(Re)drawing Lines in Our Research: Using Policy Mobilities and Network Ethnography to Research Global Policy Networks in Education

The flows and frictions of policy networks

We are all scholars of policy to some extent. Such a definitive statement is not without merit—after all, we all experience and interact with policy across education, from the “eddies and flows” (Cochrane & Ward, 2012) of its movements to the “fixities and moorings” (Sheller & Urry, 2006) of its frictions. Regardless of whether we explicitly refer to ourselves as “policy researchers,” the various dimensions of education and schooling upon which we choose to putatively focus—including pedagogy, curriculum, assessment, and student wellbeing—are continuously being (re)shaped and (re)constituted by the various material and discursive elements of policy, in both predictable and unpredictable ways. Unsurprisingly, policy remains a central preoccupation of education research, leading to a continued focus on developing, adopting, and adapting different theoretical, methodological, and analytical approaches to better understand the changing empirical contexts we face. These changes include not only the changing policies themselves but also the changing processes, actors, spaces, and relations by which such policies are developed, disseminated, contested, and enacted. It is fair to say that in this contemporary moment, all manners of policy spaces (e.g., the classroom, school, schooling system, and the State) are connected with all manners of other spaces and actors (e.g., the World Bank, the OECD, philanthropic foundations, and international “best practice”) while being disconnected from others.

Let me qualify my opening gambit somewhat further by asserting that we are all scholars of a thoroughly globalized and interconnected education policy. If, as the adage goes, all politics are local, then it follows that policy is more of a complex bricolage, assemblage, or network of relations enfolding and rearranging elsewhere(s) into here(s), and vice versa. For instance, a school's pedagogy might be informed by the schooling system authority administering teacher credentials, but it might also be steered by the professional learning modules freely offered to teachers by multinational EdTechs (see Lewis, 2022). In developing countries, education departments often create a regional forum to develop “local” policymaking capacity and cooperation, but this tends to entail interaction with donor countries and international development organizations from the Global North (see Lewis & Spratt, 2024; Spratt & Coxon, 2020). While sustainability education might appear as a converging global discourse, its integration into federal systems necessarily reflects the cultural, historical, and economic contexts and histories of its subnational polities, resulting in variegated understandings and enactments of an otherwise “similar” policy initiative (McKenzie & Aikens, 2021). Indeed, as much as all policy is “on the move” (Peck & Theodore, 2010, p. 169), it seldom moves—or does not move—in essentialized and foreseeable ways. Moreover, the presence of such policy formations, and the processes by which they are developed and disseminated, suggest that policy research requires novel conceptualizations and methodologies able to address these emerging phenomena with both descriptive and analytical power.

As several scholars have argued (e.g., Gulson et al., 2017; Lewis, 2021; Lewis & Spratt, 2024; McKenzie, 2017; McKenzie et al., 2021; Savage et al., 2021), a key response to the changing empirical processes and practices of education policy, particularly in terms of its spillage across more territorially-delineated policy spaces (e.g., the nation-state), has been the increased uptake of research frameworks informed by policy mobilities. These approaches attempt to account for the complex relations between actors, organizations, and spaces that work to enable (or constrain) policy, and identify and conceptualize how policy moves through multiple and diffuse means, including non-scalar modalities. In doing so, such approaches seek to de-essentialize national or other territorial spaces as ultimate a priori reference points, thereby facilitating explorations of “the processes, practices, and resources brought together to construct, mobilize, and territorialize policy knowledge” (Baker & Temenos, 2015, p. 825). This includes interrogating what networks and relationality comprise—that is, what connects people, objects, and so forth—so the material and discursive elements of a policy are made (im)mobile, as well as the policy itself.

This special issue responds directly to these emergent research realities to chart how policy mobility and cognate approaches, particularly network ethnography, can be employed to elucidate how policy unfolds in a variety of diverse spaces and contexts. Spanning Latin America (Matovich & Esper, 2023), China (Jin, 2023), Australia (Rowe, 2023), the UK (Gellai, 2023), and even the in-person meeting rooms of United Nations personnel (Stahelin & McKenzie, 2023), all contributors have usefully shown how network ethnography, undergirded by a policy mobility research sensibility, can draw attention to the varied actors, spaces, and relations (often obscured and unseen) through which a policy (network) is made. I have often engaged with similar processes and concerns in my own research into global education policymaking and governance over the past decade, and frequently in collaboration with some of the contributors to this issue. This has prompted my reflections here on how policy mobilities and network ethnography can open new lines of thinking and knowing—and perhaps unknowing—in our field of research.

Lines of enclosure

While defining and bounding the research problem is a key feature of education research, the fluidity and mutability of policy networks make their delineation particularly challenging. As Santori and Jin (2023) note, negotiating what to include and exclude within the bounds of a policy network is a decidedly “messy business.” Policy mobilities emphasize that policy ideas and practices do not simply flow into new environments “fully formed”; on the contrary, they undergo complex forms of rearrangement that are shaped by and, in turn, rearrange the space into which they move. In other words, policy ideas and practices are not transferred into new spaces from afar, but are (re)made and enacted through ongoing processes of rearrangement, thereby shaped through unique constellations of people, practices, and places. That said, the ability to account for this complexity and movement is simultaneously a boon and bane to policy mobilities. Indeed, while acknowledging the constant mutability of a policy network is one matter, translating this abstraction into a specific set of research practices is an entirely different matter altogether.

A key consideration when researching a policy network is thus to decide precisely where, when, and around what lines demarcation should be drawn. Although policy networks are clearly in a state of ebb and flow—typically functioning as a network itself while being connected to or a part of other networks—we must draw a line somewhere that determines which elements (e.g., actors, organizations, relations) to include or exclude. On the one hand, this involves purposefully excluding lines of connection in the policy network. For instance, in this issue, Rowe's research into Social Ventures Australia focuses more on policy developments related to P-12 (or K-12) education while intentionally neglecting associated networks concerned with higher education. On the other hand, connections might well be (and frequently are) incidentally excluded, such as tangential actors, organizations, or spaces that are harder to identify, as well as components often overlooked because they change during or after the research period. As McCann and Ward (2012, p. 328) observe, “an assemblage [or policy network] is always in the process of coming together … just as it is always also potentially pulling apart.” Faced with such dynamism and despite our best efforts to artificially “fix” what is included in our analytical gaze, these elements typically elude our attempts to exert control.

However, limits on the knowledge of policy networks arguably do not limit the utility of the concept for policy research; they merely recalibrate what can, and perhaps should, be done. Although we cannot freeze the perpetual toing and froing of a policy network, we can draw attention to the elements, connections, and flows that are most apparent (to our imperfect gaze) and influential (by our imperfect reckoning). We need to consider that any attempt to present the “findings” of our research is not the final attempt or definitive word on the matter but is invariably partial (in both senses of the word). Research is always an ongoing and relational process open to further arrangements, and De Sousa Santos’ (2007) notion of the abyssal line is particularly useful here. While drawing a line around our research is necessary, we must do so with the explicit awareness that there is always something on the other side of that line. Although there are things that are not visible or expressible in our research, either through purposeful choice or ignorance, this does not mean that they do not exist, or that these same limitations exist for others or in different circumstances. Bearing this in mind, we need to then be ready to redraw our own lines and be self-reflexive about why we draw these lines where and when we do.

Lines of enquiry

Now that the “what” (not to mention the “where” and “whom”) of policy networks has been considered, my second reflection concerns what might be called “lines of enquiry”—that is, how might one employ policy mobilities and network ethnography in empirical research. In observing the network diagrams created by all contributors, I could not help but recall Galloway's (2012, p. 90; emphasis added) argument that “only one visualization has ever been made of an information [or policy] network.” On this matter, Galloway (2012) is worth quoting in length:

Minuscule branching structures cluster together forming intricate three-dimensional spaces. Nodes are connected by links. Small capillaries merge into ever greater arteries fabricating massive hierarchies governing flows and prohibitions on flow. Yet through it all, the legibility of the map remains suspiciously one-sided, even ideologically motivated. The viewer is able to intuit certain vague cosmological ‘facts’ about the digital firmament (apparently information likes to cluster; these colour enclaves persist unmiscegenated; we love trees after all), while gleaning little about the ‘facts on the ground’ (who is connecting and who isn’t; the intra-network struggles between protocological and proprietary software; the reification of pyramidal hierarchy; monetization of unpaid microlabor). (Galloway, 2012, p. 90)

With inimitable finesse, Galloway (2012) underscores a key critique of network ethnography; namely, that such networks often become the object of study themselves, rather than merely a vehicle for representing a given social phenomenon for relational policymaking (see also Gulson et al., 2017). Across the five papers in this Special Issue, it is somewhat telling that the diagrams representing these diverse processes (both material and immaterial) as nodes and arcs all produce a figurative sameness, despite the clear differences between their respective policies, processes, actors, and outcomes.

This is not to say that network ethnography cannot produce useful insights, but perhaps more can be said about how such contextualizations and differences might be adequately captured and rendered in a way that better reflects these differences. If all policymaking can be said to occur (more or less) via heterarchies of vertical and horizontal relations, then arguably all policy processes and the policy networks that underlie them will produce similar network diagrams. We are thus left in a situation where, to paraphrase Stein, a network diagram is a network diagram is a network diagram …. And as noted by Ball and Junemann (2012), while it is possible to map out multiple relations between actors and organizations in policy networks, their most common forms of representation cannot capture that “the nature of the relations between members [of policy networks] (as represented by the arrows in … network diagrams…) is not the same in every case” (p. 10). In other words, although the arcs and nodes might appear to be the same between diagrams, they are patently not representing identical or even similar empirical situations.

Therefore, while it might be provocative, I genuinely pose the question: What is gained by producing these diagrams? If the point is to demonstrate the intricacy of the networks and myriad actors and organizations (nodes) and relations (arcs) within them, then there seems little need for them—as this does not need diagrammatic evidence. This also raises the question of what this representation offers that other forms of representation do not, especially when the software used to produce these representations invariably produces the same assortments of lines and dots, often too infinitesimally small to discern. This is particularly relevant in respect to our previous point regarding how we delineate policy networks (i.e., who or what is included and excluded, and how and when are these decisions made?) or whether we should consider all the networks enfolded within other policy networks. Effectively, what might start as (relatively) ordered images of fixed relations eventually take the form of fractals possessing insuperable detail, inhibiting their clarity and utility for justifying scholarly conclusions or more political calls for action or reform.

Perhaps they ought to be considered less as a means of reporting the analysis for a prospective reader and more as real-time in situ analyses by the researcher; that is, as an example of analytical processes, rather than merely an example of an analytical product. As with many things, intention is arguably the key in this respect. Regardless of their purpose, there is a need for researchers employing network ethnography to justify how, why, and to what end they seek to use diagrams. To return to the rationale posed by Santori and Jin in this issue, we might do well to intentionally embrace “the gray,” thereby giving greater thought and consideration to why it is we do the things we do.

Conclusion: Creating different lines of flight

In conclusion, I must first commend the contributors for this outstanding Special Issue. Good research should always provoke new thinking and reflections on what we thought we previously knew—and this issue certainly presents very good research using such a measure. At the same time, I would argue that all the articles in this issue implicitly encourage approaching policy networks and policy research “in the middle.” By this, I mean that policy mobilities and network ethnography should be considered both as methodological frameworks and as research(er) dispositions to ongoing reflexive criticality, acknowledging both new productive lines of flight and unproductive blockages. I see policy mobilities as engaging with Lury's (2012) call for live methods in social science research, which “must be satisfied with an engagement with relations and with parts, with differentiation, and be involved in making middles, in dividing without end(s), in mingling, bundling and coming together” (p. 191; emphasis added). Situations of one's enfolding in the research problem—and the acknowledgement thereof—have become prime methodological considerations. In this sense, policy mobilities very much start in the middle, albeit with middle-ness “no longer defined with respect to determinable end points [but…] an infinite and infinitely divisible space” (Lury, 2012, p. 190). It is a case of being there, wherever there is or might be, and of meeting a policy network wherever, whenever, and however you can. However, this does not position the middle as an interminable transitory phase between the beginning and end. Space, time, relations, policy, and policy research(er) are all in the middle; the middle just is. As Santori and Jin (2023) note, the ethics of network ethnography represents “both a challenge and an opportunity.” I am genuinely excited to see how these challenges and opportunities will affect our future research.