Introduction: Complex Problems, Governance, and the Need for Policy Integration
Generally, governments around the world are confronted with complex policy problems “cross-cutting” (Rhodes, 1991, p. 212) and wicked such as homelessness, climate change, environmental protection, and migration issues (Ingold & Tosun, 2020; Metz et al., 2022). These problems are transcending the traditional boundaries and faced in individual policy domains or governance levels and jurisdiction (Candel, 2023), with various geographical and administrative domains across multiple levels of government and policy sub-systems (Peters, 2015). Furthermore, the modern state and governance itself are a precondition for existing government and governance challenges, caused by first, New Public Management reforms which are sectoral, departmental and have fragmented policies rooted in different siloes and, second, the complex organisational structure and tiers of modern governments themselves (Cejudo & Michel, 2017; Lidén & Nyhlén, 2024, p. 1). Secondly, the challenges are also caused by the “nature and complexity of policy problems themselves” (Cejudo & Michel, 2017; Lidén & Nyhlén, 2024, p. 1), which often can be characterised by multifactorial and multicausal issues.
To overcome these challenges, policy integration has been seen as the ‘Holy Grail’ or ‘Philosopher’s Stone’ in the policy sciences and has become a ‘buzzword’ for practitioners and Scholars alike (Candel & Biesbroek, 2018). Policy integration is seen as the key to overcome or at least contend with the complexity to a “common understanding about a problem and responsibilities of each of the parts in addressing and solving the complex problem” (Cejudo & Michel, 2019, p. 8). Therefore, this paper argues in line with Cejudo & Trein (2023) that sufficient policy integration is necessary for resolving complex problems.
In addition, in the administrative reform literature, concepts and theories such as New Public Governance, whole-of-government (WoG) or whole-of-society (WoS) theories have been used to overcome some of the fragmentation in governance exacerbating the complexity of public policy problems (Christensen & Lægreid, 2007; Lidén & Nyhlén, 2024). Dealing with these complex problems globally, the concept of policy integration has been embedded in these modern-day governments to “help to develop solutions to fragmentation and siloism” (Lidén & Nyhlén, 2024, p. 2). The underlying assumption is that “integrated policy designs, which reduce incoherencies or even promote synergies between governance efforts, will be more effective in producing politically desired outcomes” (Candel, 2023, p. 287). In South Africa, despite increasing use of concepts such as WoG, WoS and policy integration, coordination, coherence etc., there has been a persistence of fragmentation in government actions and challenges of translating policies into practice.
In South Africa, concepts such as policy coherence, integration, and coordination have become widespread in the South African literature and official documents, policies, and speeches. Despite these efforts, policies still seem to be ineffective in implementation. This paper suggests that the problem, in part, can be found in the policy designs (content) themselves, and that this would necessitate a shift in the level of analysis, towards policy design perspective. Historically, Weimer (1998, p. 182) found that when we study public policy, “we tend to emphasise the processes by which policies are made and implemented rather than the substantive content and impacts of the policy itself ” and according to Lasswell (1951, p. 3) the historic mission of the policy sciences was to “improve the concrete content”. Following on these assumptions, this paper argues in line with Peters (2022, p. 39) that “coordination and policy integration are fundamental challenges of policy design, and as Peters (1998) argues earlier, policy integration should be pushed to the forefront of policy design. This paper identifies policy design as the proper level of analysis as found by Alers (2022, p. 50) studying indigent household policies in South Africa that, “policy failure “is a problem of poor policy design”.
Historically, from an analytical perspective, policy content has become focus of study in 1980s with the conceptual split (forming two analytical domains) between the actual process of public policy making and the abstract concept of policy design as content (Linder & Peters, 1988, p. 742). Later Howlett & Rayner (2018) characterised it as, policy designing in the realm of formulation (verb) and policy design as the actual product (noun), the content. This paper takes a policy design perspective, where complex problems are addressed using suites of policy mixes, with a focus on their interacting content (objectives, instruments, targets) (Siddiki, 2020). However, even though various approaches and methods have been developed to improve policy content, it has received limited attention in the policy design literature despite the value of thinking about policy design in this way (Siddiki & Curley, 2022) and in this paper the argument is that it is even less in policy integration. In a recent review of one of these methods, the institutional grammar (IG), Pieper et al. (2023) found zero studies using the method in Africa. In response, this paper proposes policy design-as-content methods for studying policy designs and application of policy integration for improved analytical capacity, which is instrumental in making technically sound policies to contribute to goal attainment (Wu et al., 2018, p. 3). To this end, objectives of this paper is to:
-
Review the state of policy integration in South African literature and governance (administration of the state).
-
Explore the existing policy design-as-content methods are useful for the study of policy integration.
-
Examine how these methods might be used to improve policy integration to enhance policy implementation and governance in South Africa and improve policy capacity.
To this end, the paper is structured in the following way. Firstly, the paper review and provide a brief overview of the state of policy integration in South African scholarship and administrative reforms and governance. Second, the paper conceptualises policy integration based on its two analytical dimensions policy integration process and policy product information, in which the latter makes policy amenable to policy design-as-content approaches for advancing policy integration by shifting the level of analysis to policies themselves. Third, the proposed method for analysis, the Institutional Grammar 2.0 abbreviated IG 2.0 and the integrative propositional analysis (IPA) to outline various existing insights that could be useful for advancing policy integration, followed by the discussion and some areas for future research. The neglect of policy design-as-content approaches is an untapped reservoir of conceptual insights that can provide more precise analysis to improve both descriptive and diagnostic capability through some generalisable qualities of policy content. There also emerges a new language for policy makers and scholars for designing effective policies. Perhaps, most importantly, this may provoke new research questions about the relationship between policy designs and their outcomes. This analytical and integrated understanding may also provide prospects for a research agenda for enriching Public Administration curricula as well as teaching and research.
Methodological Justification for the Study
This is a conceptual paper which according to Gilson & Goldberg (2015, p. 128) can “bridge existing theories in interesting ways, link work across disciplines, provide multi-level insights, and broaden the scope of our thinking”. Of importance is that a well-designed conceptual paper must explicitly justify and explicate decisions about key elements of the study (Jaakkola, 2020). This conceptual paper can be seen as adopting a theory adaptation research design (Jaakkola, 2020), which is achieved by introducing alternative frames of reference to propose a novel perspective to an extant conceptualisation (MacInnis, 2011). The point of departure for such papers is to problematise a particular theory, as discussed in the introduction, policy integration in the main is studied through the process of designing, instead of focusing on the policy content itself.
In this paper, as in other conceptual papers, it is important to identify two types of theory because of the difficulty to follow which provide the ‘data’ and which are framing the analysis (Jaakkola, 2020), and they are domain theory and method theory. According to Lukka & Vinnari (2014) the starting point for the theory adaptation paper is the domain theory, which in this case is policy integration and the method theories, which in this case is policy design-as-content methods to provide an alternative frame of reference by switching the level of analysis for policy integration to its content. Generally, the role of the method theory is to provide new insights into the domain theory, which can for example offer new or alternative explanation. The focus therefore is around policy integration (domain theory) in this case, and not on policy design-as-content (method theory) as such. Policy design-as-content method is chosen because of its ability to address the observed shortcomings in the existing policy integration literature and provides what Jaakkola (2020) calls supplementary value.
As a theory adaptation design, the goal of the paper is to identify new dimensions of policy integration by introducing a new theoretical lens of policy design-as-content methods, by means of (1) switching the level of analysis (policy content) using these theories to explore new aspects of policy integration, and (2) policy design-as-content methods provide strong pathways to the study of policy integration. Drawing on multiple theories/streams in the policy (policy design, policy mix, policy mapping) literature, the paper discusses the potential of policy content as an abstractable unit of analysis for studying policy integration which makes it amenable to various methods such as the IG 2.0 and the IPA methods to improve policy integration. It is important to note here that for the purpose of this paper, like other conceptual papers, the focus is on proposing new relationships among construct with the purpose of developing logical and complete arguments about these associations rather than testing them empirically (Gilson & Goldberg, 2015).
Governance Challenges, Lack of Policy Integration Research and a Missing Piece in the Puzzle
Various studies suggests that the South African government has various challenges in policy implementation and governance across various complex problems. This section provides a brief review and overview of the state of policy integration literature and governance in South Africa.
According to Naidoo (2013), the government had difficulty with delivering shared policy mandates in strategic socioeconomic development and Pieterse and Corolia (2016) argue that the increasing number of housing policies causes confusion which is exacerbated by lack of coherence and contradictory spatial outcomes. Raphasha (2015, p. 99) studied innovation policy and found that there was a lack of alignment and synergy across national innovation policy instruments, weak levels of complementarity in the overall strategic goals, and too much focus on individual policy instruments rather than the policy mix of instruments that can address the issue. This was coupled with a lack of understanding of the interactions of these instruments across different administrative levels. More recent studies on food security found that the limitations in national food security are due to a lack of coherently embedded policy governance, weak policy integration causing a drift away from the main policy goal, and a call for the government to invest more in policy integration, coordination and building capacity to reduce high levels of hunger and improve food and nutrient security (Hlahla et al., 2023). Similar findings were seen in the work of de Wee & Jakoet-Salie (2025) in their study of the South African National Policy on Food and Nutrition Security.
An evaluation of the National Health Insurance Phase 1 by South Africa, Evaluation Report (2019) found that the “value of coordination and integration between the intervention strategies had gone unrealised” and various programmes “worked in silos and integration is needed”. In a study on the public health system in the Free State, Malakoane et al. (2020) found that there was a lack of service integration owing to poor policy coordination, which was further exacerbated by a silo mentality and “bypassing” existing policies causing the fragmentation of health services”. Muthathi & Rispel (2020), reporting on respondents in their study, found that problems were also experienced owing to overlapping roles and responsibilities, which were extended by poor communication, weak or strained relations, and lack of accountability, caused by a lack of policy coherence, creating disjuncture in implementation. A recent 10-year review of the National Development Framework by the National Planning Commission (2023, p. 45) found that policy challenges included ”fragmented and lacked coordination as well as integration and did not address developmental challenges like poverty alleviation, and income inequality”. In addition, the NPC (2023, p. 53) also found poor socioeconomic transformation, which was diagnosed through a “wide range of factors, including policy incoherence”. This, in turn, led to “wasteful duplication and a lack of coordination across departments”. The NPC (2023, p. 52) made a recommendation for “consistency in government policy communication”, which this paper argues is to be achieved through policy integration to procedurally improve policy coherence, coordination and integration across policy sectors and different levels of government.
What is apparent from the brief review is that the state of administrative action and governance is facing various challenges and according to Lidén & Nyhlén (2024) policy integration is seen as a pre-condition for effective institutional arrangements, decision-making, coordination, and cooperation. However, in the case of South Africa there seems to be a lack of research focusing on policy integration despite its bourgeoning in the international literature. de Wee & Jakoet-Salie (2025) also makes the case that the concept ‘policy integration’ is widespread in official speeches, policies, and government documents, but lacks in the South African literature.
In their study, de Wee & Jakoet-Salie (2025) found in their research that the term ‘policy integration’ has been used less than 10 times through a basic search on the Sabinet repository focusing on the two journals Administrative Publica and Journal of Public Administration. This in contrast to review of the international literature on policy integration by Kaplaner et al. (2023, p. 2) indicating that between 2018–2021 the concept of policy integration has been used in policy science and other journals on the Web of Science over in over 300 articles. This is concerning for South African public administration and governance because administrative reforms such as whole of government, joined-up government, and whole of society approaches needs policy integration to facilitate the plurality of role players in achieving the different policy goals, implementing the various instruments, and affecting the different targets in multiple geographical areas. The studies that focus on policy integration process with a government focus including, on various levels of analysis/focus, including, policy indicator consistency and coherence by implementing management and governance outcomes policy integration from a cooperative governance and intergovernmental relations perspective and analysing cross-sector policy objectives and how they are aligned or synchronous, but they did not have tools to properly measure or analyse across policies (Funke & Roux, 2009, p. 19). A study focusing on policy content by Zembe et al. (2022) was descriptive in nature of the content only. A study focusing on policy content (not integration) by Alers (2022, p. 60) suggested that “municipalities’ indigent policies need to be developed in such a way that they integrate with the broader municipal plans”. Again, absent in the study is a specific recommendation to measure or diagnose how this can be done. Accordingly, almost none of these studies focus on content and de Wee & Jakoet-Salie (forthcoming, drawing on Cejudo & Michel, 2017) also observed in the literature that various studies in the South African literature convolute the concepts policy integration, coordination, and coherence.
Conceptualising Policy Integration
Policy integration refers to a governance system’s ability to address a particular cross-cutting policy problem(s) in a more or less holistic manner, by fostering coherence between the goals and instruments of its constituent subsystems (Candel, 2023, p. 287, drawing on Candel & Biesbroek, 2016). In the progression in the literature, policy integration has developed into an umbrella concept that can be defined “as a political process that entails the coordination of actors from different policy subsystems, the combination of instruments from different policy sectors, as well as arrangements for their consistent implementation and evaluation to address different dimensions of a complex problem...” (cf. Cejudo & Trein, 2023, p. 30). The understanding is that policies exist within policy subsystems which are subsets of the political system with various actors and policy issues, with separate decision-making ‘silos’ (Trein et al., 2023).
However, policy integration is a latent concept and difficult to define but can be conceptualised at least in two ways as outlined by Tosun & Lang (2017, p. 4), as government-centred and governance-centred approaches. The first, focusing primarily on public policy processes and implementation through an integration lens and the latter focus on integration of policies based on organisational and institutional dimensions (including organizational reforms, etc.) (Trein et al., 2019). Knill et al. (2020) conceptualises it even more narrowly into the policy integration process and policy product integration. The two differ in the following way: policy integration process, focus on the structural and procedural setup of the policy making system and policy product integration (content) focus on the integration and consistency of cross-cutting policies in relation to other policies in the sector. The latter is the focus for this study as it is largely an understudied area generally, but particularly in South Africa. The key analytical focus of policy product integration “concentrates almost exclusively on the policies themselves and whether or not their design is capable of addressing cross-cutting policy challenges (Knill et al., 2020, p. 10). Candel (2021) argues in line that policy integration attempt to align policy variables (for example goals, targets, and instruments). The focus of this conceptual paper therefore is motivated, in part, because of a lack of studies focusing on policy design-as-content using “approaches for valid and reliable measurement” (Siddiki & Curley, 2022, p. 122).
To advance the study of policy integration from a policy design-as-content perspective, the following section clarify and identify policy design as a key analytical locus for policy integration drawing on various streams of policy research streams.
Policy Design-as-Content: Identifying the Analytical Locus for Policy Integration
For ease of understanding and conceptual clarity, consider Siddiki’s (2020, pp. 1-2, emphasis added) policy design heuristic:
… policy design as formulation focuses on the choice of policy instruments, and the individual, societal, and contextual factors that influence instrument choice. For policy design as content, the focus is on the functional, structural, and substantive features of the content in which the selected instruments are embedded.
For policy integration, it is important to describe the proposed analytical locus from a policy design perspective. In the creation of the two analytical domains, Schneider & Ingram (1997, p. 2) defined policy design-as-content as the “content and substance of public policy”, which they later characterise as “blueprint or architecture” found in wordy, formal text-based documents such as policies. They also argue that policy content has a structural logic, which they define as “patterns through which policies address problems or seek to achieve its goals (Ingram & Schneider, 1990, p. 68). Cejudo & Michel (2015, p. 6) makes the case that policy content has a “causal theory that structures a policy, the internal coherence of policies (including) … logical connection and causal articulations between the definitions of the problem, the policy instrument and expected solutions”. Of importance here is that policy designs have an empirical phenomenon internal to it, that can be studied. More on the practical side to study the structural/causal logic, Ingram & Schneider (1990, p. 68) suggests that “… just as it is possible to diagram a sentence linking together parts of speech, it is possible to diagram the structural logic of policy by showing the relationship among the elements”. Conceptually, these are important advancements as it creates an analytical domain to study the content of policies which could lend it to the study of the integration of different policies based on their content.
An significant contribution for the importance of focusing on policy content can be drawn in the propositions by Siddiki (2018b, p. 219) that “policy effectiveness is predicated on the soundness of the policy’s structural logic, as when it lacks integrity it tends to lead to policy failure” and that of Cejudo & Michel (2015) stating that when consistency in causal theory fails, the policy loses effectiveness because implementation is based on false assumptions, as it does not coincide with reality. This suggests that that the content of policies and their inherent logic is central to its effectiveness and the paper argues that the ability to map the structural logic of policies and the relations of different policy elements (components) opens up and allows for an extension into policy integration (between policies) and open up an analytical locus based on the structural logic of a policy.
In the original definition of policy integration according to Underdal (1980, p. 159) as “constituent elements brought together and made subject to a single, unifying conception”. Important for the purpose of this paper is the idea of ‘unifying conception’ which suggests that structural logic (content/conceptual) is the proper level of analysis for policy integration as decision-making logic in large part is sanctioned by what is embedded at the conceptual level—policy content.
A key insight from the early work of Ingram & Schneider (1990, p. 68) is the idea that “… it is possible to diagram the structural logic of policy by showing the relationship among elements”. There are various methods in the policy design literature for the study of policy content, but few are amenable to the study of policy content (they are discussed later). However, drawing on the insights from the structural logic of policy as a potential analytical locus, various concepts have been developed that allows understanding of how policies can be studied a mapped/diagrammed for studying the integration of policies. Two of these concepts are intra-policy compatibility and inter-policy compatibility (Siddiki, 2020). The latter is of important for policy integration.
Intra-policy compatibility refers to the incompatibility among instruments and goals embodied in a policy design often leading to intra-policy conflict which “indicates that the policy is designed by an inappropriate causal (structural) logic” (Siddiki, 2018a, p. 222) or a misspecification of the link among the causes of the problem and the solution designed to address it (Siddiki, 2020, p. 35). Inter-policy compatibility refers to the extent of alignment among policies that govern common and different subjects, targets, and policy issues (Siddiki, 2020, p. 32). If they are not compatible, they are in conflict. As such, from a policy conflict perspective, Siddiki (2018b, p. 221) suggests that to overcome “inter-policy design conflicts, deliberate attempts should be made to map out the policy landscape in which new policies, or iteration of policies are being applied”. In addition, that building a policy map can then “enable investigations into the likelihood of negative policy interactions” (Siddiki, 2018b, p. 221). This would become the bases for evaluating policy coherence based on the mapping of policy interaction which is based on the suggestion of Siddiki (2018b, pp. 221-222) that “policy mapping and interaction assessment should formally be incorporated as part of the policy analysis process” by focussing specifically on “evaluating the causal assumption” … “ensure the validity of goals-targets-instrument relationships”. (logical connections between policy elements) “underlying policies”. This is important because it suggests that policy goals and instruments in individual policies can be mapped based on their components and relations, and if the analytical locus of structural logic exists at the individual policy level, it provides an appropriate locus and amenable to policy integration.
For policy integration, it is important to understand the composition of policies and the possible and actual interactions across different policy subsystems. To understand this, it is worth drawing on the policy mix research, focusing on policy interactions, is the idea that to “create(ing) policies whose realisation is not obstructed by internal contradictions or negative interactions across the different levels of government” as this will lead to “unintended interactions between the policy’s instruments and/or goals” (Knill et al., 2020, p. 13). The argument is that there are instrument interactions (Parsons & Barling, 2022, p. 3) or, as many argue (del Río, 2010; Navin & Attwell, 2019; Trencher & van der Heijden, 2019), policy mixes are seen as a system of interconnected elements susceptible to simultaneous mutual or reciprocal interactions. These interactions can be multi-level, and across sectors with multi-policy, multi-goal, and multi-instrument mixes, which may result in unexpected effects across a wide range of contexts, making their analysis and implementation challenging (Howlett & del Rio, 2015; Maor & Howlett, 2022). As such, Howlett et al. (2015, p. 8) conclude that “complex policy mixes inherently involve interactions between the different instruments of which they are composed, either in the form of conflicts or synergies”, horizontally and vertically.
From a policy design as content perspective, the focus is both on horizontal and vertical dimensions, with horizontal focusing on the co-existence of policies across policy subsystems, and vertical focusing on the constellations in which policies cross-jurisdictions are between at least two levels of government (Knill et al., 2020). This is akin to the idea of policy mapping and creating a policy landscape that is multi-level (Siddiki, 2018a). This paper proposes that integrating policy design and policy mix research can be used to validate the possibility of studying policy content as abstractable units of analysis that can be used to describe, analyse, and diagnose policy design with potential key insights for improving public administration and governance. The aim therefore is to study the interactions at a granular level to more accurately analyse the negative and positive interaction to improve policy outcomes or beyond that, to be self-reinforcing manner (high coherence) (Gerber et al., 2009).
From a policy integration perspective to the interaction between policies, Cejudo & Michel (2017) and Cejudo & Michel (2021) use the concept macro-causal theory, which moves away from the initial causal theory definition (for an individual policy) towards, “how the different components fit together and interact” (Cejudo & Michel, 2017, p. 758). The macro-causal theory is the explanation for “how and which policies and organisations [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” set out how these different policy components interact (Cejudo & Michel, 2017, p. 758). For integration from a policy design-as-content perspective the focus should be on exploring and analysing policy statements/directives or propositions in different policies and how they support or conflict with each other (Chinsinga et al., 2012) or the interlinkages of legal documents (policies content), focusing on their mapping. For the mapping of this ‘macro-causal logic’ this paper proposes the idea of mapping a policy landscape with the content which is explored in the following section.
Siddiki (2018b, p. 221) suggested that policies should be mapped to create a policy landscape in which new policies or iterations of policies are applied. This is an important point to which this paper heed, however, the policy landscape is not defined, and efforts are made here to define it for policy integration. Based on the idea of mapping out a policy landscape of policies, this paper draws on the work of Mettler (2016, p. 370) defining a ‘policy scape’ as a political landscape cluttered with public policies… (that) is not unpredictable, neither is it static: rather it is a dynamic environment in which policies evolve and change over time”. The author continues that these policyscapes structure political order and because of the dynamic nature of policyscapes they fail to function as in their inception, both because the external socio-economic environment changes, but also, from “dynamics internal to the policies themselves” (2016, p. 370). This idea is of importance for this paper as it suggests that the internal dynamic of the policies themselves also affect the policy effectiveness, which makes the focus on the policy product (content) itself very important, though under investigated.
For the purpose of this paper a policy landscape can be seen as a product of policy mapping, interaction assessment that creates the ‘spatial distribution of policy components and their inter-sectoral and multi-level interactions in an integrated map. These include for example, goals, targets, and instruments at different sectors and different levels of government.’ These would be the integration of two or more policies’ structural logic into a landscape. This type of conceptualisation is important as the different policy components interact and may affect the effectiveness of one another.
At the most fundamental level, policy integration can be described as a political approach that involves (1) coordinating actors from different policy subsystems, (2) consolidating goals and instruments from various policy areas, and (3) aiming to achieve more coherent and consistent public policies (Cejudo & Trein, 2023; Kaplaner et al., 2023, p. 4; see also Knill et al., 2020; Trein et al., 2023). It requires different organizations and policies/programs to work under a new logic, where their subsystem logic is essentially subordinated to the overall new goal (Cejudo & Michel, 2017). Cejudo & Michel (2015, p. 11) argues that “… at every moment of the policy process decisions are made on a new policy (and) … decisions are based on the integrated policy not the individual components”. In this case the new logic would be that of the policy landscape of ‘integrated subsystem logic’.
In a framework developed by Cejudo & Michel (2017, p. 759) they indicate that various policies and their goals, instruments, target population and inclusion/exclusion of new policies are important information for decision making. What the framework outlines indicate is that policy integration according to de Wee & Jakoet-Salie (2025) is a continuous and iterative process rather than a static moment in time. This is important as the mapping of the policy landscape allows for new and old policies to be evaluated or analysed continuously in the evolution of the policy landscape. This is parallel with the idea that of Mettler in her conceptualisation of the policy landscape considering the inter-temporal nature of policies.
Existing Policy Design-as-Content Approaches: Some Important Insights to Leverage for Policy Integration
In a recent book by Siddiki (2020), the various approaches to study policy content on various levels based on analytical foci were divided. There are three levels, namely, macro-, meso- and micro-approaches, which can be used for analytical leverage in policy design assessment. The micro-approach chosen for this paper is the IG Tool and an additional tool, IPA (de Wee, 2024), which sits well in the micro-approaches.
It is important here to remind the reader that this paper just provides an overview of the proposed approaches and their potential insights that can be leveraged for the purpose of public policy and public administration scholarship in South Africa. Drawing heavily on four latest meta-reviews of the IG scholarship, some key insights with potential leverage points for policy integration are discussed (Dunlop et al., 2019; Frantz & Siddiki, 2022; Siddiki et al., 2023).
First, the IG is used by public policy scholars sought to understand the design (content) of policies and are interested in assessing how policies structure individual or collective action amongst actors (Frantz & Siddiki, 2022) as they provide the official rules of the game for society as we work together to provide public goods and solve complex social dilemmas (Heikkila & Andersson, 2021, p. 309). These complex social dilemmas or problems require collective action including the production of public goods or services or managing shared resources (2021, p. 311). Collective action happens in the ‘action situation’ (Ostrom, 2005) in which policies influence an individual’s decision to contribute to collective effort to address a shared problem (Heikkila & Andersson, 2021, p. 318). Thus, the situation where the structural and behavioural interactions of actors are defined by policies.
To deal with collective action, the IG 2.0 was developed as a revised version of the original IG devised by Crawford & Ostrom (1995), to support the representation of the nested nature of institutions (Frantz & Siddiki, 2022) or one might say different policies are embedded in one another. For the purpose of policy integration, the IG 2.0 advance understanding of how “… policies that may be embedded within other policies, govern collective action… (and) provides an abstractable unit of analysis for studying collective action” (2022, p. 300) as in any social setting— the action situation, that jointly produce outcomes (Ostrom, 2005). From a systemic perspective it can be seen as “constellations of institutional statements” (2022) characterised by intertwined encoding of instructional statements and their linkages (2022, p. 307), allowing for “fine-grained atomic representation… which enables the ability to represent richly interlinked configurate structures and extending to the extraction of arrangements of action situations or institutional settings more broadly” (Frantz & Siddiki, 2022, p. 311). This is akin to the idea of a policy landscape discussed in the preceding section. Plainly, the policy design content itself can be used as an abstract unit of analysis for describing, analysing, and diagnosing action situations, or in this case integrated policy outcomes. This work is “… aimed at leveraging understanding of policy content towards assessments of antecedents and consequences of policies bearing particular design features” (Frantz & Siddiki, 2022, p. 310).
Some studies using IG have shown some insights useful for policy integration including, but not limited to, how policy actors coordinate and compete or engage in conflict (Schlager & Cox, 2018), interactions among actors in response to rule configurations (Oliver, 2019), how institutions (policies) influence individual and collective behaviour and systematic outcomes resulting from such (Frantz & Siddiki, 2022, p. 299), rule legitimacy or levels of compliance/non-compliance (e.g., Angulo-Cazares, 2018; Siddiki et al., 2019), the granular analysis of policy text also allows for identifying specific role-players responsible for an action, and the specificity of the target, both the ones to implement and the beneficiaries of the policy. In a recent paper, Pindaru et al. (2023) used IG to study policy integration, focusing on European Green Deal, the Sustainable Development Goals and investigating better integration of environmental considerations into other policy areas and ensuring coherence when formulating new policies Integration into other policy areas.
Because of a space limitation, other insights are not discussed, however, the point is to indicate potential insights for improving policy integration based on the conceptual insights from this approach and its analytical tools. The IG allows for the coding of policy documents as a basis for statistical analysis, new insights to institutional diagnoses through dissecting policy and linking them to actions of targets, for descriptive analysis and assessment for the purpose of policy analysis, program evaluation and diagnosis, based on different institutions and their linkages to outcomes. IG-coded data is commonly linked to empirical data (e.g., existing datasets or rules in use) in order to establish correlatives between policy and outcomes (Frantz, personal communication, 2023). This might include analysis of how policies affect behaviour and social outcomes which together allow for the generation of rich insights about policy designs (content) and their outcomes.
The second method is the IPA, which studies policy design as content from a conceptual system perspective. Wallis (2016, p. 579) defines conceptual systems as “any form … policy model, model, mental model, policy”. Each of these systems can be defined as a set of interrelated propositions (Weick, 1989), which can be seen as “a declarative sentence expressing a relationship among some terms” (Van de Ven, 2007, p. 117). Thus, these propositions can be dissected at the granular level to obtain concepts. Concepts can be defined as named abstractions (James, 1909) where they exist in some systemic relationship. In the case of policy designs, concepts will be different policy elements and their relationships which create the causal logic, drawing on Schneider & Ingram’s (1988) description of structural logic. As all policy documents are composed of policy statements characterised by propositions, it is possible to analytically map all policies based on their propositional statements (policy directives) and determine their structural logic. The primacy of mapping based on Siddiki (2018a) is also important in this IPA method. Using concepts and causal relations, policy statements/directives/sentences can be analysed and mapped. Because these policy statements have causal logic or assumptions between the boxes, it also allows for integration of policies. As a result, IPA can be used to analyse and determine a policy design’s structural logic (de Wee, 2024).
In the IPA, using conceptual systems as a unit of analysis, there are ecological similarities between a real-world or physical system (in the real world, or policy environment) and the conceptual system, including policies that are abstractions of the system. The idea is that the latter can be used to alter the former. The idea needs to be considered that the world exists out of a material universe and a conceptual universe. The former is nested in the latter, and according to Warfield (2003, p. 515), “any portion of the material universe that can be separated in thought from the rest of the universe for considering various changes which may occur in various conditions”. This idea is akin to Linder & Peters (1988, p. 72), who caused the conceptual split in policy design, suggesting that design should be freed from the policy process after which it “ceases to be a real-time phenomenon” and “becomes an activity directed towards the policy process but separate in time and space”. For policy integration, the idea is that we now have accurate definitions and an understanding of the various policy components as identified by Howlett & Cashore (2007). This has been increasingly specified in terms of defining policy goals (Petek et al., 2022) and micro-dimensions of a policy focusing on targets, tool calibrations (Capano et al., 2023) and instrument shapes (Capano & Toth, 2023).
For policy integration, the IPA six-step methodology can be used to deconstruct the policies into their constituent concepts (which describe a component of the policy) and, thereafter, the causal relations between the concepts (forming propositions) (see Wallis (2019) for paper on integration using the IPA). Thereafter, using this method, the process can be performed on all policies and have laser focus on analysing, describing, and diagnosing, by mapping out the policies and building a policy landscape by integrating the emergent structural logics.
The IPA has also been used for policy analysis and design, focusing on the structural logic of policy based on the measure of systemicity, quantifying the structure and content of the policy, and comparing it with the implementation outcomes of the policy (de Wee, 2020; de Wee & Asmah-Andoh, 2022; Wallis, 2014; Wright & Wallis, 2020). Studies have focussed on the internal logic structure of policies, and a recent study by de Wee (2024) focused on mapping the policy and intra-policy compatibility and studied how the different elements of the policy were linked and how they assisted in achieving policy outcomes. These studies have diagnosed policy ineffectiveness as a result of a lack of understanding because of poor structural logic guiding those who implement the policy amongst other reasons (de Wee & Asmah-Andoh, 2022). In addition, that policy effectiveness is based on the structure that is integrated as it contextualises role-players and their place in the policy and the various policy instruments that lead to and/or work at cross purposes. The diagnostic capability of IPA is extensively discussed in de Wee (2024). Work on the IPA can be found in various paper applying the methodology and extensive work on studying the structural logic as a conceptual system in recent work by de Wee et al. (2024) where they propose the system-theoretic perspective to conceptualise policy content as a conceptual system which provide the locus for integration based on the fact that all policies are created on paper (in the main) and that they can be mapped and integrated as a system. de Wee et al. (2024, p. 10) holds that “systems insights of the IPA are that through building policy maps as an abstractable unit of analysis (using the propositions found in policies), it allows for policy makers or academics to build policy landscapes, which can help investigate possible policy trade-offs. These implications could be valuable for policy integration or policy coherence, which is predicated on the idea of policy compatibility”. These insights are important developments for policy design-as-content which can be extended to the study of policy integration.
Discussion: Potential Theoretical and Practical Implications for Policy Integration
Policy integration as an umbrella concept has developed in parallel with the governance literature as to deal with complex problem mixes of policies are needed (de Wee & Jakoet-Salie, 2025). The focus in the main has been on the policy and political process around policy integration including the focus on the coordination of actors, resources, and implementation across various subsystems with a neglect of the policy product integration (content) focus. Responding to the problem, this conceptual paper focused on contributing to the current debate by adding a policy design-as-content perspective. The theoretical adaptation from the conceptual integration is important in many ways.
First, it is argued that integrating policies means “creating a new policy in which individual components (policies) work under a new logic … (where) … decisions are based on the needs and priorities of the set of policies and organisations being (with) overall logic that would determine decisions such as targeting, budgeting, etc “(Cejudo & Michel, 2015). This ‘overall logic’ is what they would later call the macro-causal logic. From a policy content perspective, this paper proposes the idea of mapping the policy landscape (based on the content of the policy and their interaction between the policy mixes) with the aim to better understand the manner in which the policies interact and how these interactions can be managed (Capano & Howlett, 2024). What is important about policy landscapes is that they become institutions that govern the behaviour, resources, and relationships between different administrative units in a multilevel governance system. According to Migone & Howlett (2024) the complex interconnections among policy objectives, means and instruments requires clearer analysis and conceptualisation. Using the policy design as content approaches IG and IPA, this paper suggests that focusing on the structural logic and the development of policy landscape of integrated structural logics allow for a more granular and in-depth analysis of the different components of the design (Capano & Toth, 2023) and between design. Inter-policy compatibility and inter-policy conflict concepts (Siddiki, 2020; 2018) allows for conceptualising the analytical locus in which the content of policies become and abstractable unit of analysis with implications for policy and governance.
Second, in the integration of policy mixes, from its original definition, Kern & Howlett (2009, p. 395) claims that “(they) develop incrementally over many years”. For policy designs themselves, Siddiki (2020, p. 2) states that in understanding policy dynamics, policy designs “evolves over time as the context in which it situates change”. This speaks to the inter-temporal nature of policy designs to their context. According to Howlett et al. (2015, p. 2) the process of policy de-composition is useful analytically as it allows for each component changes on its own to be better observed … as well as the interaction between them which determines the overall nature of policy dynamics visible over time”. What is important here is that the change in the policies (individually) can heavily impact the coherency, consistency, and congruence of policies (Capano & Howlett, 2024) as over time policies change through layering, patching, and packaging. The argument therefore is that even if policies are well-crafted or integrated initially, they degenerate through successive round of change, and according to Migone & Howlett (2024, p. 3) “better understanding how mixes are initially composed and how they evolve is a key question for analysts and practitioners alike”.
In response to Migone & Howlett (2024) this paper argues that from a policy design as content perspective this is possible to maintain the ‘integrity’ or structural soundness of the entire policy landscape using the proposed methods. For example, work by Ambrose et al. (2024) studied policy evolution (change) of Net-meters using the policy language captured in the provisions and how they changed over time. Synthesising literature on policy change, including understanding policy layering that includes patching (new statements without readjusting previous policy), packaging (layering new statements while terminating others to readjust previous policy), and calibration (making changes to text within the statements (Ambrose et al., 2024). de Wee (2020) used the IPA to study the Drakenstein Housing policy and found that the 2010 policy differed from the 2016 policy in that it included an additional provision and took other sentences out, however, the added provision was unsuccessful as it had no clear relationship with the rest of the policy and was generally not politically supported or budgeted for.
Third, heeding to the call of Siddiki (2018a) to map the content of policies has various implications for the study of policy integration from a policy design-as-content perspective. First, the I.G 2.0 using institutional statements allows for the studying and mapping of the intra-institutional statements (structural logic) as well as building policy landscapes with the constellations of policy statements/institutional statements (Frantz & Siddiki, 2022). The IPA using propositions of policy statements does similarly, where propositions can be diagrammed/mapped which can be studied across levels of government and between policy subsystems. Mapping therefore provides policy integration with a common language for stronger policy integration using the proposed methods (although not an exhaustive list). This level of analysis provides “a more granular and in-depth analysis of different components of the design … makes it possible to assess how effective the instrument is and under what conditions” (Capano & Toth, 2023, p. 8). Allowing for this, suggests that analysis, practitioners, and academics can excavate exactly how a policy works and how the policy is concretely delivered.
Fourth, very few policy evaluation researchers have taken up this challenge in policy integration. In addition, there is a lack of frameworks to measure the success or failure of integrated policies (Candel, 2017, 2021). The argument is that there are several interactions among policies and organisations that make it difficult to uncover the precise effects of policies (Cejudo et al., 2021, p. 975). This paper argues that policy mapping based on policy content and studying policy designs as conceptual systems using the IPA could allow for this. In part, it also draws on the insistence of Siddiki (2018b, pp. 221-222) that, based on policy mapping and interaction assessment, policy evaluators should be “evaluating the causal assumption (logical connections between policy elements) underlying policies”.
Finally, from a Public Administration and policy integration perspective, additional types of capacities are required. Parsons (2004) defined policy capacity as the weaving function, which weaves together the multiplicity of organisation and interests to form a coherent policy fabric. Thus, this weaving function is required to navigate through the complexities of interconnected problems, multi-level governance, multiple fault lines and multi-organisational settings, cross-cutting issues, policy networks, organisational interdependencies, and linkages” (Parsons, 2004, p. 44). As different role players (agencies, ministries, and levels of government) design and operate policy instruments (for example, advice and training, grants, loans, regulatory or fiscal incentives), integration and coordination are essential (Peters, 2015). The idea of building policy landscapes is important because it suggests the need for coherent policy fabric in which policy sub-systems emerge beyond their sectoral or sub-system level. Future studies could also engage in studies to ascertain the validity or value of such research for overcoming public administration challenges. This type of research based on policy content provides a nuanced framework as it describes governance arrangements of a country and hence suggests the import insight of IG 2.0 and IPA for policy integration and the descriptive, analytical, and diagnostic insights for policy coordination and governance arrangements (such as WoS).
Concluding Remarks and a Way Forward
Policy problems are complex, and they need to be solved through various policy sectors which necessitates the governance of integrated policies. However, the policy integration literature has focused so much on the policy and political process of policy integration in South Africa and elsewhere, despite the importance of policy design in policy integration. In fact, some authors argue it should be pushed to the forefront of policy integration. This can be seen in the fact that policies (their content) determine who gets what, where, when and how, and therefore determines the action and behaviour of governments and governance itself. Because of the limitations of policy integration in its application in South Africa, and because it is relatively neglected, this paper suggests that the focus should be on the policy content. The paper proposed various policy design-as-content methods for studying the content of policies, for their ability to map the internal structural logic and the relations between these logics to build a policy landscape (or macro-causal theory according to Cejudo & Michel, 2017). This in turn allows for more in-depth and nuanced approaches to the study of policy integration, which is important for effective policy making (preventing policy conflict and undermining instruments and objectives) and governance. In understanding that policy integration is an iterative process, the policy landscape also provides an additional layer of analysis in policy designing as new policies and their content can be evaluated ex ante based on how they ‘interact’ with the existing policy landscape.
Wu et al. (2018, p. 3) suggest that analytical capacity is instrumental in making policies more technically sound to contribute to goal attainment, facilitate implementation, and obtain political support for actions. This is in line with Candel (2021, p. 351) emphasising the need for “policy capacities for policy integration (that) requires additional types of these capacities”. In addition, Peters (2018) finds that increased policy integration can reduce many inefficiencies in public policymaking. This perspective could help with policy transfer successful policy transfer (Lee et al., 2022) as these methods can be used to adapt new policies to their context in a larger policy landscape and advancing policy networks (Kim & Moon, 2021).
The potential of this type of research is important because the analysis and coding of policy documents can provide statistical analysis that supports studies of how different policies have different substantive features. In addition, the institutional grammar tool (IGT) and IPA provide possible new insights into policy integration by dissecting the designed content of policies and linking them to the implementation on the ground and, specifically, to the actions of targeted actors within a policy subsystem. Studies using these approaches in their analysis, creation and evaluation of policy integration can help support policy actors in fulfilling the policy goals and objectives in various complex policy issues. Therefore, using these approaches (IPA & IG 2.0) as an additional layer of analysis and its potential insights for policy integration are “aimed at leveraging understanding of policy content towards assessments of antecedents and consequences of policies bearing particular design features” (Frantz & Siddiki, 2022, p. 310).
This paper was primarily concerned with the analysis and description of policy integration from a content perspective; however, additional studies can be conducted using different research methods, theories, and approaches to fill theoretical and methodological gaps. With several opportunities for future research, some relevant focus areas include:
-
What are the structural and substantive features of effective integrated policy designs? In terms of integrated mapped content, what should it (policy designs) look like?
-
How does policy integration based on integrated structural logic shape decision-making and policy coordination, leading to subsequent integrated governance outcomes?
-
How do integrated policy designs (content) affect sectoral/subsystem actor behaviour in practice. What or how should the structural or substantive qualities between policy instruments or mixes be embedded?
-
How new policies created can interact with the existing policy landscape and actors involved, specifically, if the policy might conflict with existing tools or are compatible and reinforcing?
-
How the use of IPA and IGT/IG 2.0 as tools for diagnostic assessment ex ante can deconstruct policy designs and what the potential effects, they might have on policy outcomes?
This conceptual paper is aimed at starting a debate or at least discussions on the importance of policy content in policy integration and its potential implications for improved governance and theory. The authors are of the view that this neglected area is an untapped reservoir of insights and knowledge that is overlooked and that the idea of policy content as an analytical locus where methods are used to develop abstractable units of analyses, is a pathway to enhance policy integration with valid and reliable methods.
