What Is a Policy Framework? An Attempt at Conceptualization

The term “policy framework” is frequently used. Surprisingly, however, there are no definitions for it; nor have there been any systematic attempts to conceptualize or clarify what a policy framework is. The fact that in spite of its ubiquitous use the term is not defined might signify that its meaning is considered self-evident. In reality, however, as we will show in this article, this is not the case – far from it. Even if it were, the place of the term “policy framework” in the public policy vocabulary and as a concept in the cognitive universe of public policy is too important to dispense with the task of defining it explicitly. The present article outlines a rationale for the conceptualization of the term and an approach towards achieving it. The main justification for attempting a definition of the concept is the need for cognitive hygiene. Methodologically, this research is inspired by the thinking of Sartori (1970) and scholarship surrounding his ideas on concept formation. Indeed, the paper does not aim to “reinvent the wheel” in public policy, but strives to extract the meaning of “policy framework” from already existing definitions, implicit if not explicit, and, in this way, to systematize pre-existing scholarship by providing empirical references to current policy frameworks across thematic fields.


Introduction
The process of thinking inevitably begins with a qualitative (natural) language, no matter at which shore we shall subsequently land. (Sartori, 1970(Sartori, : 1038 The term "policy framework" is frequently used. Referring to my personal research story, I would like to highlight that it has been extensively employed in the context of United Nations Sustainable Development Goals (SDGs) in the area of higher education. Surprisingly, there are no definitions for it; nor have there been any systematic attempts to conceptualize what a policy framework is. Indeed, we can note a plurality of empirical manifestations of policy frameworks. It can be observed that policy frameworks set up broad parameters and delimitations, in terms of principles, values, operational and managerial objectives, and modalities, for broader policy endeavors and initiatives outside the organization that produces that framework, or for regulating some internal aspects of the work and operations of that organization. Thus, policy frameworks may be used in institutional settings (for example, they can regulate internal rules and practices) and also for larger policy endeavors (when they exceed institutional borders and have a wider significance). This is a very important distinction in the understanding and definition of a policy framework. For instance, a sustainability framework in a given university is an example of an institutional policy framework; it is developed by that university to organize, even regulate, its own work and policies. In contrast, the SDGs are adopted by the United Nations, but they are not to regulate the internal work of the UN; they are indeed a global policy framework, informing a myriad of other policies and policy frameworks outside the UN. One might argue that policy processes are multifaceted and dynamic and that having any kind of definition might become an obstacle in some ways -for example, it may lead to losing flexibility in policy-making, and, as a result, promoting less innovative thinking. On the one hand, policy-making is an iterative process that often requires revisiting and adjusting policies on the basis of their outcomes. A flexible conception of a policy framework allows for this adaptability. On the other hand, having a flexible concept does not imply that the concept itself is empty. Coming back to the SDGs, it might be hard to achieve the 17 Global Goals by 2030. One of the reasons is the lack of a standard model that can be utilized across contexts. Of course, by providing a definition for policy frameworks, we cannot fully resolve all global problems, but we can certainly contribute towards effective, systematic, and transparent policy-making. Further, referring to a clear concept can be beneficial in terms of improving the clarity of communication, as having a definition would provide the basis for a shared understanding of policy frameworks by various stakeholders, in academia and beyond. The present article outlines a rationale for the conceptualization of the term "policy framework" and an approach towards achieving it (Sartori 1970;Gerring 1999;Mair 2008;Collier and Gerring 2009;Maggetti, Gilardi, and Radaelli 2015).
The fact that in spite of its ubiquitous use the term is not defined might signify that its meaning is considered self-evident. In reality, however, as we will show in this article, this is not the case -far from it. Even if it were, the place of the term "policy framework" in the public policy vocabulary and as a concept in the cognitive universe of policy analysis is too important to dispense with the task of defining it explicitly. One can encounter numerous examples of broad, international or national policy frameworks in documents from international organizations such as the OECD, WHO, EU, UN, and from governments and NGOs worldwide. Policy frameworks are also mentioned on the institutional level, as applied to universities, financial institutions, security agencies, and trade organizations, etc. Whenever the term is used in such documents, its meaning is taken for granted or the matter of its definition is simply ignored. In a nutshell, the term is there as a label, but it is not defined, and its meaning as a concept is not clarified.
Similarly, scholarly literature mentions policy frameworks, but such academic papers tend to proceed to the "empirical core" of the matter without any particular attention to defining or characterizing the concept itself. A typical example in this context is an article by Woolcock (1998), who uses the term "policy framework" very centrally in the title, but not in the body of the text. There are also some exceptions, but these are not conceptualization papers, and only mention what is understood to be a policy framework in a specific context. Here, then, a question arises: Can we label something as a policy framework without knowing what policy frameworks are?
The main justification for the attempt to define "policy framework" in this paper is the need for cognitive hygiene. Indeed, concepts are "the fundamental building-blocks for any description of reality" (Goertz and Mahoney 2012: 213;Sartori 1970). Thus, without these "blocks", one is not able to de/construct theories in a way that fully coincides with the questions asked. The aim of this article is to demystify the concept of policy framework, by means of a comprehensive review of existing literature, the application of methodologies from comparative politics and references to policy studies, and the development of a detailed taxonomy. The article seeks to provide a clear, practical definition of "policy framework" that can be universally applied across different sectors and geographic spheres, thus contributing to a more nuanced understanding and application in both policy theory and practice.

Policy Frameworks in the Literature
Not only in the sphere of applied policy, but also within the broader realm of scholarly enquiry, there is very little said about the concept of a policy framework. To demonstrate this gap, we conducted a systematic literature review prior to our attempt at conceptualization (Snyder 2019). This research followed a systematic and transparent method that is possible to replicate if required.
Our systematic literature review was based on search results from the scientific databases Scopus and Web of Science. Both were selected because of their comprehensiveness and their coverage of a wide range of topics, which are not necessarily limited to the field of policy. This choice allowed for a broad search across diverse disciplines, thus gathering evidence on policy frameworks on the broadest scale. However, Scopus included significantly more results (1,096 vs 478). The documents from Web of Science were included in the Scopus search; therefore, Scopus was used as the main reference.
The search strategy was to concentrate on the specific concept "policy framework" in the titles of publications. The keyword yielded 1,096 documents. This approach has its limitations because of its strict exclusion approach; yet, having "policy framework" in title meant that all documents were focused on this very specific phenomenon. The results of the search included only documents written in English, further limited to finalized publications in the timeframe 1965-2023. Types of documents included articles, book chapters, conference papers, reviews, notes, editorials, books, and letters. The next step included the screening and selection of potential articles based on whether or not they explicitly conceptualized the term "policy framework". This screening process was a critical step in systematic literature reviews, as it ensured that only the most relevant articles were included in the review.
Analysis of the results demonstrated an obvious gap in research when it comes to conceptualizing policy frameworks, even despite rapidly growing interest in the topic. Figure 1 shows the number of research documents with the term "policy framework" in the title published since 1985. As such publications reached almost 100 per year, it is clear that the concept of the policy framework has become a part of scholarly debate across various fields. The concept of the "policy framework" has not only found its way into academic literature, but has also become integral to numerous disciplines. The breadth of disciplines engaging with policy frameworks in their scholarly work is illustrated in Figure 2. As can be seen, the majority of publications belong to social sciences (27.9 %), followed by environmental sciences (13.4 %), and economics (10 %). Also, other fields, such as engineering, computer science and business, employ the concept. However, the mentioning of policy frameworks does not imply specific discussion of their meaning, as only 7 publications attempted to briefly explain what a policy framework actually is. These are quite limited in their explanations and frequently offer only a couple of sentences regarding policy frameworks per se, as their main focus is on different policy frameworks and not specifically on their definitions. Some papers, such as that by Modiba (2022), for example, view policy frameworks as regulatory entities that are "used by large organizations such as corporate or educational institutions, or governments to notify workers about whose endorsement is required to create new policies, what law must be followed when developing new policies, how policies should be interconnected and imposed and what high-level or long-term aims that new policies should attempt to maintain" (p. 53). While comprehensive, this definition could be critiqued for its specificity to large organizations, which might limit its applicability in some contexts. It may not fully encapsulate the use of policy frameworks in smaller organizations, non-profits, or community groups, for example. Several other authors have understood policy frameworks as providing logic for policy makers (Wallner 2013), establishing a "set of steps, procedures, principles, values and standards" (Heleta 2022: 3), or, more generally, referring to "policy goals, programs, instruments, such as funding priorities and support mechanisms" (Vihemäki, et al. 2019: 314). Some authors, like Depledge (2013: 371), considered policy frameworks to be more flexible texts, which required constant evaluation. Similarly, Agarwala (2022: 2) mentioned that policy frameworks are more general documents. A very flexible definition by Birch (2016: 3) also mentioned the broad applicability of policy frameworks, which could be "policy priorities, analysis, funding, schemes, initiatives and directives, and implementation modes that cut across policy-making, ensuring that policies are compatible and complementary." In general, explanations of what is considered to be a policy framework largely overlap in their understanding of policy frameworks as guiding structures for policy creation and implementation. On the other hand, the extent of their flexibility and dynamism is less clear. Some studies emphasize procedural aspects and regulatory rules, others focus on strategic aspects, while yet others underline the importance of flexibility and adaptability. The different subject areas also influence the results, suggesting that any attempt to conceptualize policy frameworks may need to be tailored to the specificities of the policy area in question. Overall, this variation in definitions could indicate a need for a more universally applicable, yet adaptable, definition of a policy framework.

Conceptualization: Methodological Insights
This article intends to contribute to policy scholarship by conceptualizing the notion of the "policy framework" as well by attempting a taxonomy of the term. The paper does not aim to "reinvent the wheel", but strives to extract the meaning of "policy framework" from already existing definitions, implicit if not explicit, and, in this way, to systematize pre-existing scholarship by providing empirical references to current policy frameworks across thematic fields.
This paper takes its inspiration from Sartori (1970), who is often taken as a foundational figure with respect to theoretical thinking in comparative politics, due to his significant contributions to the conceptualization and classification of political systems. His approach to conceptual clarity, known as the "ladder of abstraction", has been instrumental across disciplines. The decision to base this effort towards conceptualization on his approach stems from understanding policy frameworks as parts of a multidisciplinary domain. Further, the principles that Sartori provides are of methodological value for this research as they facilitate a structured process of conceptualization.

Conceptualizing Policy Frameworks
The following part attempts to work with the concept of policy framework using some central questions of conceptualization literature which were inspired by the logic of Sartori (1970). These questions include the need for conceptualization, its semantic aspects and positioning on the ladder of abstraction, the need for the development of a taxonomy of policy frameworks, and, finally, the conceptualization itself as well as its evaluation. Indeed, this paper provides a definition of the policy framework, but its main goal is to provide a clarification of the concept, which is the main direction of this part.

Why Do We Need to Conceptualize?
One of the first steps in any research is the conceptual question: "What is it that I am researching?" (Mair 2008: 180). Following the logic of Sartori (1970), concepts are defined as "data containers" with a "fact-gathering validity" (pp. 62-63). Sartori's main advice to the researcher is to ask the question of "What is it?" before elaborating the question of "How much?" (cited in Collier and Gerring 2009: 4). The idea that the researcher needs to conceptualize the terms prior to measurement or qualitative analysis is ubiquitous in the literature (Collier and Levitsky 1997;Becker 2008;Goertz 2006;Mair 2008). Indeed, it is "impossible to conduct [research] without using concepts. It is impossible even to conceptualize a topic, as the term suggests… Any significant work on a subject will involve reconceptualization of that subject" (Gerring 1999: 359). Advancing in theory-formation or theory-testing is unfeasible without clearly defined concepts. As Becker (2008: 128) pointed out, "concepts are not just ideas… concepts are empirical generalizations, which need to be tested and refined on the basis of empirical research results -that is, of knowledge of the world." On the other hand, there are also instances when measurement can be prioritized for pragmatic reasons, such as the existence of time constraints or an interest in new data generation (Maggetti, Gilardi, and Radaelli 2015). Moreover, "the preference for either conceptual analysis or measurement may be the result of the maturity of the field -we expect new, emerging fields to be mostly dedicated to conceptual work" (Maggetti, Gilardi, and Radaelli 2015: 5). Thus, in the case of policy frameworks, one needs to prioritize conceptualization over measurement, largely due to the lack of previous scholarship in the area.
Conceptualization is a holistic process that involves working with properties and attributes (intension), "the events, circumstances or phenomena to be covered" (extension), and the label itself, which is the embodiment of both (Buller and Gamble 2002: 6). There are a number of methodological and contextual challenges that may be involved in the process; these could range from a plurality of pre-existing conceptualizations and their diversity to a lack of such conceptualizations or their limited validity.
First of all, there are numerous empirical manifestations of policy frameworks as well as references to them in scholarly literature. Perhaps taken for granted, considered to be semantically obvious, or even mundane, the label has travelled across a variety of academic disciplines and policy areas without there being wider awareness of its meaning. Having no clear concept might lead to extremes where anything can be called a policy framework without proper consideration. Questions, such as "Where do we draw the borderlines between a policy framework and a policy or other manifestations?" or "Why do we call this document a policy framework and not something else?" might lead to further confusion and misinterpretation. Indeed, there are numerous attempts to conceptualize policies, but why are there none with respect to policy frameworks?
If we were to consider the question "What is it that I am researching?" (Mair 2008: 180), we also have to define the research area. In our case, we are searching for the term that is used in the domain of public policy but is not restricted to it; thus, it also encompasses private policies. One can find policy frameworks that are introduced by financial institutions, universities, intra-governmental organizations, NGOs, IT companies, and so on.

Semantic Components of Policy Frameworks
If we were to look at the structural components, the basic element of a policy framework is a policy itself. The concept of a policy is known for its contested definitions. For instance, Dye (1992: 2) provides a very broad and abstract conceptualization, which states that a policy is "whatever governments choose to do or not to do." Yet not all policies require governmental decisions to be implemented, since not all policies belong to the realm of public policy. For instance, a private firm that decides to adopt a gender equality and non-discrimination policy is free to do so, without consulting governmental representatives. However, the choice aspect of Dye's definition can be regarded as a useful link to non-governmental policies. Indeed, institutions, whether public or private, have to decide whether or not they adopt a policy.
A more detailed way of conceptualizing a policy can be found in the definition by Jenkins (1978: 15), who sees it as "a set of interrelated decisions taken by a political actor or group of actors concerning the selection of goals and the means of achieving them within a specified situation." In a nutshell, it is a set of interrelated decision-making steps leading to the adoption of a policy. These "steps" are performed by various "powerful" stakeholders, who have to compromise in terms of goals selection and their instrumentalization.
Moreover, a policy exhibits basic components, such as goals, objectives, settings (context) and instruments (Cashore and Howlett 2007), which are also present in policy frameworks. Following the argument of Howlett and Cashore (2014: 9), policy making "evolve[s] around the process of articulating and matching up policy goals with preferred policy means at all three of the abstract (general or conceptual), program (concrete) and on-the-ground (settings) levels". Taking the empirical example of waste reduction in the operations of a higher education institution, policy making processes include abstract aims, such as contributing to the global fight against climate change, as well as conceptual means (better waste management), objectives (the reduction of waste), mechanisms (the improvement of waste management facilities, informational campaigns), and contextual setting and the calibrations of tools (the reduction of waste by 50 % by 2030 by implementing better waste management strategies and by promoting increased awareness) (Howlett and Cashore 2014).
The link between a policy and a policy framework is present beyond its "label". Indeed, policy frameworks are similar to policies in their goal-oriented nature as well as in their documented/ written form. However, a policy is not identical to a policy framework, since the latter refers to a set of statements that might not be interrelated. A policy framework is not necessarily limited to one area of application or one institution. It provides the bigger picture -it guides the policy-making process itself, ensuring that individual policies are consistent with each other and aligned with the organization's or government's overall objectives. If we were to provide a metaphorical example, a policy can be represented by a brick, whereas a policy framework will be more like a wall of bricks, which form a construction that stabilizes the "house".
The second element of a concept is a framework. From the semantic point of view, a framework can be understood from two angles. Firstly, it is defined as a term that represents "a system of rules, ideas, or beliefs that is used to plan or decide something" (Cambridge Dictionary 2020). In this way, it can refer to an institutional or policy framework. On the other hand, it also has a geometric component, which denotes a "basic structure that supports something such as a vehicle or building and gives it its shape" (Cambridge Dictionary 2020). In other words, a framework helps to organize and unite separate policies into one coherent "organism".
In the context of institutions, Ostrom (2011: 8) mentioned three key concepts, namely theories, models, and frameworks. In Ostrom's view, the study of institutions hinges on three levels of theoretical analysis, and each level offers varying degrees of specificity pertaining to a particular problem. Frameworks are considered to be the most general forms of analysis, which provide a general structure and meta-theoretical language for identifying universal elements and generating relevant questions. Theories focus on specific elements and assumptions in order to diagnose, explain, and predict phenomena, while models make precise assumptions in order to derive accurate predictions based on a specific theory. This understanding of frameworks helps us to gain an insight into their breadth and structure-generating features.
From the point of view of policy research, frameworks are frequently referred to in the process of policy analysis. These policy frameworks - Kingdon's Multiple Streams Framework (1995), Sabatier and Jenkins-Smith's Advocacy Coalition Framework (1993), and Baumgartner and Jones' Punctuated Equilibrium Framework (1993) -are well-established and often referenced within the field. However, the primary focus of this paper transcends these specific analytical tools. Instead, it seeks to delve into the conceptualization of policy frameworks, positioning them as integral entities within the policy-making process.

Ladder of Abstraction
One of the central methodological questions in concept formation literature asks whether the researcher should choose dichotomies when formulating concepts or look beyond binary classifications. In some cases, the scholar may refer to Sartori's negative identification (1970Sartori's negative identification ( : 1042, which works on the basis of antonymous dichotomies (for instance, a married person is not single). However, this might lead to further difficulties, as it is not guaranteed that we encounter exclusively and fully binary concepts in the process of research. Specifically, what if there are more "shadows" of the concept than just polar opposites?
If we follow the logic of Sartori (1970) we should look for its antonyms. First of all, a policy framework is not the same as a policy; nor is it the same as other related concepts, such as a strategy, a law, or an initiative. Yet, as we can see, such a negative identification approach is rather a blunt instrument. Nevertheless, it might be useful in the initial stages of intellectual inquiry, such as conceptual mapping -that is, though it may appear vague, it can still be a useful tool for "narrowing down" and approaching the concept from a more analytical perspective.
Furthermore, not every concept can be compared with another, partially because they might have different scopes. Sartori (1970) offers a solution, which he calls the ladder of abstraction. In his ladder, Sartori proposes three levels: low-level (narrow-gauge theory), medium-level (middle range theory), and high-level categories (global theory) (1970: 1044). In the logic of Sartori, the higher the concept is on the ladder, the more all-encompassing it is. In order to move from one level to another on the ladder of abstraction, there is a need to unpack the meaning of the core components of the concept, namely its extension and intension (Maggetti, Gilardi, and Radaelli 2015: 9). Intension refers to the features or properties of concepts, whereas extension includes the cases covered by a concept (Maggetti, Gilardi, and Radaelli 2015: 9). In Sartori's understanding, the concepts of medium-level categories have their extension and intension in balance; the highest-level concepts are characterized by greater levels of extension and minimal intentions; and the lowest-level concepts represent the opposite situation, which means they exhibit minimum extension and maximum intention (p. 1044). Thus, concepts of the lowest level are very specific and can be applied only to a limited number of cases, which change once we move up the ladder.
Sartori's paper evoked further conceptual discussions and reformulations in the literature. For instance, Mair (2008) provides further explanation of Sartori's logic, calling extension denotation/ the range of cases, and extension connotation/ "the number of attributes or properties" (p. 187). Similarly, Gerring (1999: 357-358) expands the Sartori's approach by calling for triangulation in concept formation, which should include a "proper alignment" between "a) the events or phenomena to be defined; b) the properties of the attributes that define them; c) the label covering both a and b".
Even though it remains influential in concept formation scholarship, the idea of the ladder of abstraction was later questioned and refined by other scholars, since some concepts do not fully fit into the categorization (Collier and Levitsky 1997;Collier and Adcock 1999;Goertz 2006;Mair 2008). This was the case of Collier and Levitsky (1997), who argued that it was impossible to apply the strictest forms of the ladder of abstraction to their research, which showed that the characteristics of post-Cold war democracies were dissimilar to the "conventional" contemporary understanding and definitions of democracy. Thus, post-Cold war democracies had no clear-cut position on the ladder. As a solution, Collier and Levitsky (1997) introduced the category of diminished subtypes, which was proposed through defining democracy with the help of adjectives. This allowed a move away from the purely vertical understanding of concepts and added horizontal "side" deviations. Similarly, Goertz (2006) supported continuity in the process of concept formation by calling on scholars to include grey zones between ideal types.
In addition, Collier and Adcock (1999: 562) called for the adoption of a "pragmatic approach which recognizes that concepts, definitions, and operationalization may evolve with changes in the goals and context of research", yet by no means did the authors advocate conceptual anarchy. By offering an alternative understanding of concept formation, the authors revised Sartori's classical argument for "cut-off points" and that "formation stands prior to quantification" (Sartori 1970(Sartori : 1038 in definitions. Indeed, Collier and Adcock (1999) acknowledge the role of dichotomies, but, at the same time highlight their changing nature (p. 545).
If we were to position the concept of a policy framework on Sartori's ladder of abstraction (1970), we might be justified in placing different "shades" of it on each level. The following sub-section will attempt to deconstruct the ladder of abstraction on the basis of empirical observations. The visual below provides a short summary of the abstraction, and the discussion afterwards explains its main components. First of all, policy frameworks can be situated at the very top of the ladder of abstraction. In this way, if we think about the label "policy framework" in general, we might imagine a wide range of cases which are, indeed, policy frameworks. Here, we need the most minimalistic definition that can be applied to any policy framework, regardless of its geographical relation, scope, area, and other features. The highest level of abstraction includes the broadest number Highest level of abstraction -all policy frameworks Medium level -middle range cases Low -limited of cases and the greatest level of extension; conversely, it includes the most minimal level of intension.
Next, policy frameworks might also include medium-range concepts. These are more limited in terms of empirical cases and much more balanced in their proportions of extension and intension. An empirical example could be the SDGs policy framework that originated from the United Nations in 2015, but travelled far beyond UN walls. There are thousands of "SDGs-inspired" policy frameworks that range from local adaptations to global or transnational frameworks. Here, we can think about the example of an SDGs policy framework on the level of an international organization. For instance, the OECD has its own institutional response to the UN SDG policy framework, namely "Better Policies for 2030" (2016), which uses elements of the SDGs and incorporates them into the strategies and priorities of the OECD. There are also other examples relating to EU institutions, the World Bank, and the G7. Thus, the middle part of the ladder of abstraction includes a more focused and specific group of policy frameworks, yet it is still less specific than lower-level policy frameworks.
Finally, the policy framework concept might also appear on the lowest level of Sartori's ladder of abstraction, i.e. in a form which is the most specific and limited. Its intension is at a maximum, while its extension is, by contrast, very limited. In this case, we can think about an institutional policy framework which has its own distinctive attributes. For instance, the SDG policy framework at the level of a higher education institution. Here, we are faced with a level of narrow-gauge applicability. Thus, it is not advisable to limit ourselves to the lowest levels of abstraction in scholarly attempts at conceptualization, since this could limit our understanding of a concept. Further, it can be misleading and exclusive, and might even cause conceptual chaos.
In spite of its apparent usefulness, however, the idea of the ladder of abstraction is, as mentioned previously, widely debated in the literature, since it is far from universal and has its significant drawbacks (Collier and Levitsky 1997;Collier and Adcock 1999;Goertz 2006;Mair 2008). If we were to apply criticism to the concept of a policy framework, we might also face certain empirical examples that do not fit neatly onto the ladder of abstraction. For instance, what about examples of policy frameworks that are of a "hybrid nature", such as the Policy and Procedure Framework of the University of Southern Queensland? Its characteristics might not be fully identical with policy frameworks at other universities; therefore, the rules of the ladder might not be applicable in this case. Whether or not we can call it a diminished type (Collier and Levitsky 1997) can be decided only after careful analysis of this particular case, which is not the main goal of this paper. However, it is worth noting that the ladder of abstraction might not be the only way to understand concepts.
Overall, the ladder of abstraction serves its function well in the context of policy frameworks. It helps us to recognize the variety of possible policy frameworks that are situated at all three levels. In order to overcome the limitations of the lowest levels, this article aims to conceptualize the term that is positioned in the uppermost part of the triangle (Figure 3). Therefore, this leads us to the need for a brief and minimalistic definition that will enable us to find the unifying features that are present in the whole construction.

Typologies or Taxonomies?
Typologies are understood to be the tools for moving from "simple classification towards explanation" (Mair 2008: 183). One of the central ideas of a typology is the opposition to hierarchy, namely the inclusion of concepts which are related into one category (Given 2008: 2). According to Collier (2008), "explicit discussion of concepts and sub-types, as in a typology, is an important step in mapping out the semantic field". Jaakkola (2020: 23) adds that typology papers help to reduce complexity, offer a multifaceted and "more […] nuanced understanding of a phenomenon or concept, pinpointing and justifying key dimensions that distinguish the variants". The author mentions that typologies have several goals to fulfil -specifically, "explaining differences between variants of a concept [and] organizing fragmented research into common distinct types" (Jaakkola 2020: 22). Further, typologies are recognized to be directly related to the process of systematic conceptualization, with a special focus on links between concepts and causality beyond "simple" correlations (Fiss 2011).
Yet, there are a number of challenges that accompany the development of a typology. Collier (2008) provides a summary of potential solutions which can serve the scholar facing insufficiency or a lack of definitions. These include borrowing existing terms but assigning new meanings to them and "synthesizing various existing theoretical approaches in order to coin new and useful terms" (p. 160). Thus, instead of inventing a new concept, the method of synthetizing literature is considered to be preferential.
Another danger facing typology construction is that it might be purely conceptual/ abstract and distant from empirical cases (Smith 2002). The dimensions of typologies frequently rely on an ideal type, which means that they remain on the level of "a mental construct that deliberately accentuates certain characteristics and [are] not necessarily something that is found in empirical reality" (Smith 2002: 381;Weber 1949). Being useful in terms of systematic comparisons, typologies might be "useful heuristics", but at the same time "incapable of producing the sharp policy distinctions needed to fully support the explanatory and predictive resources scholars want to extract from [them]" (Smith 2002: 381). For the reasons mentioned above, the concept of the typology cannot serve its purpose in conceptualizing a policy framework, since it can be misleading to rely on "ideal types" that are hard to test empirically.
In the case that the inclusion of empirical observations in classificatory attempts is required, a scholar might prefer to apply another approach, namely a taxonomy. Though "taxonomy" as a concept is frequently but mistakenly understood to be synonymous with typology, taxonomic definitions differ from typologies in a number of ways -according to Lambert (2005: 4), by resulting from reasoning by inference, the consideration of a multiplicity of characteristics, and empirical derivations of "taxa" (categories) etc.
Taxonomies are frequently associated with inductive data collection and grounded theory. They are helpful in the process of pioneering attempts at conceptualization, as in the case of policy frameworks. The central objective of taxonomies is not the definition of a single and restricted entity but "a population of entities with some common properties" (Hodson 2019: 211). In this way, we might be able to escape the narrowness of typologies and move towards a more diverse, and, to some extent, even divergent, population of cases with a common feature that unites them under one "umbrella label". Here, we can refer to policy frameworks from different fields, geographical scopes, institutional contexts, etc., but we cannot exclude the clear formulation of those common properties that Hodson (2019) refers to. Using a taxonomy in this paper will enable us to uncover the variety of cases and to highlight the broad nature of the concept.
Given our preference for taxonomic definitions, we can now proceed with our attempt to describe a policy framework through the list of the properties it possesses, employing references in the literature and, at the same time, avoiding "mak[ing] up entirely new meanings" (Graebner and Ghorbani 2019: 3). One of the first things to note is that any taxonomy consists of a set of variables. In the case of the policy framework concept, it usually includes several features, the most obvious of which are its area of application, its geographical scope, and the type of institution that works with it.

Policy Frameworks by Geographical Scope
Referring to geographical scope is not novel in the literature (Krahmann 2003;Hettne and Söderbaum 2006;Kehm 2007). On the one hand, there are the pure concepts of local, regional, national, international, and global. These are seen as interconnected, but, at the same time, distinctive. The connections are seen through the processes in which social actors are positioned "as inherently 'local' and [then] subsequently become regional, national and/or global" (Herod 2003: 220). On the other hand, these processes might be happening the other way round, namely, from global to local, from regional to global, from global to national, etc. Herod (2003) presents these relations as a visual of nesting dolls, which consist of multiple interdependent layers and portray "a nested hierarchy of scales, with each scale fitting neatly together to provide a coherent whole" (p. 228).
On the other hand, Herod (2003: 223) notes that "the global and the local are seen not as things in and of themselves but are viewed instead as interpretative frames for analysing situations." This fluid approach, which opposes the strict dichotomies mentioned before, means that, in some cases, global policy frameworks might also be transferred to local, national, or regional levels, and vice versa. Here, we can mention an empirical example of the SDGs, which are global in the first place, but also national, regional, and local at the same time, depending on the research perspective.

Policy Frameworks by the Type of Institution
A policy framework cannot exist on its own -that is, without an institutional context. Policy frameworks originate in institutions and are adapted to their contexts. Thus, the taxonomy of policy frameworks would be incomplete without mentioning them. Yet institutions as such are not one-dimensional entities, since they consist of many types. Thus, it would be too simplistic and broad to state that a policy framework is an institutional entity.
Notably, there is very little conceptual consensus in the literature when it comes to defining an institution and outlining its different forms. Thus, the following part will attempt to synthetize the extant literature in this area. Since it is not the main focus of the study and the discussion will serve to identify the basis of the taxonomy of policy frameworks, the following part will not aim to provide an extensive insight into the institutional literature.
The concept of an institution has been extensively discussed in the literature, and attempts to provide valid definitions date back many decades. According to Commons (1934: 69), "an institution seems analogous to a building, a sort of framework of laws and regulations, within which individuals act like inmates. Sometimes it seems to mean the "behavior" of the inmates themselves". More recently, Jepperson (1991: 146) defined institutions as regulated entities and "socially constructed systems of roles or programs that produce routines." According to Scott (2001: 56), institutions are highly relevant, since they "provide stability and meaning to social life".
Further, Scott (2001: 60) mentions types of institutions, namely those which are cultural-cognitive, regulative, and normative. Cultural-cognitive institutions are mimetic and orthodox, based on a common understanding and shared ideas, and culturally indoctrinated. Normative institutions function on the principles of social obligation and feelings of shame/ honor, which arise from a moralistic perspective. These cultural-cognitive and normative institutions are "not applied" in such policy frameworks. Policy frameworks are likely to be documented entities, meaning they are commonly produced in written form. In contrast, regulative institutions rely on coercion and rules and are based on the notion of regulation. Thus, policy frameworks might be potential guarantors of an "organization". Therefore, when conceptualizing a policy framework this paper refers only to regulative institutions. This means that this study does not consider other types of institutions -specifically, those that are based on cultural or moralistic grounds. In this way, institutions such as families or religious groups are excluded from our understanding of the contextual base for policy frameworks.
Since policy frameworks have not previously been conceptualized through institutional lenses, this paper proposes its own types of institutions, which are also based on Scott's (2001) regulative type. In order to overcome the complexity of the institutional reality, we view institutions as single institutions (such as companies, banks, universities, schools, etc.), governmental institutions (ministries, national agencies, employment offices etc.), nongovernmental (NGOs), and intergovernmental institutions (EU, UN, World Bank).

Policy Frameworks by Area of Application
Policy frameworks in the security sector might differ from those one can find in the area of higher education. Yet they are still policy frameworks. Therefore, it is worth mentioning policy areas in the taxonomy. Intuitively, one might outline major policy areas in which policy frameworks can be observed: education, culture, security, healthcare, tourism, transport, justice, IT, fiscal policy, energy, the environment, social policies, international relations etc.
However, it is also worth mentioning that a policy framework might not necessarily be mono-sectorial. It could encompass several sectors or even be universal. Thus, for the purposes of inclusion, this paper refers to the variety of policy frameworks and includes monosectoral, pluri-sectoral, and universal policy frameworks.
Since the main aim of this study is to define the term policy framework and not to provide an overwhelming and in-depth taxonomy of all possible cases, we will not elaborate the following taxonomy in an all-encompassing way. In Table 1, we present a taxonomy of policy frameworks, based on their geographical scope, types of institutions, and areas of application.  Table 1 are neither exclusive nor exhaustive, as there are many more policy frameworks that one can encounter. However, it is crucial to note that these categories are not mutually exclusive. A single policy framework may span multiple categories. For instance, a framework developed by an NGO (institution type) could have a global geographical scope and cover multiple sectors (pluri-sectoral). This flexibility is one of the strengths of policy frameworks, allowing them to be tailored to a variety of situations and challenges.

Conceptualizing Policy Frameworks in One Sentence
In the light of the discussion above, the paper proposes the following definition of a policy framework: This minimalistic definition is chosen for the purpose of positioning it on the ladder of abstraction. Since many policy frameworks are multi-level and pluri-sectoral, with their geographical scope going way beyond institutional walls and their various narratives or themes, a simple and less worded definition might serve better than an extended paragraph with numerous attributive qualities. We acknowledge that many sources in this paper refer to the realm of public policy, but with our definition we intend to include both public and private policies, thus placing the concept higher on Sartori's ladder of abstraction.
One of the first components is the reference to general structures, which is discussed by Ostrom as a main feature of frameworks (2011). Similarly, "guiding architecture" implies structural components of policy frameworks with a "guiding" feature, implying that these could be different policy tools than just those of a regulatory kind. Next, we refer to the documented features of policy frameworks or their characteristic of representing established practices, which indicates that policy frameworks are more likely to be found in written form, since they are intended for multiple use by various actors that have access to them. By institutions, we imply formal organizations with rules and governing systems (Ostrom 1991). Referring to the taxonomy, policy frameworks can be found within a single private or public institution, non/governmental institutions, or intragovernmental institutions. Further, policy action indicates the broadness of the concept, which might incorporate different policy actions, such as policy formulation, implementation, and evaluation, etc. The last part -specifically, the multiplicity of policy areas -refers to the taxonomy (mono-sectorial, pluri-sectorial, universal policy frameworks).

Conceptual Goodness?
Moving away from dichotomous and dialectical thinking, a scholar might question the validity of the defined concept. This can be one of the final steps, since it refers to the evaluation of the "conceptual goodness" of the definition, and, perhaps, to reformulations.
One possibility is to reconstruct the conceptual quality by means of the checklist proposed by Gerring (1999). According to him, any good concept has several criteria to fulfill. The author emphasizes that these cannot be reduced to a selected few and have to be fully elaborated in the process of conceptualization. Firstly, Gerring refers to the notion of coherence, which is also connected to differentiation and clarity. In this way, apples and tomatoes are both fruits but have different properties, even though they both belong to the category of fruits. According to the author, "the most coherent definitions are those that are able to identify a "core" or essential meaning (p. 42). Further, the border of the concept can be described in abstract terms, for instance "a property X is invoked to distinguish the concept from one neighboring concept, and property Y to distinguish it from another, since neither resides uniquely within the extension" (p. 43). Secondly, operationalization needs be carried out, which leads to the question: "How do we know it when we see it?" (p. 44). Here, the author does not search for classical absolute "attributes" (p. 48), but rather looks for groupings of attributes. Thirdly, the conceptualization should be valid or true. There should be alignment between the definition and the phenomenon (p. 48). This rather abstract category might be challenging, as it has to prove that the definition and the referents are aligned.
The fourth point is that the term should be conceptualized with respect to field utility. This entails the "adequacy of a single concept within a field of concepts" (p. 51). Here, the author refers to the classificatory utility of the term and its usefulness within a field of similar attributes. For instance, if we are to think about the term "American political culture", we might come across many other terms in the field (such as liberalism or Protestantism) (p. 51). This complicates the picture on the one hand but adds classificatory value on the other. The Gerring's fifth point refers to resonance. This can be explained with respect to the dichotomy of neologisms. Here, it is the task of a researcher to select wording for the concept according to the contextual settings. The sixth criterion is contextual range, namely the scope of a concept. The author emphasizes that an important feature of a good concept is its ability to "stretch comfortably over many contexts", which contrasts with "a poor concept", which "is parochial -limited to a small linguistic turf" (p. 54). The seventh criterion, parsimony, is a crucial part of the definition, as good concepts "do not have endless definitions" (p. 57). Here, the author calls for the exercise of semantic reduction rather than the use of lengthy and unnecessary constructions. The final note touches upon analytic/ empirical utility, which unites the essences of concept and theory formation. Here, the broader note is that "concepts rest within propositions, and propositions rest within research designs" (p. 60).

Context Sensitivity
Adding to Gerring's contextual range criterion, contextual variety might have a bigger impact on conceptual stretching. Theoretically, the variety of contextual responses can be explained by the model proposed by Falleti and Lynch (2009). Their "Inputs-Mechanisms-Outputs" model, which is summarized in Figure 1, provides the explanation for addressing this contextual variety (Falleti and Lynch 2009). The researchers state that "credible causal social scientific explanation can occur if and only if researchers are attentive to the interaction between causal mechanisms and the context in which they operate" (p. 1144). The authors of the model subscribe to the idea that causal mechanisms lead to "deterministic outcomes" (Falleti andLynch 2009: 1152). Crucially, they view the context as the decisive determinant of the outcomes of the causal links: "the outcome of a causal mechanism depends on its context, we need to distinguish between mechanisms and their contexts and so define both the mechanism at work and the context in which it operates" (Falleti andLynch 2009: 1152). Similarly, Beach and Pedersen (2019: 2) see mechanisms as highly context-dependent, "meaning that the same causes can trigger different mechanisms in different contexts". Yet, here, one might be careful, since the scholar might not be able to include every single context in the multiple occurrences of the concept.

Conceptual Goodness of the Definition
The following part will analyze the proposed definition in the light of Gerring's (2001) concept evaluation strategy. Table 2 below is based on the previous discussion that explained Gerring's conceptual logic (see Part 1). The table expands the criteria that were proposed by Gerring in 2001 by adding the feature of familiarity, which was included in the initial publication that dates back to 1999. The reason is that this paper is the first one to conceptualize policy frameworks and one of its objectives is to synthetize the literature in order to provide conceptual clarity of the term. Thus, familiarity serves as a helpful criterion for a newly-defined concept, since it shows that the concept is not entirely novel in the field. The definition is not absolutely new; it is synthetized from pre-existing literature. It is not a neologism and the components of the definition are well-known in scholarly discussions.

Gerring (2001) Resonance
What is the resonance of the definition? Is it "catchy" enough?
The demand for resonance is partially covered by the reference to the concept of a policy.

Gerring (2001) Parsimony
How short is the definition and the list of attributes?
Our proposed definition is brief and concise; it does not have a long list of attributes; thus, it cannot evoke terminological confusion.

Gerring (2001) Coherence
The degree of the logical consistency/ coherence of attributes The attributes are not contradictory.
Gerring (2001) Differentiation What the term is not Based on Sartori's logic of antonyms, this paper claims that a policy framework differs from a policy. Similarly, it also diverges from many other formats, such as laws or strategies. Thus, this criterion can be fulfilled only partially. Gerring (2001) Depth "How many accompanying properties are shared by the instances under definition?" One can find numerous attributes under the umbrella term "policy framework." The depth of the definition allows the grouping of different kinds of policy frameworks, based on their geographical scope, institutional type and the area of application (see the discussion on the taxonomy of policy frameworks).

Literature Criteria Content Our definition
Gerring (2001) Theoretical utility "How useful is the concept within a wider filed of inferences?" The concept of a policy framework might be helpful in theory formation inquiries. Having a definition will be a starting point for theoretical debate and also empirical testing. Gerring (2001) Field Utility Usefulness of the concept within a wider field According to Gerring, a new concept might evoke "resettling of the semantic field." Since the term policy framework does not aim at the redefinition of any other concept in its field, it may cause only low levels of potential disruption. Goertz (2006) Classic/ family concept

Classical (AND) vs family (OR) concepts
The definition is a classical concept.

Falleti and
Lynch (2009) Contextual sensitivity Causalities and outcomes differ depending on the context This paper employs a minimalistic definition for the purpose of broader inclusion. It also acknowledges different types of policy frameworks which are specified in the taxonomy.
Clearly, the table above does not include all possible pitfalls and does not aim to be allencompassing. Certainly, it has its limitations due to its mostly theoretical nature, which is primarily based on a review of methodological literature and general research in the area of policy frameworks.

Conclusion
This paper sought to unpack the complexity of the term "policy framework". It aimed to overcome limitations in the literature by proposing a definition that is based on pre-existing discussions in the field. The proposed definition is based on the original logic of Sartori (1970) and other scholars who followed his methodological inquiry (Collier and Levitsky 1997;Collier and Adcock 1999;Goertz 2006;Mair 2008), the rich literature on classifications (Elman 2005;Lambert 2005;Mair 2008;Collier 2008;Fiss 2011;Hodson 2019;Jaakkola 2020), as well as evaluative literature by Gerring (1999Gerring ( , 2001 and Goertz (2006). What has been done is to situate the term policy framework in the wider methodological and policy literature, in order to provide conceptual clarity and start a new chapter in policy research. The limitations of this research are mostly in its rather broad and multidisciplinary orientation. It takes its inspiration in comparative politics, speaks about policy frameworks that are in the area of policy studies, and provides references across different fields in which policy frameworks are present. Our proposed definition is broad, which might be a challenge in a specific context or policy area. Further, by focusing on the works of authors from comparative politics we were not able to refer fully to the rich body of literature in policy processes, which is one of the main limitations of this paper.
Looking forward, we see potential in further research to test and refine our definition in different policy contexts, as well as to explore how different types of policy frameworks influence policy outcomes and which features of policy frameworks seem to be more efficient than others. It is also worth considering an exploration of voluntary policy frameworks in future research. For instance, the Sustainable Development Goals (SDGs) offer a compelling case study of a non-legally binding, yet widely influential policy framework. A nuanced analysis of the "soft" policy instruments underpinning such frameworks could shed light on the mechanisms that allow them to become integrated into and shape larger governance structures. Unraveling this dynamic could provide valuable insights into how voluntary frameworks can effectively contribute to policy-making and implementation despite their non-legally binding nature, thus broadening our understanding of policy frameworks and their role in global governance. Overall, we hope this paper will stimulate further discussion and research on the topic, contributing to a more nuanced and robust understanding of this important concept in the field of policy studies.