institutional religious landscape in Norway. Furthermore, in 2012, unbelief was the highest predictive factor for religious unaffiliation. Although not all the unaffiliated are unbelievers, more than 70 percent do not believe in a God or a higher power of any sort. Individual secularization is confirmed in recent research in Norway as well. Overall, religious beliefs and practices are indeed declining (Botvar and Schmidt 2010; Urstad 2010). In such a manner, unaffiliation is an observable outcome of individual secularization. An additional point related to secularization is that, parallel to the increase in religious unaffiliation, there has also been pluralization in Norway (Furseth 2015). Due largely to immigration, there has been an increase in religious organizations. However, even if there has been an increase in available religious or secular alternatives, secularization continues. This is in contrast to theories that claim that an increase in supply creates increased religious involvement (see Lannacconne 1997).
2) The second area is related to the results of social influence, understood as family socialization and the influence of friends. What I find is that the absence of religion in a person’s upbringing is highly significant in predicting unaffiliation. This is also seen in relation to religion and religious views: growing up in religiously-oriented environments is associated with higher probabilities of adopting a religious view (Voas and Crockett 2005). In this case, less religious socialization predicts less affiliation. It is also clear that having friends with other religious views clearly correlates with religious unaffiliation. Looking at the correlatives, friends are actually a higher predictor of affiliation than parents. The effect of being exposed to other views on life may challenge religious ideas and conceptions, such as belonging to religious organizations. This is in line with the “younger” Peter Berger, emphasizing that exposure to pluralism may in fact lead to further secularization (Berger 1969). Social influence can therefore be seen as two-fold: secular background and exposure to other views on life may challenge religious ideas and conceptions, such as belonging to religious organizations. Moreover, having an unreligious upbringing may even enable individuals to have more friends with pluralistic views, as religiosity is a non-factor. In cases where friends are more homogeneous, religion may be a distracting or even an excluding factor for social interaction (King and Furrow 2008). Although I find correlations regarding social influence, I cannot say anything about causality. Having more data, over a period of time, may provide better understanding of possible social mechanisms, thus underlining a need for further research.
3) The third area for interpretation concerns gender. A noticeable, unexpected result in the Norwegian case is that gender is insignificant when it comes to religious unaffiliation. Contrary to results from the USA, where the unaffiliated are more prone to be men, the proportions of unaffiliated men and women in Norway are about the same. Gender shows no effect when predicting unaffiliation. This is somewhat surprising because, when addressing religion and religiosity in general, gender has proven to be highly significant, also in Norway (Botvar and Schmidt 2010; Furseth 2004, 2010; Urstad 2010). What I am going to suggest, quite carefully, is that we are witnessing an area in which the cultural lag in regard to gender and religion is closing. There are some points that may support this suggestion. Both men and women have shown a parallel tendency in religious thought and practice during recent decades: they have become more secular (Urstad 2010). Another point is that in other areas as well, gender differences are subsiding: Norwegian women’s participation in education, politics and the labor market is among the highest in the world (Maimunah, Mohd, and Abd 2011; United Nations 2015). Moreover, women’s engagement in education and the labor market has been linked to disaffiliation from traditional life, including religion, and has been noted as an important factor relating to secularization (Aune, Sharma, and Vincett 2008; Furseth 2010; Marler 2008). In the book Why are Women more Religious than Men? Martha Trzebiatowska and Steve Bruce (2012) also argue that it is only a matter of time before women close the gap in secularity due to their increasingly involvement in the public sphere and other domains traditionally held by men. The difference in religiosity is, however, complex and has to do with the sum of many differences, not only involvement in the labor market (Trzebiatowska and Bruce 2012, 170). The bottom line is that they expect that future development in secularization will also, to an greater degree, involve women. Whether or not this is the case, the findings from my study may prove fruitful as a point of departure for further research on religious unaffiliation.
|Reg1 Region East (Oppland, Hedmark, Akershus, Østfold)||24.4||25.6|
|Reg2 Region South (Vest-Agder, Aust-Agder, Telemark, Vestfold, Buskerud)||19.2||18.1|
|Reg3 Region West (Sogn og Fjordane, Hordaland, Rogaland)||20.9||19.8|
|Reg4 Region Mid (Nord-Trøndelag, Sør-Trøndelag, Møre og Romsdal)||13.8||14.2|
|Reg5 Region North (Finnmark, Troms og Nordland)||9.4||10.0|
|Reg0 Oslo (capital)||12.3||12.2|
|1||In this paper, the term “religiously unaffiliated” will be utilized instead of “the nones” as it has less normative connotations.|
|2||https://lovdata.no/dokument/SF/forskrift/1988-12-01-996 – KAPITTEL_1.|
|3||A part of the data used in this article is acquired from Tros- og livssynsundersøkelsen 2012 (Norwegian Survey on Beliefs and Faiths 2012), conducted and financed by KIFO, Institute for Church, Religion, and Worldview Research, represented by Ida Marie Høeg, Ulla Schmidt, and Sunniva Holberg. KIFO is not responsible for the analysis or the interpretation of data presented in this article.|