Assessing the practical implementation of the EU’s values in EU–China dialogues

Literature debating the role of the EU’s values in its external relations has neglected to adequately define and empirically explore the practical promotion or mainstreaming of these aspects in diplomatic dialogues with third countries, at the micro-level. Departing from an often abstract focus by scholars on policy outcomes at the macro-level, a concentration on micro-level processes enables an explanation of how value mainstreaming is actually taking place and the elements informing this. It encompasses the role of individual EU officials, the mechanisms guiding their activities, and the impact of interlocutors from third countries. Addressing this gap, this paper defines EU value mainstreaming and conducts a discourse analysis of a comprehensive sample of interviews with EU officials operationalising EU–China dialogues, arguably the hardest test case. It is found that value mainstreaming is rarely taking place in practice due to a nuanced combination of factors. These include EU officials’ perceived lack of responsibility for undertaking such activities, anticipated obstruction by Chinese interlocutors, and counterproductive mainstreaming approaches.


Introduction
The International Relations strand of European Studies is saturated by literature abstractly debating the role of the EU's values, as specified by Article 21 of the Lisbon Treaty, 1 in its external relations at the macro-level. Most commonly, this has taken place through the lens of Manners' (2002) normative power concept. However, scholars have neglected to provide an adequate definition and empirically explore the practical promotion or mainstreaming of the EU's values in diplomatic dialogues with third countries, at the micro-level. Whereas the macro-level focuses on assessing the role of values in policy outcomes, the micro-level concentrates on the processes governing the operationalisation of the EU's values. The arena thus enables an explanation of how value mainstreaming is actually taking place in practice and the factors informing it.
Tackling this gap, this paper defines EU value mainstreaming and empirically assesses its implementation-with a particular focus on human rights-by EU officials in EU-China dialogues, which reflect the forefront of the bilateral partnership at the EU level. China represents arguably the hardest test case for EU value mainstreaming in practice due to unmatched tensions in the EU's external relations between the fulfilment of its economic interests and its constitutive values (Maher 2016, p. 965). The delivery of value mainstreaming with China also has unparalleled stakes for the EU. An increasingly confident and globally assertive China threatens the universality of the EU's shared values. Exploring these under-researched dynamics, this paper addresses the following research question: How is value mainstreaming taking place in EU-China dialogues in practice?
The paper will first develop a definition of EU value mainstreaming in dialogues with third countries, establishing officials' expected behaviour in this area, before describing the role and impact of values in EU-China dialogues specifically. Thereafter, it will summarise findings from a discourse analysis of 48 elite interviews, chiefly with officials from the European Commission and the European External Action Service (EEAS) operationalising EU-China dialogues, before finally reflecting upon conclusions and implications.

Defining EU value mainstreaming
For decades, scholars have been debating the role and impact of values in the EU's international identity and external action. This fascination has been informed by the European project, since its post-WW2 founding, aspiring to uphold and internationally promote its values, which closely align with the universal values of the United Nations (UN) (Manners 2002, p. 51). The EU has emerged particularly as a selfproclaimed champion of international efforts to realise human rights (Manners 2008, p. 41). Early debates, encapsulated by Duchêne's (1973, pp. 38-43) seminal characterisation of Europe as an unorthodox civilian power versus Bull's (1982, pp. 149-151) vision of Europe as a more traditional, aspiring military power, were dramatically catalysed by Manners (2002) conceptualisation of the EU as a normative power. Spawning a range of counter-concepts 2 and swathes of literature, Manners (2002, pp. 238-239) maintained that the EU's international identity was primarily normative in the post-Cold War era, taking precedence over its economic or military dimensions and guiding its external action.
However, Normative power Europe literature has been overwhelmingly restricted to abstract debates about the extent to which the EU meets this ideal type, based on assessment of whether values or interests have been prioritised in policy outcomes.
While providing some valuable insights into the normative impact of EU foreign policy, these works arguably neglect to capture the complexity of the EU's external action. As well as being analytically problematic-values and interests overlap in practice 3 -they provide limited insight into the micro-level processes determining the practical role and impact of values. This includes uncovering how individual officials are mainstreaming EU values with third countries and how interlocutors are impacting these activities.
This literary gap derives from such activities being chronically undertheorised and underspecified. Manners' (2002, pp. 244-245) discussion of the mechanisms through which the EU diffuses its normative power are abstract and lack further development by scholars. The most pertinent mechanism to EU value mainstreaming through bilateral dialogues, procedural diffusion, merely captures the role of institutionalised relationships as a channel for diffusion (Manners 2002, p. 244). It provides no insight into the expected behaviour of EU officials, who are responsible for mainstreaming EU values. Establishing and defining this expected behaviour is essential for assessing the practical promotion of the EU's values in dialogues with third countries like China and encompassing all of its nuances.
At its core, defining EU value mainstreaming necessitates an establishment of the EU's own official discourse in this area. The principal reference is Article 21.1 (Title V) of the Lisbon Treaty (2009), which states that the EU's external action will be 'guided by the principles which have inspired its own creation, development, and enlargement, and which it seeks to advance in the wider world [before specifying the EU's founding values]' (EU 2010, p. 28). However, the language is vague regarding mainstreaming activities by EU officials in practice. While guided could imply officials actively seeking to integrate EU values into all areas of cooperation with third countries (a hard interpretation), it could also refer to EU officials' activities merely being informed by EU values, passively determining engagement (a soft interpretation). Fortunately, supplementary documents addressing the promotion of human rights further clarify this, with a hard interpretation of Article 21.1 emerging.
For example, the Strategic Framework and Action Plan for Human Rights and Democracy (2012), which guides the EU's human rights (and democracy) promotion, states that 'the EU will promote human rights in all areas of its external action without exception' (EU Council 2012, p. 2), before listing all major sectors of cooperation with third countries. Consequently, the document explicitly suggests that EU officials have a responsibility to actively mainstream human rights into all dialogues with third countries, irrespective of sectoral substance. This language is reaffirmed throughout other key documents, such as the EU's Global Strategy (2016),which notes that 'we must… systematically mainstream human rights …across policy sectors and institutions' (EU 2016a, p. 11) and the EU Annual Report on Human Rights and Democracy in the World 2016 (EU Council 2017, p. 7), which describes how 'human rights issues should not be confined to human rights dialogues … [but] included in the agenda of other meetings, including political or other dialogues'. The EU's official discourse in this area can thus be characterised as systematic value mainstreaming in all dialogues with third countries.
These official standards are importantly reinforced and further specified by broader policymaking and academic definitions of mainstreaming. The most prominent and pertinent material surrounds highly popularised gender mainstreaming practices. In this context, the UN defines gender mainstreaming as: The process of assessing the implications for women and men of any planned action, including legislation, policies or programmes, in all areas and at all levels. It is a strategy for making women's as well as men's concerns and experiences an integral dimension of the design, implementation, monitoring and evaluation of policies and programmes [emphasis added]. (UN 2019).
Aligning with this in a European context, the Council of Europe similarly frames gender mainstreaming as '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 [emphasis added]' (COE 2018). 4 Crucially, these definitions emphasise focused actions throughout the planning/design stages, as well as during implementation, further specifying the comprehensive activities which are implicit in the EU's official discourse. Academic definitions of gender mainstreaming notably reinforce these additional attributes, tending to either directly cite 5 or closely approximate the above, widely accepted policymaking definitions. For example, De Waal (2006, p. 10) defines gender mainstreaming as a 'deliberate, planned [and] intended strategy' necessitating 'deliberate and focused interventions at every level' (see also: Van Eerdewijk and Davids 2014).
Following this analysis, EU value mainstreaming in dialogues with third countries can be defined as active and focused actions throughout every stage of the process to systematically promote values in all dialogues with third countries. To realise these standards, it can be hypothesised that EU officials from the institutions operationalising EU-China dialogues-the European Commission and the EEAS-should be undertaking the following actions. Firstly, throughout the planning and design of dialogues, officials should be actively devising ways to incorporate EU values, and such considerations should heavily inform the themes discussed with the Chinese side. Secondly, from this foundation, values should also be either directly or indirectly 6 discussed with the Chinese side during dialogues.

Contextualising the role of values in EU-China relations
To effectively assess EU value mainstreaming in EU-China dialogues and contextualise the findings, it is critical to describe the role of values in the bilateral relationship and summarise existing scholarship in this area. Beyond substantial bilateral relations with individual member states, the EU level of EU-China relations is centred around the EU-China Comprehensive Strategic Partnership. Encompassing more than 60 bilateral dialogues, spanning a diversity of policy sectors, the partnership is built around three pillars: Political Dialogue (1), Economic and Sectoral Dialogue (2), and People-to-People Dialogue (3) (EEAS 2017). While intended to institutionalise political cooperation and emulate the success of the central trade relationshippresently worth more than €1.7 billion per day-the results have been uneven in practice (EU 2019b).
Multilaterally, successes like the co-facilitation of the United Nations Framework Convention on Climate Change (UNFCCC) Paris Agreement are matched by persistent disagreements on international crises, global development, and connectivity (i.e. China's Belt and Road Initiative [BRI]) (EU 2017;Geeraerts 2019, p. 288;Maher 2016, p. 966). Similarly, bilateral success has been limited to specific policy sectors like environment/ climate change and trade (despite a high-profile dispute over China's unfair trade practices), while cooperation in many other sectors remains relatively unproductive (see: Farnell and Crookes 2016, p. 4;Geeraerts 2019, pp. 285-286).
The EU's guiding China documents explicitly acknowledge these inconsistent results and provide insights into the frustrations surrounding them. In this fashion, the overarching Elements for a new strategy on China (2016) calls for 'genuine implementation of the Chinese slogan win-win cooperation' (EU 2016b, p.4) and promotes engagement with China which is more 'practical and pragmatic' (ibid., p.5) as well as principled in nature. The latter aspect directly referencing the 'principled pragmatism' (EU 2016a, p.8) concept coined by the Global Strategy, which envisions a dual agenda-setting role of values and material interests in the EU's external action. Reaffirming these grievances, the strategy's supplement, EU-China-A Strategic Outlook (EU 2019c, p. 2) describes how the EU's unity and coherence in EU-China relations is being repeatedly compromised (see also: Raube and Burnay 2018, pp. 297-299). Lucrative Chinese investments in member states have particularly impeded EU common positions on bilaterally contentious issues like human rights (see Reuters 2017). 7 The uneven results of the strategic partnership and the tensions underpinning it, arguably derive in large part from a disagreement over values (Geeraerts 2019, pp. 281-282;Mattlin 2012, p. 186). Although China broadly repudiates the universality of the EU's values, human rights reflect the most controversial aspect, and as a result, the key focus of this paper (Maher 2016, pp. 962-963). This derives from not only its highprofile role in the relationship, fuelled by Xi Jinping's ever more repressive governance (AFET 2015, p. 13), but also the value's comprehensive nature for the EU. Human rights are deemed to be indivisible from its other values, especially rule of law and liberal democracy, which serve to enable them (Manners 2008, p. 51). They are also notably bound into the social pillar of sustainable development (EU Commission, 2016). The EU-China conflict over human rights derives from differing interpretations (Men 2011, p. 549). Whereas the EU attributes equal importance to individual civil and political rights and socio-economic rights, the Chinese Party-State prioritises the latter, justified by collectivistic value systems and its development needs (Guo 2013, pp. 48-50;Zhao 2015, p. 48).
Responding to this value-gap, the EU has pursued a pragmatic strategy of constructive engagement since the mid to late 1990s, where the facilitation of closer relations with China (embodied by the strategic partnership) is perceived as a means of passively stimulating Chinese reform (see : Casarini 2006, p. 21;Michalski and Nilsson 2018, pp. 440-441). The EU-China Human Rights Dialogue represents the sole dialogue dedicated to promoting an EU value. Scholars are consistently critical of both constructive engagement and the human rights dialogue, arguing that values are being sidelined in favour of EU material interests with China (Fox and Godement 2009 p. 8; Mattlin 2012, pp. 189-194;Raube and Burnay 2018, pp. 297-299). However, such assessments are overwhelmingly based on results-orientated observation of policy outcomes, while neglecting to explore the underlying processes informing the mainstreaming of values in EU-China dialogues, at the micro-level.
The notable exception to this is Kinzelbach's (2015) scholarship, where she reconstructs the EU-China Human Rights Dialogue 1995-2010 at this micro-level of analysis. Nevertheless, this paper's scope is much broader, covering value mainstreaming comprehensively across EU-China dialogues, in post-Lisbon era EU diplomacy with China. Moreover, unlike Kinzelbach's (2015) work, which eschewed direct quotes to maintain anonymity with a small group of Brussels-based diplomats, this paper's larger transnational sample unveils and analyses the language of EU officials operationalising EU-China dialogues. This offers novel insights into how EU value mainstreaming is taking place in practice.

Methods and sampling
Addressing the largely neglected micro-level of EU value promotion with China, this paper used a range of qualitative methods. At its core, the elite interview method was adopted due to its specialisation in extracting rich accounts from powerful individuals (Stephens 2007, p. 205). A total of 40 semi-structured elite interviews were conducted with officials from the European Commission and diplomats from the EEAS (collectively EU officials) that operationalise EU-China dialogues, based in both Brussels and at the Beijing-based delegation, chiefly during 2017. These interviews, spanning all three pillars of EU-China dialogues, were highly comprehensive, with most dialogues managed by a single official. These accounts were compounded by a further 8 contextual interviews, carried out with officials from the Council of the EU, member states, China, and NGOs. Interviews lasted 1.25 hours on average and questions principally focused on the practical promotion of EU values more broadly, the promotion of human rights specifically, the institutional mechanisms governing such activities and the role/impact of the Chinese side.
Discourse analysis was then applied to the transcripts to maximise the data's richness, based on the steps outlined by Milliken (1999). The transcripts were also triangulated with over 60 sector-specific and strategic policy documents, ensuring their reliability and the identification of the EU's (aforementioned) official discourse on value mainstreaming. Finally, basic descriptive statistics were used to systematise the findings, displaying the relative strength and prevalence of each discourse across the interviews and providing context for the key examples provided.

Findings
Contrary to this paper's established definition of EU value mainstreaming and the behaviour it prescribes amongst EU officials-active and focused actions throughout every stage of the process to systematically promote values in all dialogues with third countries-interviewees' accounts suggested that value mainstreaming was rarely taking place in EU-China dialogues in any form. At the most general level, this was explicitly confirmed by interviewees. When asked about the objectives of their dialogues with China, officials described sector-specific issues, while normative dimensions were rarely mentioned, let alone any active, focused, and systematic mainstreaming activities. This was reaffirmed by interviewees' responses to specific questions concerning the extent to which human rights were being incorporated into dialogues.
While this lack of value mainstreaming activities is somewhat unsurprising, superficially supporting existing claims in EU-China literature that the EU's values are being sidelined in favour of EU material interests, discourse analysis uncovered more nuanced explanations for these practices, not captured by existing scholarship. It revealed how specific shared perceptions amongst officials, encapsulated by five key discourses, could explain why these activities are not taking place in practice.
Firstly, the organisational discourses highlighted how internal institutional dynamics and understandings of roles and responsibilities were acting as rationales for individual mainstreaming behaviour. Discourse 1 pertained to a view that value mainstreaming was implicit to EU-China dialogues and required no further efforts to realise, while discourse 2 cast such practices as lacking relevance to all officials' dialogues and instead the responsibility of designated individuals/dialogues. The second category of discourses captured officials' external perceptions. These external discourses invoked the role of the Chinese side and its conduct to justify mainstreaming practices. Discourse 3 related to an anticipated obstructive response by the Chinese side to value mainstreaming as a principle barrier to such activities, while discourse 4 cast mainstreaming as pointless, as well as problematic, due to China's opaque institutional structures informing interlocutors being likely unable to impact the relevant Chinese policies. Lastly, discourse 5 surrounded a perception that China is not listening to the EU and it should be. It captured frustrations that China was not aligning with EU values, informing ineffective approaches to mainstreaming in the rare instances when such activities were taking place.
Notably, interviewees often displayed multiple discourses, with 85% displaying more than one discourse and 50% three or more. This highlights that although some discourses were more prevalent than others (see Table 1 below), value mainstreaming behaviour is being impeded by a nuanced combination of the five discourses as opposed to being rooted in any single discourse. Additionally, all discourses were identifiable in the accounts of both EEAS and Commission officials, irrespective of them being based in either Brussels or Beijing.
It is pertinent to highlight that while the EEAS coordinates the EU's external action, its relationship with the European Commission is one of cooperation between two autonomous institutions (see: EU Council 2010, p. 1, 7). As a result, it does not have the capacity to legally enforce value mainstreaming practices in Commission-led dialogues. However, in sharing the above discourses, EEAS diplomats, through their coordinating role, are de facto reproducing these discourses amongst their Commission colleagues. This was reaffirmed through the author finding that coordination of systematic value mainstreaming was absent from EEAS-led inter-service meeting mechanisms in Brussels and Beijing, irrespective of size/format. Finally, to provide context for the findings, the EEAS operationalises all Pillar one political dialogues and the European Commission manages all others. Each dialogue typically takes place annually for one day, alternating between Brussels and Beijing, encompassing 10-20 individuals (mostly officials) on each side and chaired by senior officials commensurate to the level (i.e. political status) of the dialogue.

Organisational discourses
Discourse 1, surrounding a view amongst officials that EU values are implicit to EU-China dialogues and require no further action to realise, was expressed by 48% of officials. A key example was found in the account of a Brussels-based DG Trade official. They described how EU values such as human rights were 'under the surface' of their dialogues with China, which 'are just sort of built on, part of our values…and our understanding that this is what is good'. 8 Their language implies that EU values are not being systematically mainstreamed as they are already passively informing the dialogue ('under the surface'). At a deeper level and reflecting a consistent aspect of the discourse found in other accounts, the interviewee's language also suggests that the inherent role of values derives from (a) the sectoral substance being based on EU values (i.e., 'part of our values') and (b) that they, as an EU official, believe in the values ('our understanding that this is what is good').
Another key example of discourse 1 could be seen in the account of a Beijingbased EEAS diplomat. They described how 'values are built into our conscience … there are not rational decisions to politicise certain areas' and that any engagement with China on areas like climate change and development was inherently normative in nature. 9 Strongly approximating their Commission colleague, they reaffirm the implicit role of values in their activities, due to the EU's inherent normativity and their personal belief in these principles ('values are built into our conscience'). The implied lack of active, focused, and systematic mainstreaming practices is also overtly confirmed by the notion that there are not 'rational decisions' to integrate EU values like human rights. Additionally, the characterisation of value mainstreaming as 'politicising' exchanges alludes to concerns over provoking obstructive conduct by Chinese interlocutors (discourse 3).
Pertaining to the notion that value mainstreaming lacks relevance to all officials' dialogues and instead reflects the responsibility of designated individuals/dialogues, discourse 2 was shared by 58% of officials. The discourse emerged as the second most prevalent justification for eschewing value mainstreaming amongst interviewees. A key example was identifiable in the account of a Beijing-based DG Trade official. They described how their normative 'toolkit', derived predominantly from the World Trade Organization (WTO), as opposed to the EU's Article 21 and reflected a 'different enforcement system' to their EEAS colleagues. 10 Here, the interviewee explicitly suggests that they are not systematically mainstreaming EU values due to the inapplicability of such practices to EU-China trade dialogues, which instead reflect an exclusive EEAS responsibility. This implied division of labour, expressed in many other Commission officials' accounts, is justified by the interviewee on the basis that trade dialogues are informed by an alternative rulebook, linked to the EU's material interests (a WTO 'toolkit' which reflects a 'different enforcement system' to the EEAS).
Another notable example was found in the account of a Brussels-based official from DG Environment. They described value mainstreaming, with particular reference to human rights and sustainable development, as 'in principle, a role for the EEAS' and that EU-China environmental cooperation was 'very successful' as a 'technical discussion'. 11 Again, discourse 2 is overtly expressed through the suggestion that dialogues in this area are limited to sectoral substance, with value mainstreaming considered an exclusive EEAS responsibility and omitted. Through describing this implied division of labour as a matter of 'principle', the language also alludes to the power and influence of discourse 2 amongst EU officials. Moreover, the notion that the dialogue's success is informed by restricting it to 'technical discussion(s)' displays the concerns over triggering obstructive conduct by the Chinese side through value mainstreaming, associated with discourse 3.
Lastly, an example from a Beijing-based EEAS diplomat, where they described the role of human rights in EU-China judicial cooperation, effectively displayed how the EEAS shares and thus enables discourse 2 amongst Commission officials: 'we have to separate this from the human rights dialogue …If you already have a dialogue on 9 Interview with a diplomat from the EEAS on 31.05.2017 at 12.00pm, in Beijing 10 Interview with an official from DG Trade on 09.06.2017 at 4.00pm, in Beijing 11 Interview with an official from DG Environment on 08.03.2017 at 11.00am, in Brussels human rights, why do you need another?'. 12 Contrary to the ethos of value mainstreaming across EU-China dialogues, the interviewee implies that human rights are being sufficiently addressed in the dedicated dialogue and should be restricted to this accordingly. Overall, discourses 1 and 2 highlight the need for formal institutional mechanisms to clarify the roles and responsibilities of all EU officials to systematically mainstream EU values with China.

External discourses
The remaining discourses related to officials' external perceptions, capturing the impact of China's role and conduct on mainstreaming practices. Firstly, discourse 3 related to the perception that the anticipated obstructive response of the Chinese side to value mainstreaming reflected a principal barrier to such activities. Shared by 63% of interviewees, this emerged as the most prominent discourse and suggested that the actions of the Chinese side were considered by EU officials to be the superseding impediment to value mainstreaming in EU-China dialogues.
The discourse primarily manifested in a fear that raising the most controversial values in dialogues-like human rights and rule of law-will upset the Chinese side and compromise exchanges. Significantly, officials sharing this chief facet of the discourse was not dependent on first-hand experience of obstructed value mainstreaming. Instead, it appeared to often derive from experience of China obstructing contentious sectoral issues and more pertinently, the negative secondhand experiences shared by those few individuals which have mainstreamed values with China, particularly through the EU-China Human Rights Dialogue. In effect, China's conduct in the human rights dialogue appeared to be inhibiting EU value mainstreaming in all other dialogues. Unlike any of the organisational discourses, this suggests that discourse 3 may be being actively cultivated by the Chinese side, which seeks to foster a climate of fear, deterring value mainstreaming in dialogues.
Providing a key example of discourse 3, one Beijing-based DG Trade official stated that 'if we integrate human rights [into trade dialogues], it might antagonise' the Chinese side, while addressing a 'purely business' approach will attain 'more traction'. 13 The interviewee explicitly suggests that they are not mainstreaming human rights in dialogues ('if we integrate'), while a 'purely business' approach implies that value mainstreaming is being broadly eschewed in favour of discussing sectoral substance, that is more likely to engender productive outcomes ('more traction'). Crucially, their language implies that this approach is informed by a perception that such activities will provoke Chinese interlocutors and stymie the dialogue ('it might antagonise' the Chinese side). A Brussels-based DG Energy official provided another key example, describing how: It would not work, it would really stop cooperation in many areas if we go there with our energy issues and we [raised human rights] …I am glad some people do it, I think we should have this human rights dialogue …but in general, as a general strategy to address it in every dialogue we have, I think it's impossible, it would not work. 14 12 Interview with a diplomat from the EEAS on 08.06.2017 4.00pm, in Beijing 13 Interview with an official from DG Trade on 09.06.2017 at 4.00pm, in Beijing 14 Interview with an official from DG Energy on 15.02.2017 at 10.00pm, in Brussels.
Discourse 3 is again expressed in the interviewee's suggestion that human rights mainstreaming is not taking place in their dialogues due to concerns over upsetting their counterparts. The power of the discourse is reflected in the strength of their language, which portrays such practices as wholly incompatible with the attainment of sectorspecific objectives in not only energy dialogues but any dialogue beyond the human rights dialogue ('It would not work …as a general strategy …I think it's impossible'). Notably the example displays how the human rights dialogue is invoked by officials as a cautionary tale of the consequences of mainstreaming controversial values with China.
Discourse 3 also importantly extended into the accounts of those EEAS diplomats involved with the human rights dialogue itself. All eight of these interviewees expressed an alternative facet of the discourse, wherein China's obstructive behaviour was presented as the principle impediment to their active attempts to promote the value. As one diplomat noted: The quality of exchanges went down significantly [in recent years], so this is why the [EEAS] colleagues have the impression of talking to a brick wall; because you say something, they reply something else, they read out of the paper and they make sure the paper is 40 pages long and it's going to take half the day and that they're just going to read it very slowly …it's one of the diplomatic techniques of how to obstruct a meeting. 15 The extent of China's obstruction is portrayed by the suggestion that every comment/ question by the EU is being dismissed and genuine two-way communication is being thwarted ('you say something, they reply something else'). This is reaffirmed by the notion that engagement with China is akin to 'talking to a brick wall' and that it is pursuing a time-wasting strategy. Notably, the extent of this obstruction appeared to inform these EEAS diplomats being sympathetic if not explicitly condoning restraint towards value mainstreaming in the Commission-led dialogues they coordinated. This supports the notion that the challenges of the human rights dialogue are deterring value mainstreaming in all other dialogues.
Closely related to discourse 3 and shared by 18% of interviewees, discourse 4 related to a view amongst officials that mainstreaming controversial values with China is not only likely to engender an obstructive response but pointless, as opaque Chinese interlocutors will likely be unable to impact the relevant Chinese policies. The discourse again appeared to be often informed by second-hand experiences of value mainstreaming and interlocutors refusing to discuss contentious sectoral issues in dialogues. Although this discourse emerged as the least prevalent, since a sizeable number of interviewees articulated concerns about China strictly controlling dialogue agendas, discourse 4 may be more widely shared and of higher impact than indicated by this paper's findings. While the discourse thus links to the working reality of China's ultra-hierarchal institutional structures, wherein Chinese officials are heavily restricted in the topics and agenda that they can engage with in dialogues, these instructions by the Party-State hierarchy are purposeful and strategic. This supports 15 Interview with a diplomat from the EEAS on 23.03.2017 at 11.00am, in Brussels the notion that China is seeking to actively deter EU value mainstreaming, accentuating perceptions of institutional limitations on the Chinese side amongst EU officials.
In a key example, one Brussels-based DG Trade official explicitly articulated discourse 4 in implying that the nature of their interlocutor was informing an avoidance of value mainstreaming: 'we deal with the Ministry of Commerce …the focus would be a bit different if we were dealing with the State Council'. 16 Interestingly, a Brusselsbased official from DG Justice and Consumers involved in a dialogue where the State Council is the interlocutor, stated that 'the State Council, are the State Council, I mean... I don't think they're really into promoting, value-based issues'. 17 This demonstrates how sharing discourse 4 is not contingent on first-hand experience of trying and failing to mainstream values with China. It also supports the existence of active Chinese efforts to convince EU officials that their counterparts cannot engage with value-related issues, with the State Council being one of the few institutions which can do so.
Discourses 3 and 4 highlight that for active and systematic value mainstreaming to be achieved, formal institutional mechanisms to enforce such activities are not enough. Officials also require greater understanding of Chinese interlocutors and clear strategies to confidently raise-either directly or indirectly-EU values in EU-China dialogues, countering a climate of fear, which impedes such practices.
Finally, discourse 5, surrounding the notion that China is not listening to the EU and it should be, was shared by 53% of interviewees. Reflecting the third most prevalent, the discourse captured frustrations on the EU side that China was not complying with the EU's values and appeared to inform ineffective approaches to value mainstreaming, in the rare instances where such activities were taking place, chiefly in the context of the human rights dialogue. The discourse seemed to underpin dynamics where EU officials repeatedly promoted EU values without success, while dismissing any alternative positions by Chinese counterparts as invalid/inferior. Importantly, the frustrations deriving from China's non-compliance appeared to engender increasingly didactic approaches to value mainstreaming by officials, which may be counterproductively increasing China's obstruction to the discussion of EU values (feeding discourses 3 and 4). These frustrations also seemed to inform a pessimism amongst officials that common ground on values with the Chinese Party-State could be found, reinforcing didactic approaches and a neglect of more creative strategies. Discourse 5 is arguably rooted in an uncritical belief in the inherent superiority of the EU's values and their universality, which displays a lack of reflectivity and could be interpreted as neocolonial by third countries.
The strongest and most pertinent examples of discourse 5 were found in the accounts of those individuals connected to the human rights dialogue itself. EEAS diplomats appeared to use references to international (UN) documents to empower the discourse, asserting the superiority of the EU interpretation of the value and international law. As a Brussels-based diplomat stated: One argument that we have [been] using more and more is actually… because they say, "human rights, Western values, you are talking to us like a teacher would speak to a pupil and you know… those days are over basically". But we 16 Interview with an official from DG Trade on 10.03.2017 at 3.00pm, in Brussels 17 Interview with an official from DG Justice and Consumers on 01.03.2017 at 10.00am, in Brussels are saying "look, you have signed up to a number of conventions, you have committed to the international human rights system and standards…so basically this is what you have to respect… the obligations that you have under these international instruments". Because then they can't throw back at us the fact that you know its Western values… well they will still do it. 18 Articulating discourse 5, the interviewee describes how the EU repeatedly asserts its interpretation of human rights ('one argument that we have [been] using more and more'), based on its perceived superiority/universality. The didactic nature of this approach is reflected in the interviewee paraphrasing how the EU demands China's acceptance of this interpretation ('this is what you have to respect') and their interlocutors feeling like they are being treated as subordinate partners ('you are talking to us like a teacher would speak to a pupil'). The language also portrays how officials are inattentive to any alterative interpretations of human rights asserted by their Chinese counterparts ('they can't throw back at us the fact that you know its Western values') and finding any common ground therein. The pessimism associated with the discourse, which limits the pursuit of more creative (i.e., less didactic) approaches, is also displayed in the notion that 'they will still do it'.
Crucially, most officials displaying discourse 5 (62%), were not actively mainstreaming values in their dialogues with China. In this sense, akin to discourses 3 and 4, sharing discourse 5 did not require first-hand experience of trying and failing to mainstream values with China. Instead, it appeared again to be catalysed by negative cautionary tales from colleagues, particularly those involved with the human rights dialogue. For these officials, the discourse informed a view that value mainstreaming could only be a conflictual exercise with poor outcomes. Additionally, through sharing the discourse, these individuals would likely practice value mainstreaming in an ineffective, didactic manner, if they were to undertake such activities, akin to some of their colleagues operationalising the human rights dialogue. This reinforces the notion that clarifying officials' roles and responsibilities in this area, through establishing formal institutional mechanisms, would be insufficient without the orchestration of effective mainstreaming strategies in parallel.

Conclusion
This paper has sought to develop a definition for EU value mainstreaming and empirically assess how it is taking place in practice at the neglected micro-level of the EU's external action. It focused on arguably the hardest test-case: EU-China dialogues. Based on analysis of the EU's own official standards and pre-eminent definitions of policy mainstreaming, this paper defined EU value mainstreaming as active and focused actions throughout every stage of the process to systematically promote values in all dialogues with third countries. To realise these theoretical standards in the context of EU-China dialogues, it was hypothesised that EU officials should be actively and systematically integrating EU values into the 18 Interview with a diplomat from the EEAS on 09.04.2015 at 10.00am, in Brussels planning and operationalisation of their dialogues. Diverging from this expected behaviour, a discourse analysis of interviews with European Commission and EEAS officials revealed that value mainstreaming was rarely taking place in practice, in any form. While superficially supporting pre-existing claims by scholars that the EU is prioritising its material interests with China, this analysis also provided much deeper insight into why this was the case. It emerged that such dynamics were rooted in a nuanced combination of perceptions shared by officials, encapsulated by five key discourses.
The organisational discourses, surrounding perceived institutional roles and responsibilities as rationales for mainstreaming behaviour, displayed how officials were avoiding mainstreaming due to it being considered implicit to their roles (discourse 1) and/or them considering it the exclusive responsibility of other individuals/institutions (discourse 2). The remaining discourses related to officials' external perceptions, encompassing the role of the Chinese side and its conduct. Discourse 3, emerging as the most prevalent shared view in interviewees' accounts, captured China's anticipated obstructive response to such activities as a principle barrier. Relatedly, discourse 4 related to how officials perceived value mainstreaming as pointless, as well as problematic, due to Chinese interlocutors' opaque competencies, while discourse 5 surrounded a perception that China is not listening to the EU and it should be and informed ineffective approaches to value mainstreaming in the rare instances when it was taking place.
These findings have significant implications. Failure by the EU to systematically mainstream its values threatens their continued universality. The BRI will see China emerge as the principal bilateral partner for over 60 countries, at a time when the Chinese Party-State is growing increasingly confident in asserting its alternative interpretation of the UN's universal values. This is reflected in its transition from denial to defence of internment camps in Xinjiang, nationalisation of its Social Credit System and its explicit termination of freedoms guaranteed under One Country Two Systems in Hong Kong. However, this paper's findings do not suggest that the EU will sufficiently rise to this normative challenge and defend the rules-based world order, integral to its global interests. Without the introduction of formal institutional mechanisms to ensure that the EU's values are actively and effectively designed into all EU-China dialogues, the EU is instead, inadvertently contributing to the universality of the Chinese Party-State's values, as opposed to its own.

Declarations
Research involving human participants and/or animals The research underpinning this paper involved human participants, encompassing 48 elite interviews with officials from the European Union, China and individual EU member states, as well as practitioners from non-governmental organisations.
Informed consent The informed consent of participants was secured through them being provided with a document that transparently and clearly explained the purpose and aims of the research. Prior to interviews taking place, participants were given the opportunity to query the author about the research, before signing the form and providing their explicit consent to partake in the project.

Conflict of interest The author declares no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.