Introduction

Policy scholars began considering framing in policy processes in the 1990s, supplementing existing theories, paying more attention to the ways meanings, symbolic language, emotions, identities and socio-cultural practices are included in the policy process, particularly in framing policy problems and solutions. Over the years policy scholars have drawn on a range of traditions to conduct frame analysis in public policy research, including psychology (e.g., Tversky & Kahneman, 1981), sociology, especially in social movement scholarship (e.g., Benford & Snow, 2000; Gamson, 1992; Snow & Benford, 1992), organizational studies (e.g., Dewulf et al., 2009; Lewicki et al., 2003; Van Hulst & Yanow, 2016), and political and environmental communication (e.g., Lakoff, 2010; Nerlich, 2010; Nisbet, 2009). These threads have influenced a wide variety of approaches to framing in the policy process literature.

Previous overview articles have categorized framing approaches in policy sciences in several ways, most often comparing them across four interrelated dimensions: (1) a frame as a category, an objectifiable stable general cognitive or social schemata, or framing as a process by which individuals or groups make sense of policy issues or situations in interactions with others (Dewulf et al., 2009; Van Hulst & Yanow, 2016); (2) framing as implicit or strategic, (3) the level of analysis: individuals (actors), situations (coming about in relations between actors) (Dewulf et al., 2009; Van Hulst & Yanow, 2016), or collectives (in, for example, media or society) (e.g. Baumgartner & Mahoney, 2008; Dewulf et al., 2009; Van Hulst & Yanow, 2016); and (4) framing analysis to study meaning or to measure frames and frameshifts (Gamson & Modigliani, 2009). Some authors have pointed out the relevance of combing these different ways of studying frames and framing (Baumgartner & Mahoney, 2008; Jones, 2018). Yet, many categorizations have excluded some approaches, mostly staying within epistemological communities.

We set out to explore the range of approaches as applied in policy studies and to assess if and how they can be complementary. Our aim is three-fold. First, we aim to provide a more robust heuristic for identifying the differences and similarities of framing approaches while recognizing their unique and complementary contributions to the field. Second, we aim to articulate what each of these approaches contributes to our understandings of policy processes. Third, we want to support mutual appreciation and cross-fertilization across approaches, which can be hampered by a lack of understanding of the approaches’ theoretical and epistemological underpinnings. There is considerable opportunity for the approaches to inform each other.

To meet these aims, we conducted a qualitative literature review of 68 articles published between 1997 and 2018 in four policy studies journals: Critical Policy Studies (CPS), Journal of European Public Policy (JEPP), Policy Sciences (PS), and Policy Studies Journal (PSJ). From our analysis, we distinguish five approaches to framing in the policy sciences: the sensemaking, discourse, contestation, explanatory, and institutional approaches. In the discussion, we suggest that these five approaches raise four key questions that animate all sorts of framing studies in policy analysis: (1) Do frames influence policies or are policies manifestations of framing? (2) What is the role of frame contestation in policy conflict? (3) How can the study of frames or framing reveal unheard voices? And (4) how do certain frames/framings become dominant? By introducing these questions, we offer a fresh way scholars might discuss frames and framing in the policy sciences across approaches, to highlight the distinct yet complementary ways they illuminate policy processes.

We first present a brief overview of previous efforts to categorize framing approaches. We then present our methods to select and review articles that analyze frames and framing related to policy processes across a range of perspectives. Following that, we present the five approaches in the policy sciences, and how they answer the four questions above. We conclude with a discussion of the broad similarities and differences across these approaches as identified in our study, how they answer the four questions posed above and suggest ways they may be more fruitfully in conversation with each other.

Previous overview articles on policy framing

As noted above, previous review articles distinguish framing in policy studies based on four broad themes: the distinctions between (1) a frame and framing (2) implicit and strategic framing (3) individual, collective and situational levels of framing, and (4) measuring frames and unpacking meaning.

First, various authors distinguish a frame from framing and use that distinction to point out different paradigms underlying framing research. Dewulf et al. (2009) distinguish between cognitive and interactional approaches to framing in negotiation research. The cognitive approach focuses on frames as stable, general schema: “Frames as knowledge schemas refer to structures of expectation about people, objects, events and settings” (2009, p. 158). These cognitive schemas are something that all people use to understand and make sense of the world. Or as Van Hulst and Yanow (2016) describe it: these frames ‘are structures ‘floating’ out there; they are objectively measurable phenomena that can travel unchanging across time and place. In contrast, framing refers to “alignments that are negotiated in a particular interaction and focus on how communication defines specific aspects of what is going on in interaction” (Dewulf et al. argue (2009, p. 158). By framing the situation, people are able to make sense of the situation “in the sense of Bateson’s classic example of the monkeys framing their actions when they are biting each other: ‘is this fight or play?’” (Dewulf et al., 2009, p. 158). In this sense, framing is a form of meta-communication (Van Hulst & Yanow, 2016). Hence, frames are the stable, measurable schemata by which individuals or groups make sense of policy issues and situations; while framing is the processes by which individuals or groups make sense of policy issues or situations in interactions with others.

Second, within the distinction of frames and framing, review papers point out that framing can also be “a dynamic and, potentially politically aware engagement” (van Hulst & Yanow, 2016, p. 2), meaning that next to implicit sensemaking of policy issues and situations, framing might also include more strategic and rhetorical negotiation of the meaning of the issue or situation. This resembles the everyday use of the word framing. Individuals or groups—strategically—highlight certain elements of reality and leave out others. This selectivity is part and parcel of sensemaking of policy issues, but it may be strategic when focused on convincing others to adopt a particular way of looking at policy issues or situations. This strategic assumption is behind studies that examine frames and their (strategic or political) influence on policies. For example, Baumgartner and Mahoney (2008) study framing and argumentation as ways by which public policy advocates “hope to see debates defined in a manner favorable to their position, and with their allies they work to achieve this” (p. 436).

Third, prior reviews distinguish three levels where frames or framing can be studied: individual, collective (Baumgartner & Mahoney, 2008) and situational (Dewulf et al., 2009; Van Hulst & Yanow, 2016). In their review of the literature, Baumgartner and Mahony (2008) introduce a distinction between two faces of framing in relation to policy issue definitions—individual and collective—and argue that to understand collective shifts in issue definitions, one needs to move beyond the study of framing at the individual level. They understand framing as a way to promote a particular definition of a policy issue. As they argue, at the individual level, studies focus on policy advocates and their framings of policy definitions. At the collective level, they assume that no single policy advocate can determine the issue definition and so scholars should study collective issue definition over time on the basis of documentary sources (media analysis, policy documents, etc.). Later, they started quantifying framings, and their salience, resonance and persistence (e.g. Baumgartner et al., 2008). This approach seeks to explain the extent to which frames influence public opinion or policy change. A third level at which interactive framing can be studied is: the situational (Dewulf et al., 2009) at the level of face-to-face interaction (Van Hulst & Yanow, 2016). When studying processes of sensemaking and meaning-making of a policy (issue) this is most appropriate. These scholars do not focus on framing by individuals alone (for example, politicians, journalists, lobbyists, or spin doctors). Nor do they study framing at the collective level (for example by media or the general public). Instead, it is in the interactions between individuals and groups of people in real time situations or in the media that these scholars study how framing and reframing of the policy issue (or situation) takes place.

Finally, previous reviews make a distinction between framing studies that either focus on measuring frames or unpacking meaning. The measuring studies aim to quantify dominance (e.g., salience, resonance and persistence) of frames over time (e.g., Baumgartner et al., 2008 and Baumgartner et al., 2014). When scholars are more interested in unpacking ‘meaning’ they focus on the cultural and symbolic meaning expressed in each frame or in framing processes (e.g. Van Gorp, 2007). To some extent, all approaches to framing in policy studies attend to the cultural and symbolic aspects of framing (Gamson & Modigliani, 2009) but the emphasis varies. Hence, the choice to unpack meaning or to measure frame resonance, persistence and salience generates different types of studies.

We build on these prior attempts to document various approaches to frame/framing analysis to explore how scholars in policy studies journals develop and apply framing approaches in their academic practice and to answer the key questions that animate framing policy analysis in the field.

Methods

We sampled articles about policy framing published between 1997 and 2018 in CPS, PSJ, JEPP and PS. PSJ and PS are established and renowned journals in policy studies, one from the US and one from Europe. We included CPS and JEPP to capture a broader range of approaches, particularly critical and post-positivist approaches to framing analysis. We conducted a purposive review of articles that use framing/frame analysis in these journals (Knopf, 2006; Weed, 2005).

We selected all articles in these journals containing “frame” or “framing” in the title or abstract. We excluded articles that only used these terms in the text, to avoid colloquial rather than analytical usage. We read the articles to confirm that they incorporated a frame or framing analysis and rejected those that used frame/framing colloquially. We provide a quantitative overview of these articles in Fig. 1. JEPP, PSJ and PS were all founded prior to 1997. CPS was founded in 2006, so the figure only shows articles on or after 2006 for CPS. Our “sample” for the review includes 68 articles in total with 22 in CPS, 19 in JEPP, 12 in PS, and 15 in PSJ. This sample is US and EU focused and therefore does not cover framing approaches from other places, and/or the policy concerns of actors in other settings not covered in the journals.

Fig. 1
figure 1

Number of articles by journal (1997–2018)

Full size image

We qualitatively coded the articles in excel for (1) theoretical perspective (theory used and analytical framework); (2) ontology (what constitutes reality or fact) and epistemology (how to know reality) when mentioned by the authors; (3) methodology/methods (focus of analysis on language or beliefs and how scholars analyze frame change); and (4), the main empirical findings on frames or framing. While coding, we did not create a set of predetermined, discrete responses to each of these codes. Rather, we added text to each code. For example, under “method” we coded an article that uses the explanatory approach this way: “discursive intertextual analysis employing the cognitive tools of conceptual metaphors (works by Lakoff and Johnson) and cognitive 'blending ' (works by Fauconnier and Turner).” And for an article using discourse approach, “an interpretive approach to textual analysis, where meanings are not taken to be essential but negotiated; comparing speeches of Blair & Brown.”

Based on the coding, we realized the articles fit within broader methodological approaches to frame/framing analysis with particular aims and empirical focus. Therefore, we inductively sorted the articles into an initial set of approaches guided by our analysis and previous review articles (sensemaking, discourse, contestation and explanatory). We then reviewed the articles within each approach to clarify the overall approach and its distinctness. We found that most articles mixed theoretical influences but showed real distinctions across approaches in their methodology, methods and subject matter. As we read the articles, we had multiple conversations to clarify the features of each overall approach, and to discuss articles that were difficult to categorize. We found that most articles fit neatly into one approach, while others were hybrids of two approaches. Therefore, we moved articles into the approach that offered the best fit while noting the hybrids (see the findings section below). Through this process, we realized we needed to add the institutional approach givens its unique use of institutional theory and in-depth case studies of governing institutions over long time periods. Ultimately, we sorted all articles into the five approaches we review in the following section. Table 1 shows there are 21 articles in the explanatory approach, 14 in both the sensemaking and discourse approaches, 12 in the contestation approach, and 7 in the institutional approach. Finally, we began writing up the articles and added key points of variation not included in our initial coding scheme including: aim of study, definition of frame/framing, notions of causality, assumptions about power, empirical focus, specific data collection and analysis methods, and data sources. These descriptions emerged iteratively through writing and discussion until the authors found mutual agreement.Footnote 1

Table 1 Articles by approach to policy framing analysis and journal
Full size table

Results: approaches to framing and frame analysis in policy journals

Here we present the following five approaches to framing that we distinguish based on our analysis:

  1. 1.

    The sensemaking approach uses ethnographic methods to study policy actors’ interactive framing in face-to-face conversations to understand how they collectively make sense of a policy issue or situation to enable coordinated action.

  2. 2.

    The discourse approach employs discourse theory and analysis to show how discourses are (re)produced in policy settings and thus include or exclude particular framings at the collective level. It explains how discourses (via framings) structure and discipline policy actors, and are reinforced by policy practices and knowledge production processes.

  3. 3.

    The contestation approach analyzes competition between individual or collective actors who strategically employ frames or framing to win a policy contest. It aims to understand how some frames/framings become more convincing or dominant over time through symbolic interpretation and how this influences policy stasis or change.

  4. 4.

    The explanatory approach similarly examines strategic contestation of policy frames but views frames as explanations of policy stasis/change via shifting problem definitions. It mostly employs quantitative methods to study what frames influence policy stasis/change.

  5. 5.

    The institutional approach conducts in-depth longitudinal case studies to understand how frames or framings are embedded in and influenced by institutional contexts, and how it enables or hinders institutional integration.

In our detailed description of each approach below, we discuss their overall characteristics, theoretical assumptions, ontology and epistemology and preferred methods. We illustrate each approach with exemplars. There are also some hybrids across approaches that we mention and explain why they are hybrids.

The sensemaking approach to framing

The sensemaking approach to framing emphasizes that people use frames to understand reality in daily life. Reality is vague and ambiguous, and by selecting elements one can create a story to make sense of this reality. Hence, framing is considered a sense making processes through which we give meaning to the world around us, in interactions with others. This approach to framing is rooted in Rein and Schön’s work, which builds on Weick (1979) in organizational theory. In this approach, a frame is not an ‘object’ that can be observed across contexts, or something that can be developed for strategic purposes; it is implicit and unintentional but may nonetheless have consequences for policy (Van Hulst & Yanow, 2016). Making sense of a policy problem in a particular way also leads to thinking about solutions in a particular direction.

The scholars in this approach aim to understand what policies mean for different groups of people in society. Frames and framing are considered processes by which sense making takes place, either by individuals or in interactions between people. We did not find many studies that applied this approach in its ‘purest’ form. Very often there is a mix between the sensemaking and contestation approaches, which emphasize more how framings differ (or are even contested) across groups. For example. Burlone and Richmond (2018) studied the meaning of the Canadian Act Respecting End-of-Life Care from 2015. Through inductive and manual coding in several rounds, they documented four different ways of understanding this policy. On the one hand, opponents to this act framed the policy issue mostly in moral terms related to sanctity of life in combination with instrumental-rational framings that alleged the danger of potential abuse of end of life care, which could be a “slippery slope.” Opponents also framed the issue morally but centered on the argument that dignity and individual autonomy take precedence over other values. In this group, some also framed the issue in more rational-instrumental ways and used the slippery slope/abuse argument as a cautionary statement against the artificial prolongation of life. Interestingly, framing between opponents and proponents exhibited some overlap.

Theoretical perspective

In the framing as sensemaking approach, framing is considered a process. Frames come about in interactions between people, changing continually as meaning is established and re-established. Interactive frames refer to alignments negotiated in interaction and focus on how communication helps to make sense of what is going on, in the sense of Bateson’s classic example of the monkeys framing their actions as fight or play (Dewulf et al., 2009). For example, in a micro-level analysis, van Lieshout et al. (2012) examine face-to-face conversations in a municipality to study “scale framing politics” of a project, including “projections into the future, upscaling and downscaling [to higher or lower problematic levels], and scale coupling” (p. 177), through which actors (unintentionally) magnify or diminish issues or shift accountability and responsibility. Again, this study examines how the issue means to different actors. It also reveals how metaphors may work as rhetorical devices even though the study does not start from the premise that actors strategically use them.

In framing as sensemaking, some scholars attend to the socio-cultural and symbolic use of language and its influence on policy making present in the other framing approaches. From this view, “frames are a central part of a culture and are institutionalized in various ways” (Goffman, 1974). Through framing, information and knowledge are selected, and policy issues are named and labeled (Van Hulst & Yanow, 2016). Some studies show how sensemaking of an issue by different communities of meaning—an often implicit understanding of a policy issue—may have political consequences (e.g.Van Ostaijen & Scholten, 2017). Framings in this approach also include emotions and resonance with cultural norms as represented in the use of words, metaphors and so on. In their work, Schön and Rein (1994, p.33) pay attention to this socio-cultural aspect of framing. Whereas they define a frame as “structures of belief, perception, and appreciation which underlie policy positions” (p. 23); they also acknowledge what they call ‘generative metaphor’—or cultural meta-frame—that creates the perspective from which one interprets a policy problem (Schön and Rein 1994). Sensemaking studies often considers this socio-cultural aspect. For example, Stucki & Sager’s analysis of the smoking ban in Switzerland (2018) shows that campaigners for the ban use the metaphor of a “holy war” against smoking and of “fundamentalist non-smokers” next to other framing elements (p. 379–380). In this study, it is hard to tell if this metaphor was created intentionally but Schön and Rein would recognize it as a rhetorical way of framing (1994, p. 32). Similar to organizational studies, scholars using the sensemaking approach assume that framing is implicit and unintentional and consider it a way of making sense of the world. In addition, very often there is reference to (un)intentional effects that this sensemaking may have (e.g., Straus, 2011). In another example, Dudley and Richardson (1999a, 1999b) studied how the policy framing of European steel policy changed over time.

Ontology/epistemology

The sensemaking approach adopts a relational, anti-foundational ontology, and a social constructionist epistemology. It emphasizes the construction of meaning in micro-level face-to-face interactions, where different policy actors align (or not) meanings in every-day interactions. For example, in Van Lieshout et al.’s (2012) study of scale framing, actors make sense of the scale of the policy issue in situ to come to agreement on who is to be held accountable for decisions at what scale. In another example, Aukes et al. (2017) focus on inter-textual interactions in their study of how policy entrepreneurs resolve frame conflicts by extensive meaning-making work in their interactions with other policy actors.

Methodology/methods

Studies in a sensemaking approach often use interviews but also more anthropological methods—such as ethnography, shadowing, observing, focus groups or sensemaking sessions. These methods allow researchers to closely study how actors engage in relational framing processes to establish the meaning of specific objects, concepts, or issues (e.g., Straus, 2011). These studies do not seek to explain policy change, explicitly refraining from explanatory methods, but are geared toward understanding—as in Weber’s verstehen—not explanation. Policy change is a consequence of making sense in situ. Policy frames are considered the results of interactive framing processes between actors in specific situations, especially conversations (Dekker, 2017). Scholars will study stories people tell each other, their negotiations and turn taking, and use that inductively to develop an accounting of frames (Dewulf & Bouwen 2012).

Discourse approach to framing

The discourse approach is closely related to discourse theory and analysis. It emphasizes the ways discourses that are produced and reproduced in society in particular settings frame different aspects of reality for participants in public policy. A discourse is a “shared ways of apprehending the world” (Dryzek, 1997, p. 8), a shared lens through which people in policy settings understand various meanings such as their own and others’ role in the setting, the knowledge that should be produced to inform action and so on. As Paterson (2010) explains, “Actors must use pre-existing rhetorical structures to make sense of the world around them, to be understood, and to be considered relevant” (p. 132). Discourse, thus, frames these aspects of the setting for participants. For example, a medical discourse may frame relations in a way that makes certain “subject positions” available to participants in a hospital—a doctor, a nurse, or a patient—and provides assumptions about what each is able to know or do. Patients are not expected to diagnose illnesses, while doctors and some nurses are.

These meanings reinforce, and in turn are reinforced by, practices. As Hajer (1995) describes it, “according to Foucault there is no a priori thinking subject trying to express or transcribe his or her preconceived idea in language. The subject operates in the context of a whole group of regulated practices according to which his or her own ideas are formed” (49). Not only do discourses frame meanings about appropriate practices, but certain practices also sustain those meanings. For example, in a hospital, a nurse’s daily routine reinforces meanings attributed to “nursing.”

This approach foregrounds knowledge and power in policy analysis. Social relations in public policy—organized by discourses such as neoliberalism, capitalism, and patriarchy—are undergirded by related knowledge systems that help keep systems of power in place. For example, Paterson (2010) shows how a discourse of “new midwifery” in reproductive health sought to replace modern obstetrics toward empowering birthing mothers as primary decision makers. Yet it evolved to be complicit with modern obstetrics and its hierarchical, expert-centered form of health care, albeit with midwives as “experts.” Therefore, policy processes consistent with certain discourses, and related knowledge systems, can reproduce dominant social orders even when challenged by alternatives. Yet, discourses are not monolithic, all-powerful systems. Counter-discourses often reframe subject positions, knowledge production processes, and relations of power. The interplay across discourses and counter-discourses can be the focus of analysis, although this approach tends to center a single discourse and how it frames policy or policy settings.

Theoretical perspective

Studies in the discourse approach are informed by Foucault’s theoretical and analytical perspective, positioning framing processes within broader discourses in society (Braun et al., 2014; Needham, 2009; Paterson, 2010; Tervonen-Gonçalves, 2012).Footnote 2 Scholars emphasize the power of discourses rather than policy actors to frame policy.Footnote 3 For example, Braun et al. (2014) examines how different discourses frame involuntary sterilization procedures as unjust. Given the weight they place on discourses as drivers of framing processes, these authors tend to downplay actors’ agency to strategically frame policy issues. The approach generally assumes that policy actors are not conscious promoters of discourses that frame meaning but take a discourse’s assumptions for granted, not questioning the discourses that inform their understanding of the world but use them unreflectively. Some scholars analyze what specific actors say or write and the effect on the policy process (e.g., Avigur-Eshel, Avigur-Eshel, 2018; Paterson, 2010).

In this approach, frame is typically used as a verb: “to frame.” And it is discourses that do the framing. For instance, Needham (2009) analyzes how a discourse of personalization provides a new logic within which “care” is framed in health provision in the UK. Frames are not objects of study as in the explanatory or contestation approaches, but discourses are, and the focus of analysis is on how they produce effects in the way they frame thinking or acting in a particular setting. For example, Needham (2009) shows how two discourses on care in the UK health system evolve: one “where participants are given choice in competitive markets in order to incentivize service providers” (204) and a soft notion of care “…linked to… ‘personalization’, in which patients receive a more tailored service” (205). She finds that these notions combine, positioning service users and providers ambiguously: staff must be “key partners in relationships of care, but also [serve as] barriers to reform” (216). This approach generally does not focus on overt policy contestation, although it can. For example, Paterson (2010), focuses on two competing discourses coalitions (Hajer, 1995) on midwifery. While interested in contestation, her emphasis remains on the power of discourse to frame health care, rather than the agency of actors to intentionally frame health care.

Yet, discourses shift over time through the ongoing interplay of language and policy practices (Glynos & Howarth, 2007). Analysis explains how discourses change through a complex interplay of language and practice. Scholars often explain what is at stake in discursive shifts for different policy actors in terms of how power changes, whether justice is being served and for whom. As the midwifery example shows, the implications for policy and policy practice are profound and tied to knowledge production processes and power relations. Discourses on midwifery can perpetuate hierarchical forms of expert-driven healthcare knowledge—by framing them as logical and natural—or they can challenge them and provide alternatives—by framing dominant approaches as unjust. Changing discourse is not easy because old forms of knowledge and power can be incorporated into new discourses, transforming their critical intent into complicity with old power relations.

Ontology/epistemology

The discourse approach adopts a relational/anti-foundational ontology and a social constructionist epistemology. The implication is that causality does not take the form of some factor (A) directly or indirectly impacting some factor (B), as in A causes B. Rather causality takes the form of constituting meaning (Yanow & Schwartz-Shea, 2006). This means that as a discourse becomes dominant, it constitutes meaning for actors in ways that shape their thought and action. As discourses shift, meaning is constituted in new ways, thus transforming actors’ thought and action. Policy change results from changing discourse as it is adopted and spread over time and place. In Paterson’s midwifery study, for example, feminists in the 1960s and 1970s began to challenge a medicalized discourse on reproductive health, developing discourse that reframed childbirth as a natural process, women as having intuitive knowledge of childbirth and bodily autonomy, and questioning the authority of (mostly male) doctors. Yet as this discourse became institutionalized in national policy, “midwifery” also became medicalized: the “expert midwife” became a professionally trained woman who provides services in a hospital rather than at home. This reconfiguration of meaning can also create new alignments in discourse coalitions (Metze & Dodge, 2016). In Paterson’s case, midwifery advocates split between those who supported or contested the “new midwifery” discourse (its meaning and practices).

Methodology/methods

To analyze policy discourses, scholars (re)construct the ways discourse frames meaning using a discourse analysis approach adapted to policy studies (e.g., Hajer, 1995). Analysis focuses on the ways discourses frame issues and/or are refashioned in policy processes. Conflict is not overt but is uncovered through careful analysis of language and practice in context. The level of analysis tends to be at the meso-level (discourse in particular settings) rather than macro-level (“grand” discourses in society), as one might see outside the policy field. Furthermore, scholars do not develop falsifiable hypotheses that can be tested, as in the explanatory approach. But focus on what actors do when they draw on discourses that frame policy in certain ways (similar to sensemaking). Braun et al. (2014) summarize the methodology:

… an interpretive approach … does not take frames, problematizations, reparations or apologies as ‘factors’ that may or may not effect [sic] certain desired or undesired ‘outcomes’, but scrutinizes what actors actually do when they employ certain frames, problematize certain practices, issue an apology or establish a specific reparation scheme. We do not search for correlations between meaning and action but rather give an account about meaning in action (Bevir & Rhodes, 2006; Wagenaar, 2011), and since meaning in action occurs over—often long periods of—time, such an account has to be an historic account… (p. 204).


The discourse approach focuses on what language does when actors draw on discourses to employ certain frames. Scholars scrutinize the internal logic of discourses and how these frame thought and action, underlying assumptions, contradictions and inconsistencies, and how those change to shape policy and policy practice. The aim is to understand particular cases, not generalize to a population of cases, although authors may generalize to theory to build knowledge about the ways discourse operates to frame policy. Typical data sources include policy texts including documents (e.g., hearing transcripts, policies) or statements that policy actors make in speeches, or other forms of talk. Often, they include interviews or observations.

Contestation approach to framing

The third approach is contestation of framing, which includes hybrids with the explanatory approach (McBeth & Shanahan, 2004) and an approach that focuses on the contestation of media frames among world powers (Chaban et al., 2007). The approach is rooted in the argumentative turn literature in political science (e.g., Fischer, 2003; Fischer & Forester, 1993; Fischer & Gottweis, 2012; Schön & Rein, 1994; Schön & Rein, 1996; Stone, 1989; Yanow, 1993), and framing in social movement scholarship (e.g., Benford & Snow, 2000), and is concerned with framing as strategic action policy actors use to influence policy outcomes. It emphasizes framing contests between competing coalitions and their conceptualizations of problems and solutions in policy debates. As Hawkins and Holden (2013) put it, “…Political controversies emerge where mutually incompatible policy frames compete to define a given issue and to dictate the policy responses to it” (p. 55). Competing coalitions seek to (re)frame the debates in terms favorable to their side, thus gaining important discursive or symbolic resources. Actors may draw on broader discourses as cultural resources to frame debates,Footnote 4 but are active participants in shaping frames, rather than being driven and constrained by discourses (as in the discourse approach).

Theoretical perspective

The contestation of framing approach defines a frame as a “meaningful category” that is inter-subjectively derived through social processes. Framing refers to the symbolic use of language. Hawkins and Holden (2013) cite Hajer and Laws (2006) this way: “Framing refers to the processes through which actors attempt to impose order upon an ambiguous social world open to a multiplicity of interpretations (Hajer & Laws, 2006, p. 254).” (p. 56). This statement points to the ambiguity of meanings associated with frames. For example, “fat pig” can mean something quite different when used on a rural farm versus an urban bar. The differences in (similar) frames across time or context can be discerned in the creative combinations of framing elements unique to a particular time or place (e.g., Dodge & Lee, 2015). Framing in the contestation approach is action-oriented, dynamic and relational; suggesting that frames do not transcend context neatly but will always have nuanced and creative differences.

From this view, framing is individual and collective strategic action meant to influence the policy process (Hawkins & Holden, 2013; Matson & Chambers, 2009; Sinha & Gasper, 2010), with implications for conceptualizing power. Here, power is the ability to influence framing in a political context and related practices. As Sinha and Gasper (2010) put it, the focus is on “…deliberate, informed use of discursive power in seeking to shape social policy outcomes and change fundamental power equations” (291). Through framing, actors make selections about what categories are meaningful and which are not, thus (re)producing those categories as meaningful (and powerful). This interactive process involves “… selective inclusion and exclusion of elements, based on perceptions of relevance” (Sinha & Gasper, 2010, p. 292).

In this research, framing is linked to discourses in two significant ways that differ from the discourse approach. First, framing is not determined but informed by broader discourses, which serve as cultural resources actors intentionally draw on to creatively reframe issues (or institutions, emotions, adversaries, etcetera). When policy actors promote a different policy solution, they cannot “…simply invent a socially influential new frame”. Instead, “… one must relate to and work with existing culturally powerful resources, amongst other constraints” (Sinha & Gasper, 2010, p. 292). Second, creative framing does not only derive from social discourses but also refashions them. Framing contests shape the discursive context as actors draw on discourse as a resource for creatively framing policy (Hawkins & Holden, 2013).

This approach has several variants, and the theoretical influences are multiple, but most use Rein and Schön as a foundation, along with argumentative turn scholars noted above. Rein and Schön argue that policy controversies become intractable when disagreements over frames make agreements on facts impossible because frames shape which facts seem relevant. For example, in the fracking controversy, pro-fracking and anti-fracking policy actors framed relevant facts differently: about economic benefits versus environmental destruction, respectively. Rein and Schön argued that frame reflection could aid resolution of policy conflicts by getting policy actors to become conscious of the competing frames and their implications. Most scholars contributing to this approach also integrate social movement theory (e.g., Benford & Snow, 2000), which brings strategic action into focus. Social movement scholars focus on the ways social change individuals and organizations engage in framing intentionally to change policy and disrupt the status quo.

Similar to the discourse approach, the contestation approach centers the symbolic use of language and images to make both emotional and factual appeals and persuade audiences in policy controversies (e.g., Mattson & Chambers, 2009; Verhoeven & Duyvendak, 2016; Wolf & Van Dooren, 2017). It examines the meanings conveyed in framing disputes related to public policy— “how policy issues mean” (Yanow, 1996). It attends to how policies are interpreted, agreed upon, and changed through contestation. For example, Mattson and Chambers (2009) found that images played a prominent role in disputes over “wildlife water development.” The image of the bighorn sheep rallied hunters, land management professionals and related elites in support of water development, against those claiming it would harm wildlife. With inadequate processes for resolving disputes, these parties failed to articulate common goals.

Ontology/epistemology

As with the two prior approaches, the contestation approach adopts a relational/antifoundational ontology, and a social constructionist epistemology. The implications for the contestation approach is that policy change is not explained as a process by which frames influence policy actors or by which policy actors use frames to encourage others to change their minds. Rather, framing contests shape the terms of debate or the discursive context (Dryzek, 1997) in ways that make some policy arguments seem favorable while others become hard to utter. For example, Hawkins and Holden (2013) explain that.

…the framing of [alcohol policy] opens the way to certain policy responses, while precluding others; it identifies legitimate participants in policy debates and shapes coalitions of interests (Schattschneider, 1960). For example, the depiction of excessive alcohol consumption as a public order issue leads towards a different type of policy intervention (involving different agencies and different measures of success), than if it were framed as a public health issue or a broader social problem (p. 55).


As Hawkins and Holden show, activists explicitly attempted to shift the frame on alcohol policy in the UK. As one respondent explained, “’a key advocacy task for us was to change the frame of the alcohol problem and to actually get politicians and the general public and the media thinking about [whole] population approaches to alcohol policy’” (p. 59). The alcohol industry responded with a counter frame that shifted focus on individuals who drank alcohol irresponsibly, and on cultural attitudes about drunkenness that needed change. The authors conclude that shifting the debate can influence the policy process and win public support.

Changes in framing are found in shifts in the terms of debate over time. The point is to show how the discursive context shifts, and the implications of this for the acceptance of certain policy ideas and practices.

Methodology/methods

Methodologically, frames have explanatory power for understanding policy change, studied in qualitative ways. Drawing on both document analysis and interviews, scholars trace framing dynamics over time. As Hawkins and Holden (2013) explain:

…The task of the policy analyst is to try to understand how, under what conditions, and through which processes specific frames emerge and are maintained. Identifying the predominant framing of an issue renders policy debates comprehensible and deepens our understanding of the processes through which specific policies emerge (p. 55).


Analysis involves coding, but frames are not understood as unchanging; the focus is not on counting the frequency with which frames are used in debates, but rather the creative combination and use of (new and old) elements within an overall frame and how those change. Again, Hawkins and Holden (2013) “… sought to identify the positions which different actors adopted on the issues at hand, the terms in which their arguments were couched and the evidence they cited for their positions” (p. 57, italics added). The purpose was “…to examine different dimensions of the frames presented by respondents, to identify connections and overlaps between the framing of issues by different sets of actors and to identify cleavages and coalitions between actors advancing specific framings of the issue” (p. 57).

Some scholars trace framing contests over time; others focus on elaborating a dominant frame and how specific individuals or groups contest the dominant frame to unseat it. Some explore the emergence of new frames, others assess frame shifts—including the ways meaning is refashioned. In all instances, analysis focuses on the contestation of framing across groups. Scholars identify the frame that succeeds in setting the terms of debate, how that shifts, and the implications for the favorability of certain policy ideas or practices.

Explanatory approach to framing

The explanatory approach to framing considers frames as rhetorical devices. Frames are defined as “narratives and symbolic devices by which policy actors manipulate policy issue characteristics” (Jeon & Haider-Markel, 2001, p. 216). Frames are central organizing ideas within which advocates of one or another persuasion attempt to give their own meaning to an issue (Gamson & Modigliani, 1989, 2009). Hence, framing is considered a strategic action with framing effects on public opinion and policy making. In many studies, issue framing, frames, framings, and frame shifts are counted and measured (Boydstun & Glazier, 2013; Doan & Kirkpatrick, 2013; Lewis, 2006; Rahn et al., 2017; Rose & Baumgartner, 2013). For example, in their study, Boydstun and Glazier (2013) show—by conducting first an inductive and then deductive analysis of media framing of the Iraq war—how framing shifted over time from a predominant use of “fear” frames to an increasing use of “charity” frames.

Theoretical perspective

In these studies, the theoretical perspective on framing is grounded in psychology, communication or media studies (e.g., Tversky & Kahneman, 1981) that further conceptualize and empirically study “societal and political effects of media framing” (Boydstun & Glazier, 2013, p. 706). Some studies build on theories from other fields to create a framework that conceptualize frames as more generic, generalizable structures. For example, Boydstun and Glazier (2013) develop generalizable frames from prospect theory (loss vs. gain frames) and social identity theory (self-referential vs. other-referential frames). These generic frames can be inductively established by first manually coding a sample of data to distinguish arguments and sub-arguments. Then, a code book can be built and applied to larger data set (Boydstun & Glazier, 2013; Rose & Baumgartner, 2013). The generic frames are than part of a framework in which relations between the frames and policy change, for example, are made. This enables one to formulate hypotheses about that explanatory power of frames and study that empirically across contexts.

Scholars in this framing approach sometimes call themselves “scientific” to contrast with the interpretive approaches to policy framing (Jones, 2018). This also means that they position themselves as ‘positivists’ (see below on the ontology and epistemology of this approach). Many also follow Eleanor Ostrom’s distinction between a framework, theory and model (also common in the narrative policy framework or NPF).Footnote 5 A framework is composed of the “universal elements that any relevant theory would take into account” and their relations (Ostrom, 2005, 28; also see Ostrom, 1990, p. 192), such as the concepts or variables in the NPF or the advocacy coalition framework (ACF). Theorizing occurs when social scientists select elements from the framework to make more specific assumptions about them (Ostrom, 2005: 28). These assumptions can be empirically tested, and if these are observed, the theory is taken to have explanatory power (Ostrom, 1990, p. 24). In contrast, a model “make[s] precise assumptions about a limited set of parameters and variables” to create a “closed system” of related variables, making it possible to “deduce specific predictions about likely outcomes of highly simplified structures” (Ostrom, 1990, p. 24; 2005, p. 28–29). The objective of frameworks, theories and models in framing studies is to derive a universal theory about how specific frames (deducted from theories) influence policy with predictive power. Framing is studied to ascertain possible framing effects on public perception, policy agenda setting or political decision making, as is media studies. For example, in their well-cited study, Boin et al. (2009) explain that “the purposeful utilization of crisis-type rhetoric” significantly “alter[s] levels of political support for public office-holders and public policies” (p.83). Building on Schattschneider, or Cobb and Elder, scholars explain how issues get on a political agendas through media attention. In addition, in refence to Stone (1989), many studies acknowledge the role of “public perception” and the political representation of an issue as part of policy formation.

Ontology/epistemology

These studies often combine a foundationalist/objectivist ontology and a positivist epistemology. But many also combine a social constructivist ontology and a positivist epistemology, acknowledging that all knowledge people generate about the world is mediated and to some extent contextual. However, by applying positivist methodological procedures and methods—e.g., by creating a framework, theory and model, and by conducting empirical research to falsify or confirm it—studies produce generalizable knowledge about framing and framing effects on public opinion or policy. For example, Baumgartner et al. (2008) provide a well-known study of the decline of the death penalty and the discovery of the innocence frame. Often, authors in this approach aim to assess what kinds of frames used in one context will be found in another. For example, Boin et al. (2009) identify three relevant generic and contesting frames in times of crisis: opportunity, threat, and insignificance (p. 86). Not only do shifts in framing lead to new policies—but newly emerging frames can spurn policy integration or conflict expansion.

Methodology/methods

These framing studies often quantify frames and their effects on policy or public opinion, and test hypothesis using quantitative methods. For example, Gruszczynski and Michael (2012) apply evolutionary factor analysis (EFA) developed by Baumgardner et al. (2008), to measure elite rhetoric in congressional hearings about authorizing funds to divert water from the Missouri River. They identified four evolutionary frames and established the dominance of one: the use of proposed water infrastructure. Another way to measure framing and its influence is combining datasets, such as public opinion studies with framing studies or surveys. For example, Rahn et al. (2017) applied a method that resembled consumer surveys, and then an Internet survey, with 1,630 respondents from seven Midwestern states, which presented frames to these respondents. The results showed that a consumer choice and food freedom frame is more effective than the public health risks frame, because of its dominance among the policy establishment.

Institutional approach to framing

The institutional framing approach is unique in three ways that we elaborate below: (1) framing is studied in narrowly bounded institutional spaces, (2) framing is deeply embedded in and influenced by institutional contexts, and (3) framing analysis is focused on institutional processes, e.g., institutional cohesion or integration, more than, for example, societal contestation in policy conflicts. It shares similarities with the contestation approach in analyzing different actors who contest the framing of public policy and its implementation (e.g., Cerna & Chou, 2014; De Ville, 2013), although some focus on the dominance of a particular frame or framing (e.g., Kohler-Koch, 2000). It also shares some similarities with the sensemaking approach, for instance, in conceptualizing framing as sensemaking to clarify an ambiguous situation or simplify a complex situation (e.g., Kohler-Koch, 2000; Cerna & Chou, 2014), or to build consensus or otherwise better organize to meet (or design) organizational goals (e.g., Smith, 2003).

Theoretical perspective

As with the sensemaking approach, framing is generally considered a process through which people give meaning to reality in daily life, or institutional life, although it emphasizes how the institutional context significantly shapes framing processes and vice versa. It generally adopts Rein and Schön’s definition of framing as “a way of selecting, organizing, interpreting and making sense of a complex reality so as to provide guideposts for knowing, analyzing, persuading, and adapting” (Cerna & Chou, 2014, p. 79, citing Schön & Rein; 1994, p. 263; see also Kohler-Koch, 2000). Generally, the main concern is how policy actors operating within complex and ambiguous institutions select appropriate frames for the situation and thus orient action and policy in a common direction (or face serious risks). Some describe frames similarly as “referents for action” (De Ville, 2012; Morth, 2000), as they orient action in a particular direction. However, some scholars also define frames as definitions of a policy issue (Candel & Biesbroek, 2016; Smith, 2003), following Baumgartner and Jones (2009) and Jones and Baumgartner (2005), among others, showing affinity to the explanatory approach. What they have in common, however, is a concern with how framing processes and institutions interact to form (or not) specific governance arrangements, particularly policy cohesion or integration.

This literature tends to elide theoretical debates about whether framing is strategic or implicit, although several suggest framing is strategic when focusing on the ways institutional actors purposefully select frames from several alternatives that best make sense of an ambiguous situation. Cerna and Chou (2014) explicitly argue that “… the ‘constellation of actors involved in the creation of a governance architecture engage in strategic and discursive interactions with the aim of defining a collective frame of reference for their action…’” (p. 78, citing Borrás & Radaelli, 2011). Although not explicitly, they equally focus on the ways frames are selected to make sense of ambiguous situations, for example, as a process of discriminating between various options (Kohler-Koch, 2000).

While this approach echoes the sensemaking approach, it has three distinct features related to its grounding in institutional theory as an alternative to rational choice models including new institutionalism (e.g., March & Olsen, 1996), discursive institutionalism (e.g., Schmidt, 2010), historical institutional (e.g., Thelen & Steinmo, 1992) or related literature on governance, EU integration, agenda setting or other policy process literature.

First, framing contestation occurs in narrowly bounded institutional spaces in government or governing institutions among policymakers and administrators rather than the public sphere where diverse coalitions may include governmental actors, activists and organized interests. Contestation occurs in subtle differences in policy statements made in local versus national policy (Scholten, 2013) or between regulators and EU member states (De Ville, 2012) rather than in society broadly.

Second, framing practices are consciously conceptualized as embedded in, and highly influenced by, the institutional arrangements in which the framing contests (or dominance) take place. This may mean that people located in powerful positions and institutions have greater resources to articulate frames and win framing debates. As Morth (2000) put it, “…organizational structure generates frame competition”, for instance, in the form of specific roles and functions of Director Generals (DGs) in the European Commission and whether the issues they tackle are considered the purview of one or multiple DGs. More generally, all are interested in the interplay between framing and institutional context, viewing frames as embedded in institutional practice (Kohler-Koch, 2000), constrained by institutional settings (Smith, 2003) or mutually constitutive with institutions (De Ville, 2012).

While most elide overt discussions of power, specific institutional arrangements clearly have power to enable (or constrain) action. Ideas (framing) play a prominent role in this process because an accepted frame confers legitimacy to a particular point of view, guiding action in a particular direction. For example, Morth (2000) argues that “Frames legitimate certain decisions and activate certain issues, actors and special types of knowledge. Thus, frames concern power—the power to define and conceptualize” (173–4). While only Morth explicitly discusses power, most similarly view institutional arrangements as placing constraints on action though frames that signal the legitimacy of policy actions or forms of governance.

Finally, a main theoretical aim of these papers is understanding institutional processes—such as building new institutions, for instance of the EU. The institutional focus varies between (1) policy cohesion or integration across governing scales or offices (e.g., Candel & Biesbroek, 2016; Morth, 2000; Scholten, 2013), (2) institution building at the EU (Kohler-Koch, 2000), (3) institutional evolution and change (e.g., Smith, 2003) and (4) the power of institutional regulations to serve as framing devices for policy debates (De Ville, 2012; Cerna & Chou, 2014).

Most scholars using this approach are interested in the interactions between ideas (framing), institutions, and policy (e.g., Cerna & Chou, 2014; Smith, 2003). Therefore, institutions are at the forefront of analysis and theorizing, and framing analysis is used to better understand institutions. Due to the focus on institutions, framing generally is viewed as something that changes over time as institutions are built or evolve, and so reframing is an important explanation for institutional change. For instance, Morth (2000) conducts an in-depth analysis of frame competition within the European Commission, demonstrating how the view of defense equipment and industry policy was once dominated by a “market” frame but was challenged by specific DGs who developed a modified market frame that incorporated industrial, scientific and technological components. This approach generates some interesting findings. For example, Smith (2003) sees institutional evolution not as a linear or successive process but a process by which new frames are overlaid on top of old ones; as institutions develop, they become more elaborated.

Ontology/epistemology

The institutional framing approach tends not make explicit ontological or epistemological statements or engage in such debates. It seems to implicitly adopt either a relational, anti-foundational ontology and a social constructionist epistemology or an objectivist ontology and positivist epistemology. While the distinctions are fuzzy, the former (relational ontology/constructionist epistemology) emphasizes how the adoption of a frame not only enables policy actors to advance their interests but, contrary to rational choice theory, “organizes the work” of the institution, for example, within the European Commission (e.g., Morth, 2000). Such organizing may influence policy outcomes but causality is not direct, having more to do with pointing an organization in a particular direction. Furthermore, as Morth (2000) argues, frame competition "functions as an important identity-building and sense-making process in multi-organizations" (173). In contrast, the latter (objectivist ontology/positivist epistemology) sees frames as referents or problem definitions that enable actors to advance their own interests, in ways consistent with the dominant frame, and that directly affect policy or organizational outcomes (e.g., De Ville, 2012). Both sidestep such considerations to emphasize how decision makers select a frame to make sense of ambiguous situations to lend clarity and generate consensus.

Methodology/methods

Most of these studies offer in-depth empirical case studies over extended periods of time that are deeply contextual, and employ some form of process tracing, consistent with their view of framing as a process that unfolds over time. The case analysis and process tracing methodology allow scholars of the institutional approach to understand how (re)framing processes affect institutional development and change over time and vice versa. One article develops a processual framework of policy (dis)integration (Candel & Biesbroek, 2016).

These studies tend to adopt a meso-level of analysis, focusing on institutional dynamics in specific institutional settings, especially the EU. However, Kohler-Koch (2000) adopts a micro-level analysis focusing on the impact of a single individual on institutional development through an authoritative speech while Scholten (2013) adopts a scalar approach that examines institutional dynamics across local and national governing scales. These studies attend to context through rich description of the politics surrounding the evolution of institutions.

Consistent with this approach, most analyze a combination of authoritative texts (speeches, reports, policy statements), secondary sources of institutional development and other documentary evidence as their data sources, with some adding interviews (e.g., Morth, 2000) and/or quantitative analysis of media attention in newspaper articles (Scholten, 2013). These studies are oriented to generating theory about institutional dynamics and change such as institutional cohesion, integration or evolution, and may also develop insights useful for informing policy development.

We made explicit the assumptions across these approaches in Table 2, which serves as a heuristic to choose framing approaches based on researchers’ questions, preference for studying frames or framing, and methodology/methods.

Table 2 Convergences and divergences across approaches to framing in policy studies
Full size table

Discussion: reframing the discussion across approaches

Our review shows that many framing studies mix theoretical insights and methods and are rather pragmatic. Their empirical questions guide choices for defining frames or framing, and the overall approach. We contend that no approach is better than the other in an absolute sense, but that different approaches are used for answering different types of questions. In this light, Table 2, as presented below, can facilitate a dialogue across the five approaches about their unique contributions and generate cross fertilization between them (see also Dodge, 2015; Durnova & Weible, 2020; Jones, 2018; Jones & Raedelli, 2015; Shanahan et al., 2014). Although very different, we also saw that the five approaches aim to address similar questions that evolve around the role of framing in policy contestation or change, but that they study frames and framing and the work framing does, in different settings or at different levels. Based on this we formulate four animating questions for the field and reflect on how each approach engages with these questions in complementary ways.

Do frames influence policy or are policies manifestations of framings?

The five approaches to framing analysis all emphasize the importance of language and meaning and attempt to understand how frames influence or are part of policy processes. However, they differ on whether they view frames as influencing policy (change) or frames/framing as internal to (changing) policy processes. The explanatory approach, and some articles in the institutional and contestation approaches, views frames as schemata that directly influence public opinion or policy making, while the sensemaking, discourse, and some articles in the institutional and contestation approaches view frames or framings as inherently part of policy making. To put it in positivist terminology, in the former group, frames and framing are considered independent variables, and public opinion, policy, policy change, policy institutions are dependent variables, while in the latter group, frames, frame shifts and framings are inherently part of policy making and policy change. They cannot be independent variables because they are internal to policy making and how policy change comes about. Hence, the main difference is whether frames are independent factors affecting policies or are internal to policy processes. Durnova and Weible (2020) similarly point out that policies in “mainstream” policy analysis are ‘rules in form’ or ‘rules in use’ and in the interpretivist approach are ‘manifestations of meaning’ (2020, p.4).

These different starting points have consequences for what framing is/does for policies. The former group productively answer how frames influence policy by demonstrating how frames measurably affect the terms of the debate, set the agenda, or affect policy decisions; how this informs policy practices, policy formation and implementation; and/or how framing effects lead to policy stasis or change, including institutional change. The latter group contributes by building knowledge of how framing as sensemaking in interactions, the formation of discourse coalitions around framings, or the process by which particular policy ideas and arguments—as forms of policy stasis and change—become dominant on the policy agenda or in policy practices. In this case, policy makers and policies are part of framing, discourses and discourse coalitions rather than affected by them. They provide more detailed insights into ‘how’ policies mean for different groups in society and how this meaning changes over time (Yanow, 1993).

There is room to combine these approaches and build knowledge in the field through conversations about their complementary insights.

What is the role of frame contestation in policy conflict?

Contesting frames and framings are studied through several different approaches. The sensemaking and most institutional approaches involve subtle contestations over framing a situation through interactions but emphasizes alignment (cooperation) to coordinate action. The contestation and explanatory approaches focus on overt contestation. The former aims to understand the dynamics of contestation through frames shifts, referring to the creative (re)combination of framing elements either in the interactions between (conflicting) actors, or in contested framings in media and/or policy documents. (This clearly aligns with studies that view policies as manifestation of meaning, described under the first question above.) In depth analysis of what framing elements are recombined, aligned or contested reveals dynamics of and reasons for discursive struggles about policies and what they mean. The latter aims to empirically (numerically) explain how competing frames of problem definitions become more or less salient, persistent and resonant over time, that is, beat out the competition—and as such have a framing effect on policy making. Both contestation and explanatory approaches analyze framing contests across competing coalitions of actors as they make sense of a policy issue. Analysts are concerned with which actors are connected in networks of shared meanings or beliefs, and how they either attempt to seduce others to understand the issue in a similar way or grow the resonance and salience of frames.

In the discourse approach, contestation involves direct challengers seeking to dislodge dominant discourses, but who often confront intransigence in entrenched and powerful actors, practices and knowledge production process that are hard to change. Counter-discourses are often changed instead and incorporated into dominant discourses; less overt processes enable dominant discourses to subtly persist even as new policy actors challenge them. This approach aims to understand why contestation of less visible discourses is difficult, and what meanings are unheard, unrecognized or excluded (see below).

Understanding these differences offers a fuller notion of contestation in policy making processes and in policy practice: from the situational interactions of policy practitioners (who compete to frame a situation to coordinate action) to overt competition over framing (as competing coalitions struggle over meanings or problem definitions). These overt forms of contestation can all be underlined by the intransigence of dominant discourses that challengers may seek to transform but instead may see counter-discourses transformed to fit dominant meanings.

How do frames or framings reveal unheard voices?

All approaches to framing aim to make visible aspects of policy making that are left invisible when only studying actors, governmental institutions, procedures, networks, and governance arrangements. Each draws attention to the role of arguments, language, and/or meaning, albeit in different ways. In the explanatory approach, frames are made visible and scholars study how this linguistic dimension—often studied in the form of media framing—influences policy making.

The sensemaking, contestation and discourse approaches all aim to make frames visible, too, but pay special attention to analyzing what interpretations, voices and ideas are left out or ignored in policy and decision-making processes. More specifically, the sensemaking approach aims to makes visible unheard voices and understandings by individuals and groups who grapple with challenging policy situations. In this approach, studies illuminate stories that are being told, and how sensemaking in interactions can create a social or democratic fabric (Hendriks et al., 2020). The focus is on nitty gritty practices and policy making processes; policy actors’ voices are given primacy and may be presented to practitioners themselves to reveal for them what they might not see in their framing. In this sense, these studies can be insightful in conflict-resolution and serve as a basis for frame reflection and reframing. In the contestation approach, voice is also given to marginalized groups and their discursive struggles, especially across broader social networks and coalitions, which may lead to social and policy conflict (escalation) (Laws et al., 2014; Wolf & Van Dooren, 2017). In the discourse approach, marginal discourses are made visible even as they are disciplined (constrained) by dominant discourse, for instance, that suppress certain policy voices and choices that might be more liberatory. It shows how these processes are tied up in establishing and maintaining systems of power, sometimes via their embeddedness in systems of knowledge production that support the status quo of power relations.

These studies reveal unheard voices at different levels of analysis. The explanatory approach makes collective frames and their influence on policy visible and as such shows the influence of language on policy or institutional stasis or change, while the sensemaking approach gives primacy to practitioners’ voices in situational framing processes. The contestation and discourse approaches both expose marginalized framings either across competing networks/coalitions at the meso level or between challenger and dominant discourses in broader society.

How do frames become dominant?

Most approaches grapple with understanding or explaining of how frames or framings become influential or dominant and lead to or are an expression of policy change. The explanatory approach explains the dominance of certain meanings over time by measuring salience (how often a related set of arguments is used), resonance (how many arguments cluster together to form a frame), and persistence (how long it lasts) (Baumgarter et al. 2008, p. 138). Studies build on concepts from media studies and explain policy changes via (numeric) shifts in policy framing. They offer a level of breadth in the study of policy change not possible in qualitative studies. Qualitative approaches, however, can also explain dominance but as qualitative shifts in symbolic meaning. These do not dominate numerically, but by offering a mental shift or a reframe that orients people in a different direction. A single image, in this sense, can change the conversation even if it is not numerically dominant.

In the discourse approach, broader societal discursive structures are studied and how these become dominant as discursive elements are strongly tied together over longer periods of time, across different sources with specific consequences. It offers a qualitative analysis of discourse dominance as an historically formed and institutionalized linguistic structure entangled with political reality that frames or even ‘disciplines’ subjects. It excels at uncovering broad societal meanings that infuse policy discussions and decisions, and how they constrain understandings in policy domains, even when those domains appear to be undergoing significant change. It helps us see that change is slow, and even when change happens, some elements of dominant societal understandings persist and are resistant to change due to constrains imposed by dominant discourse and related practice and knowledge production processes.

As with the discourse approach, both the contestation and sensemaking approaches center shifts in meaning overtime and focus on how certain framing elements may be connected or disconnected from other ideas, which may or may not result in new ways of understanding a policy issue. They conduct deep dives into meaning-making processes associated with frame shifts and the ways these are creatively connected to—or disconnected from—various framing elements over time. Neither approach aims to measure how certain frames or framing elements gain in popularity among members of the public or policy makers; but shows that when a frame is indicated to dominate a discussion, it makes some arguments seem more credible and acceptable, and therefore more likely to appear in final policy choices (or other relevant outcomes), even when a majority of policy participants prefer the losing frame.

In these studies, frames/framings become dominant and more influential on policy change either by becoming more salient, resonant or persistent to the degree that they generate policy effects (e.g., shifts in policy or public opinion) or through qualitative shifts in symbolic meaning that reorient the terms of discussion—or, as in the discourse approach, prevent these shifts through the persistence of historically formed and institutionalized linguistic structures.

We have shown that by looking into these different framing approaches we can better understand and appreciate the diversity of frame and framing studies in the policy sciences. Whereas self-declared positivists can draw from interpretivists insights about the framing as “moving targets” in framing processes, self-declared interpretivists can more clearly develop plausible conjunctures and derive patterns from their empirical studies that speak to the frameworks, theories and models of positivist scholars. Dialogue is the hallmark of good research because it helps us to see our blind spots, opens new avenues of research, and also generates fresh debates. We suggest four questions as areas of importance for future dialogue and communication across perspectives despite the ontological/epistemological and theoretical assumptions underlying each approach that makes divergent conclusions sensible from each perspective.