Bridging Rhetoric and Practice: New Perspectives on Barriers to Gendered Change

ABSTRACT

This article presents a new methodology, Gender Knowledge Contestation Analysis, and uses it to examine the processes under way when transformative gender equality policies, such as gender mainstreaming are implemented. Drawing on data gathered in the European Commission, the findings show the processes linking high-level rhetorical policy statements, strategic policies, and daily working practices. This analysis enables exploration of the mechanisms through which indifference to and nonawareness of gendered policy problems are collectively constituted and methods through which they can be challenged. Findings thus deepen our understanding of barriers to the implementation of gender mainstreaming and the steps required for its effective implementation.

KEYWORDS:

• Gender mainstreaming
• gender knowledge
• contestation
• European Commission
• resistance

Introduction

The implementation outcomes of gender mainstreaming (GM) present feminist scholars with a puzzle. Despite widespread high-level commitment to GM in over 100 states, the policy has not yielded the transformative results envisaged by its advocates. In fact, empirical analyses of GM implementation have found that strategic commitments to gender equality are usually accompanied by patchy implementation or dilution of the policy, so that significant reductions in gender inequality after its implementation have thus far proved elusive (Daly 2005). However, although this mismatch between GM rhetoric and actual implementation has been extensively documented in many locations, our understanding of how these outcomes come about remains limited (Krook and Mackay 2011).

This limited understanding has stimulated arguments that the methodological approaches thus far used to examine GM implementation are too blunt to capture gendered change and may be obscuring the actual processes underway when the GM policy is adopted (Benschop and Verloo 2006, 31; Meier 2006, 185). Building on these premises, this article uses the puzzles presented by GM implementation problems to highlight some of the conceptual limitations in our existing approaches to gendered policy analysis. These include difficulties extending the insights of discursive analysis into implementation and our underdeveloped comprehension of the links between rhetorical policy and actual practice.

First and foremost this article argues that these problems stem from the difficulties of fully operationalizing our existing perspectives on the social construction of gender in the state as a process (Connell 2006). To meet these methodological challenges, this article outlines a new approach: Gender Knowledge Contestation Analysis. This approach draws on two closely related literatures: Interpretive Policy Analysis (Colebatch 2009; Yanow 2000) and the Sociology of Knowledge (Callon and Latour 1981; Latour 2005). These literatures eschew stark analytical distinctions between written policy and implementation, and together they provide a menu of conceptual tools to trace the mechanisms through which particular ways of thinking about and acting upon policy problems are collectively established and maintained. The concept of knowledge is key within these literatures, because it provides a way to extend analysis of issue construction beyond written policy or rhetoric, into practice while also capturing how that practice is constituted (Callon and Latour 1981; Wagenaar 2004; Yanow 2000). This empirical approach thus elaborates a perspective that bridges between so-called rhetorical policy and implementation. Furthermore, by using our analysis of these policy processes to focus specifically on gender knowledge within them, we can fully explore how durable gendered assumptions are enmeshed in local understandings of “mainstream” issues and local practices. As such, the article seeks to show an approach that could be fruitfully applied in other contexts to analyze gendered policy and policymaking processes and to provide useful insights to improve GM implementation.

This Gender Knowledge Contestation Analysis is applied to examine GM implementation in the European Commission. As part of the European Union (EU), the European Commission is an important case study for the analysis of GM and of gendered change because it has in many ways acted as a pioneer in commitment to gender equality, boasting a uniquely comprehensive suite of policy commitments to gender equality.

The European Commission’s complex institutional structure also yields unique opportunities to examine and compare the kind of patchy or disappointing outcomes widely observed in GM implementation in other organizations and states. The Directorate General for Research (DG Research) is one such ideal site. The equivalent of a ministry or department, DG Research is responsible for overseeing and implementing EU science policy and during the policy period from 2002 to 2006, and it had one of the most developed suites of GM policy instruments and procedures in the European Commission (Lombardo and Mergaert 2013). In addition, it is structured into subunits called Directorates, which are each subject to the same GM obligations and should maintain the same implementation procedures. In practice, however, GM implementation varies between Directorates, thus providing a fruitful site for analysis of such variations.

Data presented in this article draw on a larger study, undertaken for a doctoral thesis, encompassing implementation in a Directorate widely regarded to have excelled in GM implementation and one where implementation was more limited. Data collection was undertaken between 2008 and 2009 and comprised analysis of over 60 documents and 30 interviews with members of the policy community and frontline staff, including unit heads and operational staff known as Scientific Project Officers (SPOs). In this article, data concerning the less successful case are presented.¹

The article is structured as follows. First, several important characteristics of the GM policy are presented along with a brief explanation of existing findings and methodological approaches. Based on the puzzles that these findings present, the most important tenants of a Gender Knowledge Contestation Analysis are introduced before presenting empirical data from DG Research gathered using it.

The analysis presented illustrates three broad findings. First, it depicts the processes that link rhetorical policy commitments to implementation practices, translating abstract commitments to action into workable prescriptions for activity. Analysis shows how these processes constituted an institutionalized non-engagement with gendered policy issues in DG Research before the adoption of GM. Second, this analysis shows how GM implementation required significant conceptual elaboration, translating abstract commitments to GM into clear prescriptions for activity. Third, data from a Directorate (Directorate D²) widely regarded to have resisted GM are presented, enabling examination of the barriers to GM implementation. These findings thus present an analysis that links rhetoric and implementation, sharpening our understanding of the processes maintaining indifference to and non-engagement with gendered policy issues and the mechanisms through which they can be challenged.

Theorizing gender mainstreaming

Formal definitions of the GM policy vary but within the EU and the European area more widely, the most influential and frequently cited definition of the policy is that of the Council of Europe’s Group of Specialists on Gender Mainstreaming.

Gender mainstreaming is the (re)organisation, improvement, development and evaluation of policy processes, so that a gender equality perspective is incorporated in all policies at all levels and at all stages, by the actors normally involved in policy-making. (Council of Europe 1998, 15)

This policy and the assumptions contained within it, have a very specific intellectual heritage, drawing closely on feminists’ theorizations of the state’s role in the production of gender inequality. Premised on an understanding of gender as a set of relational and hierarchical meanings (Hawkesworth 1997, 650; Scott 1986, 1,054) feminist scholars working in organizational sociology, political science, and policy analysis have developed sophisticated understandings of the state’s³ role in the production and maintenance of gender and gender inequality.

Feminists’ examinations of policy content, for example, have revealed andro-centric assumptions structuring interpretations, values, conventions, and practices in multiple policy fields extending well beyond the most “obviously” gendered policy fields. These analyses have illustrated how bias toward male interests and assumptions regarding gender differences between men and women can be found in all public policies so that their impacts may reinforce male advantages (Hawkesworth 1994). Similarly, analyses have revealed a widespread tendency to overlook the specific interests and needs of women in policy (Hawkesworth 1994), highlighting how local practices and discourses de-politicize women’s inequality through management speak (Acker 2006, 452; Connell 2006), or how women’s lives and political interests are often underconceptualized, oversimplified by bureaucrats, or simply unknown (Lovenduski 1998, 340).

These perspectives thus reject understandings of gender inequality as “natural,” instead highlighting the construction of gender and the attribution of gendered meanings (gendering) through state policy, highlighting the importance of practices and processes constituting assumptions within organizations (Acker 2006; Connell 2006). By asking policy practitioners to take gender into account in all their activities, the GM policy seeks in theory to make policymakers identify and displace these practices and assumptions (Squires 2005).

Linking rhetorical policy and practice

By reviewing these intellectual underpinnings of GM we can see that an awareness of meaning or issue construction forms an essential component of feminist perspectives on the states’ creation and maintenance of gender inequality. These literatures have shown gendered meanings and assumptions embedded in the policies that states and bureaucracies implement and in the related practices and discourses of organizations. However, when we examine existing research on GM we can observe a predominance of analysis focusing on written policy content, with less emphasis on the local processes constructing gender, which existing literature has theorized.

Thus, much existing research on GM has analyzed and compared strategic or rhetorical policy content (e.g., EQUAPOL Daly 2005) using, for example, frame analysis (e.g., Hafner-Burton and Pollack 2000; MAGEEQ Verloo 2007), or has assessed how faithfully it has been implemented. Findings usually show local reinterpretations of GM that dilute the structural diagnosis of gender inequality originally central to it, rendering the policy less transformative (Daly 2005; Verloo 2007) and more easily combinable with the existing local practices and assumptions, the very ones that feminist analysis has highlighted as problematic (Benschop and Verloo 2006; Caglar 2010).

The smaller number of studies that have looked more closely at implementation, alternatively, have delved a little deeper into some of the dynamics that might account for this tendency toward “dilution,” often uncovering problems of unintelligibility. This often takes the form of incomprehension and uncertainty over exactly what GM would entail or assertions of its local irrelevance. Here, studies document staff’s assertions that gender inequality is a problem elsewhere (Andresen and Doelling 2005; Benschop and Verloo 2006; Connell 2006; Lombardo and Mergaert 2013) or find interviewees openly expressing the view that women’s policy needs are unknown (Schmidt 2005). How these assertions constitute a tacitly accepted policy position is of significant interest if we wish to unpick the processes hobbling the effective implementation of rhetorical commitments to GM. Why can’t persons charged with implementing policy, after its rhetorical formation, work out and subsequently act upon women’s interests? Why are they so often convinced gender is irrelevant here?

Although these findings suggest that reinterpretations of the gender equality issue and the GM policy align with ideas previously in place within an organization, analysis of the connections between rhetorical policy content, implementation outputs, and preexisting practices and processes constructing gender, are underexamined. In fact, comparative feminist analysis of differing content in written GM policies and documentation of implementation difficulties seems to be premised on an implicit aspiration that policy can be moved fairly smoothly between locations and from paper into practice, without significant changes in content.

This perspective however limits our findings to pronouncements of failure and it does not operationalize our awareness of the complex processes through which gendered assumptions and meanings are locally constructed and negotiated in organizations. Rather, it detaches analysis of gendered meanings or interpretations in policy from the very local processes and structures shown in existing feminist research to constitute gender in organizations.

A gender knowledge contestation analysis: Analyzing meaning and construction in action

Interpretative Policy Analysis and the closely related field of the Sociology of Knowledge, by contrast, argue first that all policy, from rhetorical to the implementation stages, must be conceptualized in terms of fluid meanings, contestation, and interpretation, focusing our attention on the ongoing struggles over meaning and problem diagnosis which policy thus inherently involves, at all stages (Callon and Latour 1981; Colebatch 2009; Schoen 1973; Yanow 2000). Second, these literatures assume that policy cannot be analyzed as a discreet phenomenon, arguing instead that organizations or bureaucracies implementing policy will have their own preexisting structures, practices, and competing policies. Policy dictates will, therefore, always be understood and acted upon through the filters of the preexisting organizational practices and policies. Where policies are particularly innovative, these processes of policy implementation can become “a battle for broad and complex transformation” (Schoen 1973, 101).

Reformulating our perspective on policy implementation in accordance to these literatures means we must understand policy not simply as stated (written) intended action, but in terms of (1) actual implementation activity and (2) the collective processes through which interpretation occurs and through which actors compete to control and influence it (Callon and Latour 1981; Callon, Law, and Rip 1986; Colebatch 2009, 15; Latour 2005). To this end, these literatures in combination provide a theoretical perspective that we can operationalize to capture the processes through which local interpretations of policy problems are institutionalized and implementation is controlled. By fusing them with gender theory to undertake a Gender Knowledge Contestation Analysis, we can focus on the gendered assumptions that GM, in theory, seeks to displace and the new ones that it institutionalizes.

Analyzing policy in terms of knowledge implies two steps. First, it entails interrogating organizations’ activities in terms of how various participants and audiences understand and implement a policy. Instead of examining policy as though it is text, this approach entails asking “frontline” staff to describe the activities actually undertaken when implementing a policy and their rationales. This reveals the myriad constraints acting upon frontline staff when they work—the committees of persons, preexisting aims, forms, categories, and resource constraints, within which they (must) work. Often, this shows how frontline staff’s understandings of policy are much more limited and less idealistic than the abstract commitments made in rhetorical policy (Wagenaar 2004).

Second, these literatures have emphasized the collective nature of the processes governing the emergence of knowledge, showing how the institutionalization of practices and processes (e.g., rules, forms, and categories) systematically steers action into the service of particular interpretations of a policy problem, while systematically marginalizing others (Callon, Law, and Rip 1986). These processes serve to reduce confusion and contestation within an organization, replacing it instead by shared, assumed aims: “a set of practices is placed in a hierarchy in such a way that some issues become stable and need no longer be considered … have become a matter of indifference” (Callon and Latour 1981, 285 my emphasis). This provides a perspective on the processes through which ways of viewing problems are institutionalized.

A Gender Knowledge Contestation Analysis thus entails asking staff to describe their work and their understanding of policy (their knowledge), while also examining policy processes to analyze how rhetorical commitments to GM are filtered into action. To focus the analysis specifically on how gender is understood and perceived (or not) in these processes, this article uses the analytical concept of gender knowledge (Andresen and Doelling 2005; Caglar 2010; Cavaghan 2013) “explicit and implicit representations concerning the differences between the sexes and the relations between them, the origins and normative significance of these, the rationale and evidence underpinning them and their material form” (Cavaghan 2013, 72). The concept of gender knowledge thus provides an analytical framework to examine the policy processes spanning written documents and into practice, focusing our attention onto explicit and also implicit gendered assumptions, (non-) perceptions of gender and very importantly, the rationales undergirding them and the processes through which they are held in place.⁴

Data and analysis

The following sections deploy this approach, presenting data that show how frontline staff’s comprehension of the local policy agenda and their rationale for action are composed of and embedded in the practices they undertake when implementing it. The link between these practices and higher-level written policy is then explored, illustrating the priorities and restrictions that are recursively institutionalized in DG Research and how they are insulated from reinterpretation.

DG research policy process: Defining and shaping practice

Scientific Project Officers are DG Research’s frontline staff. They work in thematic Units, developing specialist expertise in their Unit’s (inter-) disciplinary field. Fieldwork interviews with Scientific Project Officers on the open topic of “science and research policy” or “DG Research’s work” quickly became a discussion of The Framework and of projects.

Interviewer: “Can we just start with you telling me broadly about what you think the priorities of science and research in the EU are just now?”

Interviewee: “Well, I can tell you about the Framework Programme, about the projects.”

Interviews showed that these phenomena, the Framework and projects, structured most of the Scientific Project Officer’s daily work, which comprised administrating projects and contributing to the development of future policy. This involved writing calls to invite applicants to apply for funds; screening applications; evaluating applications; negotiating projects; reporting on project progress and impact, according to set criteria; and finally feeding findings and experience back into the development of future and present policy. These kinds of activities thus constituted staff’s descriptions and practical experience of what “science and research policy” within DG Research is.

Further discussions of practice and of the policy process revealed how these daily processes are predicated on predefined notions on the use of science and research and the appropriateness of various policy interventions. Within DG Research these are established in lengthy policy processes where policy is recursively renegotiated.

Defining the mission statement and policy tools

The most overarching aims for scientific research and DG Research’s legitimating basis, its Mission Statement, are drawn from the Treaties (Rome, Single European Act, Maastricht, Lisbon) that have successively stated the shape and aims of the EU.⁵

Thus, DG Research’s Mission Statement dates back to the very early stages of the EU’s development, encompassing the 1957 Treaties of Rome and the 1974 Council Resolution on the coordination of national policies and the definition of projects of interest to the community in the fields of science and technology (European Parliament, undated). These documents define DG Research’s role and its policy tools, restricting it to the coordination of research or interventions where the EU can add value as a facilitator, for example, by encouraging harmonization of methods or qualifications and coordination of research outputs. These Treaties also define DG Research’s main policy instrument, a multiannual funding package with a four- to five-year duration, the Framework Programme (FP).

The FP is thus a pivotal document in defining shared understandings of activities and aims within DG Research. Each updated version of the FP also reiterates paragraph 163(130f) from the Single European Act, which established that DG Research’s activities should “strengthen the scientific and technological bases of industry and encourage it to become more competitive at the international level” (my emphasis). These articles also establish four activities of the FP: supporting research; promoting cooperation; disseminating and optimizing results; and stimulating training and mobility.

This legitimizing basis, fundamental aims, and restricted menu of policy tools were negotiated by Member State government representatives in the Council during Treaty reforms, prior to the rest of the policy process in DG Research. As such, these exert a significant constricting effect on the latter stages of policy interpretation.

Defining strategy: Recent variations

In contrast to DG Research’s Mission Statement, the FP is revised and renegotiated every four to five years. Its content establishes the strategy through which the preestablished Mission Statement will be achieved, according to the present scientific and economic situation. It allocates funds to broad research topics and defines the current scientific and technological objectives and priorities for the EU.

In accordance with the European Commission’s power to draft policy before passing it to the Parliament and the Council, this negotiation is largely undertaken by SPOs and Unit Heads in large-scale consultation with stakeholders. Thus, while DG Research’s Mission Statement, “competitiveness,” is prestabilized through reference to EU Treaties, its thematic research priorities and interpretations of the appropriate route to competitiveness are allowed to vary over time.

During FP6 “coordination” and “added value in research” were articulated through the notion of the European Research Area (ERA), a policy goal established in the EU’s wider economic strategy in the Treaty of Lisbon (2000) in response to “structural weaknesses” in Europe’s science and research policy. To this end, activities “to effectively use all human resources in the population as effectively as possible” were included in FP6. This was accompanied by a recognition that commitment to competitiveness entails a need to tackle “social aspects,” noting “the EU is now facing … more problems significantly affecting the economy, society and citizens for which science holds the key to a large extend,” thus a rhetoric of science serving society formed part of DG Research strategy during FP6. On this basis, FP6 established attention to the social implications of research, effective science governance, justice, and sustainable economic and social development as horizontal themes across all topics of research, linking it to the Mission Statement of competitiveness. The negotiation of the FP thus pertains only to the elaboration of current strategy in the service of preestablished notions of DG Research’s raison d’être—competitiveness—within a limited menu of policy tools.

Defining implementation

Once the content of the FP is confirmed, Scientific Project Officers and the Heads of each Unit negotiate and draft a separate a local Annual Work Programme, in conjunction with their specific research public. This translates the strategy and thematic priorities of the FP into specific individual proposals for research projects, which SPOs then administer through heavily structured and tightly documented processes (detailed earlier), undertaking the projects stipulated in the Annual Work Programme, thus fulfilling the strategy defined in the FP and the Mission Statement referred to in the FP, which is drawn from EU Treaties.

Theorizing the jump from rhetoric to implementation

The above findings gathered by using this analysis thus provide a perspective on the links between rhetorical written policy and frontline practice, which we can then use to identify barriers to the implementation of rhetorical commitments to GM during the FP6 period. Several points can be drawn out and emphasized here. This interrogation of DG Research’s policy process shows that rhetorical or strategic statements of aims are unworkable on their own; they require elaboration into workable prescriptions for action and practice.

Reviewing the stages through which this is achieved in this case reveals an ongoing process of updating and reinterpretation, which is structured and restricted. Each stage of policy interpretation or formation takes place within the confines of the stage previous to it: strategy is updated every four to five years and actual projects are formulated annually, to in turn be implemented using forms and processes designed to implement projects according to present strategic aims. Through the recursive establishment of this agenda, many competing notions of the role of science and the route through which it could contribute to society are obscured, minimized, and ultimately marginalized: a shared assumed set of aims is established, while others have been rendered a matter of indifference. The resulting implementation processes are also in turn designed to harness Scientific Project Officers’ activity into the service of DG Research’s competitiveness Mission Statement. As interviews with SPOs showed, these processes do indeed comprise how SPOs describe and understand their work.

Institutionalized non-engagement with gender

Interrogating the content of these policy processes using the gender knowledge concept reveals an absence of any explicit awareness of gender inequality issues or gendered policy problems at all stages. Only one sentence in FP6’s predecessor, FP5, mentioned gender equality, stating that DG Research should: “encourage the participation of women in the field of research and technological development” (European Parliament and the European Council 1998). The rest of FP5, the associated Annual Work Programmes, and implementation documents associated with it, did not systematically elaborate what such action could entail. Thus, although the notion of competitiveness is embedded in DG Research’s treaty base and has been elaborated into workable activities and practices through successive FPs and Work Programmes, this rhetorical commitment to GM in FP5 stood alone within the FP and was wholly underconceptualized in the other documents making up DG Research’s body of policy.

This institutionalized nonawareness of and non-engagement with gendered aspects of policy problems conforms to wider patterns observed in the existing gender and public policy literature discussed in the initial sections of this piece, and it is the starting point from which efforts to mainstream gender began. Interviews with staff in the Gender Unit⁶ repeatedly described an institutionalized nonawareness of gender inequality both within DG Research’s working environment and the scientific community it served prior to GM implementation. Subsequent policy documents seeking to kick-start GM in DG Research succinctly described this institutionalized nonperception of gender inequality as a policy of “no data, no problem, no policy” (European Commission, Directorate General for Research 2009,7). During the FP6 period, however, increased political commitment to gender equality among personnel in the Council, the European Parliament, and the Commission produced a much more concerted effort to develop and implement GM in a meaningful manner.⁷

Identifying barriers to gender mainstreaming implementation

We can further reformulate the insights provided by this analytical perspective to explicitly identify barriers to the implementation of rhetorical commitments to GM. First, it shows that gendered aspects or impacts of scientific research or associated societal problems formed part of the matters “relegated to indifference” through DG Research’s policy process prior to GM implementation. By corollary then, staff working within this policy remit and understanding EU science and research in terms of local implementation practices are likely to experience significant difficulty understanding how gender could be relevant here, precisely because their actions and the policy agenda are based on a clear remit. Contribution to EU competitiveness is the fundamental legitimation for DG Research’s activities, and prior to FP6 gender equality formed no part of this collectively conceived raison d’être. In abstract terms these two phenomena, institutionalized indifference to and nonawareness of gender inequality and the dominance of a clear and fully elaborated policy process, which renders gender irrelevant in favor of the pursuit of competitiveness, represent what GM implementation would have to overcome.

In practical terms, however, this analysis shows the large-scale and complex policy processes constituting this collective indifference, and it highlights how rhetorical policy commitments must be translated into clearly stipulated processes. In this location, such a process of translation is likely to require the insertion of elaborations about what GM would entail within each stage of DG Research’s suite of policy documents: Mission Statement/EU Treaties, strategy/the FP, and implementation documents.

Institutionalizing gender mainstreaming

Elaborating a gendered perspective

Interviews with the lead member of personnel tasked with setting up a Gender Unit in DG Research show her describing an initial struggle to work out what to do under the guise of “gender mainstreaming.” “My past was more gender blind … I think almost nobody knew about the gender issue.”⁸ The suite of actions eventually undertaken, however, elaborated an understanding of gender in science and research, which spanned all stages of the policy process in DG Research detailed earlier.

The 1999 Commission Communication “Women and Science” (European Parliament 2000) was a key document in this process. Largely sidestepping abstract, normative justifications for gender equality interventions, it instead legitimizes GM by pointing out the heritage of gender equality commitments already present in the EU’s Treaties, mimicking the establishment of DG Research’s Mission Statement. It points out commitments to gender equality enshrined in: the 1957 Rome Treaty, Articles 2 and 3, 13, and 141 of the Treaty of Amsterdam, the 1996 Commission Communication on Mainstreaming, the Luxembourg Employment Strategy, the Structural Funds, and DG Research’s FP5. This Communication was supported in Parliament (European Parliament 2000) and backed with a Council Resolution (Council of the European Union 1999).

Drawing on external evidence, this Communication notes high numbers of women entering the sciences but argues they are underrepresented at the higher echelons of research because of institutionalized nepotism and sexism. Describing this underrepresentation as a waste of talent, the Communication links this to preidentified structural weaknesses in the European Research Area, a key element of DG Research’s policy strategy at the time. The communication then elaborates a GM strategy with three themes: the need for “science by, for and on women” (original emphasis). The Communication also emphasized the absence of knowledge and data concerning women’s participation in science and its impact upon them, arguing this ignorance hampers the EU’s ability to effectively grasp, or legitimately dismiss, the issue. Remedying this information deficit was therefore argued as one of the most important aspects of new GM activity.

In addition, the Communication commits to quotas on all decision-making committees, mainstreaming gender into research when devising the FP, and funding research on women’s experience. These commitments were in turn underpinned by the official establishment of the Gender Unit and two Women and Science Working Groups: one charged with involvement at the level of the FP (the Helsinki Group) and another containing a “gender contact point” in each Directorate of DG Research. This latter group was construed as a vehicle for collective learning about gender in science and to support the local contact point in elaborating a perspective on the local relevance of gender within their Directorate. Finally, the communication allocated responsibilities to SPOs to collect gender impact data and implement Gender Action Plans on all their projects.⁹

This Communication¹⁰ thus constituted a comprehensive elaboration of a Gender Equality Mission Statement pertaining to the whole of the EU and the Commission. Through reference to the ERA (which itself is premised on reference to the Lisbon Strategy) it linked this to current strategic policy within DG Research and the EU. It also detailed mechanisms to institutionalize GM in the creation of the DG Research’s strategic policy in the FP, specifying personnel responsible for overseeing implementation and instituting prescriptive implementation and impact assessment procedures for all SPOs. These impact assessment procedures subsequently ensured that data on the participation of women and the gendered impact of science upon them created a route to churn information back into the policymaking process.

Gender knowledge and gender mainstreaming implementation in Directorate D

We now analyze how these efforts to institutionalize GM faired in actual implementation. Fieldwork interviews conducted with staff in Directorate D, a Directorate widely regarded to have resisted GM implementation, showed a continued ignorance of gender issues in science, incomprehension of gender’s relevance “here,” and nonawareness of GM tools. When questioned, personnel charged with implementing GM expressed uncertainty and anxiety about the policy, arguing their Directorate’s unsuitability for GM, or that the policy was just too hard.

“You have no idea how difficult it is.”¹¹

“It’s not so easy because [our] field [it] is a very masculine world … because of the subject, because you have a lot of engineers, a lot of [names discipline], but always men a lot of men, because there are not so many women again, in the field.”¹²

“Hard issues, gender [shakes head].”¹³

These perspectives thus mirror the puzzlement that the staff member charged with setting up the Gender Unit described when first tackling the policy, and wider findings in existing literature showing mainstream staff expressing difficulty understanding GM or arguing the policy is not relevant “here.”

Questioned on the details of women’s participation in the Directorate, interviewees could also not supply, locate, or signpost to, the relevant information. Instead, staff discussed the issue on the basis of personal anecdote or opinion.

“I think maybe it’s 60% men to 40% women in technical projects.”¹⁴

On producing the DG Research’s own monitoring information (European Commission, Directorate General for Research 2008),¹⁵ which showed exceptionally low levels of women’s participation in the Directorate’s committees, the figures were refuted by key staff on the basis that Directorate had not been properly consulted¹⁶ and that better evidence would be required.

“Well, difficult, if you ask me does this exist? I say yes but … I’d have to see firsthand evidence of this taking place.”¹⁷

The members of staff charged with implementing the GM gave also unclear and factually incorrect descriptions of GM policy in their Directorate.

“I think we have to favor women if there’s nothing else differentiating?”¹⁸

Others described rather odd activities, which bore little resemblance to centralized DG Research GM policy. One staff member described the production of communication materials that were more feminine, as part of GM activities.

“We succeeded to have this kind of brochure and it’s nice, it’s nice … it’s completely different, its more feminine in the end if you want to call like this.”¹⁹

Another described gender mainstreaming as a state of mind and also as a developed apparatus, alluding once more to anxieties about the policy, this time in terms of fairness.

“Gender here is understood as a state of mind, we think of it every day, every day because we are told about it all the time!”… “We have rules and guidelines to maximize the presence of women which we do, you know, without being stupid or penalizing men in programs, projects, evaluation panels and expert [committees].”²⁰

Another frankly stated that staff did not know how to deal with the obligatory gender action plans:

“People didn’t know how to react to gender in Gender Action Plans.”²¹

Staff were also not sure of the structures established to mainstream gender (the Gender Unit or the Women and Science Group), with some of the statements revealing a conflation of DG Research GM policy and wider equal opportunities and gender awareness training.

“I think its purpose is to, to spread gender sensitivity in the European Commission.”²²

“When you work for the Commission, even if you are a contractor, you have something on the gender issue.”²³

Protecting preexisting practice

Delving into the local policy process in Directorate D reveals mechanisms through which this ignorance of GM policy or gender issues in the local discipline has been constituted. Statements that gender is irrelevant here were undergirded by the articulation of a special status arguing that this Directorate was in fact exempted from the consideration of broader cross cutting issues inserted in FP6, on the basis of the Treaties.

“This spend must be the responsibility of the EC Treaty, education under DG EAC (DG Education and Culture). This is where these issues need to be dealt with … As to how much we can play a pro-active role, I think depends on the conditions we have in our legal basis.”²⁴

Examining the texts of the Treaty, the FP, and the Work Programme, pertaining to this Directorate however, reveals a less clear-cut exclusion of broader issues. In common with other areas, an explanatory memorandum on the ERA detailing crosscutting aims, such as social issues and gender equality, preceded the FP dealing with this discipline. The Lisbon Agenda and the Commission Communication on Women in Science are also explicitly referenced in the opening text of the FP, and socioeconomic issues are also stated as a part of the Work Programme’s aims:

A particular effort will be carried out to take into consideration ethical, social, legal and wider cultural aspects of the research including socio-economic research and innovation resulting from the possible deployment, use and effects of the newly developed technologies or processes and scenarios covered by the three thematic priorities and the other activities.²⁵

These commitments are mirrored in implementation documents. The Directorate’s 2003 Work Plan, for example, states:

Reinforcement of the role of women in science in research both from the perspective of equal opportunities and gender relevance of the topics covered will be attempted where possible.²⁶

Yet these commitments have not been translated into action—staff interviewed argued gender was irrelevant, could not describe GM activities, and clearly did not understand the policy.

Prima face, these results may appear to be the product of a passive response. The methodological perspective deployed in this analysis, however, insists that we ask how outcomes have been constituted in policy processes. A close examination of policy processes in this Directorate indeed shows how persons and documents have been deployed to stop the influx of alternative gender knowledge and thus to maintain local non-engagement with and ignorance of gendered policy issues and prescribed GM actions.

First, the allocation of personnel chosen by the Directorate to oversee GM in the Directorate, the “gender contact point,” boded poorly for the possibility for active GM implementation. Interviews revealed that this individual was not an SPO, did not have a permanent contract, and that their job role did not include the project administration process or any knowledge of the local scientific discipline. Thus, this individual lacked any participation in or knowledge of local policy or implementation practice, making any kind of participation in them, let alone the promotion of a challenging policy like GM, extremely difficult. These actions therefore effectively hobbled an important route for the flow of information between the Gender Unit and Directorate D and prevented the penetration of alternative gender knowledge and the development of a local elaboration of GM, which might displace the previously institutionalized “no data, no problem, no policy” perspective. This staff allocation was the direct result of management decisions within the Directorate.

Local instructions contained in implementation documents also provided an explanation of GM, based on preexisting gender knowledge, which assumed its irrelevance and which actively contained and minimized it. Analyzing the local Work Programme in conjunction with the guidance compiled by Directorate D, for evaluators²⁷ reveals how instructions explicitly reinterpreted the clear commitments made to GM in the FP and the Work Programme out of the actual project administration process again.

In addition to the paragraph articulating commitments to consider socioeconomic, legal, or ethical issues/impacts, and gender, the Work Programme in this Directorate also later states, in bold, that the specific, as opposed to any horizontal, objectives of the programme (i.e., technical activities specific to this directorate’s discipline and directly contributing to competitiveness) will take precedence over all others in the evaluation of projects.

Guidance for project evaluators also lists eight criteria as “principal issues of relevance” along with instructions on how to score project applications in relation to them. These eight criteria do not include socioeconomic, legal, or ethical issues/impacts, and gender. Evaluators were also not supplied any guidance on what gender aspects could or should be incorporated into project plans and were specifically instructed to place technological discipline specific aims, which are described as “the sine qua non” of the Work Programme and the FP above all “additional points,” such as gender or other social or impact issues.

The tightly defined documents structuring discipline-specific working processes within this Directorate thus explicitly minimized impact, social issues, and gender by actively excluding them from consideration or reward in project administration processes; project applicants following these instructions would be unlikely to include any impact, social issues, or gender in their applications. Evaluators were also instructed in such a way that they were highly unlikely to ascribe positive significance to gender issues in the processes of evaluating and scoring projects while specific (technical) aspects of research are given attention and attributed with great significance and reward. In fact, statistics show projects including gender may even have disadvantaged projects and that Scientific Project Officers were not consistent in ensuring that Gender Action Plans were completed or in compiling their findings (European Commission, Directorate General for Research 2008).

Thus, the choice of allocation for the local “gender contact point,” the reassertion of the primary importance of technical activities contributing to competitiveness in the local Work Programme and the de-prioritization of gender and socioeconomic dimensions in project evaluation processes maintained the marginalization of alternative gender knowledge in this Directorate and actively reasserted preexisting practices. GM thus remained unintelligible, viewed as difficult and/or irrelevant. Women’s policy needs remained construed as a matter of legitimate indifference.

Discussion

The results garnered by this Gender Knowledge Contestation Analysis shed light on the complexity of policy processes and the messy, collectively constituted relationship between new, written, rhetorical policy commitments and change “on the ground.” DG Research’s relatively formalized policy process nicely illustrates the necessity to translate abstract ideals (e.g., “competitiveness,” “gender equality”) into a clear menu of workable practices. In this case several clearly distinguishable levels of policy elaboration were necessary to make a workable GM policy: a locally convincing Mission Statement providing legitimation for the policy; a strategy explaining what GM means in broad terms, for example, “science by, for and on women”; and precisely formulated policy implementation processes and practices.

A Gender Knowledge Contestation Analysis also sheds light on why GM can be so difficult to establish. It illustrates how large-scale and long the collective processes establish local working practices and local knowledge. This reveals the difficulties of enabling alternative knowledge to enter the policy process and compete.

As an organization, DG Research boasts 50 years of experience and strategic and practical knowledge concerning effective ways to act in the service of the abstract notion of competitiveness. The establishment of this agenda is achieved through recursive marginalization of competing agendas, not engaging with competing perspectives and thus not developing the knowledge or the practices to act on them. This in turn sheds light on well-documented tendency of mainstream staff in many organizations to puzzle over what GM might possibly entail and/or to argue that gender is irrelevant here. We can logically expect alternative (gender) knowledge bases within an organization to be next to nonexistent when new circumstances require them.

Prior to GM implementation in 2002, policy processes in DG Research, as in many other organizations, had long institutionalized a non-engagement at all stages in the policy process with any possibility that issues of gender inequality might exist in science, or that these might be relevant to EU science policy. No data on women’s participation or gendered impacts existed, thus no problem was visible and no commensurate policy existed. The gender knowledge that was locally institutionalized here then was implicit, underdeveloped, and predicated on little substantive information or thought. The characteristics and content of the gender knowledge underpinning GM are quite different, premised on a theoretical understanding of gender as a socially constructed phenomena and a conscious problematization of gender inequality. These perspectives have been developed outside DG Research and its local knowledge processes.

Effectively displacing the comprehensive non-engagement with gendered policy problems, observed in DG Research and in wider literature, thus requires the creation of policy processes that disrupt the marginalization of alternative gender knowledge and facilitate completion between it and locally preexisting, likely implicit and underconceptualized, gender knowledge. Multiple institutionalized points in the policy process that enable the incorporation alternative gender knowledge and its elaboration into local practice are therefore required for successful GM implementation.

Even with these in place, however, GM implementation remains vulnerable (1) to the institutionalized local incomprehension of the GM agenda it seeks to displace and (2) to loyalties to preexisting practice and assumptions of what constitutes legitimate and credible activity. Results in Directorate D showed that the marginalization of alternative gender knowledge leaves GM open to interpretation based on preexisting practices and that maintaining preestablished practice requires much less comprehensive access to policymaking processes than the assertion of new knowledge and practices does.

GM policies must therefore be fully elaborated and underpinned by institutionalized learning and policy development processes. Sustained over time, these processes hold the potential to institutionalize collective awareness of gendered policy problems, displacing and challenging the notion that gender is “not relevant here” with an ongoing process of learning about and engaging with the latest gendered policy problems.

Conclusions

This article began arguing that to really understand how GM can be successfully implemented a method is required that can operationalize the construction of gender as a process, as theorized in existing literature. Reviewing GM implementation problems highlighted in existing studies, it suggested a specific need to focus on the processes constituting: conceptual drift or dilution of transformative content in GM policies; indifference to or ignorance of women’s political needs and interests; and incomprehension of the relevance of gender “here.” To understand these phenomena, I argued for a method that examines the interrelationships between rhetorical policy, policy as implemented, and (preexisting) practices. Thus, I elaborated a Gender Knowledge Contestation Analysis that enables us to examine processes constituting collectively held assumptions regarding gender and its ir/relevance. The concept of gender knowledge is key here because it enables us to capture and compare gendered meanings not only on paper (as frame analysis might) but also in practice, while also examining constitutive interrelationships between the two. This methodological perspective thus bridges the rhetoric-practice gap that existing research on GM implementation has yet to effectively tackle and provides practical insights that we can use to inform effective GM implementation and other attempts at gendered organizational change.

A Gender Knowledge Contestation Analysis therefore supplies useful insights into the gendered aspects of policy processes. It enables us to analyze gender where it is obscured and to unpick the processes obscuring it; as such, it is of significant interest for the growing study of “resistance” (Lombardo and Mergaert 2013) to gendered change. Further future challenges might include applying the approach to analysis within organizations with a greater predominance of informal working practices or more fluid policymaking practices. Here, articulation with frameworks, such as feminist new institutionalism, which has theorized ruptures between rhetoric and practice and which focuses more clearly on informal processes (Krook and Mackay 2011), might prove of significant use, while concepts, such as “bounded rationality” might enable deeper examination of recurrent questions of structure and agency in gendered change.

Additional information

Notes on contributors

Rosalind Cavaghan

Rosalind Cavaghan is a Marie Curie Fellow at the Department of Political Science, Nijmegen School of Management, Radboud University, The Netherlands, where her research focuses on the European Union (EU)’s responses to the financial crisis. She completed her PhD on gender mainstreaming in the EU Commission at the University of Edinburgh. Prior to academic research Dr. Cavaghan worked as a policy consultant in Scotland, undertaking policy evaluations and contract research for public sector clients.

Notes

1. For an account of successful implementation see Cavaghan forthcoming 2016.

2. A pseudonym.

3. Or indeed, in this instance developing forms of the supranational state.

4. It should thus be noted that gender knowledge explicitly does not denote expertise concerning gender. Gender knowledge is an analytical concept that enables the examination and comparison of competing gendered meanings in process and in action.

5. European Parliament “Legal Basis Policy for Research and Technological Development.” For a good introduction to EU institutions, policymaking processes, and legal basis, see Bomberg, Peterson, and Corbett (2012).

6. A structure established to implement GM in 1999 during FP6 negotiations.

7. For discussion of the political circumstances enabling this, see Cavaghan forthcoming 2016.

8. Interview, Gender Unit, B.

9. For further details see Cavaghan 2016.

10. Backed by several internal implementation documents that have not been discussed for the sake of brevity.

11. Interview, Directorate D, A.

12. Ibid.

13. Interview, Directorate D, Z.

14. Interview, Directorate D, E.

15. This report shows 5 percent female participation on evaluation panels, 9 percent on the Programme Committee, 12 percent on Advisory Groups, and 10 percent among staff (European Commission, Directorate General for Research 2008, 7).

16. Interview, Directorate D, Z.

17. Interview, Directorate D, N.

18. Ibid.

19. Interview, Directorate D, A.

20. Interview, Directorate D, Z.

21. Ibid.

22. Interview, Directorate D, A.

23. Ibid.

24. Interview Directorate D, N.

25. Directorate D Work Programme 2003.

26. Ibid.

27. Common evaluation procedures for evaluation proposals attached to the Work Programme, Directorate D, Work Programme, 35.