Introduction

In the last decades, governments and international organizations have stressed the need of confronting the most pressing global challenges “through integrated, coordinated, and coherent strategies” (SDGS-UN, 2022). From “integrated water resources management” (UN-Water, 2022) to “Integrated food security strategies” (FAO, 2002), to “Integrated approaches to combating poverty and social exclusion” (EUNL, 2016) and “Integrated Care systems” (NHS, 2022), policy integration has become the gold standard for most action plans of governments and international organizations.

Nonetheless, it is not uncommon to see that these grandiloquent strategies come down to issuing a new law or creating inter-ministerial committees, or to a handful of agencies acting in the name of a greater goal but without further implications in their daily bureaucratic routines. What does it entail for a government to actually implement an integrated policy? There is a renewed scholarly discussion on the topic, from which it has been argued that, when governments attempt to integrate policies, they put together policy instruments from different sectors and agencies in order to solve a complex problem. Still, in practice the line between integrated and less integrated policies is blurry. For instance, what does it mean for a child living in poverty, a community suffering from food insecurity, or a region facing flood risks to have a government “integrally” addressing their problem? What difference does it make for an unemployed person that a law reforming the existing unemployment policy was passed in order to attain a specific integration objective? These sorts of questions call for a shift in the focus of analysis from the design to the implementation of policy integration. Indeed, policy integration does not only require selecting specific policy instruments to create a comprehensive response, but also to make sure that those instruments remain integrated during the implementation.

In this paper, we analyze how governments that launch integrated policy strategies keep them integrated during their implementation. We study the implementation processes of two integrated care policies for children in Chile and Uruguay. These policies were designed as integrated strategies, aimed at addressing the evolving care needs of a diverse and changing population, but followed different implementation trajectories over time. Chile was able to keep the policy integrated during the implementation, whereas Uruguay was not. Chile managed to overcome bureaucratic resistance and political turnover to maintain an integrated set of services for children. In contrast, Uruguay’s inability to solve administrative opposition and volatile political support over time resulted in a sector-based implementation. We argue that these contrasting trajectories are not explained by differences in the policies’ design, but by variations in their policy regimes: how institutional arrangements, ideas, and interests interacted with the policy to keep it integrated during the implementation.

Thus, by moving beyond the design approach to policy integration, in this paper we study how institutions, organizations, and understandings of public issues engage with the integrated policy as it evolves within a political setting. We build our analysis around three constitutive elements of a policy regime: ideas, interests, and institutional arrangements (Jochim & May, 2010). We also explain how these features of the policy regime manage to keep the policy integrated during the implementation process. In this way, we contribute to the literature on policy integration by explaining the crucial role of policy regimes in keeping the policy integrated over time during the implementation, when policies face bureaucratic resistance and a sectoral logic of decision making.

In the following sections, we link the study of policy integration with the literature on implementation and policy regimes. Then, we explain our cases and methods. Next, we analyze how these integrated strategies were implemented by looking at the design and evolution of the components of the policy regime in each country and make a comparative analysis. In the final sections, we discuss our theoretical implications for the literature on policy integration and conclude with our contributions and new research questions.

Policy integration: beyond design and into implementation

Policy implementation is the process of articulating different (policy and institutional) components into a single integrated policy strategy in order to comprehensively address a complex problem (Candel & Biesbroek, 2016; Cejudo & Michel, 2017; Tosun & Lang, 2017). From a design perspective approach (Givoni, 2014; Kern & Howlett, 2009), policy integration requires diagnosing the complexity of the problem (number of goals, policies, and levels of government involved), so that policymakers be able to design a policy mix where complementarities among policy instruments are enhanced and inconsistencies minimized (Howlett et al., 2017).

Yet, the design approach is not enough to understand the dynamics of the process of policy integration (Cejudo & Trein, 2023). We need to explain how and why governments manage to move from “expressed intentions” to “implementation practices” (Trein et al., 2021). This is a classic question in the scholarly research on policy implementation, but so far unaddressed in the policy integration literature where the implementation challenges are presumably more (May, 2015). Indeed, the study of policy implementation has traditionally been devoted to explaining the reasons behind the gap between the intended policy goals outlined in the causal theory and the actual results (Hupe, 2011). Both from the top-down (Pressman & Wildavsky, 1984) and the bottom-up perspectives (Lipsky, 2010), it has become clear that the formal adoption of a policy does not automatically translate into the attainment of the policy objectives and that the difference between policy success and failure greatly depends on the implementation arrangements where policies unfold (Sager & Gofen, 2022).

When talking about integrated policy strategies, the premise remains the same: the mere announcement of an integrated policy strategy does not lead to integrated implementation. Yet, there is a crucial distinguishing feature of policy integration (from an implementation perspective): It entails implementing a concatenation of causal theories, not only the causal theories that guide each policy involved but also the overarching causal theory that guides the implementation of the set of those policies as a whole by delineating the role the institutional arrangements play in that process (Cejudo & Michel, 2017). Indeed, in contrast to the implementation of a single or several policies in tandem, implementing an integrated policy requires the simultaneous implementation of different causal theories. Some of these theories correspond to the intended operation of each component of the integrated policy (i.e., the educational or the agricultural), there is another (a macro-causal theory) that sets out how these different components fit together and interact. The macro-causal theory is the explanation of how and “which policies and organizations [should] work under a new logic, subordinating their objectives to a new overall goal, and making their decisions based on the needs and priorities derived from the complex problem” (Cejudo & Michel, 2017, p. 758).

Thus, explaining the implementation of integrated policy strategies requires analyzing both the deployment of their elements (i.e., its goals and the policy instruments used to attain them) and the institutional structures and organizational processes needed for their articulated operation (the implementation arrangement at work) (see Sager & Gofen, 2022). For this analysis, we use the literature on policy regimes since this approach integrates into the analysis the multiple programs and services that directly or indirectly affect a problem, and thus the causal theories at play in integrated policies, simultaneously coexisting (May & Jochim, 2013).

Connecting policy regimes and policy integration

The notion of policy regimes and their capacity for “holding together” the policy during implementation offers a coherent analytical framework that brings together analytical elements already identified in the literature on policy integration but used independently to explain different aspects of implementation. In the following paragraphs, we explain the elements of policy regimes that according to May and Jochim (2013) perform “integrative functions” and link them to equivalent arguments set forth in the literature on policy integration.

Policy regimes are “governing arrangements for addressing policy problems, made up of three elements: ideas, institutional arrangements, and interests” (May & Jochim, 2013). First, ideas set the boundaries of the regime because they determine the problem, the solution, and the governmental actors with responsibility on the matter and “shape the direction of supporting institutional development and propel the actions of concerned interest” (Jochim & May, 2010, p. 312). This notion has also been used in the literature on policy integration. According to Candel and Biesbroek (2016) the extent to which a cross-cutting problem is considered as such shapes the governance system’s macro-political venues and decision-maker’s actions. It is through these ideas (also referred to as policy frames) that governments put forward a common understanding of the problem to be addressed among the relevant actors (Cejudo & Michel, 2021). Moreover, the rules, norms, and organizational structures are shaped by ideas, and, at the same time, they affect how ideas are formed.

Second, in the literature on policy regimes institutions are the means through which ideas become actionable, for they specify the agencies that are responsible for the implementation of a policy. The institutional arrangement also structures the information flows among such agencies and sets the organizational relationships that would determine the implementation of the policy as much as the priority that the policy should have in the daily bureaucratic workload (May & Jochim, 2013). However, institutional arrangements are not static. This notion has been discussed in the literature on policy integration when paying attention to “the empirics of how policy mixes actually evolve post-enactment” (Howlett, 2019, p. 36). Indeed, the temporal dimension of the process of integration has been incorporated in the literature when examining the role of the authority that is continuously exerted over policies and organizations to guarantee their interaction, as well as the institutional design that allows for information flows to activate policy feedback processes affecting the integrated policy (Cejudo & Michel, 2021). This dimension has also been studied when looking at the deployment of organizational resources to enable (or prevent) integration (Biesbroek & Candel, 2020).

Third, the interests at play in a policy regime may serve as integrative or disintegrative forces (Jochim & May, 2010). While the institutional setting provides the context within which the policy operates to attain its goals, the organizational design defines the responsibilities and resources of the implementing agents. Together, institutions and organizations allocate decision power (Sager & Gofen, 2022) and thus affect the interests of different stakeholders. Depending on the relevance of such stakeholders, their sense of urgency and the degree of agreement regarding the shared purpose of a policy regime. In turn, recent scholarship on policy integration has analyzed policy integration as a political process (Cejudo & Trein, 2023), the role political actors in mobilizing the framing of a certain issue as a complex one, and devising an integrated solution (Candel, 2019), as well as the role of coalitions (Kefeli et al., 2023) and policy entrepreneurs in its implementation (Lambelet, 2023).

Once we have bridged the policy regimes and policy integration literatures by identifying the elements that perform integrative functions at the implementation stage, we next use them as the analytical categories for studying how the implementation of care policies for children in Chile and Uruguay occurred. We explain their different implementation trajectories by analyzing their policy regimes and the role their constitutive elements (ideas, institutional arrangements, and interests) played in the implementation of these integrated policy strategies. We explore how both countries progressively developed the policy regime, using budgets, organizational structures, legal instruments, and information to attempt to provide care services according to the specific needs of each child.

Cases and method

Why care policies are a relevant unit of analysis for the study of the implementation of integrated policies? In recent years, an explicit attempt to integrally address care needs has proliferated throughout the world. Governments have recognized the importance of promoting people’s well-being through proper care, particularly in the early and late stages of life. Conceiving care as a component of well-being has represented an important shift in the understanding of what welfare entails, especially in Latin America, where caregiving has traditionally been considered a family responsibility.

Since the turn of the millennium, a wave of policy interventions targeted to children that mix health, education, and social assistance instruments have been launched in the region. From “Red Cuido” in Costa Rica, “De cero a siempre” in Colombia, “Creciendo Juntos” in Argentina, and “Cuna más” in Peru, the efforts of countries in Latin America have been aimed at integrally providing care (Esquivel, 2017). Not all these initiatives have been equally successful in their attempt to integrally address children for securing their care and early development. Many of them simply lack the design features already identified in the literature for achieving policy integration; others are properly designed but struggle to maintain the integrated approach over time.

Among these initiatives, the program Chile Crece Contigo (ChCC) in Chile and the component of childcare within the Integrated National Care System (SNIC, for its initials in Spanish) in Uruguay are recognized as two of the most comprehensive care policies in the region (Marzonetto, 2019). Both are cases that were specifically designed to place the users (children and their households) at the center of their interventions: they attempt to articulate different policy instruments around children to provide them with integral care services. They both are also cases of policies that built an ad hoc institutional arrangement involving several sectors (health, education, labor, and social security) for the design and simultaneous implementation of multiple policy instruments (from regulations to direct provision of services) targeted to address specific needs of the population. Yet, during the implementation phase, each policy observed a different path, one being more successful than the other in maintaining an articulated array of goods and services to meet each child’s needs. Thus, they are suitable cases for understanding variation in the implementation of integrated strategies.

Our analysis is focused on the process of integration, from the initial formation of an idea that justified the intent of integrating different sectors to its actual operation over time.Footnote 1 To this end, we first undertook an analysis of scholarly research related to childcare services in Chile and Uruguay. By being flagship strategies of their respective governments and benchmark care policies in the region, both cases are very well documented. We revised official reports from which we derived the specific design features of each intervention, their aims and objectives, and an understanding of how each strategy was framed. An analysis of the acts and regulations enacted from 2006 to 2022 allowed us to trace the evolution of the specific components of each policy and the changes at the organizational level. We also benefited from the internal and external policy reports about these programs, which provided us with a recollection of the facts and a reflection of the strengths and shortcomings of the policies, sometimes in the words of the implementing agents themselves. Finally, for a more critical view of the programs and a better sense of their appraisal by society, we resorted to the revision of press and external evaluations.

We then relied on first-hand accounts of specialists and public officials involved in the design and operation of the strategies, either through semi-structured interviews or their own written records, with two aims. The first aim was to validate the interpretation of the factual analysis of the process of integration conducted in the documentary analysis. Second, the purpose was to understand the events and decisions that did not happen despite being politically expectable (for example, the termination of ChCC in the administration of Piñera or the resistance to information sharing between Uruguayan agencies). For a better grasp of the process of the Uruguayan case, we interviewed a top public official of the SNIC that provided us with a very detailed and comprehensive view of the system’s operation. We also interviewed a member of the Advisory Committee whose vision mirrors that of the civil society involved in the implementation. A more nuanced understanding of the design and implementation of ChCC came from the interview of a top public official that took part in the design and the first years of implementation of the program, and from a public official directly involved in the operation of the program. Finally, our recollection of the facts also benefited from a comparative perspective of both cases in an interview with two academics specialized in these programs.

The information gathered from both primary and secondary sources allowed us for a thick case reconstruction (Beach and Pedersen, 2013) that goes from the initial stages of each policy to the early 2020s, which we structured around the constitutive elements of the policy regime. Thus, to explain the implementation of each policy, we structure the cases describing how the process of integration took place: how the problem and its expected solution were understood and how responsibilities were allocated (idea), how decision makers channeled information and authority and relations to secure an integral response (institutional arrangement) and how political support facilitated the design and implementation of the integrated policies (interests). Finally, we conduct a comparative analysis of the role that ideas, institutional arrangements, and interests played in each case to keep the policy integrated.

Chile Crece Contigo

In Chile, by the early twenty-first century, the provision of childcare, and in general of social protection, was fragmented; there were several specialized agencies and programs aimed at addressing social problems by targeting very specific populations (Raczynski, 2008). This problem was understood as one of the reasons for the inability of social policy to reduce extreme poverty and geographical inequalities. By 2005, at least 30% of Chilean children under five years old “did not reach their expected development milestones” (Molina et al., 2018).

As part of its overhaul of the social protection regime, the government of Michelle Bachelet launched in 2006 an ambitious program, called Chile Crece Contigo (Chile Grows with You, ChCC). ChCC is a comprehensive strategy targeted at children from the prenatal period to preschool age (four years). This program seeks to address the problem of child poverty and its consequences on child development and social inequality (Cunill, 2014; Milman et al., 2018). It offers both universal and targeted services from the health, education, and social development ministries in order to give children access to public programs (Milman et al., 2018). By doing so, ChCC aims at promoting early childhood development, based on the premise that interventions at this point of the life cycle have a significant effect on education, health, and other outcomes (Torres et al., 2018). It started with a limited scope, but it was quickly expanded geographically (more municipalities), incorporated older children, and offered new services, overcoming public health professionals’ guild and ministerial resistance (E1_ChCC).

As an integrated strategy, it encompasses the work of several agencies, each in charge of implementing with targeted interventions tailored to the diverse and evolving needs of each child. Thus, for its implementation it was necessary to keep not only each agency doing its work, with the necessary resources available (from political support to budgets and information) but making sure that their combined work guaranteed an integral set of services for children. Families would be automatically enrolled and would not need to search for specific programs with varying enrollment requirements. As explained below, this objective was the core idea (policy frame) behind the design of ChCC and its implementation.

Ideas

The core idea behind ChCC was that child protection should not depend on the availability of scattered programs but on children’s effective access to them. In line with President Bachelet’s professional background, ChCC was to be based on a biopsychosocial model, whose premise is that biological, psychological, and social factors play a significant role in health, and thus must be jointly treated. Accordingly, the goal was to deliver an integrated package of interventions tailored to the needs of each child. This idea served as the organizing principle of ChCC: it would offer an articulated array of services that, together, secure health, education, and early child development to eligible children and give preferential access to specific populations that were usually excluded from social programs (Molina et al., 2018).

However, the idea of a system was not a general demand in the public agenda but pushed by certain specialized groups with long experience working in child protection. “For the citizens, the issue of the children was not an issue… the family was the one in charge and if at some point they require help, then the child somehow would get it from the family” (E2_ChCC). On the contrary, for many, this was the government's overstepping into private family matters. Thus, at the outset, not all the relevant ministries endorsed the idea of an integrated solution for child protection. The Ministry of Education was originally not part of ChCC, and consequently, the intervention had to be limited to children between 0 and 4 years old. ChCC began its operations with the Health and Social Development ministries, which started targeting their existing interventions (such as immunization and nutrition) to the program’s target population, by increasing their coverage to reach all the children in need (such as childcare or preschools), and/or by operating new interventions to support child development (World Bank, 2018).

Over time, the core idea evolved, since the understanding of what child development entailed changed. For instance, by 2017 a growing demand for generating integral protection for older children emerged, which made it possible to extend ChCC to children from 0 to 8 years, just as it was originally intended (E1_ChCC; E2_ChCC), and to include mental health and learning support. Moreover, recent evidence of a rising number of child sexual abuse at all socioeconomic levels (La Tercera, January 2, 2018), along with cases of child abuse in public institutions (see La Cuarta, May 19, 2019) led to a UN Committee on the Rights of the Child to recommend “establishing programs that promptly identify risks of violations” (UN, 2020).

Although these changes resulted on the expansion of the policy (toward new groups of the population and with new interventions), the original core idea remained guiding policy decisions: The idea to articulate the set of interventions around every child’s needs continued despite the inertial pull toward a sector-specific implementation (see Trein & Cejudo, 2023). Indeed, to include older children, ChCC would need school information to target them and refer them toward other interventions, and thus it was essential to overcome the Ministry of Education’s resistance to share it (E2_ChCC).

This same core idea persisted even in the following administrations. Over time, new interventions were included under the umbrella of child development, beyond basic health and educational matters, in an integrated way. As President Bachelet explained: ChCC had “a modular perspective [in which] the pieces of a whole gradually fit together into a larger system” (Bachelet, 2017).

Institutions

The implementation of ChCC was more than the simultaneous operation of complementary programs according to their own causal theory. Indeed, for the core idea to be materialized, the regular operation of the agencies and programs had to respond to the priorities set by ChCC (according to the macro-causal theory), with information and flexibility to provide differentiated services to children, tailored to their socioeconomic status or risk vulnerability. Through the creation of new rules, programs, budget allocation criteria, information flows, and a coordinating body, a new institutional arrangement was designed to channel attention, information, and relationships to integrally secure every child’s development.

Several agencies, as well as regional and local governments, were charged with specific responsibilities. They had to provide their services according to their own logic of operation, but also following guidelines devised by a coordinating body established within the Ministry of Planning (later moved to the Ministry of Social Development). Resources were allocated to all agencies, along with specific performance indicators. This characteristic of the institutional design is what enables the Ministry of Social Development to channel the attention of other agencies and levels of government toward ChCC goals and, thus, to keep the policy integrated over time. Indeed, to guarantee their compliance with ChCC guidelines, the budget for ChCC is assigned to the Ministry of Social Development, which then allocates it to the relevant ministries and municipalities according to their performance on indicators defined by the coordinating body (instead of by every sector) to monitor the implementation of ChCC as a whole.

Linking budget to performance not only channeled agencies’ attention toward ChCC goals but also their information. Indeed, the Ministry of Social Development needs information to guarantee timely, good-quality services and to provide differentiated services to children, tailored according to their needs (World Bank, 2018). Since this information is also used for budgetary decisions based on agencies' performance, ministries, and agencies in charge of the implementation of ChCC are willing to share it. The institutional arrangement was designed to retrieve updated information from every agency and program involved by conditioning their budgets to the provision of accurate information and their targeting according to ChCC priorities (Molina et al., 2018). In this way, the institutional arrangement not only defined responsibilities but also established linkages among the implementing authorities. In other words, it had to balance the demands of the macro-causal theory with the routine operation of the causal theories of each component.

This institutional arrangement evolved. From the outset, it was expected that the experience of each ministry and the information gathered during the implementation would enable policymakers (the Ministry of Social Development) to adapt processes and recalibrate ChCC components. Being able to adapt ChCC’s design to the changing reality was particularly relevant as ChCC aimed at expanding its reach (gradually including older children and teenagers) and scope (going as far as providing mental health services and extending parental licenses). This process was consolidated when, by the end of the second Bachelet administration, a new Inter-ministerial Committee for Social Development and Children was created in the Ministry of Social Development, and an Undersecretary for Children obtained authority for managing ChCC (Bachelet, 2017). In 2022, a new Law for the social protection of children was passed (Law 21,430), which further institutionalized the logic of an integral approach to child care.

Over time, the components and scope of ChCC changed, but the coordinating body remained in charge of designing and administratively dealing with such changes. For instance, the process of expansion required adapting the information systems to incorporate older children (and new sources of information from the Ministry of Education) and to bring new actors (agencies) to the existing causal theory. Yet, even when the composition and name of the coordinating body varied over the years, its functions remained: facilitating coordination among involved actors, providing information for budget forecasts and identifying eligible children, and providing technical support to municipalities for the implementation of each new intervention. In this way, despite new components, target populations, and instruments being incorporated, the policy remained integrated over time, and operated according to the macro-causal theory.

Interests

For all these changes to happen, continuous political support was crucial beyond the initial impetus provided by the Bachelet administration. This support was generated by Chile’s consensual way of policymaking. The government convened a Presidential Advisory Council, which included not only government and civil society experts on social protection and child development: “In Chile, there is a need for consensus that arises even at the presidential advisory commissions (…) which are plural. There you would find (…) right-wing [actors] (….), and [they] would be seated at the same table with advisors that come from other think tanks and universities that are left-wing” (E3_CU). This inclusive approach facilitated building support for the program outside the government. When bills were sent to Congress, “they are way more plural and try to have a wider perspective” (E3_CU).

Just as important, support within the government was guaranteed by the inclusion of officials from key agencies, especially the head of the Budget Directorate, who “translated” the purposes of ChCC into an economic logic: providing care to children would facilitate women’s labor inclusion. By doing so, it expedited the Ministry of Finance’s financial and administrative support (Farías, 2008).

Building a strong social protection system and a more equal society was one of the main priorities of the governing coalition. President Bachelet’s political backing behind the launching of ChCC overcame the usual barriers to coordination among ministries. More importantly, the political commitment that created the political capacity to implement ChCC was also evident in the congressional support to the Law that institutionalized it (Law 20,379), which was passed unanimously three years into the program and the financing of a continuously expanding budget.

The law formally recognizes the coordinating body, its administrative instruments, the services guaranteed by ChCC, and the criteria for prioritizing target populations. Even when a candidate from a different coalition (Sebastián Piñera) won the presidency in 2010, ChCC remained in place, although with less visibility and without dramatic budgetary increases. Despite the (unavoidable) association of the program with Bachelet, the credibility of top officials at the Ministry of Finance already involved in ChCC was critical in convincing the new president of the program’s positive effects on the economy. Thus, ChCC’s evolution continued: a new program for newborns and an expansion of domiciliary visits were launched during his administration.

Overall, ChCC has not faced strong political opposition. A second Bachelet administration and a second Piñera administration kept their commitment to ChCC and supported budgetary increases and program expansions. Even after the dramatic 2021 elections won by Gabriel Boric, who has offered an overhaul of many long-standing policies, he has announced that ChCC will be one of the instruments to protect social rights for all Chileans.

Thus, 15 years after its creation, the capacity (and adaptability) of an overarching idea about child development, an institutional arrangement that steered the individual components of the program, and continuous political support managed to keep the policy integrated. Despite expansions and resistance, ChCC keeps providing services to children tailored to their individual needs.

Sistema Nacional de Cuidados in Uruguay

As in the rest of the region, in Uruguay the provision of child services was fragmented and insufficient. By 2010 there were at least four types of public services for children under three years old that, together, reached around 30% of the population (ECLAC, 2011). Each of them responded to different regulations and supervisors and implemented different programs (Plan Nacional de Cuidados 2016–2020) thus reproducing social and gender inequalities.

A coalition of feminist organizations had been pushing for over a decade the idea of an integrated response to address the care needs of children, the elderly, persons with disabilities, and their caregivers, over the course of a decade. Many of these organizations were part of the bases of the Frente Amplio political party. When Frente Amplio’s Tabaré Vázquez took office for second time in 2015, he adopted this idea into his government’s plans. The result was a National Integrated Care System (SNIC) (E3_CU). The goal was to connect various care programs and services to promote people’s autonomy and secure their basic and instrumental daily life needs (Plan Nacional de Cuidados 2016–2020). As part of this plan, parental leave, regulations standardizing the quality of childcare among different providers, and increases in coverage were included.

As an integrated strategy, such an ambitious intervention involved several agencies and policy instruments. For its implementation, it was clear that several decisions would need to be made to make those components work as part of a system. An organizational structure was created to steer their actions toward that purpose; it encompassed three coordination bodies (for political deliberation and decision making, for monitoring implementation, and for getting input from civil society). In the following paragraphs, we study how ideas, institutions, and interests influenced the implementation of the early child component of the SNIC.

Ideas

Since the 1990s, feminist organizations within Frente Amplio were articulating a social demand for redistributing care responsibilities between families, the State, the community, and the market, and, within families, between men and women. Once Frente Amplio’s Tabaré Vázquez assumed office in 2015, the new government incorporated some of these demands into its quest for an institutional transformation with a strong emphasis on social justice, including the aim of “paying off the historical debt that Uruguayan society owes to women'' (CNPS, 2009, p. 127, own translation). This commitment translated into an understanding of care policy as the fourth pillar of social protection in the country, based on a gender- and rights-based perspective (E2_SNIC; E3_CU).

The underlying idea was that this system would secure a human right: the right to care for others (which overly depends on women) and to receive care. Thus, the focus was on both caregivers and care receivers (children under 12 years old, especially those under three years old; people with disabilities, and elderly people). Yet, this decision did not imply the creation of new policy instruments. Rather, the system was understood as a “cultural change” that stemmed from the articulation of existing interventions. “It was important to convey the idea (…) [that] what was being done was not creating a policy from scratch but, instead, integrating into a public system action under the umbrella of a care policy (…)” (E2_SNIC].

Putting this idea into practice required the involvement of different agencies and ministries (Salvador, 2019), many of them with a long experience in child policies and likely to oppose change. “It was important to respect institutional mandates and functions (…) otherwise it would mean to take them away from institutions that have been doing these actions for many years. [Building on existing interventions was deemed as the ideal solution to avoid] institutional conflicts and breaking institutional memories (…)” (E2_SNIC).

Thus, the macro-level theory of change was built on the idea that the sum of different pieces would amount to an integral response, rather than on an institutional re-engineering knitting together different interventions around the same child (or household) to produce a new integral response. New regulations were enacted specifying the types of services and subsidies that were now part of the System, but the distinct contributions of the agencies involved were not established (see Resolution N° 005/017). To stay away from institutional conflict and act efficiently “the option was to say: each one stays doing what they are doing, but now we are going to do it within the framework of a new policy, and this entails the incorporation of a conceptual framework that, in some cases, is new.” (E2_SNIC).

The core idea behind SNIC was an abstract purpose that did not come along with a theory about how the different pieces of the system (and the missing ones) would operate as part of an integral system. Without modifying their individual micro-theories of change, some agencies changed their operation to adapt to the imperatives of the new system: existing early childhood services extended their coverage, new childcare centers were created and regulation on quality services was enacted (from which managerial tools derived, such as quality assessments of childcare centers).

Institutions

The SNIC rested on three coordinating bodies specifically created for implementing it. The National Care Board (Junta Nacional de Cuidados), an inter-ministerial body chaired by the Ministry of Social Development (MIDES), was in charge of deciding over political and strategic matters, including serving as a deliberation forum over budget allocations. For the Ministry of Economy (responsible for the public budget and with a seat on this board), this meant that budgetary demands for implementing a new policy were posed by a single entity (the board), thus replacing multiple subsequent negotiations with different agencies for only one with the board. For the agencies that were part of the SNIC, this meant that their budgetary demands had to be channeled through the board unless it was for programs unrelated to the System (E2_SNIC). However, even when a budgetary demand was mediated by the board, the budget remained allocated to individual programs rather than to the System as a whole.

The National Care Board was supported by an administrative body, the Secretariat of the National Care System. Yet, this Secretariat was not an inter-ministerial body, but part of MIDES. Its main function was to coordinate, monitor, and evaluate the programs that make up the system. In practice, the Secretariat also implemented communication tasks, not in terms of diffusion of the SNIC but of the actions that could favor the cultural change the system required. Yet, once in operation, these objectives were not easily achieved. The Secretariat did not have staff support, and it had to rely on administrative units at MIDES and the specific implementing agencies dealing with target populations (Memoria Quinquenal, 2021).

The institutional design also included the establishment of technical commissions to discuss, plan, and carry out actions on specific topics or target groups. Their work was informed by the expertise and demands of a third coordinating body: an Advisory Committee, composed of members of civil society organizations, unions, academics, and private care providers. The technical commissions created new programs and tools, including a program for personal assistants for dependent people, and new tools for identifying target populations, such as criteria for assessing the levels of dependency.

This institutional design was not enough, at least in the initial stages, to channel attention, information, or relationships among the ministries and agencies involved in the pursuit of the SNIC’s goals. These coordinating bodies were only provided with authority to align several interventions around the guiding idea of the system, but without “advocating reorganizations on how it is implemented” (Junta Nacional de Cuidados, 2015, p. 12, own translation). Indeed, despite the creation of an information system to have a continuous flow of information to manage, monitor, and evaluate each agency's performance, they were held accountable for their performance with their own sector-specific indicators and goals, partly because the type of information and the degree of disaggregation is different, limiting its usefulness for the system’s decision making. For children and their families, this meant that access to public provision remained a matter of luck. “What the system did was to create and increase services (…). [F]or the population with dependency, the program of personal assistants was created. And so, if a family had a member with dependency, they could benefit from that program… but also, if a family had a child that attended a childcare center (centers CAIF) (…) the CAIF might alert the family and say ‘you can apply to that program’. I am not saying that this worked always like that though” (E2_SNIC). More important, accessing the array of services and programs remained a task for the families. As the literature on non-take-up shows, people’s decisions to enroll in a program are mediated by their capacity to know of the existence of the program or service and understand its rules, to interact with the bureaucracy and undertake the corresponding administrative burden and to perceive the programs and services as a solution suited to them (Ranci & Arlotti, 2019). This latter condition is extremely challenging in care policies in Latin America, given the prevailing idea that caring is a woman’s responsibility that should be undertaken at their home.

Moreover, the generation of information was thought of as a gradual process that would start by mapping the availability and quality of childcare centers and, later, the demand for them. When the agencies involved provided information about their location and coverage, they did so to facilitate peoples’ decisions regarding the available care services, not to inform the System’s managerial decisions. As a result, in its first three years of implementation, the Uruguayan care system managed to increase public childcare coverage, but not the overall coverage. Instead, “children that received care at private centers have shifted to the public sector” (Salvador, 2019, p. 21). This shows that the system did not have disaggregated information that would allow to “specify definite criteria for assigning the new beneficiaries” (Amarante et al., 2019, p. 9), particularly those excluded from childcare. It did not have the authority to modify the operation of public providers to steer them toward securing access to those without childcare.

Other information sources are being developed (such as an instrument for assessing the quality of childcare centers, and the National Care Registry). These aimed at generating data that would allow to “provide comprehensive care to the population and (…) an integrated analysis of the supply and demand” (Memoria Quinquenal, 2021, p. 181, own translation). Until then, the existing information does not allow the System to know who, when, with what quality, and at what cost is a specific child being taken care of, and more importantly, if the care that children are receiving is enough to satisfy their needs. Without information of this kind, and without authority to modify programs’ design or operation, the coordinating bodies were unable to overcome the sectoral logic of different components of the system.

Interests

When, in 2005 Tabaré Vázquez (center-left) took office for the first time, he opened up the political opportunity to discuss the ideas that eventually gave birth to the SNIC. In 2010, a new government (from the same party) established a task force to build the SNIC (Salvador, 2010), but it was only in 2015, at the beginning of Vázquez’s second administration, that care policy was placed at the center of the agenda.

This initial support was enough to pass a law that launched an ambitious system with new goals, a national plan, and a set of coordinating bodies. Yet, it did not hold back bureaucratic resistance. Ordinary activities, resources, and information remained under the scope of each agency, partly because supporting the SNIC entailed a new understanding of their institutional aims and goals, not necessarily compatible with what they had been doing for years. Thus, the political efforts of the National Secretary and the Minister of Social Development (head of the Board) were focused on overcoming bureaucratic competition. Securing political support from other branches of government was not deemed necessary since President Vazquez was governing with a majority. This backfired when the government changed because the opposition parties had not taken ownership of the SNIC; on the contrary, it was considered an unnecessary solution for problems of “rich countries” (E2_SNIC). “The Frente Amplio had a majority in Congress, it is a party that responded to and channeled the demands of its bases but did not have a direct dialogue, a strong dialogue with (…) the right… it did not need to because it governed with a majority, and all social policies were approved. In Congress, he was not challenged by the demands that came from the right. During the implementation, he had no dialogue with the right” (E3_CU).

When in 2020 a new president, Luis Lacalle from the National Party, took office, interests realigned and the SNIC stopped being a priority. In the name of efficiency, the new government decided to merge it with the National Disability Program. As the former head of the SNIC put it: “For the sake of efficiency (…) they placed a policy at the same level of a program (…) and this accounts for a total ignorance (…) it supposes that you can carry out two such different things under the direction of the same person” (En Perspectiva, 2020). For many of the organizations involved in the design of the system (and with a seat on the Advisory Committee), “This was the first sign the government gave in terms of downgrading the importance of the system and its autonomy” (La diaria, July 8, 2020). The Care Board did not meet for more than a year, the National Care Plan’s release was delayed, and the budget for key programs was held back (Red Pro Cuidados, 2021). In general, there was a backlash against the original goal of the SNIC, and despite opposition from many organizations, over time different agencies that had managed to collaborate in specific tasks, started working again in a sectoral logic. “Fundamentally what lost impetus was the notion of a system, and there is a return to the logic of services, to services understood aside from the umbrella of the care System (…). Thus, some services remain, such as the CAIF centers and they are even seeking to expand (…) but not as part of a logic of care” (E2_SNIC).

Comparative analysis

Chile Crece Contigo and Uruguay’s National Integrated Care System are two integrated policy strategies following different pathways for their implementation. We have shown how these differences can be explained by looking at each component of the policy regime steering the implementation of the integrated policy over time. An explicit comparison of the evolution of the policy regimes during the implementation is shown in Table 1.

Table 1 The evolution of policy regimes
Full size table

Using a processual approach to study policy integration (Cejudo & Trein, 2023) we explain how, behind these contrasting experiences, there are two policy regimes with different capacities to keep a policy integrated. The literature of policy regimes (Jochim & May, 2010) allows for the identification of its constituent components and the role they play during policy implementation. Our contribution is to show how the attributes of the components of a policy regime manage to keep a policy integrated over time.

In Chile, during the implementation process, the core idea guiding ChCC allowed relevant actors to understand their responsibilities and how their actions articulate with the rest of the system, and thus contribute to attain ChCC’s goals. In Uruguay, the core idea, even if shared among the involved actors, did not steer their design and operation decisions, given its lack of specificity. This was reflected in the institutional arrangement of both cases.

Both in Chile and Uruguay the core idea materialized in an institutional arrangement that depicted their understanding of the institutional obstacles for an integrated implementation. The Chilean government wanted to overcome the sectorial logic guiding agencies’ performance, so the emphasis was put on how to secure that every agency involved did their part. Thus, the coordinating body was given the capacity to influence every agency’s decision making. Each agency involved in ChCC relies on ongoing information flows that account for their performance, because their budget depends on it, and so they willingly share their performance information. This allows its coordinating body to adapt the intervention to the changing needs of the target population. In Uruguay, the main institutional obstacles were perceived to come from bureaucratic resistance, but instead of giving control to the coordinating body, the emphasis was put in an institutional design that gave the agencies involved the sense of autonomy and of not being at the service of the SNIC, but part of it. As a result, agencies are not accountable to the SNIC and their information flows, performance and budget do not inform other agencies’ decision making. In short, in Chile decision makers exercise authority for keeping the policy integrated over time, whereas in Uruguay they do not.

Political support was crucial in both cases for the launching of the integrated strategies. It also shaped their operation and continuity. In Chile, sustained political support over time (from both government and opposition parties) kept open the flows of information and resources as well as effective coordination processes between different policy sectors, even after four changes in the governing party. In Uruguay, initial support from a government with a majority in Congress and a solid backing from party organizations was enough to create SNIC, but not to overcome bureaucratic resistance to integration. Moreover, once there was a change in governing party, the system stopped being a priority.

This is not a mere administrative difference. As a result of the contrasting trajectories of integration, the resulting regime in Uruguay is one where the integrality of the interventions an Uruguayan child can get is contingent on their family’s capacity to interact with the government: to their knowledge of the existing programs, their eligibility criteria, the process for applying and the luck to access all that they require. In Chile, this is a task for the government to carry out.

The comparative analysis of these two regimes shows the potential of the regime perspective approach to policy integration to study other sectors and countries. In Table 2 we summarize our approach to understanding how policy regimes structure the implementation of integrated policies. It offers a set of questions to understand how the three components of the policy regime are devised and their effect on the implementation of integrated policies.

Table 2 How policy regimes structure the implementation of integrated policies
Full size table

Discussion

This paper contributes to the empirical analysis of how the process of integration occurs and offers a framework for studying how policy regimes guide the implementation of integrated policy strategies. This understanding contributes to two other crucial debates in the literature on policy implementation.

First, we contribute to the ongoing debate about how policy integration is achieved. Existing explanations usually held a static approach to policy integration by focusing on its political determinants (Trein & Maggetti, 2019; Vince, 2015), contextual features (Biesbroek & Candel, 2020), or attributes of the policy mix (Maor & Howlett, 2021) at one point in time. Our analysis addresses that question but incorporates a dynamic view of the implementation process of integrated policies. We show how and why a policy is kept integrated over time and provide an analytical tool with concrete questions (with observable answers) about how this occurs (as shown in Table 2). Moreover, we provide a more nuanced understanding of the political dynamics that take place across policy sectors, for instance how implementing actors seize “the opportunity to redefine policy boundaries and to change policy provisions and frames while implementing them” (Sarti 2023).

On the other hand, we are in a better position to understand how and why policy implementation of integrated strategies succeeds or fails, without assuming that everything can be explained by the design of the integrated strategy. Indeed, our analysis challenges the top-down, linear approach to understanding policy integration and makes it clear that we need empirical research about actors and organizations’ interests, decisions, and capacities to understand its implementation and explain its results. Indeed, our analysis builds on the idea that policy integration entails the concatenation of two causal theories: one substantive (regarding the public problem) and the other more instrumental (that of the macro-causal theory) that provides the causal explanation of how everything around the policy (services, programs, implementing agents, information flows, budgets, etc.) should work together (Cejudo & Michel, 2021). Yet, by looking at how such “macro-causal theory” evolves over time, we show that even the coherent design of the different components cannot compensate for the lack of a theory specifying how they work together. Even in cases where integration occurs thorough bottom-up processes (Sarti 2023; Lambelet, 2023), integration was achieved when there was a theory that explained how the different components of a policy worked together.

Conclusions

Our analysis has shed light on the process of implementation of two integrated care policies in Latin America. The contribution of this analysis is not only in understanding these empirical cases but also in the capacity to explain how we may approach the study of the implementation of integrated policies, a key challenge in the policy integration literature (Trein et al., 2021). By recognizing the dynamic nature of policy implementation, we have made explicit the way ideas, interests, and institutional arrangements not only shape implementation but evolve to keep the policy integrated over time. We have explained what function these attributes of the policy regime perform to keep the policy integrated, and we have untangled the process of implementation to be able to observe how this function is performed in specific policies.

We have made three contributions. By unbundling Jochim and May (2010)’s analysis of the role of policy regimes during implementation, we have shown the specific way in which each constitutive element of a policy regime serves to keep policies integrated. Using the cases of Chile and Uruguay, we have also illustrated how the interactions among the components of the policy regimes create feedback loops that affect the implementation of the integrated policy and impact the evolution of the policy regime. Moreover, we have offered a more nuanced understanding of the forces that integrate or disintegrate a policy when looking at the way policy regime ideas, institutions and interests pull together or steer away the integrated policy from sectoral-based decision making toward an integral logic of operation. Together, these contributions advance a better characterization of the challenges that integrated policy strategies face during implementation and the role of policy regimes in overcoming them.

Our analysis opens new lines of research in the study of policy integration. First, a more detailed analysis of the process of implementation of integrated policies is needed in order to explore what are the policy capacities required to make the institutional arrangement perform as an integrative force. Second, further research on implementation may shed light on the conditions for the effectiveness of integrated strategies, beyond the focus on design that has dominated the literature. Indeed, if the policy integration literature seeks to inform policy decisions, it needs to move beyond understanding contextual factors or identifying obstacles to integration and explain the decisions and the instruments required to implement integrated policy strategies. Only by looking at how the process of integration works, it is possible to draw policy implications for practitioners interested in addressing complex problems.