Christian churches and social welfare in secular times: How goal congruence shapes religious involvement in morality-based social services

Abstract We study the extent and nature of Christian engagement in morality policy implementation by means of a comparative case study in Germany. In particular, we observe that the nature of engagement varies between unconnected and corresponding types of activities, and we explain this variation with the policy-specific goal congruence between religious organizations (ROs) and the state. Goal congruence, in turn, can be linked to Catholic and Protestant moral doctrines that tell us about ROs' position on morality issues. The study contributes to the literature on faith-based welfare by highlighting the role of moral doctrines as drivers of ROs' social engagement.


Introduction
Although religious organizations (ROs) are faced with serious secularization pressures that diminish their member base and position of power, they still affect policyimplementation in various ways. A very prominent field concerns the welfare sector (Göçmen, 2013;Bäckström et al., 2016;Molokotos-Liederman, 2017;Manuel and Glatzer, 2019). There are highly inspiring insights on the involvement of ROs in the implementation of classic social policies such as elderly care, child care or poor relief (Naumann, 2005;Morgan, 2009;Göçmen, 2013;Blome, 2018). Meanwhile, very little is known about the engagement of ROs in services where religious and secular values clash, and which we call "morality-based social services" (e.g. services related to abortion, end-of-life care or sex work). Not only the extent but also the nature of such services offered by ROs has largely been neglected. This paper aims to close this research gap by studying the involvement of ROs in morality policy implementation. Accordingly, this paper asks: To what extent are ROs engaged in morality policy implementation? What is the nature of this engagement? Is it in line with legal intentions? And how can we explain the extent and nature of this religious governance engagement?
We argue that while the extent of ROs' engagement depends on opportunity structure, the nature of such engagement is largely dependent on the congruence in goals between the RO with regard to a specific morality policy and the actual policy (as deduced from the respective legislative text). Whether or not the RO supports secular liberal policy is determined in turn by the moral doctrines that guide church positioning on public, and especially on morality-based social services. So far, moral doctrines have hardly been explored in the context of religion or social policy, which is why this approach provides a new perspective on religious influence on social policy implementation and social service provision in particular.
We test our assumptions for two regional branches of the major churches in Germany, the Evangelical Church in Germany (EKD) and the Catholic Church in Germany. Concerning the nature of the engagement, we find that while Protestant engagement is largely in line with legal intentions, Catholic engagement is not. This is especially the case for social services related to abortion and sex work. Those "unconnected activities", as we call them, are mainly the product of incongruent policy goals, which means that religious actors may become active on behalf of the state but do not engage in line with legislative intentions. Differences between Protestant and Catholic governance engagement point back to their distinct moral doctrines, which affect the general openness of the confessions towards social service provision on behalf of the state. Finally, we find unsystematic variance in the extent of Christian ROs' engagement across Berlin and Bavaria. This indicates that the regional opportunity structure is at best a secondary factor for explaining the extent of ROs' engagement in the provision of morality-based social services, as it might be the case that unwillingness to implement due to incongruent goals will lead to ROs rejecting to be engaged in the implementation of a specific morality policy.
Our study provides a first empirical glimpse of the broad spectrum of religious actors' engagement in the implementation of morality-based social services. We offer insights into how Christian ROs may shape morality policies after these have been adopted in the parliamentary arena. These insights enrich the literature on religion and morality policy, which has so far rarely considered the implementation stage. Moreover, we provide a fresh perspective on the motivations for such engagement, linking an important governance concept-goal congruence-to theological moral doctrines. This conception pushes forward not only the young literature on religion and morality policy implementation, but also the scholarship on religion and welfare services, which is faced with an increasingly diverse group of beneficiaries (e.g. migrants, victims of domestic violence or sexual abuse, terminally-ill people).
The paper is structured as follows. First, we summarize the literature on religion, welfare states and social service provision. Next, we propose two possible outcomes of religious governance engagement-unconnected and connected engagement-and present the research design. Finally, we analyze our empirical findings and discuss them in a concluding part.
2. Three literature streams on religion, welfare and morality policy ROs are congregations, denominational organizations and religious non-profits (Chaves, 2002), and many of them, particularly the last two, are active in social service provision. Over the last two decades, increasing numbers of scholars have explored the role of ROs in social service provision in great detail. We consider three different literature streams that assess and explain the impact of religion and ROs on classical and non-classical social services, and on social welfare in general.
The first stream examines the general influence of ROs on the adoption of morality policy. This research originated in the U.S., where for instance, scholars found that the share of fundamentalist Protestants in the population affects the liberalization of moral issues such as gambling and liquor drinking (Fairbanks, 1977;Morgan and Meier, 1980). In Europe, researchers began studying the link between religion and morality policy only during the last decade. Scholarship focuses on the impact of the Catholic Church on policy output because fundamentalist Protestantism is not as common in Europe, and Catholics are considered the most influential and most conservative religious group. This literature shows that the political influence of churches varies across cultural and institutional contexts, as well as across morality policy issues (Engeli et al., 2012;Hennig, 2012;Grzymała-Busse, 2015;Ozzano and Giorgi, 2016). Generally, the research draws on the notion that the church-state relationship determines the extent to which the political and the religious spheres interact. For this reason, the church-state relationship defines the institutional opportunity structure, determining the number of access points to politics for churches. This also includes the existence of a confessional party that can act as the church's political agent in parliament, upholding Christian values during the policy adoption phase. Outside the political arena, public support for the church, which can be derived from the proportion of Christian adherents in the population, constitutes a cultural opportunity structure (Knill and Preidel, 2014, 385). As empirical research has demonstrated, opportunity structures, most prominently churchstate relation, are not only a key explanatory factor for religious influence on the policy adoption phase , but also explain the extent of religious involvement in the implementation phase (Göçmen, 2013;Euchner, 2018).
The second literature stream is central to our research, as it examines the contemporary role of ROs, so-called faith-based organizations (FBOs), in the provision of social services (Göçmen, 2013;Bäckström et al., 2016;Blome, 2018;Manuel and Glatzer, 2019). Like the first literature stream, this research originated in the U.S., where FBOs have always played a significant role in the social service sector. Here, the adoption of two policy programs − Charitable Choice in 1990 and the "Faith-based initiative" in 2001 − had launched a devolution of social service provision to the private sector and especially, faith-based organizations. This inspired academics to study FBOs and their relations to the state in greater detail (Monsma, 2000;Cnaan et al., 2004;Nagel, 2006). Similarly to the U.S., the dismantling of the welfare state in Europe led to an increase in third-sector service provision, a sector in which ROs play a major role. As a consequence, the scientific community started exploring the concrete involvement of FBOs in the European welfare markets, their organizational particularities and the challenges they face in the light of new societal developments (i.e., secularization, modernization, religious diversification) (Davie, 2012;Bäckström et al., 2016). A key finding is that welfare state retrenchment and historic church-state relations have been important drivers of FBO engagement. Furthermore, scholars stress that FBOs are able to systematically improve the living conditions of disadvantaged and poor people (Beaumont and Cloke, 2012). In this context, scholars also observe that FBOs' engagement is very much motivated by theological considerations (Netting et al., 2006;Schneider et al., 2011;Hien, 2019). In particular, studies find that FBOs' engagement in morality policy implementation is structured by the moral doctrines of their respective faith (Landry et al., 2011;Joyner, 2017). This may lead to FBOs' unwillingness to participate in the proliferation of secular moral ideas. For instance, the Baptist Church and the Catholic Church in Honduras do not provide comprehensive sex education but are committed to an abstinence-only education (Joyner, 2017, 16). And as Schneider (2016) outlines, Catholic Charities in Washington D-C had to step out of foster care services due to Catholic teachings. Their believe system does not allow them to send children to same-sex foster parents.
In the third literature stream, scholars examine the relevance of Christian churches to the emergence of Esping-Andersen's (1990) three-tiered typology of welfare state regimes in Europe (Manow, 2008(Manow, , 2018Van Kersbergen and Manow, 2009 (Kahl, 2009;Van Kersbergen and Manow, 2009). Roughly, the literature can be summarized as follows: while Catholicism and Reformed Protestantism rejected the establishment of the modern welfare state, Lutheranism largely supported it and became an integral part of the welfare institutions themselves. The rationale behind the confessions' differing attitudes can be attributed to political class coalitions (Manow and Van Kersbergen, 2009) and conflicting theological concepts of poor relief between Catholicism and Reformed Protestantism on one side, and Lutheranism on the other (Kahl, 2009). Research shows that these religious doctrines are shaping social policy until today (Morgan, 2009).
Linking these three literature streams, we investigate the extent and nature of Protestant and Catholic involvement in the provision of morality-based social services in Germany. Specifically, we argue that the extent of engagement can be attributed to opportunity structure, and that in order to understand the nature of engagement we need to take into account the moral doctrines of the respective Catholic and Protestant ROs. The paper thus combines insights from different studies on the link between religion and social policy-making, providing a unique glimpse of the actual nature of these services.
3. Un/connected Christian ROs' engagement, goal congruence and social doctrines We are interested in Christian ROs' engagement in morality-based social services with regard to both the extent and the nature of this engagement. Our arguments are based on the literature streams discussed above. Specifically, we argue that the institutional argument prevalent within the governance literature (first literature stream) is relevant for the extent of engagement, and that the doctrinal argument guiding the welfare state literature of FBOs (second and third literature stream) is a plausible explanation for the nature of Christian ROs' engagement.
First, concerning the extent of engagement, we rely on the literature stream concerned with explaining Christian ROs' influence on morality policies. This literature assumes that the opportunity structure provides ROs with crucial access points to the political arena. In particular, a favorable opportunity structure will facilitate ROs' impulses in the policy formulation and adoption phases. We argue that opportunity structure also affects religious governance at the policy implementation phase. There is research to back this claim. Euchner (2018), for instance, shows how variation in church-state relations between two German federal states has led to differences in the regulation of Islamic religious education in public schools. In detail, a close cooperation with Christian denominations in the field of public education made the integration of Islamic religious education as a regular subject less likely. In these states, such as Baden-Wuerttemberg, IRE is a model project that is run without the participation of the Muslim community. Hence, Muslims are confronted with an unfavorable opportunity structure in Baden-Wuerttemberg that prevents them from getting active at the implementation stage. In a similar vein, Göçmen (2013) shows that the historic church-state relation affects the institutionalization of religion in the welfare state and hence, leads to differences in FBOs' engagement in Sweden, France, Germany and Great Britain.
The importance of church-state relations is furthermore emphasized by Manow and Van Kersbergen (2009). In their book, they convincingly demonstrate that historic church-state cleavages led to the formation of religious parties, which structured the subsequent development of welfare regimes. In particular, where the church-state cleavage prevailed and Christian Democratic Parties were in government, churches served as cooperation partners in the construction of the public welfare infrastructure. Christian Democrats provide the church with access points to politics. In this way, ROs can present themselves as potential implementing actors of morality policies to the federal governments, where morality policies are usually specified. 1 The effect of lobbying is reinforced when a considerable share of the population formally adheres to the values and teachings of the RO, as popular support increases the RO's legitimation to act as public partner in the social sphere (Tatalovich and Schwartz, 2009). A strong mobilization potential moreover aids the impression that the RO can meaningfully provide a particular social service, as it can mobilize staff and volunteers (Fink, 2008). In short, a favorable opportunity structure-characterized by a close state-church relationship in welfare and education, easy access points to the political sphere via religious parties and a comparable large number of religious adherents-should facilitate religious governance engagement in morality policy implementation, compared with those systems in which the opportunity structure is less favorable.
Second, concerning the nature of engagement, we distinguish between unconnected and corresponding Christian engagement. Unconnected engagement covers a situation in which ROs offer morality-based social services but the engagement does not coincide with the legal intention. Connected engagement encompasses instances in which religious actors offer morality-based social services in line with the goals as formulated in the policy. We deduce a possible explanation for unconnected/ connected engagement from the literature on private governance and policy implementation. This literature has shown that the engagement of private actors in the provision of public services is a question of congruence of ideas, values and goals (Knill and Lehmkuhl, 2002;Considine and Lewis, 2003;Buffat, 2014), even more so if these goals are ambiguous (Thomann et al., 2016). We transfer this idea to ROs, assuming that the nature of services depends on the core moral values and the doctrinal salience they assign to a policy issue (see Figure 1). Specifically, we argue that the type of engagement can be explained by the goal (in)congruence between the legislation and the RO regarding a specific morality policy: goals can be congruent or incongruent (0/1), which in turn decreases or increases the RO's positional distance from the state, and thus its willingness to implement services on behalf of the state. In a situation where secular values clash with religious values and the legislative policy goal is not congruent with the RO's policy goal, the religious organization will be less inclined to implement policies on behalf of the state. Instead, the RO will revert to its own value logic of governance engagement (Thomann et al., 2016), which leads to unconnected forms of engagement.
To theoretically embed the overarching question of how moral doctrines structure the nature of engagement, we turn to the literature on the religious roots of welfare states. As mentioned earlier, scholars in this field have pointed out the pronounced differences in confessional social doctrines and the way they have influenced welfare development. One important argument made by Kahl (2009) is that confessional variations in how to reach salvation influenced how the respective confessional traditions defined their role in poverty relief: according to Catholic social doctrines, good works are defined as the only way to enter the state of grace, consequently, charity and poor relief are seen as important Catholic duties. In contrast, Lutheranism emphasizes that only by believing in God people qualify for salvation, and therefore poor relief is not seen as a Christian responsibility. This is why, unlike the Catholic Church, Lutheranism approved the establishment of the welfare state and voluntarily handed over church competencies to the public authorities (Kahl, 2009).
With regard to legal understanding, Catholic and Protestant belief vary as well. Körtner (2019), who focuses on the ecumenical social ethic in Germany, provides a very good overview in this regard (Körtner, 2019, 137). The Catholic Church distinguishes between the divine law and the human law. Human law has to conform to divine law, and human law per se has no legal independence. Consequently, Catholicism does not accept the existence of an independent secular law, but insists that such "worldly" law must be oriented towards the divine law. Divine law has two sources: the Bible and nature. In particular, natural law flows from the nature of the human being and the world, does not need to be justified in any other way, and defines moral law. In addition, since the Second Vatican Council, the Catholic Church has claimed to represent divine law outside as well as inside the church, and it demands that secular law be adjusted to conform to divine law (Körtner, 2019). As Körtner (2019, 138) points out, magisterial acts tend to support conservative interpretations of the last Vatican Council, which in turn causes the moral theology of the Catholic Church to be more conservative as well.
In contrast, Protestant, and particularly Lutheran, legal understanding distinguishes the worldly from the spiritual sphere (ibid). As earthly justice and divine justice are separate, so are earthly law and spiritual law. In particular, as predestination depends solely on faith and not on good works, Luther defined Evangelism as a promise of God's grace and mercy, not as a moral law. This means that compliance with legal rules does not lead to salvation, but faith does, and any sin has to be judged by God alone, and not through worldly justice (ibid). This allows Lutheran traditions to be much more open to different interpretations of moral ethics, as they are not said to be predefined by Evangelism.
We expect that the differences in theology regarding moral doctrines will structure the evidence of the nature of Christian ROs' engagement within morality-based social services. While Protestant organizations will find it easier to engage within the state's liberal framework, Catholic organizations will be reluctant to implement state policies, trying to win back authority over the moral discourse. Thus, we expect to see a tendency towards corresponding engagement on the part of Protestant ROs, while we expect unconnected engagement from Catholic ROs.

Comparing Catholic and Protestant ROs' engagement within different opportunity structures
We investigate whether there are differences between Protestant and Catholic provisions of three morality-based social services (i.e., services related to end-of-life care, abortion, and sex work) in Germany. Germany features a corporate welfare state regime with a long and stable tradition of third-sector involvement in social service provision (Göçmen, 2013). ROs are especially active, providing between one fifth and one quarter of all care services in Germany (Hien, 2019, 206). Germany is therefore an auspicious case for studying Christian ROs' engagement in morality-based social services.
We chose to compare ROs' engagement within two federal states: Berlin and Bavaria. The comparison holds promise for answering the question of the extent of religious governance engagement, as Berlin and Bavaria differ substantially in their cultural and institutional opportunity structures. First, in line with von Blumenthal (2009), we assume that the federal constitutions provide the basic idea of the role that religion and ROs ought to play in public life. Berlin's constitution hardly mentions religion, while the Bavarian constitution is full of references to Christian faith. Therefore, the church-state relation is close in Bavaria, where the constitution grants ROs with access to a range of public affairs, and distant in Berlin, where the constitution only acknowledges religious freedom (von Blumenthal, 2009, 128). Second, there is a political agent, i.e. the Christian Democrat Party in the Bavarian government, while Berlin is governed by a coalition between the social democrats, the left and the green party. Third, the number of Christian adherents in the population is low in Berlin, where only 23.3% are registered members of either the Catholic Church or the Evangelical Church in Germany, and high in Bavaria, where still 65.4% declare such a membership (Evangelische Kirche in Deutschland, 2020). As summarized in Table 1, the cultural and institutional opportunity structures are hence quite favorable for Christian ROs in Bavaria and less so for those in Berlin. We expect that this will affect the extent of ROs' engagement, which should be higher in Bavaria than in Berlin.
For explaining the differences in the nature of ROs' engagement, we compare Catholic and Protestant ROs associated with the Catholic Church and the Evangelical Church in Germany (Evangelische Kirche Deutschland [EKD]). The EKD joins Lutheran and reformed regional Protestant churches who hold largely progressive views on morality policy issues. In contrast, the Catholic Church has defended conservative moral views. For instance, the Vatican has recently rejected the possibility of blessings for homosexual couples.
We select abortion, sex work and end-of-life care as three typical morality policies that fall within the realms of classic social service provision. 2 The recognition of ROs as implementing actors of these policies takes place at the federal level. The policies moreover have a considerable conflictive potential and can be regarded as manifest morality policies (Knill, 2013). Manifest morality policies are issues that are regulated without implications of material loss or gain for certain groups. Instead, the focus is on individual values and beliefs. These values and beliefs are often interlinked with religious ideas and dogmas, and are therefore well suited for an investigation that is premised on the clash between secular (state) and religious values.
We base our research on an original data pool with information collected between 2017 and 2019. The actors under consideration in both federal states are the regional Protestant church EKBO (Evangelical Church Berlin-Brandenburg-Silesian Upper Lusatia) and the Catholic archdiocese in Berlin; the regional Protestant church ELKB (Evangelical Lutheran Church in Bavaria) and the four Catholic archdioceses in Bavaria; and their respective regional welfare organizations, Diakonie and Caritas. Forty-three semi-structured interviews were conducted in Germany, twelve of which involved street-level workers and leading representatives in Berlin and Bavaria respectively (see Table A1 in the appendix for a comprehensive list). These experts provide insights regarding the priority and salience of single morality issues, the related policy goals and the specific services offered. Moreover, they give accounts on the moral doctrines underlying the nature of these activities. All interviews have been audiotaped, transcribed and coded with the software Atlas.ti (Version 9).
The dependent variable-religious governance engagement-features two aspects: the extent and nature of engagement. For the extent of engagement, we look at the share of Christian religious service facilities (Protestant and Catholic) compared with other, non-Christian facilities (state and NGO) within the particular morality policy field. We include both services provided by the major welfare organizations, Diakonie (Ev.) and Caritas (Cath.), as well as by smaller Christian service organizations that are somehow associated with the respective Catholic or Protestant churches in Bavaria and Berlin. For the nature of engagement, we concentrate on the question of whether the engagement is in line with legal intentions (corresponding/unconnected). Here, we compare how ROs claim to provide these services with the guidelines in the legislative texts of the policies (instruments). We argue that the nature of the engagement can be explained by the goal (in)congruence between the legislation and the Christian RO. Again, the policy goal is deduced from the respective legislative texts, while the religious goal is identified in the interview material. We also use interview quotes to illustrate how goal (in)congruence links back to moral doctrines. Table 2 illustrates the share of Christian ROs' engagement associated with the two major churches in morality policy implementation in Berlin and Bavaria compared with other first, second or third sector providers (including ROs not associated with the two major churches). With regard to each morality policy field, our findings are as follows: (1) in end-of-life care-the extent of confessional engagement does not vary between Berlin and Bavaria neither for the Protestant nor Catholic ROs, indicating that opportunity structure provides no reasonable explanation. (2) Pregnancy counseling-the extent of confessional engagement does not vary between Berlin and Bavaria, i.e. the opportunity structure seems less relevant. (3) Sex work-the extent of Protestant ROs' engagement is higher in Berlin than in Bavaria; the opportunity structure has a counterintuitive effect. Table 3 illustrates the goals and instruments of state policy versus the goals and instruments of the religious community with regard to each morality policy. State policy goals and Protestant policy goals largely coincide, and so do the instruments applied, that is: the nature of the services. The Catholic policy goals, however, differ from those formulated in state policy, and so does the nature of the services offered. This is especially the case for abortion and sex work policy, while goals are relatively similar with regard to end-of-life care. In the following section, we explore the engagement of Christian ROs separately for each morality-based social service, and in greater depth, focusing on the variation of the extent of this engagement between Berlin and Bavaria, and variation of its nature between Catholic and Protestant organizations.

Implementing end-of-life care policies: corresponding engagement of ROs
In Germany, passive assisted dying-administering medication with the aim of relieving pain and accepting the possibility that the medication might cause the death of the patient-has been legal since 2009 (Preidel and Nebel, 2015, 56). Even though the Federal Constitutional Court ruled in the spring of 2020 that the prohibition of active assisted dying is unlawful, a respective law has so far not been passed by the German parliament. As a consequence, active assisted dying is still criminalized and the implementation of the rules on assisted dying are largely concerned with the conditions that facilitate the "natural way of dying" ( §216, §217 German Penal Code [dt. StGB]). In line with this approach, hospice and end-of-life care have recently been granted increased state subsidy (Hospice and End-of-life care Law [dt. Hospiz und Palliative Gesetz], Dez. 1, 20151, , BGBl. I p. 2114. With regard to the extent of engagement, we see that the share of Catholic stationary hospices (9% in Berlin versus 10% in Bavaria) and Protestant hospices (18% in  We are against active assisted dying. This is explicitly stated in the living will forms we offer patients. At the same time, we would like to strengthen the right of self-determination for patients; they should be able to say "I do not want that machine to keep me alive". (2DE) A representative of Caritas explains the Catholic position in a similar way, placing, however, a slightly different emphasis on a "natural end" rather than self-determination: We do not use the term "assisted dying", as it sends the wrong signal. In the center [of our work] stands the care for someone at the end of his or her life. One does not have to prolong a patient's life by technical and artificial methods; there is a natural end. (3DE) This natural end, according to Catholic belief, is not in the hands of human beings. Human beings are not qualified to take such decisions. Instead, Catholicism emphasizes the rule of God over human existence, as a Bavarian representative of the Catholic Church explains: God is the Lord over life and death. Helping suicide is not the way, in my opinion, because I myself also lay there with colic and said: "I would like to die." And the next day everything was different, the pain was gone. (1DE) As we have seen, in end-of-life care, the nature of engagement does not vary between Catholic and Protestant organizations. Both Catholic and Protestant ROs act in line with legal policy intentions by displaying a corresponding form of engagement, where the goal is to provide end-of-life care that facilitates the natural way of dying. Catholics strongly reject active assisted dying on the grounds of their doctrine that emphasizes divine power over questions of life and death. After the 2020 court ruling, the question of whether Protestant facilities should in future provide aid for assisted dying is already the subject of heated debate within the Protestant community (Zoch, 2021).

Implementing abortion policy: unconnected Catholic engagement and corresponding Protestant engagement
Federal law 34 codifies a restricted access to legal abortions based on a social indication. As a first pillar, the law prescribes obligatory counseling before an abortion, with the goal of protecting the unborn life. This "pregnancy conflict counseling" is mandatory for women considering an abortion within the first 14 weeks of their pregnancy; thereafter, abortion is only possible in the case of a medical or criminological indication. A second pillar of the Pregnancy Conflict Law is voluntary pregnancy counseling, provided to women who have questions about contraception, parental leave or child care. Pregnancy (conflict) counseling is offered in specialized centers that are run by local authorities and different third-sector organizations, among them Protestant and Catholic organizations.
Concerning the extent of engagement, we see that Protestant facilities constitute a higher share in Berlin than Bavaria (31% versus 9%). On the contrary, but slightly less markedly, Catholic facilities represent a higher share in Bavaria than Berlin (18% versus 7%). The overall share of religious counseling centers is however very similar (Berlin 38%, Bavaria 27%), with the share of religious counseling centers being slightly larger in Berlin than in Bavaria. Therefore, again, the expectation that the unfavorable religious opportunity structure in Berlin would hamper religious participation does not seem to be borne out in the case of pregnancy counseling.
Meanwhile, the nature of the services offered by each confession's ROs differs along confessional lines. Importantly, Catholic counseling centers are not implementing abortion policy to the fullest extent: when the law on pregnancy conflict was introduced in 1995, Pope John Paul II insisted that Catholic counseling centers would not participate. Catholic institutions should not facilitate access to abortion in any way; on the contrary, they should campaign for the abolition of legal abortion and protect the life of the unborn child. Therefore, Catholic pregnancy counseling centers in Germany had to abandon recognized pregnancy conflict counseling and they no longer take part in implementing the federal law on pregnancy termination (4DE). As an alternative, Catholic counseling centers offer general counseling for women in distress from pregnancy, while refusing to issue the certificate necessary for a legal abortion. A Catholic representative expresses support for the Pope's decision: I think it is consistent not to issue a counseling certificate. If one rejects abortion yet sees the misery of these women, then the Catholic Church should help alleviate the misery of women. (4DE) Another Catholic representative explains the doctrine behind this position: For us, the highest commandment is: humans are not masters of life and death. They are not allowed to say who has to die. And in this area, it happens that people come into situations where they basically need help to make a decision.
[…] The world is how it is. We cannot change this; we can only make the principles clear. These principles are violated every day. And to be there for the people after they have violated the principles, that is what we do. (1DE) In contrast, the Protestant ROs are actively engaged in pregnancy conflict counseling, issuing consultation certificates and offering help for the time before and after abortion. In this way, the Protestant ROs are important implementers of the law on pregnancy conflict; their engagement in the field of abortion corresponds with legal intentions. Although the Protestant moral doctrine, like the Catholic one, acknowledges the Christian commitment to protect the life of the unborn child, the fundamental attitude is that the life of the child can only be protected in participation with the mother, not in opposition to her. A representative of the Protestant Church in Berlin explains: On the one hand, of course, [there is] the woman's right to self-determination. One […] cannot protect the future life against the will of the mother and her interests. [Their lives] are already symbiotic and belong to each other. But on the other hand-and the Federal Constitutional Court has also put this very firmly-the state must also protect life in the making.
[…] You have to weigh these perspectives against each other. (6DE) In sum, we see that the congruence in policy goals between the Protestant church and the national policy has stimulated a corresponding governance engagement of Protestant ROs within the field of abortion. On the other hand, disapproval of the law by the Catholic Church leads to unconnected, and sometimes even contesting, forms of engagement of Catholic centers, which only implement a part of the law.

Social services related to sex work policy: unconnected Catholic engagement and corresponding Protestant engagement
Sex work has been a legal and recognized profession in Germany since 2002. 5 In 2016, the law on the protection of prostitutes 6 was passed, requiring sex workers to register with the local authorities and conduct health checks on a regular basis. The goal is both to control the industry and to protect sex workers from economic exploitation. While the registration and health checks are conducted by the local municipality, non-state sex work centers can be consulted by the state administration for individual cases (cf. ProstSchG §8). These non-state counseling centers, among them religious ones, are expected to be in direct contact with sex workers, and therefore to be well informed about their needs.
Overall, the number of counseling centers for sex workers is noticeably smaller than in the two previous policy areas. Between Berlin and Bavaria, the share of these counseling centers varies substantively for the Protestant ROs (0% versus 50%) with both of the Bavarian centers being located in Munich. Meanwhile, Catholic ROs constitute a share of 25% in each federal state. It follows, in turn, that the highest share of counseling centers in Berlin belongs to the non-confessional third sector (75%). Here, opportunity structure seems to be a reasonable explanation. Additionally, sex work is regulated differently at the local level, with Berlin allowing street prostitution and Munich restricting city areas so that sex work has to take place at the outskirts of the city. In Berlin, this might have advantaged the formation of third-sector grass-roots organizations which, as well as providing care services, are also active at the political level, opposing discrimination against sex workers and demanding their professional recognition.
The difference in the nature of these services lies in their confessional affiliation. The two anti-trafficking projects associated with the Catholic Church, Jadwiga (Bavaria) and In Via (Berlin), are both dedicated to helping trafficked women who are forced into sex work. In both states, ROs associated with the Catholic Church are therefore exclusively tailoring their services to the victims of forced prostitution, and they do not consider unforced sex workers as their target group: Of course, there is voluntary sex work, but that is not the part that interests us, and these people don't need counseling either. They are settled, earn money from it, and that's it. They don't need protection. (4DE) Focusing on human trafficking, the Catholic ROs are engaged in an important policy field, as the prostitution protection law was, inter alia, targeted at reducing forced prostitution and protecting sex workers from economic exploitation. However, the Catholic focus on human trafficking in the field of sex work can also be linked to the rejection of the Catholic Church of the legal recognition of sex work as a "normal" profession through the prostitution law adopted in 2001 (Euchner, 2015). Generally speaking, the Catholic Church supports the abolition of sex work and the criminalization of the purchase of sex. The position is grounded in the Catholic ideal that links sex to marriage.
Sex work does not help people; it is not good for humans. It is much better to go together on a common life journey and to find fulfillment in sexuality within marriage. (4DE) Catholic sexual morals are closely bound to the institution of marriage. Within marriage, sex serves two causes: a consolidation of the relationship, and reproduction. Sex work serves neither of the two causes. Quite the contrary, sex work is perceived to deconsolidate marriage as it often involves adultery.
In the view of the Catholic Church, the connection of man and woman takes place within marriage […]. Thou shalt not desire the wife of your close ones, that is one of the Ten Commandments, and prostitution is one step further.
In contrast to the Catholic shelters specializing in human trafficking, the counseling centers run by the Diakonie also provide information and shelter for regular sex workers. Accordingly, many Protestant ROs hold the basic assumption that sex work does not necessarily contradict human dignity.
Concerning prostitution, we have made it very clear that we are not getting involved in the moral debate at all. […] Our guiding question has always been: how can we do the best we can for these women without judging them?
As this quote illustrates, Protestant ROs do not perceive sex work as involuntary or wrong per se. The main principle is that activities that are not based on a sex worker's free will have to be prevented.
First of all, it is important that those concerned can discover their own attitude, so that they themselves know what they want; for this we have to create the framework conditions. The point is that the person concerned is supported in finding his or her own way.
[…] This must be the goal, leading people out of dependency relationships. (5DE) For protestant counseling in sex work, we thus determine a corresponding approach regarding the nature of engagement. Here, counseling aims at the empowerment of sex workers by raising their awareness of their rights and obligations, thereby helping to protect them from exploitation (3DE). Catholic centers, in turn, concentrate on counseling for the victims of human trafficking, thereby engaging only partly in line with the legal intentions as laid out in the prostitution protection law, which defines all persons providing sexual services as its target group.

Conclusion
While there is research on religious communities' engagement in the provision of classical social policies (e.g., child care, elderly care), we know very little about ROs' involvement in the provision of social services related to morality issues (e.g., counseling in terms of abortion, sex work or end-of-life care). This paper contributes to filling this research gap through an innovative within-country comparison between the German federal states Berlin and Bavaria, and a novel theoretical framework that highlights the role of moral doctrine as a driver of ROs' social engagement. We discover that the extent of ROs' engagement varies unsystematically between Berlin and Bavaria, indicating that classical explanatory factors, such as opportunity structure, cannot explain the variation in the extent of engagement. However, the nature of the engagement (corresponding and unconnected) points to systematic differences between Catholic and Protestant denominations, which we explain by divergent moral doctrines that either increase or diminish the goal congruence between ROs and policy-makers. For instance, in terms of services related to abortion or sex work policy, Catholic ROs act on behalf of the state but do not follow legal intentions, due to the moral doctrines guiding their behavior. Meanwhile, in the field of end-of-life care, Catholic and Protestant ROs act on behalf of the state and in line with the state's intention, as they support the state's policy goal to provide a "natural way of dying".
As we show in this study, the positions of the Protestant ROs and the state law largely coincide, while Catholic positions stand in much stronger contrast to state policy aims. Such differences in the response of the Catholic and the Protestant organizations can be traced back to these confessions' moral doctrines: Protestantism differentiates between worldly and divine justice, and emphasizes that the highest authority stems from scripture, the Bible itself. Catholicism, by contrast, demands that secular law corresponds to divine law, which is entailed both by scripture and Catholic tradition, as interpreted by the Catholic teaching authority (the Pope and the bishops).
Can our findings be transferred to other settings? We are aware that Germany, with its cooperative church-state system, is a specific case. Nonetheless, we suggest that the country can serve as a laboratory for studying the influence of opportunity structure on Christian ROs due to its federal structure and the large variation in religious characteristics between its regions (e.g., Berlin and Bavaria). Even though we do not find systematic variance of ROs' extent of engagement between Berlin and Bavaria, it would be interesting to see a reevaluation of the argument in other states that have varying opportunity structures for church engagement at the regional level (such as Switzerland, for instance). The U.S. would equally offer a highly interesting arena to explore our research questions. The diverse religious landscape across and within states would allow for additional comparisons between religious communities (considering the rise of Evangelical churches). Moreover, studiying the U.S. would grant insights in a liberal welfare state setting, where additional support of non-state actors is desperately needed for the poor and disadvantaged people.