How to explain migration policy openness in times of closure? The case of international students in Switzerland

ABSTRACT Since the 1990s, Swiss immigration policies have placed increasing restrictions on non-European Union (EU) immigrants. However, in 2011, based on the initiative of Jacques Neirynck, the Swiss Parliament approved a law facilitating the admission and integration of non-EU nationals with a Swiss university degree. How can this policy openness in times of closure be explained? Drawing on the narratives of stakeholders during parliamentarian debates, and interviews with key political actors, we propose a unique explanatory approach combining: (1) the convincing narratives of steering crafted by parliamentarians, (2) an appropriate temporal and geographical context, and (3) the biographical capacity of the policy initiator to effect policy change. This model will be useful for studies of migration policy change in general.


Introduction
As the increase in globalised human mobility continues, contemporary states face a major dilemma: how to control migration flows while simultaneously attracting highly skilled workers? As a result of this tension, migration policies are becoming increasingly selective. The term 'designer immigration' (Hawthorne 2012) refers to policies that control the selection of migrants according to how their skills may be beneficial for the knowledge economy. Using this line of reasoning, international students are often viewed by governments as ideal immigrants, not only because they can fill highly skilled labour gaps that many Organisation for Economic Co-operation and Development (OECD) countries face, but also because their assumed cultural assimilation into the host country is seen as having a positive effect on their labour market performance compared to skilled immigrants from overseas (Robertson 2011;Chiou 2017).
In the context of the global race for talent at the onset of the twenty-first century, Anglo-Saxon countries put policies in place to encourage international students to remain after completion of their tertiary studies (Bedford and Spoonley 2014;Geddie 2015;Chiou 2017). Such policies, also known as 'study-to-work' policies (Mosneaga 2015), 'student switching' (Robertson 2011) and 'two-step migration' (Hawthorne 2012), were implemented much later in the European Union (EU). In 2009, the Council of the European Union introduced the blue card, which aimed to make Europe a more attractive destination for highly educated persons from outside the EU. The EU blue card, however, has remained relatively unattractive compared to the policies that individual member states implemented to attract international students (EUbusiness 2016). As the conditions of study-towork transition policies in the EU differed from country to country, in 2016 the European Parliament approved a directive harmonising entry and residence rules in order to facilitate an easier path for students and researchers from non-EU countries. The directive entered into law the day after its publication (22 May 2016) in the European Official Journal. Member states now have two years in which to enact individual national legislation. Currently, students are allowed to stay at least nine months after completing their studies in order to seek employment or start a business, and have the right to bring family members to Europe who are entitled to work during their stay (Official Journal of the European Union 2016).
In Switzerland, international students also became a favoured source of highly skilled migrants as policies facilitating their admission and residence came into effect in 2011. Following Jacques Neirynck's initiative in 2008, the Swiss Parliament approved amendments of the Foreign Nationals Act, and of the Regulation on Admission, Residence and Employment concerning non-EU nationals, that can be summarised as follows: (1) international students are no longer required to submit a declaration confirming their intent to leave Switzerland immediately after completing their tertiary education; (2) graduates of Swiss universities are provisionally admitted to stay in Switzerland for six months after graduation to seek employment matching their qualifications; (3) graduates are not subjected to the priority rule giving Swiss and EU citizens priority over jobs if the prospective employment is of 'high scientific or economic interest' for Switzerland; and (4) the years spent by graduates studying at a Swiss university may be counted retrospectively to obtain long-term residence status (Federal Office for Migration (FOM) 2010; Vaitkeviciute 2017). These changes represented a significant opening-up of migration policy differing from the former restrictive policies aimed at avoiding competition from non-EU nationals. How can this policy liberalisation be explained? Surprisingly, there are no empirical studies answering this question (Riaño and Piguet 2016). Despite the increasing global trend of implementing policies that encourage the retention of international students upon completion of their studies, an empirical understanding of how these political debates evolveand why they ultimately result in policy change (or not)remains scant (Hawthorne 2012;Haugen 2013;Mosneaga 2015). Overall, studying why migration policy changes from a restrictive to a more liberal stance remains underexplored (Menz 2016). This paper aims to fill a part of this research gap by achieving a deeper understanding of why policy openness in times of closure occurs.

Switzerland as case study: policy openness in times of closure
Switzerland is an interesting case study for the analysis of study-to-work policies for international students for four reasons. First, at nearly 25% of the total resident population (FSO 2016), it has the second-largest foreign-born population among the OECD countries and depends heavily on highly skilled foreign labour. Switzerland is a small country, and has experienced a deficit of skilled workers since its industrialisation in the nineteenth century. Second, nearly 31% of all tertiary students (including Bachelor, Master and PhDs) enrolled in Swiss universities in 2016 had obtained their secondary education abroad and were foreign nationals (FSO 2017). This represents the second-highest share of international students of all OECD countries after Luxembourg (44%; 2016). The numbers of international students increased more than threefold between 1990 and 2016 (from 9200 to 33,000), corresponding to an increase in percentage among the entire student population from 12. 8% in 19908% in to 30.7% in 20168% in (FSO 2017. Third, unlike Anglo-Saxon countries, where international students have become a highly profitable industry, most Swiss universities do not charge an international student fee and tuition fees are already low because of generous state funding. This suggests that studying international student mobility in Switzerland may offer an alternative understanding to the marketisation perspective, currently used for English-speaking countries. Fourth, as opposed to Anglo-Saxon countries that have traditionally facilitated the permanent settlement of migrants and their naturalisation as citizens, Switzerland has been reluctant to recognise itself as a country of immigration and to grant citizenship rights to immigrants. Since the 1930s, Switzerland's immigration policies have largely been oriented around a protectionist narrative giving this paper an interesting case study with which to examine the shift in immigration policies to a more liberal approach.
Switzerland's immigration regime is currently characterised by dual foreigner's rights. In 2002, a bilateral agreement between Switzerland and the EU came into effect that gave EU nationals the same living and working rights as the Swisswith the exception of voting rights. Since then, legislation regarding foreigners no longer applied to EU citizens but only to individuals from countries outside the EU, known as 'third-country nationals'. In 2008, the Foreign Nationals Act came into effect, which restricted the entry of third-country nationals to highly skilled individuals. International students from third-countries were only allowed to stay in Switzerland until they had completed their tertiary degrees. There was no direct path allowing them to seek employment and apply for a work permit after graduation. However, in exceptional cases, students who were considered to be of 'high scientific value' were granted work permits.
In the years following the implementation of this law, academics and entrepreneurs in Switzerland expressed dissatisfaction about its seemingly arbitrary application and the difficulties they experienced when trying to hire students from third-countries. In April 2008, Jacques Neiryncka naturalised Belgian immigrant, a parliamentarian from French-speaking Switzerland, and a professor of electrical engineering at the Swiss Federal Institute of Technology in Lausanne (EPFL)submitted a draft bill to the Swiss Parliament to modify Article 27 of the 2005 Foreign Nationals Act. The aim was to 'facilitate the admission and integration of third-country foreigners who have graduated from a Swiss university' (Curia Vista 2017). However, this idea was not entirely new. Between 2000 and 2007, Neirynckalong with two other parliamentarians representing French-speaking cantons, Luc Barthassat and Didier Berberat, submitted three motions to Parliament to facilitate the hiring of foreign PhD students. All three parliamentarians were critical of the increasing influence of the nationalist Swiss People's Party in shaping restrictive immigration policies. Although the earlier initiatives helped to pave the way, Neirynck's proposal was unique in that it: (a) included all foreign graduates (not just PhD students), (b) was co-signed by representatives from all political parties and linguistic regions, and (c) contained a wider range of persuasive arguments than earlier proposals.
What is the process of submitting and approving a draft bill in Switzerland? The powers of the state are divided into three independent branches: the executive, the legislature, and the judiciary. The Federal Councilthe supreme executive and governing authorityis composed of seven power-sharing federal councillors elected by the Swiss Parliament. Legislative power is vested in the Swiss Parliament where the most important political decisions are made. However, the last word on the country's laws often rests with the people. The Parliament passes new legislation and amendments and is composed of two chambers: the National Council, which provides representation based on population size, and the Council of States, which has a fixed number of seats per canton. As the National Council and the Council of States have equal power they must both agree before a law is passed (Swiss Confederation 2017).
The 246 members of parliament (MPs) represent the interests of different language communities (German, French, Italian, Romansh) and political parties in Switzerland. Any member of the National Council or the Council of States can submit an initiative to introduce a new law, amend an existing law, or add a new provision to the Constitution. Given the volume of legislation MPs handle, it can be difficult to reach an informed opinion on any given issue. Therefore, before a draft bill comes before Parliament, it is discussed by parliamentary committees with the goal of agreeing on a common position, which can then be supported by MPs. The committees are able to discuss legislation in detail, clarify specific issues, hear experts from the administration or from the interest groups concerned, and address questions to federal councillors. They also serve as a testing ground to see whether a consensus can be reached for certain proposals across party lines. Draft laws are debated up to three times by each chamber to make sure that identical decisions are reached. Although this can be a difficult process, each chamber is often ready to make compromises and adjustments. By the time a draft bill is debated in Parliament, most of the work has already been done by the committees (Swiss Confederation 2017). Accordingly, the political committees of the National Council and the Council of States met between 2008 and 2010 to discuss Neirynck's draft bill to facilitate the admission and integration of non-EU graduates from Swiss universities. These meetings included the preliminary examination and subsequent vote to proceed with the initiative in both chambers of Parliament (2008), followed by hearings with stakeholders (students, university and administration representatives), the debate in both chambers on a preliminary draft (2009), and the resolution of differences and final approval by the Swiss Parliament (2010). Finally, the new law was enacted on 1 January 2011. However, the new law was not widely communicated to the broader public by the federal administration. The Swiss Federal Institute of Technology in Lausanne assumed the role of disseminating the information to their own students, presumably in a bid to increase their own attractiveness with prospects of future work permits (canton administrator, personal interview, 21 December 2016). This exemplifies an increasing trend whereby universities are becoming important actors in selecting and attracting immigrants (Brunner 2017).

Studying migration policy change: the role of narratives of steering
Birkland states that public policy involves 'the decisions (including both actions and non-actions) of a government or an equivalent authority' (as cited in Sabatier and Weible 2014, 9), such as law, regulations, statutes, executive decisions, and government programmes. How, then, should policy change be theoretically addressed? The role of ideas in shaping policy-making has gained scholarly attention in recent decades (Bleich 2002;Schmidt and Radaelli 2004;Menz 2016). Cairney (2012, 15-16) states that 'policy-making is not just about people exercising power to pursue their interests. It is also about the role of ideas'. Menz (2016, 627) takes this further by stating that, 'policy actors instrumentally use discourse to shape the policy debate, influence the agenda and legitimise certain policy choices'.
However, this ideational approach has received limited attention from migration scholars. For instance, Menz (2016, 627) recognises that it offers 'considerable analytical leeway in accounting for change, something other institutional approaches struggle with'. Similarly, Boswell, Geddes, and Scholten (2011, 4) draw attention to 'the potential for narratives to provide a coherent and compelling account of complex phenomena, in a way that can engender support and motivate action'. They stress the importance of exploring how 'policy narratives' or 'narratives of steering' constructed by diverse types of actors influence immigration policy agendas. Their concept of narratives of steering refers to 'the factual beliefs espoused by policy-makers and others engaged in political debate about the causes and dynamics of the problems they are seeking to address, and about how policy could impact these dynamics'. According to them, 'the success of a narrative in influencing policy agendas depends to a large extent on its consistency and, through this, its capacity to identify, define and constitute a policy issue or problem' (Boswell, Geddes, and Scholten 2011, 4).
In this paper we focus on the argumentative power of narratives of steering to convince policy elites of the need for policy change rather than on their cognitive underpinnings. Following this conceptual framework we have formulated the following research questions: . What narratives of steering were used by Neirynck in the Swiss Parliament to gain support for the initiative? . What narratives of steering were used during parliamentary debates by parliamentarians and stakeholders to support or oppose Neirynck's initiative? . Why did the narratives in favour of accepting Neirynck's initiative find broad consensus among Swiss Parliamentarians?

Methods: text analysis of parliamentary debates and policy elite interviews
The research questions stated above required a qualitative approach to capture the narratives, representations, and the political dynamics. The two methods used were analysis of the minutes of parliamentary debates and policy elite interviewing. The minutes of the debates conducted over the 2008-2010 period by members of the National Council and the Council of States were analysed to grasp the narratives mobilised in favour and against the bill. In Switzerland, all parliamentary debates are recorded and transcribed in detail, constituting an invaluable source of data. These minutes are in German and French and are made available upon request for scientific purposes. The final report prepared by the Political Commission of the National Council (PCNC 2010), which summarises the parliamentary debates and the reasons for accepting Neirynck's initiative, was also examined. The Curia Vista database (2017), which contains a summary of parliamentary discussions since 1995 and can be consulted online, provided additional information on the timeline of the bill, and on similar motions previously submitted. Furthermore, an in-depth interview (in French) was conducted with Jacques Neirynck regarding his motivations and views on the political process. This was complemented by an interview (in French) with the head of the Department of Employment in canton Vaud inquiring about how the policy change was communicated to the cantons and to the broader public. The minutes of parliamentary debates, as well as the interviews, were analysed using deductive coding (i.e., themes from the conceptual framework) and inductive coding (i.e., themes emerging from the minutes of the parliamentary debates and the interviews) techniques of qualitative text analysis. All statements having the same code were grouped together in order to establish the main strands of narrative running through the text. Subsequently, we analysed each one of the narrative strands against and in favour of the initiative.
Although access to political elites for interviewing generally involves many challenges (Rice 2010), Jacques Neirynck was open to being interviewed at his home. We combined biographical and semistructured techniques for the interview, beginning with questions relating to the history of the initiativefrom the moment when the problem was identifiedto the acceptance of the initiative by the Swiss Parliament. This biographical approach proved successful as he was able to recap the sequence of events. This was followed by semi-structured interview questions. A total of nine topics regarding the initiative were discussed, including: how the idea began; to what extent it adds to the existing law; how support was gained from other Parliamentarians; the relatively moderate opposition during parliamentarian debates; canton Zurich's opposition during the consultative procedure (considering the high percentage of highly skilled workers concentrated in this canton); how to explain support for the initiative in a political system characterised by restrictive migration policies; feedback on the implementation of the law; and the future of the initiative during a severe shift to the political right.

Narratives of steering by the initiator of policy change
What was Neirynck's motivation to initiate the policy change? During the interview, he described his personal experience as a professor and parliamentarian as motivating factors: My motivation is my personal experience.
[…] When one is a parliamentarian, people who have problems contact you. And there were several foreigners who had studied here, who had graduated here, who were asked to leave within fifteen days. And, indeed, the existing Foreigners Act stated that a foreign student had to sign his commitment that he was to leave Switzerland once he obtained his diploma [describes case studies]. There was also a Tunisian who had forged a career at the EPFL, who had a doctorate, who had worked as a research assistant, etc.; he was a computer scientist and the University of Applied Science and Arts Western Switzerland was lacking computer scientists, and so he was appointed. But the Canton refused to give him a work permit, a residence permit. So that's my motivation! From that time onwards, I put this parliamentary initiative together, on the basis of a text that was written by a lawyer from Freiburg who handled such issues. (J. Neirynck, personal interview, 15 April 2016) At the time, Neirynck was one of only two professors at the Swiss Federal Institutes of Technology who were also parliamentarians, and therefore in a key position to receive feedback from international students and companies experiencing difficulties obtaining post-graduation work permits from immigration authorities. Thus, while presenting his initiative before the political commission of the National Council, Neirynck introduced himself as a 'representative of the students associations, academic institutions and economic circles' (Minutes of Swiss Parliamentary Proceedings 2008, positioning himself as a bridge between these groups. A vital personal point revealed during the interview was his arduous road to becoming a Swiss citizen, an experience that led him to conclude that immigration authorities are not able to adequately judge the 'scientific value' of immigrants: I was hired [as an EPFL professor] in Switzerland as a Belgian citizen, but it is interesting to note that I was subject to the jealousy of the people who made the decisions at the cantonal services regarding my naturalisation, and instead of taking 12 years to become naturalized it took 24 years, with two refusals. So, the problem is that some decisions important for the scientific development of the country are made by officials who have absolutely no idea of the consequences of their decisions. (J. Neirynck, personal interview, 15 April 2016) Now we examine the narratives of steering used by Neirynck before Parliament to legitimate changing the existing immigration law. His main argument was that 'the law as it stood was a suicidal act that was contrary to the [economic] interests of the country' (Neirynck 2016), and that Switzerland's lack of highly skilled personnel required a shift in policy that attracted international students and encouraged them to stay: To stay on top of the economy, Switzerland must develop expertise in specialized fields, ranging from biotechnology to wealth management. This implies massive recruitment of technicians, engineers, researchers, and financial analysts, well beyond our national borders. The world beyond the European Union is an indispensable recruiting pool for a small country such as ours, constrained by a declining demography. Switzerland would benefit considerably by attracting young talent, by training it according to its norms and by doing everything to encourage them to stay in our country. Unfortunately, Article 27 of the Foreigners Act does exactly the opposite. (Minutes of Swiss Parliamentary Proceedings 2008 By preventing foreign graduates from staying in Switzerland after graduation, Neirynck repeatedly argued that the Foreigners Act was effectively wasting public funds used to train foreign students, causing a loss of valuable human resources trained in Switzerland, to foreign economic competitors: The cost of training an engineer, a doctor, a scientist is between half and one million Swiss francs. Section 27 of the Foreigners Act costs us therefore, at the very least, some tens of million of francs every year. But this public money, lost for Switzerland, is not lost for everyone. Because of this, Switzerland provides to our competitors abroad highly qualified personnel in the fields of industry, medicine, banking, and insurance for free. (Minutes of Swiss Parliamentary Proceedings 2008 Furthermore, he lamented that Swiss administrators failed to recognise that highly skilled graduates are potential entrepreneurs who contribute to the Swiss economy through job creation: Another argument put forward by the administration relates to the fact that it fears an influx of foreign students who would eventually settle in Switzerland to become unemployed. For the administration, the number of jobs in Switzerland would remain the same, independent of the quality of the people involved. Any new residence permit pushes a Swiss towards unemployment, especially if the foreigner is well qualified. It is thus an individual that needs to be a priori excluded. This argument ignores the de facto situation: a highly skilled graduate creates jobs, often by launching his own start-up. In the Ecublens Science Park, there are currently 120 start-ups in demand. I remind you that two-thirds of our researchers are foreigners. So, are we going to kill this branch of development? (Minutes of Swiss Parliamentary Proceedings 2008 Finally, another key point in Neirynck's argumentation is that students from non-EU countries will not return to their home countries to serve as agents of economic development but will rather go to Switzerland's economic competitors: We must refute at the outset an argument often invoked that foreign students are trained in Switzerland with the sole purpose of returning home to contribute to its development. This argument is misleading since nothing obliges a young scientist to return to his home country. Young researchers from developing countries do not return because it is impossible for them to exercise their profession. They often move to the United States. In addition, some students, especially the Chinese, do not come from countries in development, but from developed countries, that compete with Switzerland. Thus, Switzerland trains many managers for the global economy, and spends public money for this purpose, but does not allow itself to recruit them and to profit from such an action. (Minutes of Swiss Parliamentary Proceedings 2008 What kind of narrative does Neirynck construct about international students? It is clear that students represent a way to increase Switzerland's economic advantage in a globalised world. However, Neirynck's narrative of profit and economic advantage, which inevitably follows the liberalisation of immigration law, is rather narrow. Students are imagined as young, mobile, trained in fields such as engineering, biotechnology, medicine, and wealth management, and fully assimilated into Swiss culture. Moreover, he only applies the masculine form when referring to international students, a feature that is also characteristic of parliamentarian debates.
Neirynck's narrow representation of international students is reflected in the highly selective wording of the final legislation. Only students whose prospective jobs are of particular scientific or economic interest to Switzerland are able to obtain a work permit in Switzerland. Scientific and economic interest is defined as follows: [activities of sientific interest ] 'are, as a general rule, activities in the fields of applied research and product development, applying new technologies, or using the acquired know-how in activities of high economic interest' [...] a 'high economic interest may be present if there is a well-defined need in the labour market for the completed discipline of study, which is highly specialized and corresponds to the job description, and if filling the position directly results in creating additional jobs or new mandates for the Swiss economy'. (FOM 2010, 3) This wording seems to fit Neirynck's view that students of economic value are above all engineers, bio-scientists, doctors, and wealth managers.
Finally, Neirynck's narrative about students not returning to their home countries illustrates a shift in views about students from non-European countries that is present in much of the parliamentarian debate. Until the 1980s, Swiss development policies viewed foreign students as agents of innovation for so-called 'countries in development' upon their return. Swiss universities were thus expected to play a capacity-building role. This view, and the concern over 'brain drain', a term prominently used in public discourses in the 1980s, is, however, no longer present in the twenty-first century. Foreign students are now primarily viewed as agents of innovation for Switzerland's economy.

Narratives of steering by parliamentarians and stakeholders
The parliamentary debates concerning Neirynck's draft bill were shaped by supporting and opposing voices. Opposition came early in the debate from representatives of federal and cantonal administrations using a protectionist logic claiming that native Swiss workers would be at risk of unemployment. This narrative was used over subsequent hearings, however, the majority of parliamentarians agreed to proceed in favour of the draft bill. Overall, 15 out of 23 cantons and 4 out of 5 political parties represented in the Federal Parliament backed the initiative. Only one party, the far-right Swiss People's Party, opposed it. Finally, when the draft bill came before the Swiss Parliament in 2010 for the final vote, it was approved by nearly 80% of parliamentarians at the National Council and by 100% at the Council of States.
The minutes of the parliamentary debates show that, in terms of procedure and language, political debates in Switzerland are disciplined and respectful. Debates are preceded by an introduction of the issues by the elected president of the political commission. This is followed by opinions and questions from the participants and concluded by the commission's president who presents potential alternative actions. A vote on these alternatives follows and the session is subsequently finished. This disciplined debating culture is clearly an expression of Switzerland's democracy, largely based on consensus, which makes Switzerland a particularly interesting context for studying political discourse and policy-making processes.
Which narratives of steering did parliamentarians and stakeholders use in an attempt to direct the opinions and voting behaviour of MPs either in favour of or against the draft bill? We start with the arguments against. As previously mentioned, opposition was mainly expressed by representatives of the federal and cantonal administrations using a protectionist-oriented narrative of steering. In their view, since 'it is imperative to protect native Swiss workers from foreign competition and unemployment, the current law should not be changed' (Minutes of Swiss Parliamentary Proceedings 2008.
Accordingly, a representative of the FOM argued that 'the new law represents a change of paradigm in Swiss immigration policies because, so far, the principle has been that native workers should not face competition from third-country nationals' (Minutes of Swiss Parliamentary Proceedings 2008. Thus, in order to protect native workers from unemployment, the existing legislation should not be changed. Another FOM representative maintained that 'Swiss cantons are interested in having a strong economy, however in times of upheaval, they do not want to have foreign students under their social budget'. In the representative's view, 'universities never follow economic needs and fluctuations, and they want to hire students even when the economy is not doing well' (Minutes of Swiss Parliamentary Proceedings 2008. Thus, it was better to keep the current law's flexible, case-by-case system, which protects Swiss cantons from having to support unemployed foreigners when the economy is in decline. The same argument was made by one of the members of the Swiss Federal Council, pleading for a demand-driven approach, claiming that 'allowing students to stay and seek employment would depart from well-established principles of controlling foreign access to the labour market on a case-by-case basis' (Minutes of Swiss Parliamentary Proceedings 2008. Finally, a representative of the labour market authority in canton Zurich argued that 'the canton's economy needs foreign skilled personnel, particularly in the areas of information and communications technologies, medical technology, engineering and construction, and financing and advisory services, but the European Union provides sufficient and highly qualified personnel for those needs' (Minutes of Swiss Parliamentary Proceedings 2008. The representative also expressed concern over 'creating a privileged group over other highly skilled migrants from non-EU countries without a Swiss degree, which complicates the administrative procedure of issuing work permits' (Minutes of Swiss Parliamentary Proceedings 2008. Which narratives of steering were used by stakeholders and parliamentarians in favour of the draft bill? Analysis of the minutes of parliamentary debates show three main narratives of steering: (a) Switzerland's image suffers under the current law, (b) Switzerland needs to be competitive in a globalised economy and take advantage of global opportunities, and (c) foreign students trained in Switzerland have the advantage of already being culturally assimilated.
We start with the arguments expressed by stakeholders during the hearings. A representative of the Network for the Dialogue between Science and Politics argued that the message is quickly passed among foreign students that Switzerland is very restrictive, which damages the country's image. For these students, the law seems so strict and unavoidable that many would assess the situation as being hopeless and wouldn't even consider applying to stay. Thus, instead of going to the Federal Institutes of Technology in Zurich or Lausanne, these students would go to Berkeley. (Minutes of Swiss Parliamentary Proceedings 2008 Moving on to the arguments of debating parliamentarians, an MP argued that students who do not stay in Switzerland would inevitably become competitors. Their argument highlighted the fact that students from outside the EU now come from countries that are Switzerland's commercial competitors, particularly the BRIC countries (Brazil, Russia, India, China). When these students return to their countries of origin they will be hired by competitor firms and thus Switzerland has given these firms the tools to win over us. (Minutes of Swiss Parliamentary Proceedings 2008 Another parliamentarian argued further that Asian students trained in Switzerland could be used as key agents for Swiss companies in order to take advantage of global economic opportunities. As an example, the argument was presented that in the canton of Vaud, a large Swiss informatics company has a subsidiary in Vietnam. It would be of more interest for the company to employ Asian students (rather than Europeans), especially Vietnamese, in order to establish a good relationship with the subsidiary in Vietnam. (Minutes of Swiss Parliamentary Proceedings 2008 Finally, the idea of capitalising on the acquired cultural assimilation of foreign graduates, was presented by another parliamentarian who posed the question 'whether it was not to Switzerland's advantage to hire a student from outside of the European Union, who has been trained in the Swiss system, rather than a European student trained in France, Germany, or Spain?' (Minutes of Swiss Parliamentary Proceedings 2008. Overall, the protectionist approach defended by Swiss administrators did not resonate with parliamentarians. As stated in the Report of the Political Commission of the National Council (PCNC 2010), the commission supported the liberalisation of the Foreigners Act for the following reasons: first, it noted the considerable differences in how the individual cantons implemented the Foreigners Act, often to the disadvantage of international students and to the detriment of Switzerland's position as a powerful educational and economic centre. Second, the former law made it difficult, and sometimes impossible, for third-country studentswho represent, according to them, well over 30% of all foreign students in Switzerlandto remain in Switzerland after completing their tertiary education. Consequently, the country failed to keep highly skilled specialists who, instead of taking employment in Switzerland or creating their own business, went elsewhere. Third, educational costs in Switzerland were heavily subsidised by the federal and cantonal governments. Instead of profiting from this investment, Switzerland's competing economies were the beneficiaries. Subsequently, the Foreigners Act was changed in 2011 to facilitate the admission and labour market integration of third-country nationals with a Swiss university degree.

Explaining policy openness in times of closure
How can this policy-shift towards openness in times of restrictive immigration policies be explained? We propose three complementary explanations: the role of narratives of steering, the role of the geographical and temporal context, and the role of the biographical capacity of the policy initiator.

The role of narratives of steering
Asked in the interview why he believes his initiative succeeded, Neirynck responds that 'good sense prevailed'. His economic narrative states that in the global competition for highly skilled workers, Switzerland ultimately loses out on this valuable pool of human resourceswhich it has paid to educateto competing economies. This narrative of economic benefit appealed to parliamentarians from different political orientations. Also, his argument fulfils Boswell, Geddes, and Scholten's (2011) proposition that for narratives to succeed in influencing policy agendas they need to be plausible, compelling, and resonate with perceived interests. Moreover, Neirynck uses real case studies and dramatic statistics on public spending to persuasively argue that the former law not only affects third-country students, but ultimately Switzerland's global position as an educational and economic centre. He describes this particular immigrant group as unproblematic, highly specialised, trained according to Swiss cultural norms, and being in high demand in the labour market. By representing them as young, innovative entrepreneurs, he plausibly highlights their potential contributions to Switzerland's global competitiveness. In conclusion, Neirynck's narratives of steering are more persuasive than the protectionist narrative used by representatives of the administration.
In examining the recent liberalisation of labour migration schemes in Europe, Menz (2016, 628) also observes how policy elites successfully use the global competitiveness narrative by claiming that a 'so-called global battle for brains is underway, in which countries with restrictive migration policies lose out in gaining access to an otherwise readily available global talent pool'. The success of such narrative can also be interpreted in Cerny's (1990, 225) sense that contemporary states have become 'competition states' whereby political actors try to capture global economic potentials.
However, the Swiss case is unique in that the main narrative of steering is not simply about global economic competitiveness but also about a return on public funds investments. As previously stated, the Swiss state heavily subsidises the university costs of international students providing Neirynck with a compelling argument that Switzerland loses the brains it has trained to its international competitors. Moreover, in a federalist country where each canton is governed semi-autonomously, the narrative concerning disparities of practice within these regions finds particular resonance among Swiss MPs in enacting policy change.

The role of the geographical and temporal context
We propose that the effectiveness of narratives of steering to effect policy change largely depends on how favourable the geographical and temporal context is. Temporally speaking, global and local events had a positive effect on policy change, for example: the globally circulating narratives at the outset of the twenty-first century about the need to attract international talent to be globally competitive (Geddie 2015); the regulations implemented by the EU in 2009 to facilitate the post-graduation retention of third-country students; and the motions that Neirynck and other MPs had previously submitted to liberalise immigration policy effecting international students. Moreover, the debate on depleting poor countries of their best and brightest seemed to no longer concern Swiss policy-makers, which worked in favour of the argument to retain international students. Also, in 2008, at the time the initiative was introduced, the steady increase in the numbers of EU migrants resulting from the 2002 Agreement on the Free Movement of Persons with the EU was not yet visible to the Swiss public.
Looking at the geographical context, Switzerland is a small countryboth in size and population. The elected MPs only serve on a part-time basis, as many work as entrepreneurs, academics, and trade unionists, for example. This particular context allows initiators of policy change, such as Neirynck, to be in close contact with key stakeholders, and to know, first hand, the interests of the groups they represent. Moreover, Switzerland's political system based on transparency, stakeholder participation, disciplined debate, and political consensus make it a uniquely favourable context for a powerful narrative of steering to succeed.

The role of biographical capacity
The concept of biographical capacity, which we define as the biographical experiences equipping a person with the knowledge and personal contacts to occupy a legitimate position to push for change, further explains why policy change can occur. Occupying both roles of university professor and MP, Neirynck was in daily contact with international students sharing their negative experiences in obtaining a work permit. This enriched his understanding of the issues and motivated him to represent their interests in the Swiss Parliament. Furthermore, as a former immigrant, he was in a good position to understand the arbitrary decision-making by immigration authorities. Moreover, he had spent over 20 years in Parliament, which gave him an in-depth knowledge of the system, along with personal contacts with other MPs. This helped to further facilitate his in-house political support of the initiative. Embodying the double role of academic and experienced MP gave him the credibility required to argue before Swiss parliamentarians. Finally, through his life and work experiences in Europe, Africa, and the United States, he developed the conviction that research knows no national borders and promoted the idea of 'research without passport', as stated in the interview. Thus, as an EPFL professor, parliamentarian, former immigrant, and world citizen, Jacques Neirynck had specific motivations and legitimate expertise with which to successfully defend his initiative in Parliament. Biographical capacity is an invaluable tool when analysing and seeking to understand policy change from the perspective of a policy-maker's actions.
In brief, we propose an original three-fold approach to explain policy liberalisation towards non-EU students, which involves assessing (a) the effectiveness of the narratives of steering used by policy elites to convince MPs of the need for policy change, (b) how favourable the spatio-temporal context is for such narratives to succeed, and, (c) to what extent a policy initiator's biographical capacity places him/her in a legitimate position to push for change.

Conclusion
Using the concept of narratives of steering, this paper examined the recent policy shift in Switzerland to facilitate the international mobility and labour market integration of non-EU graduates of Swiss universities, despite prevailing restrictive policies towards non-EU immigrants. Conducting qualitative analysis of the minutes of the parliamentary debates leading up to the approval of the new policy, and in-depth interviews with key political actors, has proven invaluable in uncovering the narratives constructed by policy elites and the key dynamics underlying policy change.
This paper demonstrates that 'designer migrants' (Hawthorne 2012) are allowed special streams of entry and labour market integrationeven during periods of general closure in migration managementbecause they are viewed by policy elites as valuable assets that help increase Switzerland's economic competitiveness. Indeed, in times of strict migration management, policy-makers employ a utilitarian discourse to justify their selectiveness towards migrants who supposedly contribute most to economic growth (Piguet 2006;de Haas, Natter, and Vezzoli 2016).
However, despite striking similarities with the competitiveness discourse used by policy-makers in Europe (Menz 2016), Canada (Geddie 2015), and New Zealand and Australia (Bedford and Spoonley 2014), the narrative used by Neirynckhighlighting the forced departure of international students upon graduation resulting in a loss of millions of francsappears to be a uniquely Swiss one. Whereas in Anglo-Saxon countries, higher education institutions have capitalised on the opportunity of recruiting international students as alternative sources of revenue to ensure their financial sustainability (Choudaha 2017), in Switzerland the state covers the majority of their educational costs. By arguing that Switzerland's loss of public investment is a gain for other global competitors, Neirynck uses a powerful economic narrative to convince MPs of the need for policy liberalisation.
This study shows that narratives of steering by policy elites are significant in influencing policy change in international student migration. We pose the question, however, if narratives of steering, per se, are sufficient enough to effect policy change? We suggest that by viewing narratives of steering as not being performed in a vacuum but in specific temporal and geographical contexts, such as: location, size, political system and policy-making history of a country or a region, one can see favourable or unfavourable conditions for such narratives to succeed. We also suggest that using a perspective that views policy change as, not simply pushed forward by abstract actors, but by specific individuals whose biographical capacity placed them in a particularly advantageous position to effect policy change, given a favourable spatio-temporal context. In proposing this concept, we understand biographical capacity to be the life experiences equipping a person with the knowledge and personal contacts to occupy a legitimate position to push for change.
Although the new law facilitating the migration and retention of international students from third-countries represents an important opening-up in an otherwise closed political system, it is, however, highly selective. Only individuals with a Swiss degree, in a profession with proven shortages in the labour market, or the ability to develop new products and technologies for a globally competitive market, stand to profit from this change. Similarly, this conclusion echoes de Haas, Natter, and Vezzoli's (2016, 30) observation that migration policies increasingly aim to affect the selection of immigrants rather than volume. The growing emphasis on specific criteria demonstrates that 'the real aim of most migration policies seems to be to increase the ability of states to control who is allowed to immigrate'. Our case study clearly demonstrates this assertion.
Some questions remain for future research. First, to what extent is the window of opportunity opened by the new law merely a symbolic one? Second, considering the legislative autonomy the cantons enjoy, to what extent will the difference in the implementation of the law persist? Third, as third-country migrants are increasingly required to fit into narrow profiles, what will the implications of such selective policies be in terms of stratification of residential and labour rights? Fourth, in a country where immigration policies have created tension between the protectionist interests of nationalist parties and the needs of a knowledge-oriented economy since the 1930s, what is the future of international student mobility policies? Will the current liberalisation continue? Will it be reversed? These questions need to be empirically addressed.
In summary, our study contributes to the literature on policy change in international student mobility by proposing an explanatory model that combines three dimensions: (a) effective narratives of steering crafted by policy elites to influence policy change, (b) a temporal and geographical context creating favourable conditions for those narratives to succeed, and (c) policy elites having the biographical capacity to influence policy agendas. This theoretical model will be useful for studies of migration policy change in general. This research is also of interest for scholars seeking to explain spatial variations across countries and regions on why migration policy changesfrom a restrictive stance to a more liberal one or vice versa.