Same-sex marriage and the Lutheran Church in Finland: How rapid change in values and norms challenges the church and its decision-making

The effect of legislative changes on same-sex unions on people’s attitudes toward homosexuality is a fairly well-researched topic. There is less research on how state legislation on same-sex unions has affected churches’ attitudes and policies on registered partnership and same-sex marriage. In this paper we focus on attitudinal polarisation in the Evangelical Lutheran Church of Finland (ELCF) in the matter of same-sex marriage and why there is a gap in the attitudinal development of the church and the society. The analysis is based upon six sets of representative surveys among the Finnish population, the employees of the ELCF and the elected officials of the ELCF between the years 2012–2017. Whereas the majorities (55%) among the Finnish population and among the ELCF members (54%) are already favourable to the idea that same-sex couples should have the right to marry in the church, the ELCF has not approved of it. This situation has polarised the ELCF from within. Empirical analysis shows that the reason for the discrepancy between popular opinion and the ELCF stance stems from the fact that ELCF decision makers are older and more religious than the general membership. However, the data also shows that even among the most religious, younger age groups are already more favourable towards the same-sex marriage. Therefore, it is likely that the polarisation will lessen over time as more and more religious people accept same-sex marriage.

officiate same-sex marriages (Lindberg 2014), the Evangelical-Lutheran Church of Finland (ELCF) has refused to follow suit.
The Nordic debates on same-sex marriage and the role that majority churches have played in them have sparked the interest of some scholars. The focus has been on socio-cultural factors related to the functional differentiation. In the debates on their roles as officiators of marriages, the Nordic majority churches have been viewed as providers of a social service function and described as public utilities, since their availability is taken for granted and belief seems to be a secondary thing. There has been a desire to avoid an open split between the majority churches and the states. The role of these churches as public utilities, the political influence over them, especially in Sweden, and the value-profile of each country have contributed to letting the churches adapt to the changed situation at their own pace (Lindberg 2014;Hokka 2013).
The effect of legislative changes on same-sex unions on people's attitudes toward homosexuality is a fairly well-researched topic (e.g. Redman 2018; Clements and Field 2014;Slenders et al. 2014;Hooghe and Meeusen 2013;van der Akker et al. 2013;Takács and Szalma 2011). How the legislation on same-sex unions has affected churches' attitudes and policies on registered partnership and same-sex marriage has interested only a few scholars (e.g. van den Berg 2017; Dabelsteen 2015; Cadge 2002) even though Christian churches in many parts of the world have had to respond to the changing public opinion concerning same-sex marriage. The stance of the church cannot be ignored, especially when the majority of a population are members of a particular church.
The Nordic majority churches face the issue in an especially acute form due to their numerical significance and special relation to the state. Nordic people may be less interested in religion per se but express through their church membership a strong loyalty to the established church (Sundback 2000). In spite of functional differentiation, the Nordic majority Lutheran churches have maintained their prominent cultural role in society. They continue to enjoy privileged positions in many public institutions and the states also rely on the majority Lutheran churches during times of national celebrations and crises (Furseth 2018). Silvio Ferrari has argued that the position of the church, if not carried by legal frameworks and formal structures, may be upheld just as strongly by emotions, traditions and tacit understanding (Ferrari 2010).
In this article we look at the Finnish case, which is unique among the Nordic countries due to the reluctance of the ELCF to open its marriage ceremony to samesex couples despite the recently changed marriage law. We focus on a situation when the national legislation, people's views on the role of the church, and the church's own position on an issue do not point in the same direction. We examine how religiosity affects church members' attitudes to the legislative changes on marriage and how changes in values and norms challenge the decision making of the ELCF.

Previous research and theoretical approach
Modernisation theory and cultural heritage theory have been used to explain attitudes towards homosexuality (Gerhards 2010). Modernisation increases economic welfare, which in turn leads to changes in people's value systems. Educational attainment has been perceived as another indicator of value change. Higher levels of education lead to a questioning attitude to traditional values, especially in high-income and high-security countries, whereas education itself does not drive these changes in low-income countries (Inglehart 2018(Inglehart , 1997. Typical variables that previous research has found to affect attitudes toward homosexuality include socioeconomic status, gender, education, age and having a homosexual friend or acquaintance. Younger age groups and women are more tolerant towards homosexuality than older people and men (Ayoub and Garretson 2017;Slenders et al. 2014;Adolfsen et al. 2010;Gerhards 2010;Herek 2002;Hellevik 2002). In highly modernised Sweden and Norway, gender, political ideology, education, whether a person has lived in the capital, and attitudes toward gender equality shape people's attitudes towards same-sex marriage. Contrary to other studies, same-sex marriage was equally accepted by all age groups (Jakobsson et al. 2013).
Inglehart points out that people become increasingly open to new norms when a society attains high levels of existential security. However, attaining high levels of economic and physical security does not produce immediate societal changes but involves a time-lag: "When conditions shaping a society's younger generation's pre-adult years differ substantially from those shaping older groups, intergenerational value change occurs. It proceeds with a multi-decade time-lag between the emergence of the societal conditions conducive to the change, and the time when a society as a whole has adopted new values" (Inglehart 2018, p. 80). The declining importance of religion and the rapid spread of individual-choice norms suggest "that value change can reach a tipping-point at which conformist pressures reverse polarity, accelerating changes they once retarded" (Inglehart 2018, p. 100).
People's attitudes are shaped not only by the larger society but also by their reference groups (Cochran and Beeghley 1991;Loftus 2001). Religious affiliation and religiosity play important roles in structuring attitudes about same-sex unions (Olson et al. 2006). The opinion formation of strong identity attachments may cause negative attitudes to homosexuality. A negative association has been found in several studies between religiosity and tolerance of homosexuality (Liefbroer and Rijken 2019;Clements and Field 2014;Sheepers et al. 2002). The importance of social reference groups means that religious involvement needs to be taken into account in addition to formal affiliation and religious convictions. It has been found that the less that people are involved in their religious institutions, the more likely they are to support the principle of non-discrimination towards sexual minorities (Slenders et al. 2014;Jakobsson et al. 2013;Gerhards 2010).
Besides affiliation and religiosity, also the organizational structure of the religious communities has been found to affect the situation. In her study of the debate on homosexuality in the mainline denominations in the USA, Cadge (2002) found that the controversy is stronger in churches with a strong hierarchical structure. The more authority local congregations are granted over issues related to homosexuality K the less there is controversy. Davie and Starkey's (2019) study concerning people's response to an openly homosexual bishop within the Church of England points in the same direction: the policy-making of the Church of England is constrained by its position within the wider Anglican Communion, resulting in a tension between the changed public opinion and the church policy.
Studies reveal that same-sex marriage legislation affects public opinion but scholars have divided opinions on which direction the change takes. In their study based on data from the European Social Survey (ESS 2002(ESS -2010 Hooghe and Meeusen show that in countries with registered partnership or same-sex marriage recognition, people are more tolerant toward homosexuality than in countries without legislation. The likelihood that they would disapprove of homosexuality is higher among those who are older, less educated and religious. Hooghe and Meeusen conclude that in countries where same-sex marriage is legalised, disapproval of homosexuality declines and the development is toward more uniform attitudes (Hooghe and Meeusen 2013). Other scholars studying the European situation have arrived at similar conclusions: when same-sex marriage is legalised, attitudinal changes toward homosexuality take a more positive direction (Takács and Szalma 2011), though some note that the direction of causality is somewhat unclear (van der Akker et al. 2013). Furthermore, the association between laws and attitudes may be a result of a two-way causality: apart from leading to a more tolerant attitude toward homosexuality, the legalisation of same-sex marriages also reflects public opinion in that particular country (Slenders et al. 2014).
Redman affirms that legalising same-sex marriage has an effect on public attitudes on homosexuality but arrives at a somewhat different conclusion on the direction of such change. His data is based on the World Values Survey (WVS), integrated with data from the European Values Study (EVS). Redman shows that the effect can take different directions, positive or negative, by reinforcing pre-existing attitudes. Instead of leading to more uniformed attitudes, same-sex legislation has a polarising effect. Individuals whose attitudes were already positive toward homosexuality change their attitudes in a more favourable direction, whereas people who were negatively disposed towards homosexuality do not change their opinions. Furthermore, if there will be any change, it is likely that same-sex legislation will make negative attitudes even more negative. However, Redman acknowledges that since the younger age cohorts are more accepting of homosexuality than the older ones, it is likely that legislation safeguarding the rights of homosexuals will in time have a positive impact on the attitudes of the greater population (Redman 2018).
The results of previous studies on the effect of same-sex marriage on public opinion on homosexuality can be utilised in studying attitudes toward same-sex marriage, even though we recognise that although attitudes to homosexuality and same-sex marriages should not be equated they are related. In this paper, we focus on the attitudinal polarisation in the ELCF on same-sex marriage and why there is a gap in the attitudinal development between the church and society. First, we examine how legislative changes have affected the law and practice within the ELCF. Second, we look at how opinions on same-sex marriage has changed among people with different degrees of closeness to the ELCF: the entire Finnish population, the ELCF members in general, its employees and its elected officials.

Data and method
Our data is both qualitative and quantitative. First, in order to fully understand the situation of the ELCF with regard to same-sex marriage in Finland, we provide some essential background information on the recent legislative processes and changes concerning LGBTI-issues in both state legislation and Church Act. The examination is based on documentary sources, primarily church policies, reports, studies, legal documents, and previous research. They shed light on the key structural changes that provide the context for the second part, in which we analyse the changing opinions of people through survey data.
We have used survey responses from six separate surveys. The Church Research Institute of the ELCF has commissioned several representative surveys between the years 2012 and 2017 that can be used to examine empirically how people's religiosity and the closeness of their ties to the church affect their attitudes towards samesex marriage. 1 We will focus especially on two nationally representative surveys from the years 2015 and 2017, because they both included an identical set of items designed to explore how Finns react to the prospect of the ELCF officiating marriages of same-sex couples. The surveys in question are Gallup Ecclesiastica 2015 (N(unweighted) = 4194) and Church Marriage Survey 2017 (N(unweighted) = 1282) (henceforth GE 2015 and CMS 2017 respectively). They are both representative probability samples of Finns aged 15-79. The survey method was computer-aided web interview (CAWI). The sampling frame in both studies was the nationally representative webpanel maintained by the fieldwork organisation (Taloustutkimus). The response rates were 22% (GE 2015) and 27% (CMS 2017). 2 For the purposes of analysis, the data were weighted according to age, sex, living area and membership in religious organisations. The GE 2015 survey was conducted in October-November of the same year that the law regarding same-sex marriage was enacted by Parliament, and CMS 2017 was conducted a month after the law came into force in spring 2017. The entire period from 2013 to the first half of 2017 was a time when same-sex marriage and its implications for the ELCF were intensely debated in the public sphere.
In addition to these nationally representative surveys, there are also the Church Employee Surveys and Church Elected Official Surveys of the years 2012, 2015, which used the same items to measure attitudes. The Church Employee Surveys were sent to a random sample of Church employees, both pastors and lay staff, of whom 863 (23%) (2012) and 1390 (37.1%) (2015) responded. It should be noted that only about one tenth of Church employees work as parish pastors and a further third are employed in other type of spiritual parish work (chaplaincies, diaconal work, youth work, etc.). The rest, slightly over half of all employees are working in 1 The Church Research Institute is a separate administrative unit of the Church Council of the Evangelical Lutheran Church of Finland, from which it receives most of its research funding. 2 The rather low response rates and the use of a web panel as a sampling frame might cause some concern regarding representativity. These surveys can, however, be compared to other surveys on religiosity of Finns using different sampling procedures and/or survey methods, and there have not been found any serious and unaccounted discrepancies in any standard measures of religiosity. Other recent nationally representative surveys on religiosity commissioned by other organizations are the ISSP (2008, 2018) and the EVS (2009, 2017).
K administration, property maintenance or cemeteries. The pastors and other spiritual parish workers were slightly overrepresented in both surveys (60% vs. 46% in 2012 and 54% vs. 45% in 2015). The survey of elected officials was sent to a random sample of elected officials of parishes or parish unions. A total of 482 (33%) (2012) and 1185 of them (45.7%) (2015) responded. Both surveys were conducted electronically. These surveys enable us to examine how attitudes among church employees and elected officials differ from those of the general membership of the ELCF and the whole population.
The nationally representative datasets were analysed using the multivariate method of OLS regression in order to examine the effect of religiosity on attitudes towards same-sex marriage when several socio-demographic and attitudinal variables were controlled for. Otherwise, the analysis depends on simple comparisons of shares of respondents in different answer categories of the several datasets.

The same-sex legislation and the response in the church
According to the Finnish constitution, the state is neutral in matters of religion. All citizens have the right to freedom of religion. Since the Freedom of Religion Act came into force in 1923, the ties between the church and the state have been gradually severed. The state still holds legislative but not executive power over the daily matters of the church. The two national churches, the ELCF and the Finnish Orthodox Church, operate under public law. Although the ELCF has internal autonomy to manage its own affairs, the national Parliament must ultimately ratify the Church Act in the form proposed by the church, either by accepting it or rejecting it altogether (Kääriäinen 2011). In the past 30 years, all Nordic majority Lutheran churches, with the exception of Denmark, have become more autonomous in relation to the state. Nevertheless, in 2019, all Nordic Lutheran churches still enjoy constitutional preference, though the state's legislative authority over the church has been restricted and the state executive authority over the church has either been annulled or restricted. Among its Nordic neighbours, the ELCF is the most independent from the state (Kühle et al. 2018).
The general administrative bodies of the ELCF are the Church Synod, the Church Council and the Bishops' Conference. The highest decision-making organ is the Synod, which includes 64 lay representatives, 32 representatives of the clergy, all bishops, a Sámi representative and a representative of the Council of State. Members of the parish boards elect lay representatives and the pastors of the dioceses elect representatives of the clergy to the Synod. Their term is for four years and the Synod gathers twice a year. It is a task of the Synod to decide upon the changes in the Church Act and the Church Order, and therefore the issues concerning marriage fall within the Synod's authority. These kinds of changes require a three-fourths majority vote in the Synod (Kääriäinen et al. 2009, pp. 141-142).
The ELCF is the largest religious denomination in Finland. Even though its membership rate is declining, in 2017 70.8% of the population still belonged to it. Those members of the ELCF eligible to get married in the church can choose either a civil marriage or have their marriage officiated by the church. Church rites play an important role in the lives of Finnish people. In 2017 about 56% of those church members who entered into marriage had their marriage officiated by the church and 8% of church members' civil marriages were blessed by the church (Kirkkohallitus 2019). The amendment to the Marriage Act that entered into force on 1 March 2017 enabled same-sex couples to enter into marriage. For those church members who want to enter into a same-sex union a civil marriage is thus far the only alternative accepted by the church. Out of all marriages contracted in 2018 about 1.6% were same-sex unions (Official Statistics of Finland 2019). It is not possible to find out how many of these couples were members of the ELCF.
Attitudes toward homosexuality in Finland first started to change during the latter part of the 20th century (Juvonen 2015, pp. 71-78). Homosexuality was criminalised until 1971 and it took a further ten years for the Board of Health to remove homosexuality from its list of illnesses (1981). Joining the European human rights agreement in 1990 and the European Union in 1995 influenced perceptions of homosexuality: it was now perceived more from the human rights point of view. Furthermore, several public individuals started to speak openly about their homosexuality. In 1995 the Parliament passed the bill forbidding discrimination including discrimination on the basis of sexual orientation.
The 21st century has witnessed a rapid change towards more accepting views and attitudes to homosexuality leading to legislative changes on same-sex relationships (Juvonen 2015, pp. 79-99). In 2002 the law on registered same-sex partnership came to power followed by the Acts on Transsexuality (2002) The consequences of the Partnership Act for the church started the discussion in the ELCF. The conservative groupings opposed any legal acceptance of homosexual relations and the Bishops' Council was divided over the issue (Juvonen 2015, pp. 124-131). In 2002 two motions were made to the Church Synod. The first one proposed that there should be a clause in the Church Order denying employment in the church to any person who has registered a same-sex partnership. The motion abated in the Synod. The second motion suggested that the Church Synod should work towards preparing an alternative formula for the blessing of the same-sex partnerships. The following year the Synod asked the Bishops' Council to look into the consequences of the 2002 Partnership Act for the church (PK 2010, pp. 3-6).
The process moved on slowly. In 2005 the Bishops' Council appointed a committee to look into the matter and after four years' work the report was handed over to the bishops (KRP 2009). In 2010 the Bishops' Council presented its report to the Church Synod. Some bishops saw that the blessing of a same-sex union was possible, some opposed it. Both groups based their views on the Bible and the Lutheran Confession but came to different interpretations. The bishops came to the conclusion that praying with and for the couple falls within the realm of private life and not within the holy rites of the church (PK 2010). This was met favourably by the Church Synod. In February 2011, on the basis of the decision of the Church Synod, K the Bishops' Council pronounced a pastoral directive: pastoral care can include praying with and for people of the same sex who have registered their partnership. However, it was emphasised that such a prayer does not mean blessing the same-sex partnership. It was not allowed to use parts of the marriage liturgy, such as questions, promises, exchanging of rings, declaration of partnership, and the blessing of the union (PO 2011). Public opinion about the directive was divided. Some people found it difficult to understand why the church wished to deny God's blessing on the basis of sexual orientation. Others who perceived homosexuality as a sin saw even praying for and with a same-sex couple as diverting from God's word.
The changes in the legislation on discrimination in 2014 (Non-discrimination Act 1325/2014) forced the ELCF to specify their consequences for the church in employment situations, including the question of an employee's sexual orientation. The Church Act and the Church Order of the ELCF do not state anything on sex or sexual orientation of persons employed in the church. Therefore, the 2014 legislation on discrimination apply regardless of the type of work, including employment for ordained ministry. In practice this meant that sexual orientation could not be a reason for not employing a person or for terminating his/her employment (KHyk 2015). This is the case also with those officially registered religious communities who have not stated specific restrictions for employment in their Community Order or Confession of Faith. On issues of employment the Orthodox Church in Finland follows the statutes stipulated by the Orthodox Canonic Law (Helander 2017).
Anticipating the gender-neutral Marriage Act to come into power, the ELCF presented a legal report on its implications to the church. Even though pastors had the right to officiate marriages, they were not obliged to marry same-sex couples. Furthermore, if pastors married a same-sex couple, the marriage became legally binding even though the pastor had acted against the praxis of the church and consequently, according to the Church Act, would face penalties (OS 2016). The legal report did not calm the situation. Several pastors expressed their theological views on same-sex couples' right to be married in the church and their willingness to marry them.
In 2016 a motion was made to the Church Synod asking the Church Council to investigate whether the church should withdraw from its right to officiate marriages. The Synod discussed the motion and, since matters related to doctrine and rites belong within the authority of the Bishops' Council, decided to send it for further preparation to the Bishops' Council in cooperation with the Church Council (PK 2017). The investigation ordered by the Bishops' Council was delivered in September 2017 (Helander 2017). While the investigation was going on, a motion was made to the Church Synod proposing that same-sex couples who fulfil the other criteria for their marriage to be officiated by the church should also have the right to marry in church (Edustaja-aloite 1/2017).
The gender-neutral Marriage Act caused individual congregations to take a stand on whether same-sex couples could use churches and other congregation premises to celebrate a marriage solemnised by the civil authorities. Some congregations welcomed all couples to all their premises regardless of gender, some limited the use of the church to heterosexual couples while same-sex couples were welcome only to use other premises like the parish hall. There were also congregations who limited the use of their premises to celebrate the marriage of a man and a woman only. No single congregation has left the ELCF due to the marriage controversy. However, some of the revival movements operating within the church have, in spite of doctrinal differences, joint together and thus strengthened their profile as defenders of the traditional moral values. Such polarization between the conservative wing of the ELCF and Finnish society is evident primarily around socio-moral issues related to family and sexuality (e.g. Kanckos 2012). The conservatives in the ELCF also back their views with an ecumenical argument since the Catholic Church in Finland, the Orthodox Church, Pentecostals and the Evangelical Free Church of Finland perceive marriage as a union between a man and a woman (Helander 2017).
The question of whether the ELCF should also officiate same-sex marriages became more acute when some of its pastors started, on the basis of their theological convictions, to officiate such marriages without waiting for permission from the church. They also justified their action by the fact that the Church Act or Church Order do not state clearly that the marriage is only between a man and a woman; Finnish has no third-person pronouns specific to the gender of the person. Pastors who have performed same-sex marriages followed the ritual form of the Church Handbook. Furthermore, the majority of church members supported same-sex marriage (Helander 2017). Over 80 pastors from all dioceses have publicly stated that they are willing to marry same-sex couples. About 61 pastors have been said to have done so, but only six pastors have been penalised by their Diocesan Chapters. The penalties have varied from an oral reproof to a written notice placing pastors in the ELCF dioceses in a legally different position. Three of the four pastors receiving this notice have appealed, first to their Diocesan Chapter and, after receiving a negative reply, to the court of law. By early 2019 over 100 same-sex couples have either been married in the ELCF or had their registered partnership blessed by a pastor (Seppälä 2019).
The reasons for the discrepancy between legislation, church members and the views of the church leaders and decision makers need to be investigated further. The election of the members of the Church Synod, the church's highest decision-making body, is done by an electorate chosen in the church elections. Generally in these elections the conservative members seem to be more active than the liberal ones. Second, matters of doctrine and changes in the liturgical handbooks require a threefourth majority vote. Therefore, changes in issues which deal with values and norms face hard opposition from the conservatives who tend to uphold the traditional status quo and its related values.

Religiosity, same-sex marriage and proximity to church and its teaching
To date, six representative surveys have been conducted which have used an identical measure to probe the respondents' attitudes towards same-sex marriages being performed by the ELCF. The measure consisted of a statement, "The church should also marry couples of the same sex", with a 5-point response scale ranging from  complete agreement to complete disagreement. The distributions of responses of all surveys are given in Fig. 1. Several things stand out very clearly from the figure. First, the difference in attitudes between the whole population and the members of the ELCF is very small. This means that the cultural change taking place in society is reflected quite closely in the attitudes of the members of the ELCF. This situation stems from the position of the ELCF as a majority church in Finnish society, which embraces such a large portion of the entire nation.
Second, the closer the respondents are to the decision-making organs of the church, the more conservative are their attitudes toward same-sex marriage. Of the church's elected officials, only a quarter (25%) agreed and two-thirds (66%) disagreed with the statement in 2015. The corresponding figures in the employee survey of the same year are 34% and 53%, while among the general membership the figures are 45% and 44%. That is to say, there is about a 9-10 percentage point difference between each group in the acceptability of the church performing samesex marriages where the general membership is the most open to it while employees and elected officials are more reserved. This indicates that social proximity to the local parish, either as an employee or an elected (lay) official, works to support conservative attitudes with regard to marriage rights.
Third, the figure also shows that the change in the more liberal direction has been almost equally rapid in all the surveyed groups over a short time-span of just 2-3 years. Among the general membership, the share of those who favour same-  The church seems to be changing in the same direction and almost at the same pace on all of its levels, but with a time-lag among those with closer ties to the organisation. This raises the question of the extent to which the time-lag in change is created by religiosity as such and to what extent it is created by other, more demographic factors.
Next we examine more closely the demographic groups at the forefront of the change and to what extent the social composition of the parish employees and elected officials account for the difference with the general membership. For this purpose, multiple OLS regression was used to analyse the nationally representative datasets, where the key demographic variables indicated by previous research were used as controls. The analysis was done in steps so that the increase in explanatory power created by the inclusion of religious factors could be determined. The results of the regressions are shown in Table 1.
As the results of the regression models show, the strongest predictors of attitudes towards same-sex marriage are religiousness as measured by strength of belief in the Christian God, along with sex, age, urbanity and personal acquaintance with sexual minorities. Membership in the ELCF has only a negligible effect on the attitudes when the religious convictions and participation in religious services have been controlled for. It may also be noted that the coefficient of religious membership in the ELCF has changed from being negative to positive towards same-sex marriage between these two surveys (the unaffiliated being the reference category). More determined opposition can, however, be detected among the members of other religious communities, mainly members of the Orthodox Church and Protestant Christian minorities. In this case the coefficients are consistently negative and statistically highly significant.
Family situation also has an effect: couples and families with children are less open to same-sex marriages than singles. Level of education and income, however, have only a slight impact on the attitudes. As the comparison of the different surveys shows, the coefficients are fairly similar, except for participation in religious services, which seems to have become a slightly less important predictor of attitudes.
These results indicate that the explanations for the time-lag of the attitudes towards same-sex marriages can be sought primarily in four things: age, gender, religiosity and social involvement with the church. If the social composition of the church employees or church elected officials is predominantly male, older than the general membership, and more religious than the rest of the population, it is to be expected that their attitudes are less favourable to same-sex marriage being performed in the church.
As shown in Fig. 2, the parish elected officials are markedly older than general membership. Full two-thirds (66%) of them are over 50 years of age, while among the general membership the corresponding figure is only 42%. However, the gender balance is not so different from the general membership: 54% of the elected officials were female in the year 2014 and the same percentage was female in the whole ELCF membership (15 and older).
Elected officials are, however, much more highly religious than general membership of the ELCF. As shown in Fig. 3, more than half indicate that they participate in  church services at least monthly. Among the general membership, the corresponding figure is only about one in ten. Even if we allow a slight distortion in these figures due to social desirability, the difference is massive across all age groups. It is thus evident that the elected officials of the church are much more religious and much more socially involved in parish activities than the rest of the membership. This fact is only partly to be explained by age, as the corresponding differences can be detected in all age groups. As age and religiosity emerge as the defining predictors of the attitudes towards same-sex marriage among elected officials, it is important to take a closer look at how these predictors interact. Our previous observation that the change in favour of opening the marriage rite to same-sex couples is taking place approximately at the same pace (although with different starting points) regardless of the closeness to church suggests that religiosity may affect these attitudes differently in different age groups. For this reason it is useful to look at the differences between the mean values for the willingness to grant marriage rights to same-sex couples in different age groups together with their degree of religiosity.
The effect of religiosity within different age groups in the whole Finnish population stands out quite clearly in Fig. 4. Commitment to Christian teaching has a strong effect on attitudes towards marriage. The average difference in means between those who believe in God as taught by Christianity and the rest is 1.3 units in all age groups (the average difference ranging from 0.97 units in the oldest age group to 1.56 in the age group of 25-34). However, the figure also shows that the change has started to take place even in the most religious group: there is a bigger step towards more favourable attitudes in the most religious group in the youngest age cohort, the 15-24-year olds. This strongly suggests that there is a time-lag in the changing attitudes among the most committed believers. The direction and magnitude of the change is the same as the one that had begun about 40 years previously among those whose faith in God is dissimilar to the Church's teaching and among those who do not know whether they believe or not. The attitudes of those who do not believe in God at all probably started to change even earlier, and by now they have already reached a plateau.
Among the elected officials, the means quite closely reflect the values of this most committed group nationally, ranging from 1.97 (SD 1.42) in the oldest age group (65-79) to 2.62 (SD 1.71) in the youngest (under 35) (Elected Official Survey 2015). Among the church employees, the corresponding means are in the same range, only slightly higher, ranging from 2.39 (SD 1.60) in the oldest age group (over 50) to 2.93 (SD 1.72) in the youngest age group (under 35) (Church Employee Survey 2015). It is, however, notable that the direction of change in all these different samples is the same: the younger the respondent, the more open they are towards making marriage available to same-sex couples. 3

Discussion and conclusion
Empirical analysis shows that, as of 2017, popular opinion has reached the point where the majority among the Finnish population is favourable towards the idea that same-sex couples should be allowed to marry in church. More than 55% of the population at least partly agree with extending marriage rights to homosexuals while only 33% disagree or partly disagree. As Inglehart (2018, pp. 77-101) has pointed out, there is a tipping point at which the social pressure changes direction in the issue of acceptance of homosexuality. Our evidence indicates that this has just taken place in Finland. From now on conformist pressures are likely to speed up the acceptance of same-sex marriage in all population groups. This is most likely also starting to affect attitudes of the most actively religious within the ELCF.
Empirical analysis also shows that the reason for the discrepancy between popular opinion and the ELCF stance on same-sex marriage stems largely from the fact that ELCF decision-makers are older and more religious than the general membership. These constituencies are still mostly against the idea of same-sex marriage being performed in church. As previously pointed out, people's attitudes are shaped by their social reference groups. So long as the perception among the most active church members is that most of their peers frown upon the idea of same-sex marriage, they are also more likely to oppose it. More active involvement with the church and belonging to older age groups create an opposite social pressure where the acceptance of same-sex marriage is still widely resisted. As some scholars have shown, the acceptance of the same-sex legislation may have a polarising effect in the society, where some groups tend to hold even more tightly to their views in the face of changes of opinion that they disagree with.
However, in line with Redman's (2018) observations, there are indications that, among the most religious groups, the younger age groups are already starting to move in the same direction as the rest of the population. Our data also points in this direction. Therefore, it is likely that the polarisation will lessen over time as more and more religious people accept same-sex marriage. If present trends continue, the tipping point will be reached in a few decades' time even among the religiously committed membership. From that point on, social proximity to church will no longer create resistance against same-sex marriage. This means that while polarisation within society between the religious and the less religious members is likely to continue for a while, it will lessen over time as increasing numbers of elected officials will accept same-sex marriage. The tension between the values of the church and the society is a temporary one.
Such a tension has created a lively, at times even heated, debate within the church. It has been one of the issues that have brought religion into public scrutiny in all the Nordic countries during recent decades (Furseth 2018). This brings into sharper view the special role that the majority churches have in Nordic societies. Even though in Finland changes in public opinions on values, norms and attitudes have been rapid in the 21st century, it seems that there has not been an arena for general discussion on same-sex marriages after the changes in legislation at the societal level. Here the church has provided such an arena, providing people the opportunity and time adjust to the changed situation. However, a question arises as to how long the growing majority of church members will tolerate the views and punitive actions of a church which does not represent their values and views on same-sex marriage. Furthermore, the appeals of the penalised pastors to the court of justice challenge the ability of the church to solve the tension within its own autonomous structure. If unresolved, both developments may have far-reaching consequences on church membership and state-church relations.
Our results are in line with the recent major research project on the public role of religion in Nordic countries, which characterised the situation in terms of increasing complexity and tensions (Furseth 2018). The study takes special note of the importance of the debate concerning same-sex unions, and how it brought core values of the Nordic societies, gender and sexual equality and religious freedom, into tension. The study also indicates that in the Nordic countries sexual equality is increasingly seen as a value that cannot be subsumed under religious beliefs and values. The researchers conclude that "[t]he continued position of the Nordic majority churches as carriers of cultural values depends on their willingness to accept this situation" (Lövheim et al. 2018, p. 179).
The cultural shift concerning attitudes toward same-sex unions has thus brought the ELCF a particularly challenging choice: on the one hand, it could choose to continue in its role as symbolic carrier of national identity and common values, which, however, requires an ability to accommodate its thinking to the changing moral views of the people. On the other hand, it could emphasise its internal autonomy and religious freedom, which, in turn, entails the severing of its symbolic ties with the nation and accepting a sect-like protesting minority position vis-à-vis society in general. At present, both stances have their vocal advocates in the public discussion concerning the church's position on the issue in Finland. However, so far neither side has expressed a determined wish to sever the ties to the ELCF organization due to this issue.
The decisions that the churches make in this regard signals their willingness to continue in their role as carriers of cultural values accepted by a broad segment of the population. The churches in all other Nordic countries except Finland have made a choice that shows their willingness to continue in such a symbolic role. As the above analysis shows, there is still a long way to go to reach a similar decision in Finland. The delay is due to the hierarchical structure of the ELCF decision-making combined with demography of the decision-makers themselves. Because the issues concerning same-sex marriage are to be decided in the Synod, and by a threefourth majority vote, it is very difficult under current situation to reach a decision in either direction. As an unintended consequence of these structural constraints, there emerges internal polarization on the issue which in turn affects the image of the ELCF in people's mind. The longer the decision is postponed, the more likely it is that the public will interpret this as a growing symbolic distance and tension between the values of the church and the broader Finnish culture.