Refugees, migrants, neither, both: categorical fetishism and the politics of bounding in Europe’s ‘migration crisis’

ABSTRACT The use of the categories ‘refugee’ and ‘migrant’ to differentiate between those on the move and the legitimacy, or otherwise, of their claims to international protection has featured strongly during Europe’s ‘migration crisis’ and has been used to justify policies of exclusion and containment. Drawing on interviews with 215 people who crossed the Mediterranean to Greece in 2015, our paper challenges this ‘categorical fetishism’, arguing that the dominant categories fail to capture adequately the complex relationship between political, social and economic drivers of migration or their shifting significance for individuals over time and space. As such it builds upon a substantial body of academic literature demonstrating a disjuncture between conceptual and policy categories and the lived experiences of those on the move. However, the paper is also critical of efforts to foreground or privilege ‘refugees’ over ‘migrants’ arguing that this reinforces rather than challenges the dichotomy’s faulty foundations. Rather those concerned about the use of categories to marginalise and exclude should explicitly engage with the politics of bounding, that is to say, the process by which categories are constructed, the purpose they serve and their consequences, in order to denaturalise their use as a mechanism to distinguish, divide and discriminate.


Introduction
The tragic but brutal truth: They are not REAL refugees! Despite drowning tragedy thousands of economic migrants are still trying to reach Europe. (Daily Mail, 28 May 2016) To help real refugees, be firm with economic migrants. (Guardian, 6 February 2016) In the second half of 2015, the story of Europe's 'migration crisis'which had been dominated by stories of hundreds of people drowning in the Mediterranean between Libya and Italy earlier in the yearcame instead to be dominated by images of thousands of people arriving on the shores of the Greek islands. Given that 90% of those arriving in Greece originated from just three countries -Syria, Afghanistan and Iraq (IOM 2016)countries in which there is known to be widespread, and escalating, conflict and political unrest, it might be anticipated that the most of those arriving in Europe during the course of 2015 would be accepted as de facto refugees in need of international protection. 1 In practice, this has not been the case.
On the one hand, many of those arriving across the Mediterranean have been dismissed by Europe's political leaders as 'economic migrants' taking advantage of the situation in Syria to secure entry to the EU in order to work. Hungarian Prime Minister Viktor Orban, for example, has claimed that the 'overwhelming majority' of migrants in Europe are not refugees but are merely seeking a better life (Budapest Beacon, 4 September 2015). Robert Fico, his Slovak counterpart, declared that up to 95% of those arriving in Europe were economic migrants (The Economist, 7 September 2015). At the same time, there is a strong political and media narrative which suggests that even where people have been forced to leave their countries due to conflict, persecution and human rights abuse, they should remain in the first countries to which they arrive rather than making the hazardous journey across the Mediterranean to Europe . Their decision to do so is viewed, under the false pretext of the 'safe first country' clause, as confirmation that they are 'migrants' rather than 'refugees', and therefore undeserving of protection.
The use of the categories 'migrant' and 'refugee' to differentiate between the experiences of those on the move and the legitimacy or otherwise of their claims to international protection is reflected in the opening quotations, both of which refer to 'real refugees' despite their sources being located at opposite ends of the political spectrum. These debates have led UNHCR and a multitude of other national, international and civil society organisations to engage in efforts to educate the public on the differences between 'migrants' and 'refugees', often privileging the rights and needs of the latter (see UNHCR 2016). Others have challenged the media to use the term 'refugee' rather than 'migrant' which, it is argued, undermines the rights of those fleeing violence and conflict (see Malone 2015).
These examples and interventions demonstrate the extent to which the use of different categories to describe those on the move has become deeply politicised in the context of Europe's 'migration crisis'. But they also reflect a series of assumptions about the dynamics of migration and that it is possible, straightforward even, to differentiate between 'refugees' on the one hand and 'migrants' on the other. As noted by Zetter (2015), there is an increasing mismatch between the legal and normative frameworks that define the existing protection regime and contemporary forms of migration (including forced displacement). Whilst dominant representations of the 'crisis' typically give the impression of a linear, uninterrupted flow of people heading towards Europe (Mainwaring and Brigden 2016;Crawley et al. 2016aCrawley et al. , 2016b, many of those arriving in Europe during 2015 had been living for months or even years in countries other than those in which they were born, requiring us to engage with the complex economic, social and political realities of the 'in between'. In this context, our aim is to challenge what Apostolova (2015) has described as a form of 'categorical fetishism' which, despite significant academic critique, continues to treat the categories 'refugee' and 'migrant' as if they simply exist, out there, as empty vessels into which people can be placed in some neutral ordering process like a small child putting bricks into a series of coloured buckets. Drawing on interviews undertaken with 215 people who crossed to Greece in the second half of 2015, 2 we argue that the decisions made by these individuals and their families to leave their countries of origin, and the process by which they came eventually to arrive in Europe, were far more complex than typically presented by politicians, policy-makers and the media. Whilst we do not make claims to the representativeness of the data, the sample of people that we interviewed is one of the largest of its kind. 3 As such, it provides important new insights into the complex experiences of those arriving in Europe at the height of the so-called 'migration crisis', not least because the interviews were conducted with individuals shortly after their arrival (in some cases just a matter of days) when these experiences were at the forefronts of their mind and before they came to realise the need to narrate their stories in a particular way to fit the existing policy and legal categories (Zetter 2007;Signorini 2015). This evidence not only raises questions about the extent to which existing categories are able to capture these complex and messy social realities but also challenges us to think more carefully about the use of categories, and the process by which the boundaries between them are constructed. This is particularly important given the interest of migration scholars in engaging with political and policy debates on migration whilst simultaneously avoiding the dangers inherent in conflating policy and analytical categories (Bakewell 2008). As van Hear (2012) rightly suggests, policy categories, vernacular usages and social science understandings of migration modulate one another, with concepts and categories travelling and interacting dynamically between the three spheres. Nonetheless, it is important to be aware that if policy categories are allowed to shape, or even dominate, academic researchperhaps in an effort to engage policy interest and, in turn, enhance the 'impact' of our work-then the politics that lie at the heart of the policy-making process will also come to shape and inform academic thinking, concealing rather than revealing the dynamic processes with which migration is increasingly associated. Taking the dominant categories as the basis of our analytical approach can limit our understanding of migration and make us potentially complicit in a political process which has, over recent years, stigmatised, vilified and undermined the rights of refugees and migrants in Europe (Bakewell 2008;Zetter 2015).

The problem of categories
There are few more challenging questions for academics and policy-makers alike than where, and how, we draw the line between 'forced' and 'voluntary' migration (Richmond 1993;Zetter 2007;Betts 2013;Long 2013). The positioning of this line, and the factors, places, and experiences which come to be associated with the categories that lie either side, shape our understanding of who constitutes a 'refugee' on the one hand, and 'migrant' on the other.
Migration scholars have long problematized the extent to which it is possible to distinguish clearly and easily between different types of migrants and have argued for the need to move beyond opposing binaries. Two main themes dominate this critique. Some (e.g. Richmond 1993;Malkki 1995;Koser and Martin 2011;Zetter 2015) have argued that the distinction between 'refugees' on the one hand, and 'migrants' on the other, does not reflect the way migratory processes work in the 'real world' arguing that they fail to take account of 'the fact that people with different motivations travel together' and that individuals may change status or simultaneously fit in two (sometimes more) preexisting categories (Koser and Martin 2011;Collyer and de Haas 2012;Mainwaring and Brigden 2016). Categories constructed to make sense of migration are often based on the place that migration occurs (both source and destination countries), the causes of movement and the time or duration of migration. The construction of these categories can lead to a 'cookie cutter approach' which homogenises and over-simplifies the experiences of the people they contain (Gupte and Mehta 2007). This problem, it is argued, has been exacerbated by changes in the nature of international migration, including the rapid diversification of migrant profiles and patterns of migration (Koser and Martin 2011;Collyer and de Haas 2012).
This critique has been associated with the development of new concepts intended to help make better sense of the complexities of migration ('mixed flows', 'mixed motivations', 'transit migration') and new categories intended to bring into the purview of the international protection regime those trapped in the space between 'refugee' and 'migrant': examples include 'people in distress' (Goodwin-Gill 1986), 'distress migrants' (Collinson 1999) and 'survival migrants' (Betts 2013). Yet, as will be seen from the evidence presented below, even these categories prove largely incapable of adequately explaining the complex experiences and back stories of those crossing the Mediterranean in 2015. This is partly because of the increasingly protracted and fragmented nature of journeys to Europe (Collyer 2010), a theme to which we will return. But it is also because notions such as 'transit migration', which have come to dominate much academic and policy-thinking in relation to Europe's 'migration crisis', reinforce rather than challenge the idea that migration takes place between two fixed points (Collyer and de Haas 2012). This problem of linearity is exacerbated by a lack of consideration of the temporalities of migration (Cwerner 2001).
This brings us to the heart of our broader critique of categories: namely the process by which they are constructed and the political purpose(s) that they serve (Zetter 2007;Scherschel 2011;Long 2013;Becker 2014). The process of categorization is deeply embedded in both social science and policy, providing an important means by which we construct the social world (Moncrieffe 2007;Bakewell 2008). But this process is not neutral: rather it reflects 'subjective perceptions of how people fit into different spaces in the social order and of the terms on which society should engage with them in varying contexts and at different points in time' (Moncrieffe 2007, 1). Migration regimes, like all other ordering systems, create hierarchical systems of rights. Before 1951, 'refugees' were identified as specific nationalities no longer physically present in or politically protected by their state of origin (Long 2013). Once the Refugee Convention had been signed, the identity of the 'refugee' became clearly separated, in legal terms at least, from that of the 'migrant' in an effort to protect those fearing persecution (Long 2013).
Yet whilst the legal definition of a 'refugee' is shaped by the Refugee Convention, its interpretation and application takes place at the national level reflecting national interests and priorities which change over time. This means that the seemingly neutral and objective category of 'refugee' is in fact being constantly formed, transformed and reformed in response to shift in political allegiances or interests on the part of refugee-receiving countries and the evolution of policy and law. In other words, policy and legal categories may appear fixed, neutral or objective even but are, in fact, constantly subject to challenge across different national and regional contexts as lawyers, advocates and academics push at the boundaries of international law (Gauci, Giuffre, and Tsourdi 2015). Developments in case law and policyand the iterative process between the twocan serve to bring some people into the category of 'refugee' whilst simultaneously excluding others. Take, for example, changing country guidance in the U.K. in relation to asylum claims from Eritrea. In October 2016, the Upper Tribunal (Immigration and Asylum) rejected the existing Home Office's argument that Eritreans could return safely to their country of origin by simply signing a letter of apology and paying a 'diaspora tax'. 4 By issuing new country guidance on Eritrea, the Upper Tribunal effectively upgraded the country's nationals from 'deportable failed asylum seekers' to legitimate beneficiaries of international protection. At the same time Afghan nationals, once widely regarded as having almost a prima facie claim for protection under international refugee law, are no longer widely regarded as the legitimate beneficiaries of international protection: in October 2016, the EU signed an agreement with the Afghan Government allowing its member states to deport an unlimited number of the country's asylum seekers, and obliging the Afghan Government to receive them. 5 This process of inclusion and exclusion on the basis of nationality-based categories can also be seen in the construction of an EU 'safe countries of origin' list which identifies countries in which there is considered to be democracy together with an absence of persecution, torture or inhuman or degrading treatment or punishment, threat of violence or conflict (European Commission, undated). Those arriving from these countries are almost invariably considered 'migrants' and often subjected to detention and accelerated procedures. In the case of the U.K., this list extends to 26 countries of origin including Ukraine, Gambia (deemed safe only for males), Kenya, Mali and Nigeria.
It is clear therefore that there is nothing 'natural' or 'fixed' about the legal and policy categories associated with international migration: rather these categories are in a constant state of change, renegotiation and redefinition. The categories 'refugee' and 'migrant' do not simply exist but rather are made. Choosing to labelor equally not labelsomeone as a 'refugee' is a powerful, and deeply political, process, one by which policy agendas are established and which position people as objects of policy in a particular way (Zetter 2007;Becker 2014). This is nowhere evident than in times of 'crisis' when categories can be used to fragment the international protection regime and limit responsibility for what is perceived to be an unsustainable number of arrivals (Zetter 2007;Polzer 2008). The evidence from our research suggests that the categories dominating political, policy and media debates associated with Europe's 'migration crisis' are not simply inadequate tools for capturing the complex drivers of migration across the Mediterranean but also serve to perpetuate and reinforce a simplistic dichotomy which is used to distinguish, divide and discriminate between those on the move.

The political economy of forced migration
We begin our empirical analysis by exploring the complex relationship between economic, political and social drivers of migration and the ways in which these came together to shape the lives and decisions of those we interviewed in Greece. It is important to acknowledge that whilst 'forced migration' has often been presented and discussed as a discreet and particular form of migration, involuntary movements of people are always part of much larger assemblages of socio-political and cultural processes and practices which include a variety of people who, whilst all displaced, find themselves in qualitatively different predicaments (Malkki 1995). Underdevelopment, conflict and, by extension, economic and forced migration are closely linked (see Nyberg-Sørensen, van Hear, and Engberg-Pedersen 2002;Bakewell 2011). This 'migration-asylum nexus' not only constitutes a major analytical and policy challenge, but also suggests that political upheavals, conflicts and economic difficulties often occur simultaneously giving people multiple motivations for the decision to move (Bissell and Natsios 2001). Threats to an individual's personal integrity can also go hand-in-hand with dismantled economies and lack of economic opportunities, thus reinforcing the socio-economic factors influencing a person's decision to migrate (Castles, Loughna, and Crawley 2003;Neumayer 2005).
The findings of our research confirm that conflict, most notably in Syria, was a major factor contributing to the significant increase in people arriving in Greece during 2015. Like Schmeidl (1997) and Castles, Loughna, and Crawley (2003), we found that political violence and civil war are the strongest predictors of forced migration. Many of our respondents said that they had left their countries because the violence had become intolerable and they consequently feared for their personal safety and that of their families. The vast majority (91%) explicitly mentioned factors that could be described as 'forced migration' including conflict, persecution, violence, death threats and human rights abuse. This was the dominant theme in our interviews with respondents from Syria, Afghanistan and Iraq, and also from Yemen. Whilst some respondents had been targeted specifically because of their involvement in conventional forms of political activity or other forms of resistance (e.g. journalism, humanitarian work), most had decided to leave their countries because life had simply become intolerable. This included a large number of Syrians who were subject to almost daily barrel bombings, sniper fire and other attacks. Homs, Damascus and Aleppo were frequently mentioned as cities in which it had become impossible to live.
Many respondents described being caught between competing factions and militia (e.g. Assad's forces, the Free Syrian Army, so-called Islamic State (IS) and the PKK): '[w]e don't know who is to blame for what is happening in Syria but the Syrian people pay the price' (Syrian woman aged 47). Respondents talked at length and in detail about how their homes had been bombed and how their children lived in constant fear: '[I] was living in Aleppo. Before the war our lives were fine. We had simple lives … But things started becoming hard in Aleppo in 2013 … My children have become sick because of their fear' (Syrian Kurdish man aged 33). Fear of death was a consistent theme across many of the interviews with those who had left Syria: '[w]hatever you do there [Al-Buwaidah al-Sharqiyah, Syria], you can find no peace. Every day we were dying and born again' (Syrian man aged 36). Those from Iraq also described the effects of the escalating conflict in detail, particularly in cities such as Baghdad and Mosul: 'I was living in Baghdad. You don't know when you are going to die there' (Iraqi man aged 28).
It is, however, impossible to fully appreciate the drivers of migration without examining the ways in which political and economic factors come together to shape the experiences of those living in times of war. The longer the conflict continues, the more complicatedand difficult to unpackthis relationship becomes (Crawley et al. 2016b). As a result, people may feel that they have no alternative other than to move even if they are not specifically targeted or, as yet, directly affected. Conflict, particularly where it becomes protracted, undermines the ability to earn a livelihood and feed a family by killing primary breadwinners, destroying businesses and making it impossible to travel to work. Many of our Syrian respondents had taken the decision to move for economic reasons but their economic insecurity cannot be understood outside of the ongoing conflict that is devastating parts of the country: I was living in Deir Al-Zor. Nothing specific happened to me. The economic problems started when IS entered the city. Most of the shops closed … I was in a good financial state. Before I run out of money I decided to leave the city. (Syrian man aged 32) A similar situation was described by some of our Iraqi respondents and by those living in Afghanistan and Yemen: 'I was living in Mosul. There were no jobs available … Since IS came, all rich people took their money and left the city. There are no jobs available anymore' (Iraqi man aged 19) I left because of the war. There is no safety in Yemen. You might die any minute … In Yemen, there is no electricity, the schools have closed and it is not safe to be on the streets. War has no mercy. (Yemeni man aged 20) I was living in Nuristan province (Afghanistan). My life was good there, the money we were earning was enough. But the Taliban and the Kochis started a war. The Taliban are governing our region. Two months ago I started thinking of leaving. I started receiving threats. They were threatening me that they will kill my family and take my fortune. (Afghan Hazara man aged 23) Protracted conflict devastates the economic infrastructure of countries such as Syria, increasing the prices of basic goods and commodities. Price increases have been exacerbated by internal displacement and the movement of large numbers of people to some of the safer cities, such as Afrin and Suwayda: 'People started flowing into Afrin from across Syria because there was peace … The cost of living increased incredibly. Afrin is a very expensive city now' (Syrian man aged 33).
In other cases, individuals and families were unable to work and support their families because of rules that were imposed on them, most notably by IS. More than a quarter (28%) of respondents said that the activities of IS were a significant factor in their decision to leave, particularly in Syria but also in Iraq, Afghanistan and Yemen: When you are a civil servant in Kabul you receive death threats by the Taliban. Now there is also IS in Afghanistan, and the situation has deteriorated. IS and the Taliban are collaborating. I have received many death threats. I was forced to leave Afghanistan. Otherwise, I would have never left my country. I was scared about my daughters' and my personal safety. Before I left, the last threat I received said that either I quit my job, or myself and my whole family will be beheaded. (Afghan Tajik man aged 50) Many of these respondents had been detained, tortured or forced to watch beheadings by IS. They expressed grave concerns for the safety of their families, and particularly women (wives, sisters, daughters) who were perceived to be non-compliant with strict Sharia laws concerning their dress and behaviour. But they also described the consequences of IS for their ability to make a living, including where individuals were unable to comply with new supposedly 'Islamic' requirements in relation to their respective professions. Examples included a barber who was no longer allowed to cut hair, a man who had to close down his copy shop because photos were forbidden, and a man whose wife was no longer able to work as a gynaecologist and 'When IS came in our city they confiscated the university degrees of all women. They told them that they need to get a degree that has been issued by IS' (Syrian man aged 34).
Meanwhile, pre-war economies in places such as Syria have all but collapsed giving way to extortive siege economies, the profits of which provide financial incentives to deny truces, with devastating impact for civilians (UN Human Rights Council 2016). One of our respondents, for instance, described being visited by a police officer in Syria. He was given 24 hours to collect a significant amount of money from all the shops in the neighbourhood: 'something like a tax in order to be safe from the regime. [The police officer] warned me that if I was unable to collect it, I would have to pay the entire amount myself' (Syrian man aged 32). At that point he and his wife decided to leave. Others described kidnapping by state and non-state actors (including a range of militia groups) as an increasingly common threat to their safety and that of their families. In some cases, individuals were targeted because they were perceived to be a political threat. More commonly, however, people were targeted because they had resources and were viewed as being able to pay a significant ransom. In other words, those with resources were most at risk: 'I was living in Damascus. I was scared about my daughter's life. We were rich, and whoever is rich is in danger of having their children kidnapped' (Palestinian Syrian woman aged 43).
This evidence points to the fact that, particularly in situations of conflict, the decision to leave can only be understood within a wider political economy of forced migration rather than as a response to individualised threats of violence. But the problem of categories is not simply their failure to reflect these complex social realities. It is the inability (or unwillingness) of those employing these categories to engage with the movement of people between categories over space and time.

Moving between categories over space and time
One of the major problems with the dichotomous, location-based categorisation which dominates much scholarly and policy-thinking is that it presupposes that people move between two fixed places. As a consequence, what happens 'in between'geographically and temporallyis viewed as being largely inconsequential. This is reflected in political, policy and media narratives associated with the 'migration crisis': it is assumed that the decision to travel to Europe has been taken at the point of departure from the country of origin and that much, if not all, of what happens 'in transit' is orientated towards this objective. The evidence from our research strongly challenges this view (Crawley et al. 2016a), pointing to a series of separate migration decisions shaped by the changing circumstances in which an individual finds him/herself. Understanding how the motivations and experiences of those on the move change over space and time in response to these shifting circumstances not only brings into question the usefulness of policy categories which are fixed in time and space but also opens up new possibilities for understanding migratory processes.
This can be seen in the complex stories and experiences of those we interviewed, many of whom did not intend to travel to Europe when they first left their countries of origin but moved on for different reasons. The time between leaving the country of origin and arriving in Europe was often considerable: more than one in five of our respondents arriving in Greece had left their countries of origin more than 18 months previously (Crawley et al. 2016a(Crawley et al. , 2016b. For many of these people, the countries to which they had originally travelled were initially perceived as destination countries where they intended to settle and live. Often, during these longer periods of stay, people worked, ran a business, rented accommodation or even applied for immigration status (from the authorities or on the black market). People experienced some degree of integration into local social systems: they married, had children and made a life. Many never intended to continue onwards to Europe but decided to leave due to a combination of political and economic factors sometimes allied to severe discrimination and a lack of access to rights and/or citizenship. As a result, their onward migration can be considered a separate migration experience driven by its own motivations, decision-making, planning and aspirations (Crawley et al. 2016a). The experiences of two groups of respondents illustrate this process most clearly: Afghans who had been living in Iran and Syrians who had been living in Turkey.
One of the most striking aspects of the data provided by our Afghan respondents is the very significant period of time that most had been living outside Afghanistan prior to their departure to Europe 6 : nearly half (48%) had left Afghanistan more than 5 years prior to their interview with us and of these a significant proportion (38% of the total) had been living outside Afghanistan, mainly in Iran, for more than 10 years. Seven respondents had not been to Afghanistan for more than 20 years, and some for as long as 35 years. In addition, 7 respondents had never been to Afghanistan at all, having been born in either Iran or Pakistan. That means nearly two-thirds (60%) of our Afghan respondents had either never been to Afghanistan or had not lived there for a considerable period of time.
Irrespective of their legal status Afghans in Iran encounter a range of difficulties. They are only allowed to work within their areas of residence, and are only authorised to be hired for specific jobs, which clearly limit their employment opportunities (Farzin and Jadali 2013). Children born in Afghanistan are often unable to obtain birth certificates, travel documents or to access public services (Human Rights Watch 2013). In 2007, Iran declared some provinces, or cities of specific provinces, to be 'no-go areas' for refugees limiting the ability of Afghans to live, travel or work in such areas (Farzin and Jadali 2013). In addition, the majority of provinces have imposed residency restrictions on refugees, which forbid landlords to rent houses to non-native Iranians, in particular Afghans (Afghanistan Independent Human Rights Commission 2015). Although the Iranian government took steps in 2015 to ensure access to education and healthcare for refugees, Afghans continue to face inequality, discrimination, mistreatment, forced deportation and severe detention conditions as well as restrictions on residency, freedom of movement and employment (Farzin and Jadali 2013;UN General Assembly 2016).
These problems were reflected in the experiences of our Afghan respondents, almost all of whom described insecure working conditions and exploitation due to a lack of legal status as well as extensive experiences of discrimination and mistreatment, including physical violence. One of those who had been born in Iran told us, I could have been naturalised in Iran in the past, but I was expecting that the situation in Afghanistan will get better, so I decided not to. In Iran I was afraid to go out. They are treating Afghans as if they are dogs. (Afghan Sayyid man aged 32) A young man who had been living in Iran for two years after travelling there to work in the construction industry told us, 'I realised that I was never going to build a life in Iran, and one day I would be kicked out of the country. So I decided to leave the country in order to find a better life elsewhere' (Afghan Tajik man aged 19). Another explained that he decided to leave because of the deteriorating economic situation for Afghans and the educational discrimination faced by his children: 'I left with my family to Iran 12 years ago … But I decided to leave Iran for the sake of my children. My children were discriminated … All these discriminations against Afghans in Iran forced me to leave' (Afghan Hazara man aged 35).
Forced deportations, in most cases without notice, were also of concern to many of our respondents, particularly those who had never been to Afghanistan or who came from ethnic groups against whom there is well-documented persecution. Whilst all Afghans experience varying degrees of discrimination in Iran, the situation appears to be particularly difficult for the Hazara because they are more easily identifiable due to their distinctive physical appearance (Human Rights Watch 2013). For several respondents, the possibility of being forcibly returned to Afghanistan to face persecution at the hands of IS (and others) was an additional factor shaping the decision to leave Iran: [r]efugees in Iran are in big trouble. They live a dark life. We wanted to go back to Afghanistan. But we talked to our parents in Afghanistan and they told us that IS has increased its presence there. IS are persecuting Hazaras. They are beheading Hazaras. So we decided to go to Germany. (Afghan Hazara woman aged 28) Finally, there were concerns about the recruitment and deployment of Afghans, often minors, by the Iranian authorities to fight in the Syrian war (UN General Assembly 2016). Several Afghans, including one child aged 16, told us that they had been conscripted into the Iranian army to support the Assad regime in Syria, on the promise of Iranian residency or citizenship which never materialised.
The complex stories associated with the 'in between' make it very difficult to categorise Afghans who have travelled to Europe from Iran as straightforwardly either 'refugees' or 'migrants'. The decision to leave Iran for Turkey and Europe was motivated, in part, by a deteriorating economic situation for Afghans in general, with discriminatory treatment of ethnic Hazara in particular, but this was combined with a fear of being forcibly returned to Afghanistan. At times in their lives, some of our Afghan respondents were 'refugees' fleeing conflict, at others they were 'migrants' looking to improve their economic situation, albeit in the context of discrimination. But there were also times in their lives when they were neither refugees nor migrants they were born in Iran and had never migrated, and times when they were both.
At the same time, many Syrians who crossed to Greece in 2015 have come to be represented as 'migrants' despite widespread coverage of the conflict that has ripped their country apart. Since 2011, millions of Syrians fled their homes and sought refuge in Turkey. As of December 2016, Turkey hosted approximately 2.8 million registered Syrian refugees. 7 Until the end of 2013, Turkey implemented an open-door policy for Syrian refugees, who were hosted in camps funded and managed by the Turkish state and a number of NGOs. However, in 2014, due to the shrinking capacity of the camps and changes in the Turkish policy response, the vast majority of Syrian refugees ended up living in towns and cities creating economic difficulties and social tensions between Syrian refugees and their host communities (İçduygu 2015).
The difficulties of making a life in Turkey runs through the stories of the Syrians we spoke to in Greece, a third (34%) of whom had moved on for what might typically be understood as economic reasons: they were running out of money, found it impossible to secure employment or were working long hours for very little pay. With the passage of time, and in the absence of a resolution to the conflict in their home country, respondents told us that they had grown increasingly concerned about the impacts on their families, and especially their children, many of whom had been out of schools for years or had health issues. In the words of one of our interviewees: I went to Gaziantep … My house rent was 200 liras in the beginning and two months later the landlord asked me 500 liras/month rent. I agreed because I knew that my family would join me. The cost of living was very high there. (Syrian man aged 24) Many Syrians relied on savings, selling assets (e.g. jewellery) and remittances as their main sources of household income in order to make ends meet (see also World Bank 2015). When these sources were depleted, they found it virtually impossible to secure an income. A lack of work authorisation forced Syrian refugees to find employment in the informal economy, where working conditions were often unacceptable and wages extremely low (İçduygu 2015). Although Syrians living in Turkey are now entitled to access work permits as part of the EU-Turkey agreement intended to prevent them from moving on into Europe, work permits remain difficult to secure and without them the majority of Syrians continue to be absorbed in the informal sector and low-skill jobs (World Bank 2015).
Long work hours, unsafe conditions, lack of guaranteed payment and low wages were typical experiences for Syrian respondents working in Turkey's informal economy: '[w]e worked for 1 month in Antalya. Our job was to clear farms off snakes. 30 people were working together there, but none of us was paid in the end' (Syrian Kurdish man aged 55). Apart from difficulties in obtaining a work permit, a lack of Turkish language skills also served as obstacle to securing employment: '[m]y main problem was with the language in Gaziantep. I wanted to open a business there. But it wasn't easy without speaking the language. I spent a year in Gaziantep jobless' (Syrian man aged 32). These findings are reflected in research by Koser and Kuschminder (2016) based on interviews with more than 1000 people from Afghanistan, Iran, Iraq, Pakistan and Syria living in Greece and Turkey. Two-thirds of their respondents planned to migrate onwards due to poor living conditions and a lack of legal status.
Despite clear evidence about the difficult living circumstances and lack of protection for Syrians living in Turkey, there appears to be an expectation on the part of EU policymakers that, having survived the bombs and gunshots, Syrians should 'stay put'. If they decide to move on then the legitimacy of their status as 'real refugees' is brought into question. In expressing the need to create a life and not only to live, 'refugees' have 'shown their hand', revealing themselves to be 'migrants' in search of economic betterment. But 'refugees', like 'migrants', are also human beings with needs and aspirations that go beyond the ability to simply exist. Those who have lost their homes, livelihoods and often members of their family need not only to feel safe but also a sense of an economic and social future for themselves and their children. And in countries such as Turkey that has proved very difficult. Whilst there is generally no interest in the policy implications of the 'in between', the shifting experiences of those on the move over space and time are increasingly used strategically to exclude refugees from access to protection and rights. This use of categories to include and exclude reveals a propensity on the part of politicians and policy-makers to simultaneously ignore the complexities of migration in the construction of policy categories whilst taking full advantage of this complexity when it suits their political interests.

Challenging the politics of bounding
We conclude by stepping back from the data to consider the implications of our findings for policy, practice and academic scholarship. Our research with those who crossed the Mediterranean to Europe in 2015 challenges the construction of policy categories based on binary, static and linear understandings of migration processes and experiences. It builds upon a substantial body of academic literature that has demonstrated a disjuncture between conceptual and policy categories and the lived experiences of those of the move (Scherschel 2011;Gupte and Mehta 2007;Zetter 2007;Bakewell 2011;Koser and Martin 2011;Collyer and de Haas 2012;Becker 2014). The lives of those on the move are complex. They are not simply a sum of the categories that are constructed around them. Neither can their experiences always easily be dropped into one or other category intended to contain and make sense of the world. People can and do shift between and across categories both in their countries of origin and as they travel through space and time. This movement has proved deeply problematic for policy-makers and politicians, many of whom have chosen to blame refugees and migrants for their failure to fit, rather than problematizing the nature of categories and the process of category construction. This is not merely an issue of semantics. Categories have consequences. They entitle some to protection, rights and resources whilst simultaneously disentitling others. As Becker (2014) suggests, powerful actors establish and use categories, and the labels with which they are associated, to understand and frame a problem which in turn reflects how issues areand are notrepresented in policy debates and discourse. In Europe, as elsewhere, 'the almost arbitrary categorisation of who constitutes a "refugee" … leads not only to a gross violation of rights but to the systematic exclusion of large groups of people who would like to see themselves as "refugees"' (Gupte and Mehta 2007, 65). In the context of Europe's 'migration crisis', this can be seen most clearly in the policy of relocation, which enables individuals from some countries of origin who arrived after 25th March 2015 to be transferred from Greece or Italy to another participating country on the basis that they are in clear need of international protection. 8 The relocation scheme reinforces the idea that people from some countries are legitimate beneficiaries of asylum whilst others are not, even before their stories are heard or their experiences known.
So where do we go from here? How do we, as academics who often have an explicit interest in challenging simplistic assumptions about the experiences of those on the move but also in engaging with political and policy debates on migration, engage with the problem of categories without ourselves indulging in a form of 'categorical fetishism'? How do we translate the nuanced, complicated findings of our research into messages and ideas that encourage and enable politicians and policy-makers to shift their thinking and develop approaches which at least begin to reduce the disjuncture between migration policies and the lives of those on the move? And how can we most effectively do this in a context in which politicians and policy-makers are more concerned with numbers than with empirical evidence which challenges the objectives of containment?
The first point to make is that categories, for all their inherent problems, are both pervasive and inevitable (Moncrieffe 2007;Polzer 2008). As Collyer and de Haas (2012, 468) suggest, 'ignoring or rejecting them does not mean they go away and may blind us to the important interrelationship between scientific and political forms of knowledge production that have become inherent to the creation and maintenance of categories'. Moreover, challenging or moving the boundaries between categories simply creates new ones. But having a critical awareness of the constructedness of categories is, as Polzer (2008) suggests, a key aspect of social scientific academic enquiry. We can, and should, approach migration categories, particularly those used by politicians and policy-makers to exclude or position certain groups in a particular way, from a far more critical perspective, where appropriate questioning the way in which 'the problem' has been defined.
Secondly, we need to recognise the dangers of simply working with the categories constructed by others in an attempt to engage policy-makers in the findings of our research and the implications of what we have to say (Bakewell 2008;Scherschel 2011). Whilst many of us who have undertaken migration policy research, including within government, might question the extent to which it is possible to bring academic evidence to bear on the policy-making process, it is clear that there is an iterative relationship between academia and the policy world and that, over time, 'terminology that arises from the mutually reinforcing environments of academic and policy arenas may find its way into popular conceptions of migration' (Collyer and de Haas 2012, 473). At this point, the category may become shorthand for a series of assumptions about the underlying migratory processes and the experiences of those who are perceived to 'fit' into this category. The concept of 'transit migration' is an obvious case in point. The idea that refugees and migrants are simply 'biding their time' in other countries before 'heading to Europe' is now written through the political and policy narratives associated with the so-called 'migration crisis'. Yet this image is highly misleading (see Collyer 2010;Crawley et al. 2016a). If we want to challenge the use of categories to exclude and contain, then this will require a more critical engagement with policy categories which misrepresent or artificially constrain our understanding of migration processes.
Which brings us on to our third point: we need to avoid falling into the trap of suggesting, either explicitly or through the ways in which we organise and structure our scholarship, that those placed in one category rather than another are somehow more 'deserving'. There are understandable concerns that opening up the category 'refugee' to include those who fall within a broader category of 'forced migrants', or who were initially economic migrants but find themselves in 'refugee-like' situations, could reduce the already limited protection provided under international law for those fleeing conflict, persecution and human rights abuse (Feller 2005). Yet the move to foreground or privilege the term 'refugee' over 'migrant' does nothing to contest the faulty foundations of the binary distinction between the two categories, it simply perpetuates its logic (Uniacke 2016). Moreover, this strategy actually plays into, and reinforces, a dichotomy which discriminates against 'migrants'. Instead, we should be challenging where the boundaries between categories are placed and the differential valueand rightsassigned to those who are situated accordingly.
Finally, and perhaps most importantly, we need to explicitly engage with the politics of bounding, that is to say, the process by which categories are constructed, the purpose that they serve and their consequences. Our call is not for an end to the use of categories as a way of making sense of our social and political worlds, but for explicit recognition and engagement with the idea that categories do not simply represent or reflect the world but simultaneously create and limit it (Moncrieffe 2007). The findings of our research challenge us to look beyond the use (and abuse) of categories and focus more explicitly on the politics and relations of power that underpin the bounding process (Jones 2009). Only by bringing the boundary itself into consciousness can we start to denaturalise the use of categories as a mechanism to distinguish, divide and discriminate (Jones 2009;Apostolova 2015;Uniacke 2016) or even to render entirely invisible (Polzer 2008). Once we understand the political and policy objectives that underpin the bounding process, we can focus our efforts on challenging this process and the interests it serves with the aim of 'bringing into bounds' that which is increasingly excluded. In the context of Europe's 'migration crisis', this includes those people who originally left their countries of origin for economic reasons but who found themselves in situations of profound insecurity, as well as those who left due to conflict and the inability to make a livelihood in the context of war but were unable to secure protection, rights and a long-term future for themselves and their families. are grateful to Michaela Benson and three anonymous reviewers for their helpful comments on an earlier draft of this paper. Most importantly, we would like to thank all of those who contributed directly to this research and in particular the hundreds of refugees and migrants who crossed the Mediterranean in 2015 and were willing to share their stories with us.

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

Funding
This work was supported by the Economic and Social Research Council [grant number ES/ N013506/01].