Introduction

The concept of policy diffusion is more complex than might be initially assumed. The literature presents differences among and across disciplines, settings, and methodological traditions. Those fundamental differences are compounded by “naming” differences: how best to distinguish policy transfer, emulation, convergence, and diffusion, not only definitionally but also operationally in empirical analyses? Analysts of the topic are products of their own training, experiences, and values, and that is reflected in the research.

Thus, a positionality statement seems warranted. While we draw on numerous sources in our reflections in this chapter, we primarily base our perspective in neo-institutional theory (Scott & Meyer, 1991), and most specifically in the “world society” or “world polity” literature, as exemplified most prominently by the work of John Meyer and colleagues (Meyer & Rowan, 1977; Meyer & Bromley, 2013; Bromley & Meyer, 2015). As Pope and Meyer (2016, p. 281) argue, from Western cultural traditions have evolved shared “world models” infused with what they term “dominant scripts” or “foundational cultural assumptions,” such as rationalization, universalism, science, professionalization, progress, and individualism. In this context, across-border influences can create pressures on organizations’ traditional structures and processes, i.e., their internal technical logics.Footnote 1

Higher-education systems and institutions comprise a prominent exemplar for this perspective, in that they exist in highly institutionalized and increasingly globalized environments. External forces cannot easily be resisted in the increasingly porous, interconnected context of twenty-first century higher education. Visions of universities as faculty-driven organizations societally chartered to provide havens for learning and scholarship ring increasingly less true. Institutions’ strategic and managerial choices are becoming less a product of local campuses’ independently chosen academic priorities than of across-border trends. Those trends have been most prominently characterized in the literature as neo-liberalism, academic capitalism, and marketization (Slaughter & Leslie, 1997, 2001; Slaughter & Rhoades, 2004; Maringe & Foskett, 2010; Pusser et al., 2012). Some writings on the topic rely heavily on limited country- or region-specific lenses and research methods, but as Marginson and Rhoades (2002), Shahjahan (2012), and Shahjahan and Kezar (2013) have argued, there is much for higher-education researchers to gain from going beyond their “national container” and avoiding “methodological nationalism.” Indeed, the trends are sufficiently ubiquitous to merit characterization as evidence of an emerging “world society” (Bromley & Meyer, 2015).

From that last perspective, shifts in higher education can be driven by the spread across governmental borders of “hard” (legalistic) and “soft” (administrative and professional) pressures to act in certain putatively rational ways. Accompanying these pressures is the imposition of across-border accountability standards, enforced by both coercive and normative means (Bromley & Meyer, 2015). Intermediary organizations can play an authoritative role in influencing adoption approaches, especially via their prestige (ontological) authority, moral authority, or capacity-based authority (Meyer & Bromley, 2013). Pressures toward certain processes and standards are energized and applied by growing cadres of professionals trained and socialized to enforce growing consensus among polities around rationalized culture. A number of works in higher education policy adoption have endorsed the power of such influences extending beyond mere geographic contiguity (e.g., see Ness, 2010; Sponsler, 2010; Tandberg, 2013; Ness et al., 2015).

Settings vary, however, in their level of openness to, or insulation from, these pressures. Meyer and his colleagues (see especially Pope & Meyer, 2016) have highlighted six dimensions of diffusion processes shaping these variations. Adopting these dimensions for tertiary policies and systems is straightforward. To the extent universities and their faculties are structurally embedded in a large number of cross-border scholarly and professional organizations, their leaders are more likely to adopt policies employed by peers in other polities. To the extent tertiary policies, values, and practices do not match those of cross-border initiatives, local policymakers will decouple rationalization initiatives from their original forms, to help ensure allegiance and compliance “on the ground.” To fit with local understandings and knowledge bases (e.g., senses of national or state history), tertiary policymakers will also domesticate or “glocalize” across-border policy reforms originating outside their borders. Relatedly, to the extent externally imported policies appear discordant with local cultural framings, tertiary policymakers will pursue contingent diffusion, embracing only those policy aspects sufficiently comfortable in context. Further, to the extent tertiary policy pressures are in sync with in pressures in other policy arenas, policymakers may embrace multiple diffusion by adopting a cluster of interrelated policy choices spreading across borders. For example, in embracing the global environmental movement, institutions may reform their curricula in concert with reforming their business practices in environment-sensitive directions. Indeed, universities may adopt somewhat opposing policies within a particular domain like higher education, owing to cross-pressures across domains (Levine et al., 2013). Finally, multi-level diffusion processes can take place. These processes seem particularly relevant for the university-centered analyses presented in the preceding three chapters. As Pope and Meyer note (p. 296), “… [M]odels may diffuse at various levels of the social system. Individuals, organizations, nation states, regional groups, or intergovernmental organizations can be receptor sites for diffusion…” In this sense, “nesting and hierarchy” are factors in policy diffusion.

Certainly, the three chapters we review here can be productively examined through a variety of historical institutionalist lenses, including path dependence, power, displacement, layering, drift, and conversion (see, for example, Pierson, 2004; Fioretos et al., 2016; Hall & Taylor, 1996; Mahoney & Thelen, 2010; Thelen, 1999).

Employing a lens derived from that of Meyer and his colleagues, we examine in this essay (1) the extent to which the three preceding chapters show consistent spread of rationalist policies across governmental borders, (2) the role of intermediary factors in policy diffusion, and (3) the ways “filters” at borders shape whether and how fully individual polities embrace policies already adopted elsewhere. We assay the apparent differences across settings and provide some concluding comments on critical implications.

The Three Chapters

The Gándara and Woolley chapter (Gándara and Woolley, this volume).This chapter [henceforth GW] provides not only an intriguing empirical analysis of adoption in higher education but also a valuable review of prior work on policy diffusion across the U.S. states. While praising the value of the numerous earlier quantitative analyses of policy adoption, mostly using event-history analysis, GW make their greatest contribution in stressing several limitations of that body of work. First, prior analyses have too often focused on replications of theorized influences on adoptions, at the expense of the frequent failures of some familiar propositions. Why haven’t we seen more consistent results in across-state diffusion analyses? What are the key factors leading to non-adoption? Second, GW argue that analysts have too often assumed that regional diffusion across similar, (usually contiguous) states is the most predominant form of diffusion, ignoring national-level networks, national policy pressures, and the ascendant influences of certain states as recognized leaders in policy development, often spurred on by intermediary organizations. Third, GW emphasize the fraught nature of the concept of policy adoption, for both researchers and the larger policymaking arena. How, exactly, do adoption processes unfold in polities? At what stage in these processes are diffusion processes most influential? What aspects of these processes are diffusing (rationales, targets, metrics)? Fourth, because quantitative analyses at the level of states’ formal legislative and administrative actions cannot tell us enough regarding the “What,” “How,” and “Why,” of adoptions, GW make the case for more use of other techniques, including case-study analysis, interviewing, document analysis, and the like. Helpfully, GW cite various recent works addressing the limitations of the dominant earlier paradigm in U.S. diffusion analyses.

The second part of the GW chapter presents an empirical analysis of two states that touches on the concerns raised in the first part of the chapter. Qualitative and focused on non-adoption, this analysis of proposals to adopt performance-based funding [PBF] in those states concludes that, while external pressures were present tilting the two states toward adoption, a variety of internal characteristics and considerations overrode external influences in both states. Most notably, the states’ “flagship” institutions led efforts to resist national trends toward convergence in adopting performance funding.

Viewed from the Pope and Meyer (2016) perspective, the most striking conclusions to be drawn here connect to embedding and domestication,. Although some aspects of the PBF movement are supported structurally across the U.S. (e.g., by the influential Lumina Foundation in its grants and networking efforts), tertiary education in the U.S. is largely a state responsibility. There is no federal or professional sanctioning attached to non-adoption, limiting the penalties for resisting adoption at the state level.

It is important to bear in mind that flagship institutions are primary policy players and cultural lodestars in both states, and PBF adoption would have worked against their interests in maintaining funding and influence. In effect, adopting PBF as policy designers framed it would tend to disfavor those universities and disrupt the states’ local order. For these dominant institutional actors, efforts to domesticate the national PBF policy movement represented hegemonic threats. Further, negotiations for revising aspects of PBF policy models did not succeed, ultimately dooming prospects for specialized adoption (e.g., adopting PBF only for non-flagship public higher-education sectors, as had been done in at least one other state).

In the end, the two states resisted the convergence occurring in other U.S. states’ tertiary systems. This finding calls attention to Hall and Taylor’s observation (1996, p. 941) that institutionalized arrangements can distribute power unevenly across social groups. In such settings, it makes sense to “assume a world in which institutions give some groups or interests disproportionate access to the decision-making process… some groups lose while others win.” In highlighting the role of key actors in non-adoption, GW make a significant contribution to the literature.

The Rexe, Clarke, and Lavigne chapter (Rexe et al., this volume). In this chapter [henceforth RCL], the authors provide an exemplary review of literature on policy innovation and diffusion, then present an intriguing overview of policy shifts in Canadian provinces over a 25-year period. They note policy convergence across provinces in freezing or limiting tuition increases and in such arenas as institutional foundings, mission shifts, and reliance on intermediary actors (e.g., for quality-assurance processes). The authors also highlight the partisan roots of some policy debates in the provinces, with adoption of new accountability regimes and new programs frequently the targets of right-wing policymakers.Footnote 2

In all, the range of substantive policy shifts assayed across the provinces and across the time period is impressive, and this ambition pays off in a number of valuable observations. Of particular interest here are the distinctions the authors draw between innovations based in across-border emulation, competition, and lesson-drawing, and innovations based in within-polity variables (e.g., postsecondary governing arrangements, socioeconomic conditions, political). For across-border influences, the directionality can be horizontal, as polities learn from, compete with, and emulate same-level polities (e.g., provinces) or vertical, as polities are coerced, demanded, or encouraged to adopt policies from authorities at lower or higher levels (e.g., the federal government).

There are places where definitions are blurred in the chapter. For example, the authors say that “convergence… suggests that common structural arrangements and conditions give rise to similar policy adoptions,” but also say that “a primary factor in policy convergence is emulation, a result of lesson drawing from others experiences and therefore a result of policy learning.” Such wording blurs the meaning of convergence as to within vs. across-polity origins. That said, of course, it is impossible to make a definitive judgment on such questions because adoptions’ roots lie at multiple levels.

This hearkens us back to Gándara and Woolsey’s point that naming and measuring pose significant challenges for analysts of policy innovation and convergence. That point arises again as the authors take us through the varied policy innovations taking place over the period in the provinces. Their ambitions are descriptive and exploratory, and they succeed fully on that front. Their concluding thoughts emphasize the critical need for further considering timing, exogenous factors such as local and national economic and political contexts, the role of federal and intermediary actors and, importantly, the “dosage” of various policy initiatives.

That latter point, raised earlier by Kelchen et al. (2019), seems crucial. For example, is a province or state’s adoption of a performance-funding initiative covering less than 10% of an institution’s budget comparable to one covering 80 or 100%? Can and should such different commitments be simultaneously analyzed as “PBF adoption” in across-polity studies? Or is the polity with less than a 10% commitment far more akin to an polity with no commitment, in conceptual as well as practical terms?

Viewed from the Pope and Meyer (2016) perspective, the RCL chapter is particularly valuable for its sensitivity to the distinctions among policies and the parallel distinctions in their origins. In their concluding comments, RCL ask “What factors or contexts give rise to provincial or regional tendencies, including policy leadership?” This pattern fits the complexity of multi-level diffusion, which distinguishes the origins of specific policy shifts by their connections to other policy stances supported by other organizations or other levels of government.

The chapter’s observation also connects to a conclusion drawn in some early work on policy innovation in higher education. Hearn and Griswold (1994) speculated that, as opposed to purely educational policies, financially re-allocative policies may have roots largely outside of a state’s higher-education governance and policy contexts, suggesting that “One might even speculate that such innovations fall within the domain of populist politics, rather than that of rationalist, professionally driven policy-making (p. 183).”

As RCL note, however, they did not seek to comprehensively address internal province-level factors for adopting popular policies nor did they completely identify horizontal and vertical policy diffusion models or mechanisms. Further analysis could illuminate why and how the convergences noted in the chapter took place.

Additional analysis could also inform analysts regarding connections between the Canadian policy innovations and similar innovations in neighboring and other nations. Striking on that front are the movements in Canada toward (1) mission shifts and degree expansion, parallel to the movement in the U.S. of 2-year colleges toward offering 4-year degrees, and (2) shifts to constrain tuitions, parallel to the movement toward “free community college” in the U.S.

The Dobbins, Martens, Neimann, & Vögtle chapter (Dobbins et al., this volume). This chapter [henceforth DMNV] provides a comprehensive, informative assay of five quality-assurance system dimensions in Germany, France, and Italy: involvement of various actors, institutions, types of quality assurance (internal, external, or peer-reviewed), areas covered, and types of assessment (ex-ante vs. ex-post). Like the GW chapter, the authors argue convincingly for analyzing non-adoptions.

DMNV indicate that quality-assurance processes emerged in higher education systems of Germany, France, and Italy before the Bologna Process started in 1999, but pay particular attention to convergence and divergence from the Bologna framework. While DMNV see the emergence of new quality-assurance practices as part of “lesson drawing” for solving the prevalent higher education problems, Bromley and Meyer (2015) suggest that a value-assessment concept has emerged in every domain with the spread of globalized rational culture. This assessment is done through traditional accounting principles and techniques and through new counting and assessment methods, such as return-on-investment, cost-benefit analysis, financial accounting statements, university rankings, various types of ratings, social rate of return, and impact assessment in teaching and research.

However, the quality-assurance processes were institutionalized and to some extent harmonized across the three countries with the soft laws or law-like rules which Bologna Process has introduced in the form of European Standards and Guidelines-2015 [ESG-2015] (2015) and various Ministerial Communiqué. Although ESG-2015 does not set any quality standard, it does define quality as “fitness for purpose,” (ESG-2015, p. 7). Also, ESG has given ten standards for establishing internal quality-assurance mechanism and seven standards for establishing external quality-assurance mechanisms. As result, common trends amongst quality assurance of Germany, France, and Italy can be seen. For instance, all of these countries have internal and external quality-assurance processes for new program, teaching, and research. Furthermore, the new rationalized culture has empowered various individuals, which supports the participation of students, employers, and other societal actors in quality-assurance processes. Multiple types of quality assessments (ex-ante vs. ex-post) in teaching and research are a manifestation of counting mechanisms based on scientific principles.

Nevertheless, ESG-2015 is very broad, and thus susceptible to various interpretations and implementations. Of the three chapters, DMNV stands out for its close attention to filtering at the national level of continent-wide influences and also for its attention to implementation issues. Regarding the latter point, the management theorists Pfeffer and Sutton (2006) have argued that scholars and managers alike pay far too much attention to decision making and far too little attention to the ways that decisions are implemented. In implementation, the effects of decisions are muted or amplified in ways invisible through a simple focus on the direction and details of a policy decision itself.

Research on performance funding in the U.S. states has been challenged by the fact that some state legislatures and agencies formally adopted or publicized such a policy but never implemented it, or funded it at levels far below specified originally, or delayed implementation for significant periods. Minnesota is a frequently noted example of non-implementation of an approved policy, while Georgia is a frequently cited example of planned implementation that was eventually abandoned. This decision/implementation gap is a significant issue for research, including but also extending beyond the dosage variation noted in earlier chapters. It gets to the heart of what, exactly, is a policy, and what exactly are we studying when we characterize states as adopters in our analyses.

The DMNV chapter concludes that, while some aspects of quality assurance are similar and diffused rather smoothly across national settings, differences remain owing to country-level distinctions and also to variations in implementation approaches. The chapter’s perspectives merge nicely with the perspective of Pope and Meyer (2016) on multiple ways. First, the three nations were each influenced by the Bologna Process, which the authors note “can be defined as an institutionalized structure for the exchange of information among participating countries that is linked to all of the mechanisms of transnational communication,” namely lesson drawing, problem solving, policy emulation, and policy promoting. Further, the Bologna Process in concert with the European Commission have represented top-down influences encouraging compliance at the national level. Clearly, in both the lateral and vertical dimensions, these features highlight the structural-embedding concept of Pope and Meyer.

Similarly, DMNV focus on policy filtering in their attention to historical institutionalism, suggesting path dependence as a factor supporting the emergence of “national idiosyncrasies and deviant developments amid processes of diffusion.” Path dependencies seem especially clear in Germany and France. Germany’s Humboldtian academic self-governance tradition and federalist system have tended to ensure that academics have a significant role in the quality-assurance system (driven mainly by the Länder). In contrast, France followed its centralist state traditions and developed a highly top-driven quality-assurance system.

We see elements of contingent diffusion in DMNV’s attention to policy diffusion. In Germany, for instance, a significant role of academics in the quality-assurance process is a continuation of Humboldtian academic self-governance tradition. Although Länder or the state-level bureaucracy do actively pursue institutional quality assurance, they do so to a lesser extent than the French or Italian national bureaucracies, and they have resisted the national government’s pressures to weaken the traditional German federalist structure. For example, the roles and competence of the federal-level Accreditation Council, established in 1999, have been transferred to the Foundation Accreditation Council, to multiple accreditation agencies, or to the European Quality Assurance Register for Higher Education (EQAR). Further indicators of this resistance lie in the continuing primacy of institutions’ internal quality-assurance processes, heavy reliance on peer-review, and developed ex-ante and ex-post assessment approaches. In sum, accountability still remains largely in the hands of academic professionals.

In France, on the other hand, top-down and bureaucratically controlled quality-assurance processes reflect the nation’s history of state-centered higher-education governance. But the particulars of these processes as well as recent marketization trends in French higher education can increasingly be seen as domestic responses to the prevailing poor performance of French universities in comparative global rankings in 2003 (Dobbins, 2012). Thus, quality-assurance efforts heavily rely on external reviewers and the quality of research is assessed in output or ex-post terms (Dobbins & Knill, 2017, p. 74). In effect, domestic conditions have pressured French higher education to adopt ESG-2015’s definition of higher-education quality as “fitness for purpose,” with those purposes defined by ESG-2015 and the Council of Europe as “preparation for sustainable employment, personal development, preparing students for active citizenship, and creating a broad advanced knowledge base and stimulating research and innovation” (Camilleri et al., 2014, p. 7).

Finally, Italy lies in between Germany and France in terms of its quality-assurance model. Historically, Italian universities have exhibited strong academic oligarchy (Clark, 1983). Although the new state-level quality-assurance agency [ANVUR] is relatively independent from this historically embedded academic oligarchy, the Bologna Process’s quality-assurance efforts in Italy have been slowed by these legacies and shaped contingently by them. As of June 2021, ANVUR had not been registered in the European Quality Assurance Register for Higher Education (EQAR) because of its failure to comply with all ESG-2015 standards (Vinter-Jørgensen et al., 2020). Academic peer review remains a cornerstone of Italian quality assurance and international experts participate in that process, but bibliometric criteria are becoming increasingly important (Dobbins & Knill, 2017; Nosengo, 2013). Italian higher education has moved towards marketization (Dobbins & Knill, 2017), but the influence of ESG-2015 and related cross-border pressures has been mediated by domestication and contingent diffusion.

Additional analysis could help position the Italian change process along the conceptual lines of displacement, layering, drift, or conversion. While Germany and France can both be characterized as examples of incremental change within a rather stable tertiary institutional arena (Paivandi, 2017), albeit in different forms, Italy presents a significant yet incremental centralizing shift away from a system traditional controlled by academics.

In the end, the DMNV chapter highlights the enduring power of loose coupling and institutional autonomy in higher education. A regulatory approach setting strict quality targets from “outside” often will encounter faculty and institutional resistance and hostility, producing “window dressing” rather than compliance (Barnabè & Riccaboni, 2007; Gonzales, 2015; Dougherty & Natow, 2015). At the same time, the chapter illustrates the tensions as a nation’s acceptance of a meta-regulatory approach can contribute to “long-term erosion of traditional forms of authority” in that nation, and movement toward broader conceptions of power in the global rationalizing culture (Bromley & Meyer, 2015, pp. 72–73). In their responses and adaptations, Germany, France, and Italy each “work the margins” between extra-national and national priorities, traditions, and values.

Evidence of Policy Diffusion and Filtering in the Three Chapters

Across the countries, states, and provinces covered in the three chapters, patterns of diffusion (and diffusion-resistance) suggest some conclusions. Governmental leaders in each setting evinced understanding of policy pressures and decisions in other parallel governmental settings and in government settings above or below. Similarly, the preferences of external intermediary NGOs were clear in each case. Pressures toward convergence were felt and in many cases acted upon. In particular, the supranational formal and informal influences of the Bologna Process and other European organizations and networks played notable roles in shaping national policies covered in the DMNV study. Absent similar formalized pressures, the states and provinces studied in the GW and RCL analyses also acknowledged and deliberated pressures emanating outside their borders.

At the same time, context-specific filters were observed in each setting. Assaying parallel developments in other U.S. states, decision makers in the two states in the GW analysis eventually found insufficient reason to fuel convergence toward adopting performance funding like many of their peers (for similar findings, see Rubin & Hearn, 2018). Similarly, Canadian provinces examined in the RCL study only inconsistently adopted policies proliferating in other provinces. And, in the analysis of France and Germany, path dependencies among certain arrangements, preferences, values, and norms circumscribed wholesale adoption of externally favored policies in each nation. In the end, the work of Morphew et al. (2018) is upheld: national, provincial, and professional consensus regarding public and private priorities and levels of authority play critical roles driving institutional and system change. In Northern Europe, the strategic emphasis on pursuing research excellence is embraced across national polities, while in the North American states and provinces, there is less emphasis on agreed-upon national and regional priorities and more effort to preserve deeply institutionalized sub-national authority.

Conclusion

A number of implications emerge from the three prior studies. First, as GW argue explicitly and the other chapter authors suggest somewhat less directly, the power of quantitative methods to discern policy-diffusion influences is limited. Often, the qualitative evidence presented here allows us to see in more depth the workings of polities’ decisions regarding adoption. Too often, quantitative analyses rely on overly broad operationalizations of policies and of local and supralocal influences. Without “dosage” information on a funding policy, for example, it makes little sense to compare adoptions across borders. Without precise indicators of the extent to which certain understandings and organizational arrangements are historically embedded over decades or even centuries, measuring influences on adoption across otherwise seemingly similar settings is compromised.

Second, as each of the chapters emphasizes, non-adoption and partial or selective adoption merit more attention. Numerous tertiary-education studies note that what is encouraged or imposed formally at the macro level can very often take on a somewhat different shape in distinctive contexts. To the extent outside influences are disruptive, universities can cope by seeking to buffer their internal operations from tight external scrutiny and accountability. Often, they will conform performatively at the macro level while resisting at the micro level. Numerous analyses of national and provincial/state policy adoptions highlight this pattern.

Notably, analysis of performance funding in the U.S. by Dougherty and colleagues (Dougherty & Natow, 2015; Dougherty et al., 2016) has highlighted the ways a nationally favored “logic model” of performance funding morphed upon encountering local state conditions, morphed again as state initiatives were imposed on individual institutions, and then morphed yet again as actors within institutions (such as faculty in academic departments) encounter and respond to new expectations for accountability and behavior change. Strained through so many levels, it was little surprise that substantial decoupling occurred along the way, to use Pope and Meyer’s (2016) term.

Third, and relatedly, each of the chapters illustrates the limitations of focusing solely on the adoption/non-adoption decision. As Gándara et al. (2017) have noted, analysts have too often focused on the “why” and the “whether” and not enough on the “how.” As noted earlier, analysts like Jeffrey Pfeffer and colleagues (e.g., Pfeffer & Sutton, 2006) have stressed the significance of the implementation stage. By framing diffusion mainly at the point of a vote or fiat, the literature may constrain its power to inform and guide. At the opposite end of the timeline, long before a legislature or state bureau makes an adoption decision, intermediary organizations and policy “champions” are often hard at work spreading the case for or against adoption. Those actors can not only facilitate policy learning but limit it. And, even prior to the case for a particular policy emerges, the meta-argument for it exerts influence. For example, the Dobbins et al. chapter highlights the fact that prior to the Bologna Process, local fiscal and political pressures and the new public management movement were laying the groundwork in European nations for particular reforms to take root.

Finally, the research in this arena would benefit from greater definitional consensus and consistency. The field’s attention to policy convergence, emulation, diffusion, lesson-learning, and transfer, in particular, seem intuitively understandable on first glance but the distinctions break down upon closer contact. While the chapters here each confront these distinctions smartly, there are some differences that make across-chapter conceptual framing difficult. The problem is even more acute in the broader literature. A question dealt with by the authors in this volume and others involves the roots of convergence: to what extent is growing similarity in policy choices across polities a product of the importation of ideas (diffusion), as opposed to growing similarities in social and economic conditions internally? That is, what is endogenous and what is exogenous in diffusion?