Forced to leave? The discursive and analytical significance of describing migration as forced and voluntary

ABSTRACT

This article examines voluntariness in migration decisions by promoting the acknowledgement of forced and voluntary migration as a continuum of experience, not a dichotomy. Studies on conflict-related migration and migration, in general, remain poorly connected, despite calls for interaction. This reflects the forced–voluntary dichotomy's stickiness within and beyond academia, which is closely connected to the political implications of unsettling it and potentially undermining migrants’ protection rights. We delve into notions of the ‘voluntariness’ of migration and argue for the analytical need to relate evaluations of voluntariness to available alternatives. Drawing on qualitative research with people from Afghanistan and Pakistan coming to Europe, we hone in on three particular renderings of migration: migrants’ own experiences, scholarly qualitative observations and labelling by immigration authorities. Analysing migration as stages in a process: leaving – journey (and transit) – arrival and settlement – return or onward migration, we highlight the specific effects of migration being described as being forced or voluntary. Labelling as ‘forced’ (or not) matters to migrants and states when asylum status is on the line. For migration scholars, it remains challenging to decouple these descriptions from state systems of migration management; though doing so enhances our understanding of the role voluntariness plays in migration decisions.

KEYWORDS:

• Migrants
• refugees
• forced
• voluntary
• drivers of migration

Introduction

Migration studies often utilises oppositional typologies to categorise different groups and experiences: skilled-unskilled; temporary-permanent; and, key to this article, forced–voluntary. However, the blurring of boundaries between categories is also regularly acknowledged, and most scholars prefer to think of such dichotomies as a spectrum of experiences (see Castles, De Haas, and Miller 2014; Fussell 2012; Richmond 1993).

Regarding the forced–voluntary distinction, whilst there may be identifiable extreme cases, most migrants’ experiences of the degree of volition in their migration decisions means they fall somewhere in the blurry middle of the forced–voluntary spectrum. Even a cursory review of migrant experiences shows overlaps between migrants who are labelled as ‘forced’ or ‘voluntary’. Fieldwork observations convince us of the impossibility of describing someone's decision to migrate as entirely voluntary or entirely forced. Making a clear-cut distinction between forced and voluntary migration is not acceptable from an empirical, and consequently, analytical, point of view, even if labelling for the sake of migration management upholds the notion (see also Carling 2017; Crawley and Skleparis 2017).

However, is it enough to point out the complexity and highlight the spectrum of experiences, before moving on? Can migration studies grapple with the drivers of migration and their connection to aspirations and desires (Carling and Collins 2018) without addressing the volition involved in migration? Whilst our fieldwork observations rightly encourage us to emphasise complexity, we, as migrations scholars, cannot – and should not – simply leave the discussion there. Because the distinction between forced and voluntary, as two very different forms of migration, has a tendency to stick around in academic work and policy, migration scholars must unpack the experienced, observed and labelled dimensions of this distinction. And we can do so by examining differing analytical and discursive constructions.

Firstly, if we as academics emphasise the ‘mixed’ nature of all migration, then what are the implications for those with particular protection needs and those with the right to refugee status under the 1951 Refugee Convention (Kumin 2014; Van Hear, Brubaker, and Bessa 2009; Van Heelsum 2016)? Secondly, migration management, as practised by many European countries, requires labelling migrants as forced or voluntary to sort them into bureaucratic categories such as asylum seeker, refugee, labour migrant, family reunification migrant and international student, amongst others; where the first two categories are seen as forced migrants and the latter three are seen as voluntary migrants.

Even if the particular terminology of forced and voluntary is not necessarily used in public discussions of migration, the academic categories of forced and voluntary can be mapped onto public categories of refugee/asylum seeker and economic migrant, respectively, at least in Europe. Bureaucratic distinctions and labels have discursive power, and the labelling of individuals as forced or voluntary impacts their migratory experiences.

We argue that it is not enough to simply point out that the distinction between forced and voluntary migration is complex. We propose that whilst migration scholars need to keep acknowledging the complexity of the degree and type of volition involved in migration decisions, we also need to further unpack the forced–voluntary spectrum. Our focus is thus on migration decisions, honing in on the question of when someone – in their own experience, via scholarly observations or through an immigration authority label – might be described as having been ‘forced to leave’, and how this matters analytically and discursively.

Our hope is to advance analytical discussion of voluntariness, choice and alternatives, whilst also seeking to tackle the stickiness of the forced–voluntary migration dichotomy from an analytical angle. Undertaking this is no simple task, and we aspire to encourage discussion, rather than to provide answers. Addressing the distinctions and similarities between forced and voluntary migration is fraught with dilemma. Unsettling the common understanding of a clear-cut ‘forced migrant’ may contribute to undermining the international refugee regime. However, it has also been argued that a presumed binary between forced and voluntary migrants has been instrumental in creating the figure of the ‘illegal migrant’ and ‘bogus asylum seeker’ – those not deserving of international protection (Casas-Cortes et al. 2015).

This article is positioned as a discussion piece with applications beyond any one migration context. That said, it is shaped by our longstanding ethnographic engagements with two migrant groups: one usually labelled as forced (Afghans) and one generally labelled as voluntary (Pakistanis). It is also shaped by the time of writing, as the 2015–2016 European ‘crisis’ over refugees and migration was unfolding. We start by investigating the effect of labelling and how labels can have discursive significance. We then discuss some of the ways the concept of volition has been addressed by scholars of migration and other disciplines. This is followed by a brief overview of Afghan and Pakistani migration, setting the stage to examine how voluntariness, choice and alternatives in migration decisions are experienced by migrants, observed by scholars and labelled by immigration authorities, as well as how they intersect and what their implications are.

The significance of labelling migration

Writing about conceptualising migration and displacement, Bakewell (2011) makes a strong case for unpicking the usage of ‘migration’ and ‘displacement’ as terms to describe a process, a condition and a category. This means that a migrant might be someone who is doing migration, who is being a migrant or someone who is being described as a migrant. Moreover, different actors may use the term ‘migration’ (or ‘displacement’) in these different ways without considering the implications of the different usages (2011).

Building on Bakewell’s (2011) suggestions, we use analytical and discursive significance as a way to differentiate the implications of labelling migration as forced or voluntary. Following Foucault (1972 inter alia), discourse represents a way of thinking, which then has a significant effect on the social construction of ‘reality’ (e.g. through language). Crucially, discourse also checks the realms of possibility, what is and is not seen as acceptable (Butler 1997). In migration and refugee studies, the discourse and ‘labels’ (Zetter 1991, 2007) to describe migrants sway the selection of responses – which are ‘acceptable’ and which are even possible – to those migrants. Whether someone is discursively presented as an economic migrant or a refugee, for instance, majorly influences their treatment by immigration authorities and humanitarian actors. As Zetter (2007, 173) notes, labelling people as ‘refugees’ not only serves a descriptive purpose, but can also explain ‘the complex and often disjunctive impacts of humanitarian intervention’. In other words, a key part of the discursive significance of identifying people as refugees, vis-à-vis other migrant categories, is how they will be treated.

If we accept the discursive (and normative, see Ottonelli and Torresi 2013) power of labelling someone as a forced or voluntary migrant, then we should also question those descriptions, especially being written, as they are, by powerful governments and international institutions. Are they accurate? Are they appropriate? As migration scholars, we should not just accept the discourses handed to us, but also recognise our own roles in creating, influencing and maintaining or disrupting them. In addition, we should question whether a discourse represents ‘reality’ as we see it through fieldwork observations and other research interactions. It is here that we get insight into migrants’ experiences and their own interpretations of experiences, which are inevitably produced through migrants’ execution of their agency in relation to prevailing constraints.

In referring to discursive and analytical significance, these are the dimensions we examine, including, on the one hand, the labelling of individuals into bureaucratic categories. On the other hand, there are the observed and experienced realities that are analytically produced through data collection and analysis. Encompassing migrants’ lifeworlds and the role of agency, these dimensions necessarily become nuanced and complicated.

We acknowledge that by focusing on the forced–voluntary continuum, we run the risk of reifying extremes. To avoid doing so, we take a critical approach that questions not simply ‘causes’, but also responses, as they are mediated through migrants’ agency and, in turn, do or do not trigger states’ duties to protect individuals according to international obligations. The role of agency is crucial, as virtually no migration happens without a great deal of it in place. Nevertheless, migrant agency is often seen as suspect by restrictive immigration regimes, such as those seeking to control migration from Africa, the Middle East and Asia to Europe as well as from South America to North America. Following De Genova (2016), those labelled as refugees may be pitied as ‘victims’ of forced displacement whilst they reside in camps in neighbouring countries; yet, as soon as they show more ‘entrepreneurial’ agency by choosing to leave the camp and head farther afield, they become suspect, labelled as ‘illegal migrants’ or ‘bogus asylum seekers’, reflecting the discursive power of migrant labels and categories.

Despite their thematic dovetails, studies on conflict-related migration and migration, in general, remain poorly connected. Within the sub-discipline of forced migration studies, discussions about description – whether as forced migration studies, refugee studies or the study of conflict-related migration, war-time migration and war mobilities (see Harpviken 2009; Hathaway 2007; Lindley 2010; Lubkemann 2008a, 2008b) – reflect the definitional challenges of pinning down ‘forced’ migration as something analytically distinctive from ‘voluntary’ migration. What may be pinned down, though, is the usual context of violent conflict, which frequently proxies for a specification of which contexts are associated with forced migration and which are not. This has also been critiqued by scholars of such contexts, for example, Liberia (Lubkemann 2008a).

The literature in both migration studies and forced migration studies is somewhat limited in its analysis of the implications of volition on migration processes, despite forced–voluntary being a key dichotomy in typologies of migration (notable exceptions include Fussell 2012; Ottonelli and Torresi 2013). In forced migration studies, much of this debate has concerned the implications of expanding ‘refugee studies’ as a field focusing on those outside their country of origin due to a well-founded fear of persecution (as defined by the 1951 Refugee Convention), to ‘forced migration studies’ as a field that incorporates a much more diverse range of displaced people.¹ This has led to interesting discussions about the term ‘forced’, but the focus has been on the difference between being ‘forced’ by political factors recognisable under the 1951 Refugee Convention versus other factors (e.g. economic or environmental), and whether forced migration necessitates crossing an international border, or includes internal displacement.

Meanwhile, in mainstream migration studies, theorising around why people migrate has largely focussed on migrants assumed to be migrating voluntarily, rather than looking at the forced–voluntary spectrum as a whole. However, there is also an emerging limited body of work bringing to attention the counterpart to forced migration, namely ‘involuntary immobility’, where volition is a key concern (Carling 2002; Jónsson 2008; Lubkemann 2008b).

The difficulty in pinpointing whether someone's migration was entirely voluntary or entirely involuntary has led to a small, but compelling body of literature, which explores volition in more depth (see e.g. Bartram 2015; Long 2013; Ottonelli and Torresi 2013). In the following section, we draw on these studies as well as some non-migration research on the concept of voluntariness (see Colburn 2008; Olsaretti 2004) to examine current academic understandings of volition.

The role of volition in migration decisions and processes

‘Voluntary’ – meaning acted by choice, with free will and without compulsion – is a term regularly used to refer to migration types. However, there is a range of ways to understand it. Do we understand voluntary in terms of a totally free choice from a range of options? Or do we understand it as choosing the least worst choice from a limited or virtually non-existent range of options?

Whilst acknowledging that some people may migrate out of the ‘joy’ of mobility, for adventure and to see the world, the majority choose to migrate because they believe that physically moving to another place will lead to improvement in their and/or their family's living situation. ‘Improvement’ may take diverse forms (a better job, being closer to family, escaping a war zone, more pleasant weather, etc.). Consequently, a starting point for understanding volition in migration is the range and quality of alternatives available to potential migrants if they just stay where they are. In other words, to what extent will they be able to enjoy a reasonable quality of life without migrating? We might consider the migration less voluntary when the answer is ‘not at all’ rather than ‘to some extent’. The perception of suitable options and necessity of alternatives – and the notion of a ‘reasonable quality of life’ – are subjective. Still, it is the premise of much conceptual literature, some of which we examine here, exploring voluntariness outside of migration studies.

For example, if we take a worker in a ‘3D’ (dirty, dangerous and demeaning) job, we could argue that as long as they are not a slave then they are doing that job voluntarily. In neoclassical economics, a worker voluntarily gives labour in exchange for income because the alternative – no income – is worse. But is this acceptable? Is it enough to say that a worker in a 3D job is doing that work voluntarily because they are not enslaved, when the alternative to doing that job is perhaps destitution? For the libertarian philosopher Nozick (1974), voluntariness depends on non-infringement of rights. His case is that if another individual or the state's actions infringe on a person's rights, then we can consider their consequent choices coerced, rather than voluntary. Taking this approach, it is clear that the observer's view on what constitutes ‘rights’ will have an important bearing on whether they view actions to be coerced or voluntary.

To return to the example of the 3D worker, if we consider it the right of that worker to receive an income (for example in the form of welfare payments) by the state if they are unable to find acceptable non-demeaning work, then a lack of that alternative income might lead us to conclude that they were coerced into do that 3D job. This is a very different stance from Nozick (1974) whose view of rights minimised them to rights to life, liberty and property, and whose work suggests that he would have taken a dim view of the state taking such an expansive role as to provide welfare assistance (Kymlicka 2002). If we apply Nozick’s (1974) understanding of voluntariness to migration, and if the potential migrant's rights to life, liberty and property were being infringed, then their migration would be non-voluntary. A case in point would involve the archetypal ‘convention refugee’, for example, an intellectual forced to leave Afghanistan in the late 1970s because the alternative was being imprisoned by the Communist government for their political views. This scenario does not portray a migrant leaving because they cannot find employment and, according to Nozick, they have no right to expect an external actor to provide employment in their current place of residence. The migration would thus be voluntary, even if the alternative was destitution, because the choice was made ‘freely’.²

Like Long (2013, 160), we see Olsaretti’s (1998, 2004) more expansive view of voluntariness as an improvement on Nozick's minimalist approach to rights. Whereas Nozick holds that someone who is ‘free’ (i.e. their rights to life, liberty and property are not being threatened) means their choices are voluntary, Olsaretti (1998) argues that freedom and voluntariness do not necessarily map onto each other. She illustrates this point through comparative examples, which we adapt here using migration-relevant scenarios.

Example 1: Amira is an inhabitant of Syria, which she is free to leave. However, although she wants to leave, Amira knows that if she leaves she will not survive the hardship of the journey out of the country and she will die. Her choice to remain in Syria is not a voluntary one.

Example 2: Bisrat is a national service conscript in Eritrea, meaning he is not free to leave the country. However, Eritrea has all Bisrat feels he needs and he does not want to leave, so he remains voluntarily. (after Olsaretti 1998, 71)

Olsaretti (1998) posits that a choice is voluntary when it is made in the context of acceptable alternatives or if the lack itself of alternatives is acceptable to the person making the choice. Bisrat does not make a choice based on a lack of alternatives because he is not concerned with the lack of alternatives; he is content with the alternatives he has – this determines whether or not his choice is voluntary. The above scenarios illustrate how freedom, the presence of acceptable alternatives and voluntariness do not necessarily coincide. Amira is free but staying involuntarily; Bisrat is not free but staying voluntarily. Both Nozick and Olsaretti take a philosophical approach to volition, aiming to demonstrate that a choice is either voluntary or involuntary. This contrasts to empirical research by migration scholars, which reinforces the difficulties in drawing such sharp distinctions. However, Olsaretti’s (2004) argument – that choices must be made not only freely, but also in the context of acceptable alternatives – has been echoed by migration scholars (see Bartram 2015; Ottonelli and Torresi 2013).

To Olsaretti's work, Colburn (2008) makes an important addition: that the perception of acceptable alternatives is shaped by a person's beliefs and access to information. Acceptable alternatives are instrumental to whether a choice is voluntary or not (Bartram 2015). Migrants themselves, immigration authorities, migrant rights’ activists (and other civil society, non-governmental organisations and UNHCR) also each have ideas of what is acceptable, as illustrated by the international refugee regime. The disjuncture between what is and is not acceptable is starkly illustrated by refugee status determination decisions that posit internal relocation to ‘safe areas’ as an acceptable alternative to asylum, thereby enabling the removal of rejected asylum seekers to insecure countries, such as Afghanistan.

This is the basic premise of Bartram’s (2015) article, which points out that the majority of refugees are not physically forced to leave. Rather, finds Bartram, certain implications of staying (for example, being forced to change religion or being made a political prisoner) are recognised as an affront to human dignity and as unacceptable alternatives to fleeing, whereas other implications (for example, extreme poverty) are not seen in the same way. Thus, volition in relation to migration decisions is closely tied to available acceptable alternatives and the agency to act on those options. Whether someone's migration is labelled as voluntary or not, however, depends on the labellers’ perception of what constitutes ‘acceptable’ alternatives. Ottonelli and Torresi’s (2013) and Bartram’s (2015) proposition of voluntariness results in many more of the worlds’ migrants – international and internal – falling in the non-voluntary bracket than the 1951 Refugee Convention covers.

Another crucial insight from Olsaretti’s (1998) work that has not yet been embraced by migration scholars is the distinction between primary and secondary choices. She argues that just because a choice is made voluntarily (or involuntarily) does not mean its subsequent related decisions are necessarily voluntary or involuntary. An example is given of a worker who might involuntarily work in an exploitative industry but might voluntarily choose to work for one employer over another; the voluntariness of the second-order choice does not prescribe the voluntariness of the original choice. Relating this to migrants, we may consider those who voluntarily leave their country of origin but then involuntarily put themselves in the hands of smugglers because there are no acceptable alternatives. Or, another example would be migrants who are forced to leave but then make voluntary choices about their mode of travel (Koser 2008). Such distinctions prove crucial because one of the main critiques of ‘forced migration’ is that all migrants exercise agency in their decision to migrate. However, following Olsaretti (2008), migrants exhibiting agency in their choices of how (or when or where) to migrate does not negate the possibility that they migrated involuntarily in the first place.

Afghan and Pakistani migration

The remainder of this article draws on our longstanding empirical engagement with Afghans and Pakistanis in Europe (specifically, the United Kingdom and Norway) (see Erdal 2012; Erdal and Oeppen 2017; Oeppen 2013). To reiterate, we endeavour to unpack where and how the distinction between forced and voluntary migration matters vis-à-vis migrant experiences, scholars’ observations and labelling by immigration authorities. In this section, we outline our involvement in the field and briefly comment on the nature of Afghan and Pakistani international migration.

In the summer of 2001, as undergraduate students, we together conducted fieldwork among Afghan refugees in New Delhi (Bivand and Oeppen 2002). Our questioning in this article of categorising and labelling ‘forced’ and ‘voluntary’ migration was initially contextually and temporally informed by our studies: that fieldwork and a module called ‘The Global Refugee Regime’. The urban refugees we engaged with in Delhi, mainly middle-class Afghans who fled Kabul after the toppling of the Soviet-backed regime in 1992 or during the ensuing civil war, were all seeking resettlement through UNCHR programmes. In striking contrast to media images of vulnerable fleeing masses, these individuals showed agency and were obviously on quests for a better life – not in India, which did not allow them to gain legal employment, but elsewhere, notably preferring North America, Scandinavia or Australia. Talking to our research participants about their future plans impressed upon us the need to examine migration as a process, a cycle or a trajectory, rather than a one-off event.

Broadly speaking, Afghan migration is often labelled as forced migration and Pakistani migration, as voluntary. These general categorisations stem from the region's recent history. Afghanistan has experienced war since 1979; against the Soviet Union in the 1980s, civil war in the early 1990s, followed by the Taliban regime until 2001 and the NATO-led military presence until 2014, and on-going insecurity. The number of casualties and level of displacement nationally and internationally concretely expose the high costs of these conflicts. For Afghan populations, these tolls provide the inevitable backdrop for the discursive use of the forced migration category and are a reminder of the undeniable persecution fears of many Afghans seeking asylum in Europe. Whilst the conflict dominates discussions around Afghan migration (Harpviken 2009), it is also predated by a historical culture of Afghan migration and mobility, largely connected to international trade networks (Marsden and Hopkins 2011).

By contrast, recent Pakistani migration has been dominated by regional male labour migration to the Gulf States, involving both short-term contracts and longer term migration. Pakistani migration has also included European and North American destination countries. That said, Pakistan is experiencing low-intensity conflict in several regions, most notably in Baluchistan and, since 2009, in the northwest. This violence along with the effects of the ‘Global War on Terror’, including drone attacks, have taken a heavy toll on civilian populations, resulting in deaths, injuries, disabilities and psychological harm.

The current situations in Afghanistan and Pakistan thus do not lend themselves to a simplistic divide in which migration from one is ‘forced’ and migration from the other, ‘voluntary’. Rather, the categories must be unpacked, analysing degrees of volition in individual cases and how forced and voluntary are constructed. In the next section, we use a stylised migration trajectory as a framework to explore the analytical and discursive significance of describing migration as forced and voluntary at different stages. There are of course limitations to seeing migration as a trajectory (or cycle, Black and Koser 1999) with identifiable stages, rather than a more fluid on-going process. For the purposes of our analysis, however, we found it useful to break it down into stages.

Stages of the migration process

Examining the decision to leave on a forced–voluntary continuum, we consider stages of migration including leaving, the journey, entrance, settlement, return and remigration, onward migration. We explore where and when differing descriptions of volition matter for the individuals involved, as we observe them and how they themselves describe their experiences, but also in terms of categorisation by immigration authorities. Aiming to move beyond merely accepting that the distinction is complex, we note how and where the distinction is more or less relevant and where there are intersections of experiences, observations and labelling as forced or voluntary migrants.

Of course, there are plenty of Afghans in Pakistan and Iran, and migrant workers from both Afghanistan and Pakistan find their way to the Gulf States and farther afield in Asia. Nevertheless, our fieldwork has been about migration to Europe, so we focus upon that trajectory here.

Leaving

As reflected in our title question, ‘forced to leave?’, it is the circumstances and reasons for leaving that unlock potential protection from the international refugee regime. This is a fact despite migrants’³ trajectories being complex and migration reasons being likely to evolve in the run-up to, during and after the actual decision to leave, where the degree of volition and the alternatives available will vary, for the initial decision, as well as en route. Reasons to leave are always compound – comprising reasons that are experienced, observed and labelled – and result in varying analytical and discursive constructions. If a broad interpretation of voluntariness is applied – where a lack of alternatives is the criteria – a different discursive construction of what constitutes being ‘forced to leave’ emerges, than if it is the 1951 Refugee Convention criteria which are applied, or indeed broader ‘humanitarian leave to remain’ definitions.

Consider, for example, Amir,⁴ a young Afghan going to Europe. His decision to leave, as he experiences it, may be forced on several levels: due to the on-going conflict and insecurity in Afghanistan, but the decision to leave might have been taken by his parents, thus adding to his experience of feeling forced and without alternatives.

However, because of that on-going conflict, millions of Afghans have been displaced to neighbouring countries, particularly Pakistan and Iran. Over 35 years of conflict means many Afghans have been born and brought up in these countries. Amir's family lives in Pakistan, where they experience insecurity due to discrimination and insufficient opportunities for livelihood (Human Rights Watch 2015). Thus, at least three reasons compel Amir to experience his leaving as forced: the conflict in Afghanistan; the decision being taken by his parents; his immediate insecurity in Pakistan.

Meanwhile, within Pakistan, albeit on a different scale than in Afghanistan, there is displacement due to insecurity, low-intensity conflicts, earthquakes and flooding. Thus, when Parvez leaves Pakistan, his experience is that he has no alternatives, other than to leave, due to insecurity, the misfortune of prematurely losing his father and lacking contacts who might have helped him secure a livelihood locally.

Amir's and Parvez's experiences of their decisions to leave focus on lacking alternatives, with conflict, family circumstances and livelihood opportunities playing a role. From our point of view, their decisions to leave might be analytically constructed as forced due to lacking alternatives. Yet, when it comes to their labelling within the international refugee regime, there are certain distinctions between the two cases. Labelling by immigration authorities in countries of asylum varies, as individual histories and experiences are scrutinised; an assessment may start from an assumption that Parvez's migration is not forced and Amir's is, based on their country of origin. Another difference in the circumstances that frame their departures is the hope of the migration being labelled ‘forced’ and so being able to apply for and possibly obtain asylum, which Amir nourishes, whilst Parvez sees as unattainable.

The decision to leave might be more or less forced depending on how we interpret available alternatives. Do we expect alternatives to exist at merely the level of survival, or beyond? Analytical and discursive constructions of leaving as forced depend not just on the experienced and observed levels of voluntariness, but also on the anticipation of particular labelling by immigration authorities in Europe. It is worth noting that many of our research participants, from both Afghanistan and Pakistan, felt forced – in some sense of the word – to leave.

The journey

For Afghans and Pakistanis able to obtain a visa, movement is relatively easy, usually involving flying to Europe, as for the small number who comprise resettlement quotas. For Amir and Parvez, unable to access means to travel and enter another country legally, the journey itself becomes a dangerous, clandestine endeavour (Koser 2008). The journey is where the differences between Afghan and Pakistani migrants are the smallest. Whether the decision to leave is constructed as forced or voluntary makes little difference at this stage. For those who have no legal means to enter Europe, differences in journey experiences are related to financial resources (and other forms of capital), rather than whether their decision to leave was more, less or not at all forced.

Attention has increasingly turned to the journey itself, including migration brokers and the migration ‘industry’ more broadly (Baird and van Liempt 2016; Gammeltoft-Hansen and Sorensen 2013; Xiang and Lindquist 2014). In our encounters with Afghans and Pakistanis, such as Amir and Parvez, we found little difference in modes of arranging mobility through migration brokers (see also Ahmad 2013). The business of ‘fixing’ migration enables people to obtain fake passports and visas and to use what are essentially travel agent services that organise physical transport, border-crossings and the circumvention of border controls. Using the services of ‘fixers’ and smugglers does not make the journey itself more or less forced by nature – it is simply the means to enter Europe (Koser 2008). For migrants who are trafficked, the issue of volition is to a greater extent taken out of their hands. However, none of our research participants were the victims of trafficking, so this particular issue falls outside the scope of this article.

The journey, usually interrupted along the way, constitutes instances of mobility and immobility (Collyer 2007; Crawley et al. 2016). Routes, like those followed by Amir and Parvez, would run through Turkey, into southern Europe and Greece, and an extended stay in Turkey or Greece would not be uncommon. On the Turkish side, most transit migrants would seek to cross the border and not make efforts to regularise their status. For the majority of migrants from Afghanistan and Pakistan, entering Europe happens without valid entry visas, making their journeys precarious. They tend to circumvent border controls, planning once in Europe to, as in Amir's case, apply for asylum or, as in the case of Parvez, who lacked clear strategies but had clear aspirations, become regularised (Abdin and Erdal 2016; Carling and Collins 2018).

Migrants’ own experiences of a forced decision to leave remain significant during the journey, being tied in with hopes for eventually applying for asylum. During the journey, the degree of volition in the decision to leave is of little analytical significance. From the immigration authorities’ point of view, it makes a potential difference, but only in relation to a future asylum claim, which may or may not be filed. We see therefore how during the journey, distinctions between forced and voluntary migration effectively become insignificant. This underscores how immigration controls in Europe have an overarching effect on the discursive construction and survival of these labels. Migrants’ assessments – of their own experiences and, equally, their chance of convincing immigration authorities of the forced nature of their decision to leave – are salient for onward migration beyond transit. Thus demonstrated are the intersections of experienced, observed and labelled dimensions of volition in migration decisions.

Entrance and settlement

When Amir arrived in Sweden, he applied for asylum at the local police station. Entrance is where the distinction between those who do and do not stand a chance of being labelled as forced matters. Amir and Parvez were both in Greece for extended periods, but for Amir it was always a transit stay and for Parvez it became the destination. Both could seek asylum in Greece or pursue further travel and seek asylum elsewhere in Europe. For Amir, it was important to make sure he did not apply for asylum in Greece, in light of the Dublin Regulation (Koser and Kuschminder 2015).

Both men's stay in Greece was, in practice, similar. Their future prospects differed, however; Amir believed it possible to apply for asylum in Europe, but for Parvez, this was not relevant. Meanwhile, both felt a responsibility towards their family, wanting to remit money to settle the debt with their migration brokers. For Parvez, a future plan was to go to Spain, joining relatives who had regularised their status there. Parvez's cousin was a student in Norway, having obtained a scholarship to pursue a degree there, which sensitised Parvez to the existence of alternative routes to entry, although they were not accessible to him.

Thus, entrance is a relative construction. For Amir, formal entrance was in Sweden, where he applied for asylum. For Parvez, formal entrance might never happen or might be found in Spain in the future. In both cases, the location of existing social networks mattered. For many journeying Afghans, like Amir, the stages of transit and entrance become conflated, as there are multiple instances of ‘entering’ before the final ‘entrance’ and the filing of an asylum claim. Here, European migration management is crucial in shaping the migration process of individuals, often more so than whether or not the decision to leave was experienced – or observed – as forced or voluntary (Ghosh 2000).

Settlement is regularly conflated with the stages of transit and entrance as these actions occur repeatedly. Here, we focus on the process of settlement occurring when a destination, however it may be defined, has been reached. For Afghans who, like Amir, apply for asylum, the first period during which the asylum claim is being processed is one of uncertainty. The salience of state systems of evaluation of asylum claims is key to how experiences of being ‘forced to leave’ are performed (Kea and Roberts-Holmes 2013), in response to the understandings of persecution put forward by bureaucracies. This is a crucial intersection of experiencing, observing and labelling decisions to migrate as forced or not – in the outcome of an asylum claim, there is little room for debate: asylum is either granted or not. You are either defined as a ‘forced migrant’, or you are not.

For those who get asylum, settlement in many European countries follows a fixed strategy comprising language and skills-training programmes that aim to boost refugees’ financial and social self-sufficiency over time. For the rejected, regardless of their experience of a migration decision as forced or not, return migration, onward migration and forced return are common. That is, provided they do not ‘go underground’. For Pakistani migrants entering southern Europe, like Parvez, few roads lead to legal residency. Instead, migrants’ own entrepreneurship and networks prove defining of settlement processes, determining how they find a way to earn a living and a place to stay (Ahmad 2013).

In practice, for Afghans with rejected asylum claims, strategies and experiences are similar to those of Pakistani migrants without a regular status in Europe. The analytical construction of migrants with rejected asylum claims being ‘forced to leave’ depends not only on the actual experienced or observed volition and available alternatives. To a great deal, it also depends on the interpreted legitimacy of immigration authorities’ labelling of migrants as forced (with access to protection granted) or not. Meanwhile, migrants, like Parvez, are more often than not analytically constructed as voluntary; this is a direct result of being labelled as voluntary in migration management terms, rather than evaluating experiences and observations of available alternatives and thus a migrant's volition in deciding to leave. Authorities’ power to label migrants, based on evaluations of the nature of their decision to leave, impacts analytical and discursive constructions. Bureaucratically, the frame of what is forced is relatively narrow and specific, leaving few descriptive shades between black and white. It is a stark contrast to any experienced and observed realities, which are far more nuanced.

Return or onward migration

For both Afghans and Pakistanis in Europe, the issue of return is less related to the voluntariness of their initial migration and more related to immigration status. With a time-restricted student visa in Norway, Parvez's cousin has a return for a specific date; if she does not respect it, her status will become irregular and she risks deportation. For Parvez, living as an irregular migrant, staying may be a question of how likely he is to be detected and deported, but it is also a question of family matters and how well he can cope. Many migrants do return to their origin country, at least temporarily, and some circulate between countries over time. For Amir, who was granted asylum, return or onward migration is a question of voluntariness, within some constraints, where the evaluation of his initial migration as forced proves instrumental, for if his asylum claim had been rejected he would have become deportable.

Experiences and observations of decisions to leave as more or less forced are relatively insignificant in the face of immigration authorities’ return policies. For them, it is the evaluation of someone's decision to leave as forced, deemed valid according to set criteria, which makes a difference. However, analytical and discursive constructions as forced and voluntary may have an effect on public sympathy and support, as seen, for example, in anti-deportation campaigns (Gibney 2013).

On the whole, migrants like Parvez and his cousin would not be discursively constructed as deserving of public sympathy in the face of deportation. Meanwhile, migrants from Afghanistan, with its well-known insecurity, would likelier be seen as victims of incorrect evaluations concerning whether or not migration was forced. The distinction between forced or voluntary decisions to leave in the first place thus sticks throughout the asylum-determination process, and has a great impact on Afghans with rejected claims. Europe has also seen a rise in Pakistanis who file asylum claims, which frequently get rejected and lead to deportations, or who are intercepted as irregular migrants and become subjected to deportation.

Migration is neither a linear process nor, necessarily, a circular one (Harpviken 2014). Among those with refugee status in Europe or, more broadly, third-country nationals acquiring European citizenships, onward migration within Europe as well as to North America is relatively common. Amir contemplated the idea of one day joining relatives in California.

Meanwhile, onward migration regularly occurs among those without legal documents, sometimes seeking opportunities of legality. We see this among Pakistani migrants moving towards Spain or Afghans moving to Italy and applying for asylum there to access mobility within the rest of the Schengen Area.

For those migrants deported from European countries back to Afghanistan or Pakistan, onward migration might essentially be remigration to Europe. Often this involves migration between Afghanistan and Iran or Pakistan whilst preparing for remigration to destinations further afield (Majidi 2009). Among Pakistani migrants who are forcibly returned to Pakistan, remigration to Europe as well as labour migration to the Gulf States is common (Koser and Kuschminder 2015).

It is worth noting, too, that what might start as a voluntary migration could involve forced onward migration, or vice versa. For example, West African migrants arriving in Europe, having been forced to leave Libya in recent years, where their initial migration was more voluntary, but their secondary migration was shaped by the conflict in Libya. In such cases, the importance of countries of origin and nationality may be over-emphasised in labelling migrants as forced or voluntary, particularly for refugee status determination.

Conclusion

Embarking on this article, we stated that it is not enough to point out the complexity of the forced-voluntary continuum. Whilst many academics have rightly noted that defining an individual as a forced or a voluntary migrant is highly problematic (Castles, De Haas, and Miller 2014; DeWind 2007; Fussell 2012) – and even more so when describing migrant flows en masse – the distinction sticks in academic work and public discourse. Following Bakewell (2011) and Zetter (1991, 2007), we believe that semantics are relevant to consider, as are considerations of who is doing the labelling and what the discursive work of those labels entails. There are also important distinctions between self-identification and imposed labels; an analytical description of an individual's migration decisions along a forced-voluntary continuum is necessarily more reflective of the complexity of individual experiences, agency and contextual circumstances than a binary labelling of forced or not. Yet, analytical descriptions of the processes of migration decisions are part of a broader landscape, within which we must acknowledge that discursive constructions of migration, and the labels attached to them, carry a ‘normative weight’. How a migrant is labelled may have significant implications on treatment from receiving states (see Ottonelli and Torresi 2013).

An analytical understanding indicates that by applying criteria of voluntariness and available alternatives, many migration decisions are not seen as entirely voluntary. A discursive construction of ‘the migrant’ or ‘the refugee’ veers away from the complexity of reality and reifies the binary of forced vs. voluntary; it turns it into two boxes between which migrant experience necessarily can and must be divided. The discursive significance of describing migration as forced or voluntary matters to states, as to whether or not individuals are seen as legitimate asylum seekers, with the right to international protection. Analytically, the salience of describing migration as forced and voluntary lies in the intellectual pursuit of better understanding the role of volition in migration decisions, under differing conditions. In analytical terms, resisting the dichotomy of forced vs. voluntary, which largely sticks for migration management reasons, should lead neither to using the nebulous ‘mixed migration’ term nor accepting the discursively constructed binary of ‘forced’ and ‘voluntary’.

A state may have legitimate reasons for approaching border control, migration management and humanitarian protection as it does, and these may stray from an individual migrant's perspective. Forced and voluntary migration is also a distinction made by migrants themselves, whereby someone fleeing a violent conflict may be recognised by other migrants as having no choice. Despite feeling they had no other acceptable alternatives, but to migrate in pursuit of a life that could be lived, some migrants may effectively have had more of a choice. And here the significance of how migration is described returns: whether a migrant is labelled as forced or voluntary within an immigration system majorly impacts their access to protection.

We readily recognise that many migrants forced to flee the conflict in Afghanistan have legitimate needs for protection. We also acknowledge the unsettling idea that questioning migration categories may contribute to destabilising an already eroded refugee regime in Europe (Koser and Martin 2011). Yet, the discursive and analytical implications of describing migration as forced and voluntary need to be scrutinised. There must be better ways of safeguarding needs for protection than the world currently offers. One course of action might be for scholars to challenge the status quo more, not only by deconstructing government labelling, but also by further unpacking the assumed dichotomy between forced and voluntary migration, thereby examining voluntariness.

As Ottonelli and Torresi (2013) caution, discussions of the forced–voluntary migration continuum risk becoming circular if policy hinges on particular interpretations of volition in migration, whilst scholarly approaches to volition are also built on those very policy definitions. We therefore support the stance that ‘[i]f other principles than the recognition of migrants’ agency must come into play, they need to do it openly, rather than under the guise of an unorthodox and inconsistent notion of voluntariness’ (Ottonelli and Torresi 2013, 810).

The purpose of describing migrants as forced and voluntary analytically and discursively is different in intent, but it is also different according to scale. In migrants’ experiences as well as scholars’ qualitative, field-based observations, the analytical unit is often the individual or family. For the state, whilst protection is determined individually, the scale of the analytical unit is that of groups: be it of refugees from particular conflict areas or another mass of people feeding into the broader category of ‘immigrants’. States’ migration management efforts are not individual-level; they target collectively. Thus there is a discrepancy in the scales to which attention is being paid reflective of the primary interests of different actors: of controlling populations on the part of states; of finding safety and quality of life, on the part of migrants; and for scholars, analytically trying to understand what is going on and why.

The use of ‘mixed migration’ makes us, like many others who study migration, uncomfortable. What we see empirically is that people migrate from both Afghanistan and Pakistan for a diverse set of reasons. Different degrees of volition are involved, and in many cases across both countries no acceptable alternatives to migration can be found. If emphasising that mix means that migrants in need of protection are less likely to receive it, or must go through multiple stages of ‘proving’ that they need it, then emphasising the fluidity of the forced–voluntary continuum can have negative implications for our research participants (Bakewell 2011, 15).

However, as migration scholars we have an obligation to move beyond a statement acknowledging complexity. We must dare to grasp what has so far been analytically difficult, even if that means destabilising the fragile and dysfunctional refugee regime. The current need for humanitarian assistance and relief to people fleeing violent conflict in Europe's neighbouring areas is obvious. Meanwhile, enormous resources are invested in seeking to control Europe's borders, to limited avail. Yet, asking the obvious question – what if border control was not the primary focus, but rather the lives of those in desperate humanitarian circumstances? – is not politically permissible in the public debate in European countries.

Whilst states may have particular priorities relating to control and management, scholars of migration should be free to choose which perspectives are best to apply. A state-centred analysis is arguably no more valid, reasonable or scientific than one emphasising the experiences of migrants who are differently placed along the forced–voluntary continuum. Some are in need of protection; all are in search of a future in which life may be lived.

Acknowledgements

The authors are indebted to our Afghan and Pakistani research participants in various locations over the years, for their trust and willingness to share experiences and stories. Also, to Khalid Koser and Tony O’Connor, who in their undergraduate lectures ignited an interest in questions of international migration and refugee protection, encouraged our first fieldwork plans, and provided analytical, as well as practical guidance. The authors are also grateful to the anonymous reviewers, and to the special issue editors Francis Collins and Jørgen Carling, for feedback which has helped develop our argument.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 For a key discussion in this debate, see Hathaway (2007) and responses to that article in the same issue of the Journal of Refugee Studies.

2 For a similar discussion relating Nozick's ideas to the notion of voluntary repatriation, see Long (2013).

3 For ease of expression, in this section we use ‘migrant’ as an umbrella term to describe people moving (including refugees and others labelled as forced migrants).

4 Names used are pseudonyms, which were created for compound characters. We collated details of migration trajectories from several individuals in order to present ‘typical’ examples, where the point is not any one individual's story, but the fact that such trajectories are shared by many migrants. We chose to focus more on young male examples, reflective of current migration flows to Europe from Afghanistan and Pakistan.