Between migration and mobility discourses: the performative potential within ‘intra-European movement’

This article displays how discourses around intra-European movement are constructed to reveal the performativity of discourses. Therefore, it mainly aims to deliver theoretical contributions to the field of discursive policy analysis by empirical case study material. The overall argument is that discursive policy analysis benefits from an analytical framework that deals with a refined operationalization including ‘storyline’ and ‘poetic’ elements. This framework is applied to intra-European movement in the cases of the European Commission and the Netherlands. These cases are particularly interesting, since both authorities have competing constructions of ‘intra-European movement’, highlighting ‘migration’ versus ‘mobility’. As such, the article displays the importance of ‘poetic elements’, opens up the discursive black box of discourse analysis and unravels the performative potential of certain discourses.


Introduction
'The working of words upon actions is the basic political action'. (De Jouvenel 1963, 99) The European Union (EU) enlargements of 2004 and 2011 shed a new light on the European continent. New member states joined the EU and EU citizens could freely move around a new territory. This simultaneously resulted in policy proposals on the EU level to stimulate this mobility, while in some member states, like the Netherlands, more repressive policy proposals were promoted. Not insignificantly, while the EU speaks about 'mobility', the Dutch government refers to 'migration'. One could argue that the new European borders produced varied governmental discourses within one legislative area. Therefore, this article holds a discursive perspective to understand these contradicting governmental discourses. In order to do so, first the policy discourses are descriptively reconstructed while secondly, the performativity of these discourses on policy proposals is highlighted. Thus, this article primarily aims to deliver theoretical contributions to the field of discursive policy analysis by delivering empirical case study applications within the case of 'intra-European movement' as general signifier.
The case of 'intra-European movement' is selected because of its discursive complexity and its contested political context. In general, migration issues are politically contested, deliver wicked problems, diverse problem definitions and policy controversies (Van Leeuwen and Wodak 1999;Scholten 2013). Especially by studying two governmental authorities that hold contradicting views, this case is well suited to understand the definition of problems, ideas and meaning that guide particular actions (Stone 1988;Fisher 2003;Yanow 1996). Moreover, it questions which discourses are present around intra-European movement and to what extent are they preceding policy proposals? Therefore the research question of this article is: Which governmental discourses on intra-European movement can be identified at the European (European Commission) and Dutch national level and how did these discourses affect policy proposals (in the period 2002-2014)?
The analysis is built up in the following way. Firstly, the theoretical premises and methodological implications are described. Secondly, the empirical findings are presented on the domain of intra-European movement. Finally it concludes on these findings.

Theoretical outline
The politics of migration or mobility are analyzed in many ways but since the so-called 'argumentative turn', discourse theory emerged in migration studies and policy analysis too (Fischer and Forester 1993;Van Leeuwen and Wodak 1999;Balch 2010). This perspective understands the usage of language as a medium through which actors create the world (Hajer 1993). It puts attention on language as a performative or constitutive dimension of 'reality', understood as discourse, by actively producing society by attributing meaning, norms and power which disciplines human agencies to think, speak and act in a certain way (Throgmorton 1993;Foucault 1994;Fisher 2003). Therefore, discourses are defined as 'an ensemble of notions, ideas, concepts and categorizations through which meaning is allocated to social and physical phenomena, and which is produced and reproduced in an identifiable set of practices' (Hajer 1995, 44). This definition of discourse, embraces both 'texts' and 'practices' and explicitly emphasizes performativity (Hibberd 2005).
This study primarily holds a discursive-institutional approach within the tradition of 'discourse as practice' (Fairclough 1992;Jacobs 1998;Schmidt and Radaelli 2004;Schmidt 2008Schmidt , 2011. It examines linguistic practices within particular discursive events, including policy documents (Hastings 1999). This perspective makes it possible to observe how discourses institutionalize and affect social outcomes, or stated by Hajer (1995, 264):'The main theoretical thesis (...) is that one can observe how the institutional practices (...) work according to identifiable policy discourses that through their storylines provide the signpost for action within these institutional practices.' Within this discursiveinstitutional approach a textually oriented discourse analysis is applied (Fairclough 1992;Hajer 1995;Sharp and Richardson 2001). This does not imply a sole focus on text as such, but a dialectical relationship between social practice and discursive practice since 'discourse constitutes social practice and is at the same time constituted by it' (Van Leeuwen and Wodak 1999, 92). Changes at the social level can be constituted through changes in linguistic practices since language itself is seen as a form of action (Hastings 1999;Yanow 2003).
However, by assuming this dialectical relationship, this study does not want to keep the linkage between rhetoric and action implicit (Sharp and Richardson 2001). Therefore, it is important to explicate the operationalization of policy discourses, to separate its 'rhetorical' or 'linguistic' component from its proposed 'materialized' component (Fairclough 2012). By this, it becomes possible (1) to show some key strategies of discursive legitimation and (2) to show the performative potential in the justification of courses of action. The aspects of legitimacy and performativity have, therefore, a very central role in the operationalization and analysis.
'Legitimation' is defined as a discursive technique that explains and justifies social activity, and typically involves providing 'good reasons, grounds, or acceptable motivations for past or present action' (van Dijk 1997, 255). Within this focus, 'legitimation' can be conveyed through discourses whose outcomes reward legitimate actions (Van Leeuwen 2007). And 'performativity' is seen as the discursively regulated practices within policy communication that reify them in that very process (Feldman 2005). The 'performative potential' of discourses is how powerful discourses contribute to dominant strategies for action (Healey 1999;Howarth and Griggs 2006). By operationalizing this legitimation process it becomes possible to study to what extent policy discourses influence policy proposals and 'provide a signpost for action' since discourses 'serve as precursors to policy outcomes' (Hajer and Versteeg 2005, 178). From this perspective, discourses have a feed-forward effect on policy proposals, assuming that the construction of policy discourses affects the construction of policy proposals (Schneider and Sidney 2009;Jacobs 1998;Schmidt 2008Schmidt , 2011 because: Putting forth a diagnosis also includes a prognosis of what should be done to solve the alleged problem. How problems firstly are constructed as problems and secondly how they are framed have a crucial impact on the policy design that is developed to tackle the problems. (Jørgensen 2012, 50) Or in other words, policy design includes a clear diagnosis of 'a problem', followed by a clear prognosis and call for action what is needed to be done, since social problem conceptions involve a 'theory of causation' (Verloo 2005).
Therefore, specific attention is put on 'policy proposals': the proposed 'courses of action' that evolve from a certain discursive construction. And a policy proposal differs from 'policy practices', since it merely focuses on the initial phase of policy discourses instead of whether or not these proposals became routinized in institutional practices. These policy proposals can be measures, laws and legislations, 1 which could have been implemented. Consequentially, since the main focus is on the initial process of discursive legitimation (pre-structuration) (Hajer 1995). Therefore, this article does not focus on policy phases, such as the formulation or implementation phase, but merely on the construction processes of discursive legitimations.

Operationalization
To be able to study discourses, a well-operationalized research design is important. One of the main contributors in the field is the work of Hajer, which distinguishes three layers in the analysis of policy discourses (2003,104): -Storylines, metaphors and myths.
This Hajerian perspective involves a layered chronology in the manifestation of discourses to understand discourse structuration and discourse institutionalization (Hajer 1995;Schmidt 2008Schmidt , 2011. Former applications of this approach assume an interrelationship between these layers, include a large span of interpretation and a non-mutually exclusionary approach of these layers. It stays rather implicit how the discursive analytic framework is empirically applied and how theoretical concepts are operationalized (Hajer 1995(Hajer , 2003Balch 2010).
Therefore, the theoretical focus in this study deviates in its focus and design from the Hajerian approach. First of all, this article focuses on the initial phase of discursive legitimation. This focus is needed to open up the discursive black-box how discourses gain legitimacy (even before the phase of discourse structuration or institutionalization). Secondly, the research design deviates from Hajer's account on what is called 'the first layer' of policy discourses, consisting of 'storylines, metaphors and myths that help sustain the societal support for particular policy programmes' (Hajer 2003, 104). While Hajer lumps all together, in this article, storyline elements are separated from metaphors and myths as poetic elements. This separation and explication is needed to understand the legitimation processes better and to overcome the general under-operationalization in discourse analysis, sometimes even based on 'intuition' (Hajer 1995;Howarth and Griggs 2006;Balch 2010;Carrete and Gasper 2010). Consequently, the layer of epistemic notions is excluded, which is conceptualized as 'a state of mind' in a particular period (Hajer 2003). Mainly because this layer is too loosely conceptualized and former applications did not contribute to a refined operationalization (Hajer 2003, 106;. Therefore, the following conceptualization draws upon alternative studies to explicate 'poetic' and 'storyline' elements. First of all, discourses maintain poetic elements or tropes (Throgmorton 1993). By poetic elements we mean concepts, metaphors, myths and numbers. Concepts are repeatedly used structuring and forming terms and labels (Hajer 2003;van Ostaijen and Scholten 2014). This strongly relates to metaphors, which seduce the reader to see something as something else (Lakoff and Johnson 1980). Metaphors can contain the power of becoming small self-fulfilling prophecies, becoming a guide for future action (Lakoff and Johnson 1980, 156) or can function as 'the premises upon which decisions are made' (Edelman 1971, 68). Concepts can also have this use-value, but metaphors can function in a symbolic analogy (Schon 1979;Stone 1988;Yanow 1996Yanow , 2003Charteris-Black 2006). And the more coherent a policy discourse is developed, the more critical metaphors become (Chilton and Ilyin 1993). Myths are repeatedly used public narratives holding certain assumptions and which could occur as historic continuity or with reference to the past as source for the current (Schama 1988). Myths are socially unquestioned and constructed public narratives of a particular culture, which 'diverts attention from a puzzling part of reality' (Malinowski 1948;Yanow 1996, 191). Finally, counting by numbers is a way to classify objects (Cohen 1982). Behind the usage of numbers lay deliberate decisions about how to count as since a phenomenon is perceived at least frequent enough to bother counting (Stone 1988, 172). Numbers are the final poetic element because 'numbers work exactly like metaphors' and as such 'numbers are another form of poetry' (Stone 1988, 163-165). All the above are the 'poetic' elements of policy discourses.
Secondly, discourses contain storyline elements, which are the 'narrative' components of discourses (Throgmorton 1993). The storyline elements can be specified by objectives and subjectives (Linder 1995;Schneider and Ingram 1997). Overall, the objective is considered as the problem definition, so 'what is defined as problem' while the subjectives are the targeted populations or fields to which problems and objectives are related (Stone 1988;Linder 1995;Ingram 1993, 1997). Both are closely affiliated since the objective can imply a certain subjective or vice versa by the type of intervention or prognosis (Verloo 2005). Therefore, the type of intervention is the third element, and by focusing on the proposed 'course of action' it looks at the performativity of discourses (Douglas and Wildavsky 1982;Linder 1995;Hibberd 2005). By the type of intervention, the normative guidance for action on policy proposals can be studied (Schneider and Ingram 1997). Poetic and storyline elements do not occur in an isolated context, but get meaning in relation to each other. This leads up to the following grid (see Table 1), operationalized for a comprehensive discursive policy analysis. This grid holds a central place in the upcoming data analysis.
The aim of this grid is threefold: -Firstly, to open up the discursive black box of discourse analysis.
-Secondly, to study to what extent this comprehensive operationalization contributes to an empirical case studies analysis. -Thirdly, by putting explicit attention to the performative element of discourses, contributing to go beyond mere descriptive-analytical accounts on metaphor or discourse analysis. 2

Research methods
This study is an interpretative discourse analysis on the discourses produced in the field of intra-EU movement, which serves as general signifier or 'object of research ' (Bourdieu and Wacquant 1992;Yanow 2003). To be able to focus on governmental policy discourse, it starts by distinguishing two public authorized levels of governance, more or less (manifestly) present in the field of intra-EU movement. 3 Firstly, the EU is selected, since EU legislation made the 'free movement' of persons possible, by a 'removal' of certain physical borders. This legislation, followed by European policies implementing Article 21/22 of the Treaty of Rome and Lisbon makes the EU an inevitable starting point for analysis. For the EU, the focus is on the ways how 'intra-European movement' is communicated by the European Commission (EC) (by means of certain Directorate-Generals (DGs)), toward its controlling institutional authority, the European Parliament. Secondly, as member-state, the Netherlands is selected, since the Netherlands have historically played a foundational role in the construction of the EU and the Euro and have been a proponent of the European free movement regulations (Hollander 2013). Next to this, the 'opening of the borders' for Polish (in 2007) and Bulgarian and Romanian citizens (in 2014) resulted in numerous governmental communications positioning the Netherlands as member-state in the EU. Besides, the Netherlands was one of the countries that co-authored a letter to the EC demanding attention toward 'negative side-effects' of intra-EU movement (Mikl-Leitner et al. 2013). 4 It makes sense to select the European and Dutch governmental level since they hold different perspectives on 'intra-EU mobility'. While the EU refers to 'free movement' as the 'mobility' of 'mobile workers', the Dutch national government refers to this as the 'migration' of European 'migrants'. The EU and the Dutch national state are also selected, since both levels have played a significant role in the current governance of intra-European movement (Sciortino 2000) and because of their administrative and legislative relevancy. 5

Data selection
One frequently used method in qualitative case study research is desk research of written material (Yin 1994). In this study, by a textually oriented discourse analysis, attention has  (Lakoff and Johnson 1980;Stone 1988;Yanow 1996) ! Analogies or examples

Storyline elements
Objectives Problem definitions and policy aims (Stone 1988;Linder 1995) Subjectives The ones, populations or fields to which problems and objectives are targeted Ingram 1993, 1997) Type of intervention Course of action implied by policy instruments (Douglas and Wildavsky 1982;Linder 1995) Policy proposals been put on key texts (Fairclough 1992;Sharp and Richardson 2001). A 'textual' analysis is inevitable to grasp the discourses within this topic. Regarding this desk research, governmental documents on intra-European movement were studied, produced by the EC and the Dutch national government as communications with the European Parliament and the Dutch Parliament, within a fixed period (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014). 6 Several strategies were applied to select the most relevant documents. On the European level, first the most relevant DG's were selected. By means of the organizational objectives listed at their websites 7 and by a first interpretative study of their policies (a document search on the EC website 8 by specific selection criteria 9 ) a top five of most relevant DG's has been prioritized. By this selection process, all official documents of the DG's Home Affairs, Employment, Justice, Freedom and Security (JLS) and Internal Market were selected online 10 by the search words 'migration', 'mobility' and 'movement'. By adding all 'hits', this resulted in 430 documents, allocated as follows: To overcome a qualitative in-depth study of all 430 documents, an interpretative selection step was needed. Therefore, a close reading of all document abstracts provided a qualitative indication if a document was about intra-European movement. This enabled to qualitatively separate the most relevant 'COM' documents (Communication from the Commission) from other documents on the basis of their irrelevancy. 11 This made it possible to boil down all documents to 15 in the end, which were in-depth interpretatively studied. Because of the size of these documents, this resulted in a large data file. 12 This procedure secured the selection process with the most relevant EC documents present.
On the Dutch national level, a comparative procedure was performed with some adjustments (because of a different context). Firstly since communications of all Dutch departments are less substantial as on the EC level, it was possible to search on all communications from government toward Parliament instead of selecting on beforehand on DG's or Departments. Next to this, online searches need to be applied by the number of a dossier. 13 On the basis of the search on the Dutch translation of the terms 'mobility', 'migration' and 'movement', 14 four dossier numbers 15 rose, of which '29407' is the most relevant ('free movement of employees from new EU member states'). This dossier included 332 documents (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014). Comparable with the EC document search, all 332 Dutch documents have been interpretatively studied by their abstracts and selected on their qualitative relevancy. This resulted in 53 documents, which also contributed to an extensive data file. 16 In this way, on both levels, a comprehensive and comparative data selection process was completed. First of all, by focusing on the communicative letters of the executive board (Commission and national government) to their controlling powers (Parliaments). Secondly, by completing a comparative selection process of all documents by the digital search engines of these authorities on three key words ('migration', 'mobility' and 'movement'). Thirdly, by selecting a comparative time frame (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014). Finally, by interpreting the relevancy of data by prescanning, since all documents included abstracts. And if the abstracts were not insightful enough, the documents were studied at hand. This made it possible to control a comparable selection process of both 'nested cases' within the topic of 'intra-European movement'.

Data analysis
All selected documents were printed and chronologically (based on publication date) analyzed. By a qualitative in-depth analysis of the data, all documents were coded by the items of the grid (Table 1). First by hand-coding and highlighting all relevant words, phrases and paragraphs in the documents that fitted the grid. Secondly, all coded elements were processed in a comprehensive Microsoft Excel matrix. The matrix allocated all data separated on their year of publication and their Dutch or European 'authorship'. This gave an overview of all poetic and storyline elements in all studied documents. By re-reading the matrix and checking the data context in the documents over and over again, an intensive interpretative analysis was applied which made it possible to distinguish certain policy discourses, and discursive shifts in the documents. By this approach, this type of research can be indicated as a matter of back-and-forth reasoning (Berg and Lune 2004) or abductive research (Yeung 1997;Danemark et al. 2002). By an abductive approach the researcher goes back and forth between theoretical concepts (the deductive grid) and the empirical findings. By doing this kind of discourse analysis, the researcher could delineate discursive shifts, depending on discursive coherence or variance in the empirical data (Healey 1999).
In line with the twofold aim of the research question, this article now firstly descriptively reconstructs the findings on governmental discourses 17 followed by the findings on policy proposals. After the presentation of both findings, an analysis section is included, investigating the performativity of these discourses on policy proposals. This is first structured by the EC case, followed by the Dutch case. This structure aims to chronologically unravel the formation of discourses and its policy implications throughout time.

Findings European discourse analysis
The analysis on poetic elements show that there are 'skills shortages', 'mismatches' and 'skills shortfalls' [concepts] on the labor market which need a removal of all kinds of 'labour market bottlenecks', 'barriers', '(cultural) obstacles', 'hurdles' and 'gaps' [concepts/metaphors] (CEC 2002, 72, 694;2004, 66;2007, 24, 773;2011, 248;EC 2013, 837;2014, 10). Next to this, citizens [micro-subject] must be equipped as 'human capital' to stimulate them as 'adaptive workforce', to challenge 'labour market pressures' and to overcome 'labour mismatches' [concepts]. Therefore, 'human resource development' and 'flexicurity' are needed, for a better 'talent pool' in order to make 'life long learning' possible [concepts] (CEC 2002, 72;2004, 66;2007, 773). Especially in order to 'counteract braindrain and brainwaste' and to promote 'braincirculation' [metaphors] (CEC 2011, 248). This is needed in order to do justice to 'the fundamental principle', the 'cornerstone', the 'Four Pillars' and 'four fundamental freedoms' of the EU of which the EC is the 'guardian' [concepts/metaphors] (CEC 2002, 72, 3). Therefore, the EC needs to intervene and take action because all numbers show 'low mobility levels' (CEC 2002, 72;2007, 24) and 'an absence of a mobility culture' [myth]. Since the 'mobility rate is lower than the rate of Third Country Nationals (TCN's)' (CEC 2002, 72;2007, 24, 773;2011, 248;EC 2014, 10). Because the share of intra-EU mobility is lower (2.8%) than the share of TCN's (4%), and since presented research shows that 'only 325.000 persons move annually, while 2.9 million would like to move'. This shows a 'mobility potential' that needs to be developed [numbers/concepts] (EC 2014, 10). Because 'the freedom of movement makes a positive contribution to labour markets throughout Europe' and 'it is a powerful and positive symbol of what Europe means for the individual citizen' (CEC 2011, 248) [myth].
Most of the poetic elements contribute to the storyline elements such as objectives targeted at more 'open', 'flexible', 'efficient', 'integrated' or 'accessible' labor markets (CEC 2002, 72, 694;2007, 773;EC 2014, 10). The overall objective is to: Promote active citizenship and social inclusion, and reduce inequality, but also to ensure the development of the Single Market and the successful integration of the European economy under the Economic and Monetary Union (EMU). (CEC 2002, 72, 6) The identified problems such as 'low geographic mobility', 'low occupational mobility' and 'fragmented information' [objectives] (CEC 2002, 72;2004, 66;2007, 773) are the reasons why European 'labour markets' [macro-subject] ('Single Market', 'European markets', 'EUlabour markets') and secondly, the micro-subject of the 'EU citizen' ('EU migrant workers', 'intra-EU mobile citizens', 'intra-EU movers', 'mobile workers') need intervention by the EC. This in order to create a 'more efficient allocation of resources', 'more integrated labour markets' and 'labour markets that are better able to adjust to asymmetric shocks' (EC 2014, 10). Therefore, information systems about job opportunities need to be better accessible for everybody (CEC 2002, 72;2004, 66;2007, 773).
Throughout all periods the EC discourse is quite consistent. This can be illustrated by zooming in on one of the 'four fundamental freedoms': the 'free movement of people', 18 which is mainly interpreted as 'the free movement of workers'. The Commission communicated about this: The free movement of persons as one of the fundamental freedoms guaranteed by Community law and includes the right to live and work in another Member State. Initially, this freedom was essentially directed towards economically active persons and their families. Today the right of free movement within the Community also concerns other categories such as students, pensioners and EU citizens in general [. . .]. Free movement is a means of creating a European employment market and of establishing a more flexible and more efficient labour market, to the benefit of workers, employers and Member States. (CEC 2002, 694, 3) The above citation explicates the 'free movement of economically active persons and their families' which primarily aims a 'more flexible and efficient labour market, to the benefit of workers'. This shift in focus is built up from a free movement of people considered as free movement of economically active persons considered as free movement of workers considered as a means to create flexible labor markets. This example shows the coherence of the EC discourse on EU citizens as workers, contributing to the overall aim of Europe as flexible labor markets. This example shows how discursive power works to 'perspectivise' something as something different. This perspectivation makes it hard to neglect alternative constructions. When 'free movement' is considered as 'a means of creating a European employment market', then it becomes very legitimate and persuasive to stimulate the flexibility and mobility within this market. By 'free movement' as the means in the construction of a market makes it legitimate to stimulate 'movement', and to problematize 'low intra-EU mobility' [numbers], the 'absence of a mobility culture' [myth] and persuasively develop the 'mobility potential' [concept] and more geographic and occupational mobility [objective]. This example underlines the action-oriented potential of certain concepts and myths creating a legitimate perspective to act upon.

Findings on European policy proposals
From 2002 onwards the EC presented an 'Action Plan for Skills and Mobility' (CEC 2002, 72). This Action Plan lasted for several years 19 to combat 'occupational mobility', 'geographic mobility' and to stimulate information and transparency to achieve 'open and more accessible labour markets' (CEC 2002, 72, 20). In order to reach more 'occupational mobility', the documents calls for a 'Europass framework' 'to support the transferability of qualifications' by a 'MobiliPass'. Next to this, to achieve 'lifelong learning', 'E-learning programs' are announced. And to get more 'geographic mobility' the 'Health Insurance Card' is announced to make social security and pension rights 'portable' (CEC 2004, 66;2007, 24). Next to this a 'Language Action Plan' and a 'Researchers Mobility Portal' are included, to 'strengthen mobility for education ' (2004, 66, 19) to enlarge the mobility of workers in general and researchers in particular.
In order to improve 'fragmented' information, the 'European Job Mobility Portal', Mobility Information campaigns and the modernization of EURES ('a cooperation network for exchanging vacancies and facilitating intra-EU labour mobility' as 'one-stop mobility information portal') are presented (CEC 2007, 24, 3;EC 2014). All policy proposals, aim to tackle the earlier indicated 'objectives' such as 'low occupational mobility', 'low geographic mobility' and 'fragmented information', stated as: Mobility generates social and economic benefits. Increased intra-EU labour mobility will widen employment opportunities for workers and help employers fill vacancies better and faster. [. . .] More integrated labour markets would enable the EU interdependent economies to better adjust to asymmetric shocks. [. . .] There has been a significant increase in the number of workers that indicate 'firm intentions' [. . .] to move to work abroad. [. . .] This represents a significant mobility potential and a challenge for the EURES network. (EC 2014, 10, 2) The above citation directly links the objective of 'integrative markets', with the subjectives of the 'labour market' and the 'EU citizens' as 'worker', the problem of both 'immobilities', the metaphor of 'asymmetric shocks' with the myth of a 'mobility potential' which legitimizes the launch or the improvement of EURES.

Analysis
The discursive findings show how a specific 'mobility' discourse is built up by the EC. First, it constructs 'Europe' as liberal labor markets and Europeans as economic individuals. It shows a liberal-economic discourse of a functional demand and supply of rational citizens on a Single Market. Secondly, the EC sees 'free movement' as the cornerstone of 'development of the Single Market and the successful integration of the European economy under the Economic and Monetary Union' (CEC 2002, 72, 6). By seeing 'Europe' as one geographically borderless and consistent zone, European citizens are one and undivided economic workforce, not segregated by their national cultures. And by Europe as 'Internal Market', European citizens can move freely as Europeans, not as migrants, maybe because 'mobility' is a more 'comforting alternative' (Carrete and Gasper 2010, 7). However, this perspective acknowledges movement of Europeans as 'mobility' contrasting with the movement of 'Third Country Nationals' as 'migration'.
From this 'comforting' perspective, not hampered by nation-state borders, European citizens should 'move', because their 'movement' prospers the development of 'human resources' and the 'Single Market'. European citizens hold a great 'potential' which needs to be fulfilled by their movement. But a significant discursive shift occurs when 'free movement' becomes a need to 'move' and 'the freedom of movement makes a positive contribution to labour markets throughout Europe' (CEC 2011). When 'Europe' is seen as a market the 'freedom of movement' is the instrument to stimulate that market, counteracting low spatial (social-geographic) and occupational (social-economic) mobility. These notions construct the European discourse as an evident 'mobility' discourse. This resembles with other research stating that European discourse is centered 'around mobility and migration as labour market tools for increasing flexibility and competitiveness' (Balch 2010, 175). Therefore, 'hurdles', 'obstacles' and 'skills shortages and bottlenecks' need to be combatted (CEC 2002, 72;2013, 837;EC 2014, 10). By analyzing the Commission's communication by its poetic and storyline elements, a delineated liberal-economic discourse about rational citizens, flexible markets and the necessity of free movement as instrument for this economic reality becomes visible: It is the joint responsibility of Member States and EU institutions to uphold the right to free movement, including by countering public perceptions that are not based on facts or economic realities. (EC 2013, 837,13) Most of the proposed policies (such as the Europass framework, the Health Insurance Card, the MobiliPass and the Job Mobility Portal) are legitimized by poetic elements. For instance, the Health Insurance Card needs to contribute to 'reducing time of social security claims' and to 'streamline administrative practices and cooperation' (CEC 2007, 24, 6). The Job Mobility Portal needs to contribute to 'more transparency', while the Europass framework, which includes the Mobilipass, needs to contribute to 'the transparency and transferability of qualifications' (CEC 2002, 72, 13). Like the EURES example has shown, most policy proposals are legitimized by earlier formulated poetic and storyline elements. These proposals show how discursive power works, when something is 'perspectivised' as something different. When 'free movement' is considered as 'a means of creating a European employment market', than interventions (such as the Europass framework, the MobiliPass and the Job Mobility Portal), which need to stimulate the flexibility of this market, seem legitimate. By analyzing the EC discourse highlights this 'perspectivation', and how a persuasive discursive reality is constructed, to legitimize certain policy interventions.

Findings Dutch governmental discourse analysis 20
First of all, in the beginning period 21 (2004-2007) regarding the 'opening of the borders' 'a postponement would do more harm than that it would protect the Dutch labour market' because of a 'tight labour market' [concepts]. In this regard the Dutch government wanted to make sure that 'disturbances' and 'repression' on the labor market are banned (Letter to Parliament 2004Parliament , 2005Parliament , 2006  The fundament of economic cooperation within the EU is consisted by the Internal Market. The goal of this it to achieve a European Economic Space wherein 'the four freedoms' are secured. The Netherlands has, as trading nation, large interests with a good functioning of that internal market which is free of unneeded trading barriers. (Letter to Parliament 2005, 21, 2) In this first period, identified problems are related to 'barriers', such as 'labour market tests' and 'labour permits', a 'minimum of bureaucracy' (Letter to Parliament 2004) 'bureaucratic romp lump' (Letter to Parliament 2006) and 'administrative burdens' (Letter to Parliament 2005) [concepts] which could harm secure welfare state accessions and economic prosperity. The Dutch government wants to regulate this new legislation smoothly to maximize profits for the Dutch economy, because they do not want to: 'Shoot ourselves in our own foot to hold a too restrictive policy. Especially, in a situation with a strong labour demand, a shortage could lead to a diminished effect on the economic development' (Letter to Parliament 2006, 2). This is mainly targeted at two subjectives. One latent group (the benefits of the Dutch population) and one manifest group, the migrants, referred to as 'CEE-employers', 'CEE-employees', 'CEE-landers', 'foreigners', 'immigrants', 'allochtons' or 'CEE-migrants' (Letter to Parliament 2004. Therefore, the government announces 'flanked policies' [intervention] of 'preventive and repressive instruments' (Letter to Parliament 2006, 53, 2) to improve adequate housing and to maximize profits for Dutch economy (Letter to Parliament 2006, 53, 6).
In the second period (2008)(2009)(2010)(2011), new actors and a change in tone of voice occurs. All kinds of 'shortages', 'inappropriate usages' and 'reverse sides' such as '06-busses' and 'explosive firms' occur [concepts/metaphors]. Simultaneously the urgency of numbers shifts. First, one could observe an under expectation ('no extra rise' in 2007), later this evolves, by matters of available estimations, to an over expectation of the population size. Next to this, new concepts arose (such as 'regional attachment' regarding 'homeless shelters', 'language demand' regarding integration and 'over occupation' and 'housing nuisance' regarding housing), since they do not want 'a repetition of history ' (Letter to Parliament 2008, 99) [myth] because: With the arrival of labour migrants from CEE countries, the parallel with the sixties and seventies forces itself, when also large extensions of groups came to the Netherlands. Than it was presumed, that they would stay for their durance of the demand of labour forces, and to return back to their land of origin. The consequences of this misapprehension are known. Extensive groups of migrants came in a position of backlashes, which influenced next generations. Now, after forty years, there are still efforts to overcome these backlashes. We can't allow that in a certain time again an extensive group stayed unnoticed and which came at large socio-economic distance. (Letter to Parliament 2009, 103, 4) This results in a strong socio-cultural focus on policy proposals regarding 'abuses' and 'reverse sides' because: 'it is unwanted when large groups of newcomers have an isolated existence' (Letter to Parliament 2009, 103, 4) and 'it must be prevented that they get social benefits and stay at the sideline' (Letter to Parliament 2011, 118, 7). The objectives change from an economic to socio-cultural 'bottlenecks', 'shadow sides ', 'problems' and'sideeffects' (Letter to Parliament 2008, 2013). Problems are mentioned such as 'mobile banditism', 'nuisance ', 'deterioration', 'homelessness', 'isolation' and'exploitation' (Letter to Parliament 2008, 2013) while 'abuses' and 'irregularities' are targeted at 'migrants' and actors such as 'malafide employment agencies' and 'slum landlords' as deviant target groups [subjectives]. Therefore interventions are proposed so that: '[. . .] the government and others can combat abuses quicker and better ' (Letter to Parliament 2008, 98, 7).
The final period (2011)(2012)(2013)(2014) can be introduced by: The free movement of workers is one of the most important pillars of the EU. One Europe with open borders is important for the Netherlands. The Netherlands has known after all always an open, international oriented economy. In general, labour migrants from other EU countries deliver a positive contribution to our economy. But there are also problems, and I don't want to close my eyes for that. (Letter to Parliament 2013, 162, 5) In the final period both discourses seem to merge, since economic metaphors ('The Netherlands as open economy') are combined with more socio-cultural concepts ('problems'). This is illustrative in this period, where legal-economic concepts ('unfair competition', 'level playing field', 'graying' and 'greening' of society, 'the best and the brightest', 'a race to the bottom' and 'the costs of labor migration') are more often combined with sociocultural concepts ('nuisance', 'integration' and 'deterioration') (Letter to Parliament 2013, 1). This is legitimized because: 'from the past we know that it is of large importance that migrants are straightaway entrained in the Dutch society. They have to know their rights and duties' (Letter to Parliament 2014, 187, 4) [myth]. Therefore, on housing, local governments cannot 'look away', 'close their eyes' or 'persist' any longer to overcome 'deadlocks' and 'NIMBY feelings' (Letter to Parliament 2012, 150;2013, 162) [concepts/metaphors]. And rising numbers play an important role to keep a strong-shared 'sense of urgency'. However, ironically, non-rising numbers do not delegitimize interventions, because: 'problems are not always findable in statistics ' (Letter to Parliament 2013, 177) and 'the unavailability of good quantified insights cannot be a cause to do nothing' (Letter to Parliament 2012, 150). In this final phase the objective is to combat social problems in order to stimulate economic profits. On subjectives, it shows a return to the macro-subject of the 'labour market', next to the micro-subject of the 'CEE-migrant' which is replaced after 2011 by 'EU-labour migrants' and 'EU citizens' next to 'non-economically active migrants', 'knowledge migrants' and 'mobile EU citizens ' (Letter to Parliament 2013-2014 [subjectives].

Findings on Dutch policy proposals
In the beginning (2004)(2005)(2006)(2007), most interventions aim to enlarge 'full free movement' (Letter to Parliament 2006, 44, 51, 53) to reduce all kinds of 'barriers', such as 'labour market tests' and 'labour permits'. Therefore 'bureaucracy' (Letter to Parliament 2004) 'bureaucratic romp lump' (Letter to Parliament 2006) and 'administrative burdens' (Letter to Parliament 2005) need to be minimized. Stated by: The Cabinet wants to take several measures which cause some guarantees to combat unfair competition in the (structural) situation of free movement of workers. Therefore it will take some flanked measures, especially concerning labour conditions and housing [. . .]. In this way there will be a controlled transition to the situation of full free movement. (Letter to Parliament 2006, 2-3) Policy proposals aim to maximize 'full free movement' and to reduce unwanted 'side effects'. Therefore 'flanked' policies of 'preventive and repressive instruments' (Letter to Parliament 2006, 53, 2) are announced to maximize profits for Dutch economy (Letter to Parliament 2006, 53, 6), such as governmental fines to reduce underpayment (Law on Minimum Hour Wage, WML), or incentives to maximize equal treatment for EU mobile worker and Dutch citizens (such as the Law labor conditions bordercrossing labor (WAGA) and the Law allocation labor recruitment intermediaries (Waadi)). Next to these laws and legislation to 'reduce barriers', numerous new interventions were announced to combat economic irregularities on the labor market. In that regard, the capacity of control agencies (such as the Labour Inspection, the Alien Police, the Social Intelligence and Investigation Service) will be extended. Next to this, to 'combat' economic irregularities an 'Intervention team Covenant' and an 'Approach Malafide Infrastructure' (AMFI, to combat slum landlords) are proposed. Thus by these 'flanked' policies, economic and legal proposals need to maximize profits and regulate the labor market.
In the second period (2008-2011) the tone of voice differs because 'we don't want a repetition of history' (Letter to Parliament 2008, 99). The freedom of movement remains unquestioned, but some 'problems' and 'adjustments' need to be taken into consideration: To lead CEE-migration into good tracks, the Cabinet announces a packet of measures, aimed to prevent or answer earlier mentioned problems [. . .]. The Cabinet takes the borders of European legislation into account, The Cabinet will make an effort to realize adjustment on a certain point of this legislation. (Letter to Parliament 2011, 2) Therefore, the 'Action Plan housing and integration labour migrants' is developed (Letter to Parliament 2008, 98) to stimulate housing agencies to develop more and better temporary housing accommodations. Besides, the 'Action Plan Nuisances and Deterioration' (2008) is developed to control and regulate unwanted neighborhood consequences related to this 'new' type of housing. Next to this, some repressive measures are proposed as well, such as the 'Regional Coordination centre Combatting of Fraud' (RCF), an adjustment of the 'Fraud Law' (Letter to Parliament 2011, 118) and the 'National Steering Group Intervention teams' (LSI) to combat 'abuses related to welfare provisions' (Letter to Parliament 2011, 118). And the 'Action Plan reduction Malafide recruitment Agencies' should deliver more control on temporary employment agencies. Thus, on a range of social and economic issues, action plans are proposed in this second period.
The final period (2011-2014) covers economic proposals to combat the evasion of minimum wages by self-employed employees with the 'Method façade independency'. This is related to a broad Ministerial approach 'Method Façade constructions' to 'be better able to do maintenance and control' on the work floor by a 'Identification pass'. Next to this, a revision of the 'Law Labour Aliens' and a 'Pilot Residence Termination EU citizens' are proposed to control irregularities on the work floor and to terminate the residence permit of irregular EU workers (Letter to Parliament 2013, 172, 181). Secondly, all kinds of social legislation were proposed to regulate the housing market better and to counteract housing shortages. 22 This was labeled as an 'integral packet of measures' (Letter to Parliament 2011, 118, 132) to 'streamline' administrative services to an 'intensive cooperation' of actions (Letter to Parliament 2011, 132). This resulted in the 'Operation Plan 2014-2015' to have a better cooperation between housing actors. Finally, new legislation is proposed to improve registration procedures, in order to equip municipalities to improve the efficacy of their policies toward Central and Eastern European populations.

Analysis
By analyzing the poetic and storyline elements, three distinct periods can be characterized in Dutch national discourses. The first period holds a strong liberal focus on the 'opening' of the borders by minimizing 'administrative burdens' and maximizing the impacts of 'the four freedoms'. Measures are mostly aimed to the most profitable timing implementing new EU legislation for the Dutch economy (Letter to Parliament 2004, 1, 3) and discussions mainly focused on whether labor market restrictions were appropriate. This is the discursive period when EU mobility is constructed as labor mobility. Indicative, proposals are termed as 'flanked' policies, because legislation should not hamper the free movement aims. Therefore, 'preventive and repressive instruments' are proposed to regulate labor market and welfare state claims. Therefore this first period can be characterized as a legal-economic discourse.
The second discursive period includes socio-cultural and legal issues on labor migration, such as 'abuses', 'reverse sides' and 'isolation'. When social 'problems' evolve, all kinds of 'Action Plans' (Letter to Parliament 2011) have been proposed to combat housing irregularities and fraudulent practices. The second discursive period shows how and why the Dutch discourse became a distinct migration discourse. In this period, all kinds of explicit and implicit parallels were made by referring to EU labor in terms of 'again', 'also' and to learn 'lessons of the past' (Letter to Parliament 2009. 23 By drawing explicit references to the recent (guest worker) past, Dutch authorities legitimized comparisons with migrant history, especially referring to Turkish and Moroccan migrants who migrated to the Netherlands in the 1960s and 1970s. By calling upon this historic legacy, political urgency could be put upon these 'mobile workers' as 'migrant' group. Eventually, this made it possible to connect 'mobile work' to a policy approach which needs to attack the 'misapprehensions', 'backlashes' and 'socio-economic distances' (Letter to Parliament 2009) from the past by early interventions because 'we don't want a repetition of history ' (Letter to Parliament 2008, 99). By this historic parallelization evolving from a 'guestworker syndrome' (Friberg 2012), authorities legitimately started to include all kinds of national presumptions about 'failures' and 'lessons' from that collective past. Consequentially, all sorts of 'action plans', 'pilots' and 'an integral packet of measures' became legitimate to counteract the 'reverse sides' (Letter to Parliament 2011) and 'lessons' of that past. By historical perspectivation the Dutch case study displays how and why Dutch authorities made it legitimate to see 'mobile workers' as 'labour migrants', and 'labour mobility' as 'labour migration'. This displays how the Dutch discourse on 'intra-EU movement' became a migration discourse, which evidently deviates from the European mobility discourse earlier outlined.
Finally, both discourses earlier identified get a close issue connection in the final period, merging within an 'integral packet of measures' (Letter to Parliament 2011). The final period starts off when discourses include both legal-economic and socio-cultural elements by connecting aims to stimulate the labor market with interventions to regulate housing issues and the societal participation of migrants. All kinds of laws were proposed to control housing issues better and to regulate irregularities on labor market, in order to combat 'shadow sides ' (Letter to Parliament 2013, 175) and to make EU migration 'maintainable ' and 'compliable' (Letter to Parliament 2013, 172, 174).
Thus, the analysis unravels three distinct discursive periods regarding 'intra-European movement' at the Dutch national level: from a more legal-economic discourse (2004)(2005)(2006)(2007), to a legal socio-cultural discourse (2008-2011) toward a merging of both in the final phase (2011)(2012)(2013)(2014). It also shows that most of the proposed policies are legitimized by poetic elements. For instance, the 'flanked policies' are legitimized by the metaphor to see the Netherlands as 'trade nation', therefore prescribing not to intervene too much in the liberal market. The 'Action Plans' in the second period, are directly legitimized by guestworker myths and 'lessons from the past', aiming for early interventions toward migrants. And the 'Pilot Residence Termination' in the final period is legitimized because people have 'to know their rights and duties ' (Letter to Parliament 2014, 187, 4) to combat 'shadow sides'. As such, the discursive power of poetic (and storyline) elements to 'perspectivise' something as, causes a persuasive discursive reality, which legitimizes policy interventions. The Dutch case shows the performative potential of discourses and displays contingent shifts throughout time.

Conclusions
The main question was: 'Which governmental discourses on intra-European movement can be identified at the European (EC) and Dutch national level and how did these discourses affect policy proposals (in the period 2002-2014)?' By a refined application of discourse analytical tools, a legal-economic discourse on EC level and a legaleconomic, legal socio-cultural and a combination of both discourses on the Dutch national level have been distinguished. To meet the aim to deliver theoretical contributions to the field of discursive policy analysis, the following conclusions are considered.
Firstly, this study shows that a refined operationalization of discourse analytical tools is not only theoretically relevant, but also holds empirical importance. This study critically examines the Hajerian approach and delivers empirical evidence that a refined deductive alternative contains relevant contributions to the field. The distinction between poetic and storyline elements enables an analysis of the interplay between both elements. Consequentially, this showed that poetic elements (such as concepts, metaphors, myths and numbers) have an important instrumental and legitimizing value for storyline elements and the overall discourse. This deviates theoretically from Hajer's account, and empirically displays the importance of poetic elements as building blocks, constructing a certain storyline with the 'right' value-laden or normative perspective. When market metaphors are used positioning mobility as labor mobility, economic objectives become more legitimate.
Secondly, this study displays the discursive necessity of poetic elements. Because, when poetic elements are used to problematize (which is mostly the case in governmental communications), this is mostly followed by a policy proposal or 'course of action'. It resembles the literature (Lakoff and Johnson 1980;Yanow 1996;Verloo 2005) and shows the mediating role of poetic elements, connecting a policy objective with a policy proposal by the construction of a problem as. Therefore, poetic elements are an important signpost and mark the direction of a proposal. This notion nuances the feed-forward effect and the prescriptive effect of policy discourse on policy proposals (Schmidt 2008(Schmidt , 2011Schneider and Sidney 2009). It is not the whole discourse that has prescriptive value for policy proposals, but it needs to be specified by the focus on poetic elements. This study shows the relevance and importance of poetic elements for discourse and policy analysis. 24 Thirdly, this study shows how performativity works. First by showing how poetic elements legitimize storyline elements. And secondly, by showing how (poetic and) storyline elements legitimize a certain policy discourse and its proposed 'course of action'. This 'two-step' distinction opens up the discursive black box of performativity and displays the 'performative potential' of discourses by the legitimizing role of poetic and storyline elements. For instance, at the level of the EU, repeatedly the myth is used that the EU has a 'traditional lack of mobility culture' so that the EU can be presented as fragmented, disintegrated, inflexible and immobile labor markets. Therefore, all kind of 'hurdles' and 'bottlenecks' need to be removed to become more integrated, flexible and mobile (CEC 2011) contributing to EURES as a legitimate proposal. Regarding the Dutch case, they use the myth that they do not want 'a repetition of history' (Letter to Parliament 2009, 103, 4). Therefore, it becomes legitimate to see mobile workers as migrants and focus on migrant 'problems', 'abuses' and 'reverse sides' known from the past. Consequentially, all kinds of 'Action Plans' are proposed, to learn from 'lessons' of the past (Letter to Parliament 2009Parliament , 2011. 25 Both concise examples highlight the discursive processes how a poetic element (myth) legitimizes storylines and a certain discourse and how a discourse legitimizes a certain course of action. These examples unravel how discursive legitimation and performativity works. It shows the construction of a certain discourse, creating a certain version of 'reality' that necessitates a particular action.
Fourthly, although it was not the main aim of this article to unravel the distinctive elements between mobility and migration discourses, it does shows the specificity between the European 'mobility' and Dutch 'migration' discourses. It displays how the EC understands, 'perspectivizes' and legitimizes 'intra European movement' as 'mobility'. Because of the historical construction of Europe as 'Single' or 'Internal Market', 'free movement' is considered as 'a means of creating a European employment market', which now became the means for further European integration. This contrasts with the Dutch case showing how historic references legitimized an approach of EU laborers as migrants. And by seeing this group as migrants this implied that all kinds of policy actions (in the sphere of 'integration' and participation) could be developed on the basis of insights and 'lessons' from the past. It shows the importance of the historical context to understand the specificities of both mobility and migration discourses to 'perspectivize' something as and how this has contributed toward legitimate courses of action.
This study also contains limitations. Since this article studied how discourses affect policy proposals, it had a minimal focus on whether the presence of a certain discourse did not affect policy proposals. As such, this study did not include many exceptions to the rule of performativity. This limitation could be an interesting focus for future research, to show to what extent elements have prescriptive value for policy actions and why not. From this point of view, it would be interesting to study why and when a certain persuasive discourse is not legitimate enough to cause a 'course for action', disentangling the conditions of discursive illegitimacy. Next to this, including more and different case studies would gain broader insights about the applicability of the research grid.
Finally, this study shows that the concepts 'mobility' and 'migration' are not just different words for the same empirical phenomenon, but rather display an institutionalist-discursive 'top of the iceberg' since they represent distinct institutional perceptions, interests and authorities. And from a more political-administrative perspective, this case study selection of the European and Dutch case shows the institutional tensions and mismatches between two authorative levels that deal with the same empirical phenomenon. The discursive perspective adds to our understanding of institutional divergences and deadlocks next to convergences and cooperation between the EC and member-states.
But especially, for migration scholars this study unravels the discursive and normative dimension in migration research, since 'migration' or 'mobility' are not only objective empirical phenomena. It shows that the usage of a certain concept ('migration' or 'mobility') can be linked to a certain authorized perspective on the empirical issue. This study shows that there is a political dimension observable in 'migration' and 'mobility' research, which deserves consideration in future research.

Disclosure statement
No potential conflict of interest was reported by the author.

Notes on contributor
Mark MAC van Ostaijen is a PhD candidate and lecturer at the Department of Public Administration at Erasmus University Rotterdam, the Netherlands. His research focuses on the governance of Central and Eastern European movement within the project IMAGINATION: http://www.markvanostaijen.nl. 24. Besides, an explicit distinction between poetic and storyline elements contributes to the replicability and internal validity of doing discourse analysis, and it unravels the instrumental interplay between both elements. 25. Hajer would have mentioned these constitutive myths, repeatedly used saga which functioned as an emblem (Hajer 2003