Introduction

With ageing populations and changing gender roles, Latin America is undergoing demographic and social changes that particularly affect the division of care responsibilities between the state and the family. Within Latin America, Uruguay is prominent for being the first country to introduce an integrated care system in 2015 that covers not only children, but also people with disabilities, older adults and caregivers. While existing care policies in Latin America predominantly focus on children, this integrated type of system is also quite unique compared to long-term care (LTC) systems in Europe or other parts of the world, as most public LTC systems have a focus on older adults and were developed separately from care schemes for children, or even people with disabilities (Bettio & Plantenga, 2004; Ranci & Pavolini, 2013; Theobald & Kern, 2011). In addition, the system in Uruguay emerged after a long participatory process involving national actors from civil society, academia, politics, and administration, but also international organizations (IOs), in particular United Nations (UN) agencies, such as UN Women, the United Nations Population Fund (UNFPA), and others.

While existing studies provide some indication of the national factors that led to the introduction of the system in Uruguay, such as the influence of feminist movements and academics or the role of a progressive government (Aguirre, 2014; Aguirre et al., 2014), little is known about the role of IOs in the development of the new system, even though they were substantially involved. How did the interaction between national and international actors emerge during the process of establishing a national care system? And how did this cooperation shape the creation of the care system?

To answer these questions empirically, I use interview data and document analysis to first, explore which elements led to the cooperation between IOs and national actors in the policy process between 2005 and 2015, and second, to analyze how this interaction shaped the system’s introduction. While discussions about care policies began between 2005 and 2010 and were mainly driven by women’s organizations and academics, the Broad Front (Frente Amplio; FA) government put the issue on the agenda before the 2009 elections and finally passed the law on National Integrated Care System (Sistema Nacional Integrado de Cuidados; SNIC) in 2015. In examining the role of IOs in these processes, this study draws on Sabatier and colleagues’ Advocacy Coalition Framework (ACF) (Jenkins-Smith et al., 2018; Sabatier, 1988, 1998; Sabatier & Weible, 2007), which addresses questions of policy change and coalition formation. As LTC is an emerging policy field in Uruguay (and the region) I use their notion of a nascent subsystem, which is a policy field in the process of being established. From a theoretical perspective, the case therefore offers a better understanding of IOs’ role in advocacy coalitions, in particular in emerging subsystems where the cooperation between different actors is not yet as established.

The article is structured as follows: first, I introduce the theoretic framework by giving a brief introduction to the ACF, then review the literature on nascent subsystems and propose a threefold distinction to analyze the role of IOs in advocacy coalitions. Second, I describe data collection and the methods applied. Third, I briefly outline the case and the actors involved. Fourth, I present the analysis by first looking into the interaction between IOs and national actors and then turn to the question how this cooperation contributed to the emergence of a new subsystem. After a discussion of the empirical analysis, I finally address limitations of the study and suggest avenues for further research.

The advocacy coalition framework

The ACF was developed by Sabatier and Jenkins-Smith in the 1990s and addresses questions related to coalitions, learning, and policy change (Sabatier, 1988; Sabatier & Jenkins-Smith, 1993, 1999; Sabatier & Weible, 2007). The ACF assumes that coalitions within a subsystem try to impact the policy process through interaction and cooperation. Actors within a coalition share similar beliefs regarding the policy field (policy core beliefs) and more normative general assumptions about human nature (deep core beliefs). Based on these shared beliefs, they cooperate with each other to influence the subsystem toward their preferred policy solution. In contrast to the third level of beliefs (secondary beliefs), which are “relatively narrow in scope” (Sabatier & Weible, 2007, p. 197) and refer to specific details of the policy, the deep core and the policy core beliefs rarely change within coalitions. The framework explains policy change through policy-oriented learning, external events, internal events, and negotiated agreement (Sabatier & Weible, 2007). However, in this paper I focus on coalition behavior and coordination patterns, as well as beliefs in nascent subsystems, which I will address next.

Nascent subsystems

Sabatier and Jenkins-Smith (1999) distinguish between mature subsystems, which have been in existence for a decade or more, and nascent subsystems, which are in the process of formation. In a nascent subsystem an issue has only recently emerged on the public agenda, with little history of public policy outputs, little or no serious consideration in public decision-making forums, and where advocates have only recently become active (Stritch, 2015, p. 438). These nascent subsystems can be characterized by fluidity of beliefs and patterns of coordination (Sabatier & Jenkins-Smith, 1999; Weible et al., 2020). The literature on nascent subsystems is still limited, but authors studying them reflect mostly on the process of coalition formation and belief patterns (e.g. Stritch, 2015; Wiedemann & Ingold, 2024) or the process of subsystem emergence (Ingold et al., 2017).

In nascent subsystems, coalition formation is an important process because actors start to cooperate on the basis of shared beliefs in order to influence policy outputs. Stritch (2015) shows in his analysis on trade union disclosure that while cooperation across belief systems is possible in nascent subsystem at “relatively low levels” (p. 439), closer cooperation is limited to actors sharing congruent beliefs about a policy. Similarly, in a recent analysis of a nascent subsystem, Wiedemann and Ingold (2024) find that some “belief clusters” (p. 52) are more mature than others and that these clusters are better organized, although in general, beliefs are not so well defined and policy core beliefs are more focused on problem definition and understanding. When it comes to the topic of conflict between coalitions with different beliefs, previous research suggests that nascent subsystems in a collaborative setting are likely to receive general support with regard to policy issues because the threats and the political positions are relatively unclear, only when subsystems mature could conflict between coalitions arise (Fidelman et al., 2014; Nohrstedt & Olofsson, 2016). Therefore, the way a new policy field is established matters, as well as the role of potential “policy broker[s]” (Fidelman et al., 2014, p. 127), who can negotiate between coalitions. The characteristics of a nascent subsystem can, however, also explain the absence of opposing coalitions if the issue has been dominated by a strong coalition and potential risks of the new policy are still unclear (Olofsson et al., 2018).

With regard to the process of policy development or the emergence of a new subsystem itself, Ingold et al. (2017) provide some hypotheses for explaining the formation of such a subsystem. They test three hypotheses that deal with (1) the existence of a similar belief system, (2) the role of actors’ reputation as leaders, and (3) mutual knowledge as a basis for agreement on a policy design (p. 443). They find that in particular the role of powerful leaders and previous cooperation between actors affect the policy process.

This article will look into both the formation of a coalition based on similar beliefs and the process of creating a new subsystem through a dominant coalition. While most research on nascent subsystems deals with rather technical topics from environmental policies where the idea of a “new subsystem” is based on technical innovations, I hope to contribute by exploring the emergence of a new field of social policy, namely LTC.

The role of international organizations in advocacy coalitions

To analyze the role of international actors such as UN agencies, I propose a threefold distinction based on the literature on advocacy coalitions. First, IOs can trigger policy change more indirectly through international events, norms, and ideas; second, they can be directly incorporated into national coalitions; and third, they can act as policy brokers between opposing coalitions.

First, their role can be described more on the general level, where international ideas or norms are used by national actors in subsystem contexts or where international events and debates can trigger policy change as external factors (see e.g. Elliott & Schlaepfer, 2001). For instance, Kefeli et al. (2023) argue in their study on environmental policy in Uruguay that international salience was an important factor for strengthening a minor coalition in problematizing a topic. In a similar vein, Schabbach and Da Rosa (2021) conclude in their analysis of education policies in Brazil that international ideas and agreements contributed to policy change in the country.

Second, IOs can be considered as actors incorporated into national coalitions. Especially in countries outside of Europe or North America, the presence of IOs in national policy-making is an important factor, as studies in the regional contexts of Africa, Asia or Latin America show (Ainuson, 2009; e.g. Harris, 2019; Hsu, 2005; Tomazini, 2019). In the case of Zambian nutrition policy, Harris (2019) shows that certain international ideas were transferred to the national policy processes by IOs through the formation of an advocacy coalition. In a study on cash transfer programs in Brazil and Mexico, Tomazini (2019) finds that UN agencies were part of a broad coalition promoting cash transfer to alleviate poverty, and, similarly, Ainuson (2009) reports that international actorsFootnote 1 were involved in water policies in Ghana by financing activities of national stakeholders.

Third, IOs can be viewed as “policy brokers” (Sabatier, 1998, p. 104), meaning actors who are concerned in finding compromises between conflicting coalitions. The literature on IOs as “international brokers” (Osei-Kojo et al., 2022, p. 191) is still scant, with only limited indications that IOs can be involved in policy processes to help advocacy coalitions in reaching an agreement. One example would be Greve’s (2019) analysis of IOs’ role in public-private partnership policies, where IOs act as part of coalitions, but also as brokers, especially through offering “practical solutions and concrete recommendations to governments” (p. 208).

In summary, research on the specific role of IOs in advocacy coalitions is still limited, although there is some evidence of their role at the normative level and of their direct participation in national policy processes, either as actors integrated into coalitions at the national level or as policy brokers. By analyzing the role of IOs in coalition building within an emerging subsystem, I hope to provide a better understanding of their interaction with national actors in policy-making and explore their contribution to the creation of a new policy field.

Data collection and analysis

This study relies on three main data sources: semi-structured interviews with key actors who participated in the policy process or who have expert knowledge on the development of the care system; minutes from parliament and the senate; and documents from the organizations and actors involved.

The initial selection of interviewees was based on a review of literature on the SNIC and on documentation from workshops or seminars listing the actors. To identify the main actor groups to be interviewed, I relied on Aguirre’s (2014) distinction between political actors, academia, social organizations, and international cooperation (p. 802). I also used the distinction to construct group-specific interview guides, which were structured around the different phases of the policy cycle such as problem definition, agenda setting, policy formulation, and implementation. In terms of content, I focused on interaction and cooperation between the different actors throughout the whole process and on identifying the main actors or events that enabled the development of the SNIC. I interviewed 11 experts, covering all four actor groups. The majority of the interviews were conducted in-person during a field trip in Montevideo in November 2022. Two interviews were conducted in December 2022 and January 2023 online. The interviews lasted between one and two hours and were conducted in Spanish.Footnote 2 The audio files were transcribed and then analyzed using MAXQDA software.

The second set of sources was minutes from parliamentary debates between 2009 and 2015.Footnote 3 The minutes were used to analyze the opinions of different political parties regarding the care policy and thus helped to broaden the interviewees’ views. The minutes were coded and analyzed in the same way as the interviews.

The third set of sources was documents produced by IOs; national policy documents such as strategy papers, policy proposals, and documentation; research documents produced by the university and civil society organizations; as well as party programs.Footnote 4 The goal of the document analysis was to complement the analysis of interviews and minutes and provide a more comprehensive picture of events, actors, and processes.

The analysis of minutes and interviews was based on Braun and Clarke’s (2021, 2022; Braun et al., 2022) thematic analysis approach, which is a method for systematically identifying, organizing, and offering insight into patterns of meaning or themes across the data. Coding was applied deductively as a first step, based on concepts from the ACF literature, such as actors’ beliefs or cooperation patterns, but in a second step codes were also generated inductively. In an iterative process of coding, themes were derived from the data following a three step approach of (a) generating themes, (b) reviewing themes, and (c) defining and naming themes (Braun et al., 2022), as detailed in Fig. 1.

Fig. 1
figure 1

The data analysis process. Source: Author’s elaboration, based on (Braun & Clarke, 2022; Braun et al., 2022)

Full size image

The national care system in Uruguay

In this article I focus on the development of the SNIC in Uruguay, which was a milestone in the country’s social policy agenda, as it introduced a public care system for children, older adults, people with disabilities and caregivers at the national level. Furthermore, as the first country in the region to introduce a distinct LTC system,Footnote 5 Uruguay also contributed to strengthening the regional discourse on care policies within Latin America since the country is perceived as a “pioneer in the development of social protection” (Matus-Lopez & Cid Pedraza, 2016). The SNIC provides specific benefits for four groups: children aged 0–3, people with disabilities, older adults (65+), and caregivers. While the benefits for children are mainly an expansion of existing services, benefits for older adults and people with disabilities include a subsidy for a personal assistant (persons aged 80 + and 0–29), tele assistance (70+), day centers (65+) and residential care (65+).Footnote 6 The system’s financing is based on taxes as well as on co-payments by its users and the main administrative bodies are the Care Secretariat, the National Care Board and the Advisory Committee. Figure 2 provides an overview on the benefits offered by the system.

Fig. 2
figure 2

Benefit structure of the SNIC when it was adopted by law 19.353. Source: Author’s elaboration based on Amarante et al. (2019), Matus-Lopez and Cid Pedraza (2016), Ministerio de Desarrollo Social [MIDES] (2014, 2015) and ONU Mujeres (2019)

Full size image

The development of the system was shaped by several general developments at the national level, notably the change of government in 2005 and the involvement of the women’s movement in the debate.

The change of government in 2005, when the Broad Front (Frente Amplio; FA) government was elected, represented a policy shift from previous governments led by “center-right” parties to a “leftist political party” (Bentancur & Busquets, 2019, p. 137). The elections in Uruguay were part of a general trend in the region, the so called ‘Pink Tide’,Footnote 7 which describes the rise of left-wing governments in several Latin American countries in the early 21st century as a reaction to neoliberal tendencies in the past (see Grugel & Fontana, 2019). In the first period of government under the presidency of Tabaré Vázquez (2005–2010), the care topic was not very prominent within politics, as the government introduced reforms in other sectors such as health, social security or taxation (Batthyány et al., 2013c). In this phase the care issue was raised by civil society organizations and academic actors, but also by female politicians, and it gained momentum towards the end of Tabaré Vázquez’s presidency prior to the elections in 2009 (Aguirre et al., 2014; Filgueira et al., 2011; Friedman & Tabbush, 2018). The care topic became prominent in FA’s election campaign, with the party calling for an “integrated care system […] that allows for meeting the needs of family members at different ages and in different situations” (Frente Amplio [FA], 2008, p. 73, own translation). In the second FA government under the presidency of José Mujica (2010–2015), the political debate intensified. In 2010 the government established an interinstitutional working group consisting of different ministries and other national institutions.Footnote 8 In the working group, led by the Ministry of Social Development (MIDES), technical and political decisions on the creation of the SNIC were made (Aguirre & Ferrari, 2014a, 20–24). In 2012 the working group presented a proposal for the care system, which was approved in September 2012 by the Social CabinetFootnote 9 (Aguirre, 2014). The proposal contained the main objectives and principles for the new policy, the groups covered by the system, a definition of care dependency and its measurement, and proposals on financing and responsibilities (Grupo de Trabajo Interinstitucional, & Consejo Nacional de Política Social, 2012). The legal framework (Law No. 19.353) of the care system was then finally adopted by parliament in 2015 following Tabaré Vázquez’s election victory after he put the SNIC at the center of his presidential campaign (Amarante et al., 2019).

One driving factor for the agenda was the women’s movement and their engagement in collaboration with other actors in the political debates (Aguirre et al., 2014; Esquivel, 2017; Johnson et al., 2018). As care work is mainly provided informally by women, the call for political action was pushed by civil society organizations, in particular by a small feminist organization called “Gender and Family Network” (Red Género y Familia), which was founded in 1994 and since then addressed topics related to gender, social policies and care (Fassler, 2009). The network organized several workshops and projects to raise awareness and to address the issue of care with policymakers in 2008. In 2013 the “Pro-Care Network” (Red pro Cuidados) was established. The organization incorporated actors from the Gender and Family Network and its objective was to facilitate the exchange between individuals, organizations, and institutions; and to be able to advocate more strongly for a public care policy (Aguirre & Ferrari, 2014a). Another important actor was the University of the Republic (Universidad de la República), and within it the social sciences department and their contributions on gender-related research. Their time use surveys in particular had an important impact on the public perception of care, since they highlighted the issue of unpaid female work which was then used by academics and civil society organizations to argue the need for a public care system (Aguirre et al., 2014, p. 48).

While these national actors had an important role in establishing the SNIC, other actors were also involved in the discussions: international actors and in particular UN agencies such as UNFPA, UN Women, United Nations Development Programme (UNDP), Economic Commission of Latin America and the Caribbean (ECLAC), and others. They were involved directly through the generation of knowledge, the provision of technical assistance by means of consultancies and studies for the discussion, and financing the implementation of projects, seminars, and other formats for exchange (Aguirre & Ferrari, 2014a, p. 29). Nevertheless, open questions remain about how the cooperation between national and international actors developed and how this affected the SNIC.

International actors and national policies – the introduction of a novel care system

Turning to the analysis of the introduction of the SNIC, I focus first on the process of coalition building – starting with national actors and then looking into the role of IOs in the coalition. Second, I will present how the coalition of national and international actors shaped the debate and the development of the system.

Coalition building by national actors

From the perspective of national actors there was a clear coalition of academic actors (university and research institutions), civil society actors (mainly women’s organizations), and the government (politicians and administration). They were pushing for the care system from different directions but shared some broad beliefs about which principles should undergird the new system. In the early phase prior to 2010, women’s organizations and academic actors involved in gender-related research cooperated with each other in particular, and also maintained some personal connections on the individual level. The cooperation was based on the general idea of gender equality and women’s rights (deep core beliefs) and the need for public policies that would rearrange responsibility for (female) care work (policy beliefs), or as one interviewee reflected on the role of different actors:

And then I think that each of those who were there, especially the most involved actors who invested work, had an important role, because those who were there were feminists engaged in the issues, that is, comrades, so to speak. (Interview 2)Footnote 10

The government started to bring the issue onto its agenda due to the lobbying of these feminist networks, which were not only outside the political system (civil society and academic actors), but also members of it, e.g. feminist politicians, such as the “bancada femininas” (“female bench” (Interview 6), which is a group of female politicians from different parties who work together on topics related to gender and women’s rights (see Johnson et al., 2018, p. 58). The core beliefs on gender equality and co-responsibility were present in these groups as a former member of the administration points out: “but there is a very strong feminist vision of the need to achieve a scheme of co-responsibility that goes far beyond services, that is, it has to do with a cultural change” (Interview 1).

In terms of who was included in the care coalition, interest groups representing people with disabilities joined the coalition, but also brought their own ideas into the discussion. One important element that this group sought to include in system development was the personal assistant program, which is a service that was established in 2010 by law before the SNIC but was then integrated into the new scheme (Angulo & Mancebo, 2017). However, there were different foci, as the disability organizations emphasized other aspects related to their own needs:

This, and that is why it is, it is more difficult with the disability organizations, because what was achieved there was the formation of an alliance for disability, but they were, they are many, very varied, with very different problems. (Interview 2)

The disability organizations therefore engaged mainly to further their own interests, and in the later process after the law was passed, and especially after 2020 when the government changed and the institutionalized meetings of the advisory care committee became less frequent, these organizations were less involved in the coalition (Interview 2). In contrast to the disability organizations, there were also other groups who did not participate actively in the coalition at all, such as the workers’ union PIT CNTFootnote 11 or the group representing the interests of pensioners, even though the coalition tried to integrate them (Interviews 2 and 8).

In addition to sharing similar beliefs within the care coalition, there were also close links on the personal level. Actors often had overlapping affiliations because they were working “either through the university or through a social organization, or through a political organization”, and sometimes they were “the same person, in three places, promoting more or less the same ideas” (Interview 1).

Summing up, the coalition of these national actors was based on personal relationships and overlapping affiliations as well as shared policy and deep core beliefs,Footnote 12 which are presented in Table 1.

Table 1 Belief system of the care coalition – national actors (academic actors, civil society (women’s organizations), government (politicians, administration))
Full size table

In terms of opposition to the care coalition, there was no strong coalition against the proposed system, as the topic itself was still new. Based on the interviews it seems other political parties did not develop a clear set of policy beliefs:

The rest of the parties had no idea about this and quickly included a system for care in their platforms, in their programs, but there was no conception or clarity on the part of the opposition parties. (Interview 5)

When looking into the parliamentary debates on the care system, the other parties did not argue against having a system, on the contrary they highlighted the need for such a system, as this quote from a representative of one of the two main opposition parties, the National Party (Partido Nacional), shows:

We agree that it is necessary to establish a universal care policy that includes the elderly, people with different types of disabilities, as well as children, because the care and support received will benefit their learning and development capacity. (CR_2010_05_18, p. 18)Footnote 13

However, the opposition parties’ view on gender and caregiving differed from that of the care coalition. As the discussion evolved, opposing ideas on certain elements of the system became visible, such as introducing a cash benefit for informal caregivers instead of services (Interview 5). Even though the other parties were not against the system as such, some of them held different views about the gender perspective and the role of the family:

We personally understand that no one better than the family itself deserves to have the recognition for their self-sacrifice and the love that they permanently give in the care of their closest relatives, and with some type of remuneration that helps in the fulfillment of this noble task. (CR_2010_05_18, p. 18)

In the final debate on the adoption of the law, members of the opposition parties stressed their concern that the system might contribute to “family disintegration” (CR_2015_11_18, p. 69) due to its emphasis on offering services to care-dependent people outside the family. They argued that

This National Care System, […] weakens the basis and foundation of our society. There is not the slightest motivation for the family to be primarily responsible for the care of its member in a situation of dependency. (CR_2015_11_18, p. 70).

There was thus a very clear distinction between the care coalition and some members of the opposition with regard to ideas on the role of the family and its role in caregiving. Nevertheless, the law was adopted in the end, unanimous in the senate and by majority in parliamentFootnote 14 (Lissidini, 2016, p. 110).

Presence of IOs in the advocacy coalition

Turning now to the role of IOs, the analysis revealed two important aspects of their presence in the care coalition: (1) relationships; and (2) similar beliefs. In the following section I will elaborate on each theme.

Institutional and personal relationships

Examining the development of the care discussion, the aspect of relationships has two dimensions: previous cooperation; and institutional and personal overlaps. The first relates to previous cooperation or contact in other fields, projects, or networks where national and international actors participated. When looking into each national actor’s cooperation with IOs, the relationship between IOs and women’s organizations became visible on gender-related topics as one interviewee from an IO explained:

First, the issues of gender violence become quite clear. Then, there are issues of sexual and reproductive health where you know the discussion about abortion, possibly, everything where there is an alliance between the feminist movement and the multilateral world, the United Nations Population Fund. (Interview 11)

This cooperation, here between UNFPA and feminist networks, was mainly based on financial support, as the IO supported research, workshops, and projects by the organizations related to gender (Interview 8). As the care topic gained importance after 2005, their collaboration intensified, and the women’s organizations carried out dialogue round tables in 2008 and 2009, which were supported financially by UNFPA (Interview 8). It is noteworthy that IOs therefore started to support the idea of care policies very early on, even though the issue was not even prominently addressed at the government level, or as one IO representative expressed it: “We had a project that nobody put anything on” (Interview 10).

In cooperation with the university and the government, IOs were also actively involved in the time use surveys. The first such survey in Uruguay was conducted by the university in 2003 and covered only Montevideo and the metropolitan area. In 2007 the first national time use survey was realized and it contributed significantly to quantifying the unpaid work of female caregivers. It was conducted by the statistical office and the university, with financial support from UNFPA (Aguirre & Ferrari, 2014b; Batthyány & Scuro, 2010). In 2013 a second national survey was conducted with the participation of the same actors (Batthyány et al., 2017). UN Women was also active in the processes of realizing the time use surveys, as they collaborated in particular with the National Women’s Institute (InMujeres), as an interviewee from the IO points out:

InMujeres is part of the National Care Board and works on the issue of care. So, for example, we did all the time use surveys with InMujeres and the National Institute of Statistics, but it was my collaboration with InMujeres. It is my main cooperation partner. (Interview 10)

The second dimension of these relationships is similar to that between national actors because people working for IOs also had overlapping affiliations and there were close links between academic actors, the government, and IOs. First, some of the actors working for IOs had worked throughout their career for national institutions such as the university or the government, and also several national actors worked as consultants for IOs (e.g. Interview 2 and 11). Second, national actors were integrated throughout the whole process in publications by ECLAC, UNFPA, UNDP, or UN Women. There were two types of documents: One type was used more within the country, and IOs published them jointly with national actors and national institutions such as the government, the university, or civil society (e.g. MIDES, 2009; Rico, 2011). The other type was produced more for IOs’ regional and international activities, such as publications from ECLAC but with contributions from Uruguayan actors (e.g. Aguirre et al., 2005; Batthyány et al., 2013a). Therefore, as one interviewee points out, the topic of care on the IOs’ agendas was to some extent even influenced by these national actors: “[university actors] brought the issue of care into ECLAC. They were like the promoters of bringing it into ECLAC.” (Interview 6). On the other hand, the publications by UN agencies at the regional level also helped the coalition to gain recognition internationally, but also nationally:

It is fundamental, […] because it is also an important actor in generating dissemination, generating information. Having an activity that has the backing of an international organization makes it heard by the outside world, but it is also important for the political positions inside the country. (Interview 7)

So, these relationships on the personal and institutional level on the one hand, and previous collaboration in projects, publications, and other formats on the other hand, contributed greatly to the cooperation between IOs and national actors in the process of developing the care system.

Shared belief system

When looking at the specific IOs that were involved in the process, the two that stand out are UN Women and UNFPA. The strong role of UN Women in particular reflects the presence of the gender perspective in the coalition as well as collaboration based on shared values and beliefs. The idea of a care system with a strong gender perspective was crucial for the topic entering the public agenda as well as for how the system was ultimately designed. Even though during the policy formulation process specific elements, such as groups covered or which benefits to establish as well as financing and regulation issues, were discussed, and here there were also controversial views within the coalition (Interview 7), the basic understanding of the care system that would primarily support the caregivers, and here specifically women, was a shared vision of IOs and national coalition members, as an interviewee from UN Women argues: “more than anything else my line of work from UN Women is how the document is used, the care system: the opportunity for women’s economic empowerment” (Interview 10). The argument of reducing unpaid care work was therefore also in line with the idea of improving women’s working trajectories and women’s formal employment, which was emphasized by national and international actors (Batthyány et al., 2013b; MIDES, 2014).

There were also more abstract synergies in terms of ideas between IOs, the women’s organizations, and academic actors (Interview 4). These synergies were related to certain events that happened at the same time, such as the Conference on Women in 2007 organized by ECLAC, which then also contributed to collaboration and exchange of ideas between different actors, as an actor from civil society mentions:

Conceptual ideological, right? Ideas. […] The Quito meeting, for example, was very important and it was one of the things that I believe were coincidences, those synchronicities that appear later, the needs and our approach. (Interview 8).

In addition to the gender perspective, another important element formed the basis of the SNIC, which was the rights-based approach. A care system based on a legal framework that guarantees the right to care and establishes the relationship between caregivers and care recipients was a shared vision between national and international actors, as a representative of an IO points out:

I believe that this is part of the vision of the right to care, the right to care and to be cared for. And obviously care has to do with the life cycle of people. So, this comprehensive or life-cycle spirit is based precisely on the concept of understanding care as a basic right for all people. (Interview 10)

Summing up, the coalition had shared beliefs, in particular related to gender and rights as a foundation for establishing a public care system. Cooperation between different actors was therefore based on these beliefs, which were then also translated in the specific law that was ultimately adopted, as I will demonstrate in the following section.

A coalition of national and international actors shaping the subsystem

I turn now to the questions of how this coalition of national and international actors contributed to the establishment of the new subsystem, understood here broadly as a public care system, and how the coalition shaped the specific elements of the system, e.g. services, groups covered, and its overall organization.

As already mentioned, one of the key actions that paved the way for the political process of establishing a care system was the time use surveys. They described and quantified the issue of unpaid female work that the coalition was problematizing, helping the issue to gain recognition from political actors:

It gave data, didn’t it? That quantified what we all knew. […] in this country, what doesn’t have a number is not true. You can’t have knowledge if it doesn’t translate into numbers. I believe that the time use surveys allowed us to have elements to be able to argue more forcefully with the different powers. The executive power and the legislative power, above all. (Interview 8)

The discussion about the types of benefits was primarily characterized by support for in-kind benefits. The introduction of a cash benefit was rejected by the coalition, as it collided with the idea of reducing care work by (female) informal caregivers. The argument was driven by the need to address the issue of unpaid care work:

Because when we want to expand care services […], it is because we want to reduce care time for these women so they can do other things, so they can go to the promenade, or work, or do politics, or do whatever they want with their lives. (Interview 5)

Regarding the groups covered by the system, the main argument for creating a care system was the problem of informal care work and this was then more closely linked to childcare in the later process. Even though other issues related to care such as population aging were present, they were not prominently used to argue for a public care system. This again also impacted the later design within the government:

The feminist agenda was a driving force, but it also marked certain axes linked to childhood and linked to the right to women’s free time, which came from the processes of time use surveys that had been carried out for a decade in Uruguay. So, [the topic of older adults] was very difficult, because although they appeared as the priority population, there were also some conceptual aspects in which it was not balanced to childcare. (Interview 9)

The IOs involved mainly held a similar perspective, in particular UNFPA and UN Women, even though there were also regional debates on ageing and eldercare initiated by ECLAC that called for better protection for the group of older adults: “Protecting against the need for care in old age will be an unavoidable challenge for public policies in the coming decades.” (CEPAL, 2007, p. 111; own translation). Nevertheless, ECLAC was also very active in addressing gender issues and linking them to care policies at the regional level, such as the Conference on Women in Quito, as already mentioned (Esquivel, 2017; see also Mahon, 2018).

While UNFPA and UN Women started to work with the women’s organizations early on, ECLAC only joined the process at the national level later and generally cooperated more with the government than with the other coalition members such as civil society:

We have had little contact with ECLAC. I think it is our fault, because of our incapacity. And with UN Women we have a good relationship, we have done some things together. […] there has always been support, at least at the institutional level. If we do an activity and they are willing to say yes, they support it. (Interview 8)

The relative weight of eldercare policies within the system and its construction is also related to the coalition being unable to substantially integrate the group representing older adults, as an actor from an IO admits: “one thing to understand is that the dissemination of the national integrated care system has relatively little weight in the organizations of older adults” (Interview 11). The main civil society organization for older adults in Uruguay is ONAJPU (Organización Nacional de Asociaciones de Jubilados y Pensionistas del Uruguay), which represents pensioners. But as their core interest is pensions, other related fields such as eldercare were not very prominent on their agenda. An interviewee mentioned also that mainly “former male workers”, and “former male leaders” (Interview 6) were active in the organization and that they had little understanding of the gender issues the coalition was addressing, also reflecting diverging policy and deep core beliefs. Therefore, there was not a strong national or international actor group within the coalition that specifically addressed the needs of older adults.

Within the government and the administration there was an emphasis on childcare, even though the system aimed at all four groups, but there were differences in how much weight each group had in the construction and implementation of the SNIC. Services for children already existed before the new system was introduced, so the new policy mainly expanded those services. In contrast, services for older adults rather had to be developed from scratch, which was not only more difficult but also received less financial support from IOs: “of course, [the administration] focused on what was easier, which was childhood, early childhood. Because that is what you get money for from the international organizations” (Interview 6).

Finally, there was also a discussion at the political and administrative level about how the financial resources should be allocated. As the whole system was set up with only a very small budget,Footnote 15 especially in the later stage of implementation there were discussions on how the money should be distributed among the different groups:

We are dealing with that, […] that is, with public social spending and what percentage goes to the elderly and how much to children, everyone was saying “well, but then who is the care system going to”, because adults are already allocated, yes, but it is allocated, this, there is, pensions, pensions, but there are many, that is, those who have care needs, dependency, do not have care services. (Interview 2)

In conclusion, these factors, the general direction of the debate, the degree of involvement of certain interest groups, the allocation discussions within administration and politics, and the importance of feminist movements and IOs focusing on gender topics shaped the system and the way it was directed at the different groups, as well as what kind of benefits were introduced.

Discussion

This study raised questions of coalition building between international and national actors in emerging subsystems, specifically by examining the role of IOs in the process of creating a new care system in Uruguay, and how the coalition between these actors shaped the system and its elements. While there is extensive research on national actors in advocacy coalitions, the understanding of IOs in such coalitions is still limited. In the Uruguayan case, the role of IOs was quite active and direct, at least in the nascent stage of the subsystem, as they collaborated closely with the different national actors in the care coalition.

This involvement of international actors in national policymaking on the basis of beliefs and relationships departs from the common view in the literature that IOs formulate ideas that are transferred or translated by national actors in their respective contexts (e.g. Dolowitz et al., 2020; Stone, 2012). In Uruguay, IOs were often intertwined with national actors, and the distinction between national and international actors was blurred in many situations. Thus, the collaboration and interaction between international and national actors contributed to the generation of ideas within the coalition, albeit less in the sense of a transfer from the international to the national level, but rather as an exchange at eye level. Since Uruguay was also the first country in Latin America to introduce a national care system, the case served more as an experience from which IOs could draw ideas and lessons for other countries on the continent, rather than attempting to transfer specific ideas to the national context.

The common belief system was one of the most important elements contributing to the cooperation of national and international actors. It is remarkable that in the early phase in particular, namely during the agenda setting phase, IOs already took part by financing and supporting civil society organizations, even though the government did not address the topic very prominently at the time. This means that if they share the same beliefs (deep core and policy beliefs) on a political issue, actors collaborate even if the success of transforming these beliefs into policies is rather unclear and the subsystem is only nascent. While studies on nascent subsystems imply that “actors’ policy preferences and beliefs are not yet well defined” (Ingold et al., 2017, p. 456), which makes it hard for actors to identify allies, the activities and exchanges between academic actors, civil society, and IOs started quite early. However, the role of previous contacts and personal relationships helped the actors to establish a coalition based on common beliefs from a very early stage onwards, which is also an important factor addressed in previous research (Wiedemann & Ingold, 2024). Another aspect contributing to this could be that in very early stages of subsystem development the deep core beliefs, i.e. more normative general assumptions, are in the foreground, e.g. in Uruguay the rather general idea of gender equality, which actors might not only articulate in the new subsystem but also in other already mature subsystems, which means their general normative values are already known.

The notion of unstable beliefs is presumably true for potential opposing coalitions, as they were more or less non-existent. Even though in the later stages of policy formulation when the topic was pushed more at the political level the opposition in parliament did have some diverging ideas and also espoused different ideologies (deep core and policy beliefs), there was not a distinct coalition working against the policy. These different degrees of policy preferences and cooperation patterns across actors on a policy issue have also been found in other studies dealing with nascent subsystems. Wiedemann and Ingold (2024), for example, argue that for actors who are not so involved from the start, the new issue might be similar to an external shock they are confronted with and they have to engage with the new policy even though their beliefs are not well defined. Since the opposition parties in Uruguay only started to work on the policy issue at a rather late stage, after the topic had already entered the policy agenda at the government level, their preferences and policy ideas were not as clear as those of the coalition that dominated the debate.

The ACF assumes that in nascent subsystems alliances are rather unstable (Jenkins-Smith et al., 2018, p. 157). During the process until the law was adopted, this was not so much the case for the key actors, but for new actors whom the coalition tried to incorporate. However, after the adoption of the law, some of the actors who had been very active in the beginning (e.g. women’s organizations) had less impact on the implementation while others gained more importance. In regard to the involvement of IOs, after 2015 there was also a shift from UN agencies who had only limited financial resources to financial IOs such as the Interamerican Development Bank (IDB). Therefore, the more mature the subsystem became, the more the diversity of actors increased, which might also lead to more advocacy coalitions in the future that do not only focus on gender issues.

An interesting aspect is the role of disability organizations who participated in the coalition, even though they came from a slightly different perspective than the feminist actors. Regarding policy beliefs, it can be assumed that they did not have contrasting beliefs and values, rather different priorities and foci. An interesting explanation for the participation of this group, which also managed to bring in their own ideas into the system’s development, is offered by Fenger and Klok (2001), who differentiate between congruent, indifferent, and diverging beliefs as well as different levels of interdependency between actors. They assume that in situations of “symbiotic interdependency”, where actors profit from each other in terms of resources, they then collaborate in a “coalition of convenience” (p. 164) even if they are indifferent in policy beliefs. Their collaboration is however more unstable than that between actors who cooperate based on shared beliefs because it is based on interests. In Uruguay, the disability organizations worked together with the actors from feminist networks and academia, but their collaboration was more based on common interests than on common beliefs, and therefore less stable than the collaboration between the key actors in the coalition.

Another important factor was the connections and overlaps within and between national and international actors on an institutional but also on a personal level. As previous studies have shown, “trust and former contacts” (Ingold et al., 2017, p. 458) form the foundations for building networks between different actors, which were also important elements in the Uruguayan case. In particular the personal ties between international and national actors due to overlapping affiliations facilitated the cooperation of different institutions as these relationships between actors helped the coalition to identify allies especially in the beginning. However, when they later tried to incorporate new actors such as the older adults’ representatives or the workers’ union, with whom they were not so closely connected, there were also more difficulties, as these groups had different preferences and priorities, without a strong feminist vision.

Regarding the law that was adopted, the ideas of the alliance of national and international actors were manifested in terms of specific characteristics of the policy such as a universal, rights-based system, service orientation, or even the broad coverage to include all care-dependent groups in the system, in order avoid any type of informal care work. Even though the system itself has limitations due to restricted financial resources, the policy beliefs of the dominant coalition were translated to a major extent into law, with the support of the other parties. Previous research has pointed to similar dynamics: when there was a dominant coalition in the setting of a nascent subsystem, support for their ideas is more likely, as potential threats or risks of the new policy are unknown (see Fidelman et al., 2014; Olofsson et al., 2018).

While the relationships and overlaps helped the coalition to impact the subsystem due to access to important resources such as power, information channels, finances etc., with the change of government in 2020 they faced a new situation in which most of these channels and overlaps were severed and even some of the more institutionalized exchange formats, such as the care advisory board, were less active (Cejudo & Michel, 2023). In this new situation, the role of IOs might change as well. While they were previously a part of the care coalition, their role might change to that of policy broker as they could be perceived as neutral actors by the more established and coordinated “opposing coalitions”. However, looking at most recent developments ahead of the 2024 elections, it seems that the care coalition rather resurges with IOs being part of it, as in 2023 the “National Care Forum” (Foro Nacional de Cuidados) has been established, integrating actors from civil society, academia and IOs such as UNFPA, UN Women and ECLAC (Foro Nacional de Cuidados, n.d.).

Conclusion

This article aimed to shed light on coalition building in nascent subsystems with a focus on the role of international actors in these processes, and on the impact of such a coalition on policy development. While the case of Uruguay has to be seen within its national contexts and specific actors’ relations, this study also offers some more general conclusions and theoretical insights related to the ACF.

First, whereas ACF scholarship predominately focusses on mature subsystem contexts and national actors, this study provides insights into international actors’ activities in emerging subsystems and confirms the findings of other studies that trust and previous cooperation are important foundations for coalition building. While these factors might be related to cooperation on the institutional level, this study reveals relationships on the personal level that were even more important: in the Uruguayan case overlapping affiliations contributed to the formation of a very strong and dominant coalition based on common beliefs that was able to impact the political agenda. This is an important finding, particularly with regard to IOs, as they are often seen as outsiders who formulate ideas that are then transferred or translated to the national level. In this case, however, the overlaps and personal ties between national and international institutions in particular contributed significantly to coalition building between these actors and the formulation of ideas in a cooperative setting. Second, regarding the role of international actors, in this case UN agencies, it can be concluded that IOs can have a very direct and concrete role in advocacy coalitions, if they share similar beliefs and if national actors are able to take advantage of cooperation with international actors. Even though IOs often have the notion of being “neutral” actors, working more normatively from the outsider’s perspective, this case study shows that if their values and beliefs coincidence with activities by actors at the national level, IOs are willing to engage actively in the process and to shape national policy outcomes even at a very early stage. While the main drivers of the debate were the women’s organizations and academic actors, IOs facilitated the ability of these actors to push the topic to the agenda of the government. Third, it can be concluded that emerging subsystems and the way that policies are shaped within them are more likely to mirror a dominant coalition’s deep core and policy beliefs when opposing coalitions are barely existent and when there is low engagement from actors who might have different foci or interests.

This study is not without limitations. First, I focus mainly on the actors who dominated the subsystem, especially with interviews, whereas actors who were not as central such as pensioners’ representatives or workers’ unions could not be interviewed, and their role in the process was only analyzed indirectly. Similarly, I did not interview actors from the political opposition, and had to rely on documents, minutes from parliamentary debates, and interviews with other actors. Second, the research design of a case study limits the generalizability of its findings. These may be a particularity of Uruguay, such as the overlapping affiliations, and could be different in other cases and settings.

Therefore, it is important to conduct research on similar cases, such as Chile, Colombia or others, where care systems are debated with the involvement of IOs, to see whether there are similar patterns of coalition building, whether different interest groups at the national level interact with IOs differently, and whether this could impact potential outcomes. It is also important to look at the current and future development of the subsystem in Uruguay, as here I could only touch briefly on the change of government in 2020 and more recent developments, but especially with regard to the potential emergence of new coalitions and also possible shifts in the roles of IOs, future research could contribute to a better understanding of the transition process from a nascent subsystem to a mature subsystem.