New mobilities across the lifecourse: A framework for analysing demographically-linked drivers of migration

Migration, along with fertility and mortality, is one of the fundamental drivers of population change. Taking the lifecourse as the central concern, the authors set out a theoretical framework and define some key research questions for a programme of research that explores how the linked lives of mobile people are situated in time-space within the economic, social and cultural structures of contemporary society. Drawing on methodologically innovative techniques, these perspectives can offer conceptually significant and policy relevant insights into the changing nature and meanings of migration across the lifecourse.


Allan Findlay David McCollum Rory Coulter
Vernon Gayle ISSN 2042-4116 CPC centre for population change Improving our understanding of the key drivers and implications of population change New mobilities across the lifecourse: A framework for analysing demographicallylinked drivers of migration

INTRODUCTION
Migration is a fundamental driver of population change. It receives much attention from those seeking to monitor the scale of demographic developments, but conceptualising the importance of migration has received much less attention in population studies. This is especially true given the diverse disciplinary perspectives that have traditionally been taken in migration research, and as illustrated in a recent review of research on the relationship between migration and demographic change (Findlay and Wahba, 2013). This paper has been written in the context of the UK Economic and Social Research Council (ESRC) funding of a second phase of the Centre for Population Change (CPC). It takes as its objective an exploration of the conceptual contribution that researchers in CPC might achieve in extending their research on the relation between migration and demographic change. The paper is not therefore intended to offer an exhaustive review of the literature on the topic, but rather it aims to provide a possible conceptual framework for a contemporary programme of empirical research.
Given the demographic focus of CPC (Falkingham, 2010) it takes the relationship between migration and the lifecourse as the central concern (Bailey, 2009), and explores how the linked lives of mobile people are situated in time-space within the economic, social and cultural structures of contemporary society (Elder et al, 2003).
Before proceeding further it is important to discuss a few key terms and to narrow the range of interest that the paper deals with. First, it is recognised from the outset that human migration is only one form of mobility and the term has many limitations. In the text we often use the word mobility to acknowledge that our interests include certain population movements (such as student mobility between place of normal domicile and place of residence for the purposes of study) that some might not consider 'migration' in the strictest traditional interpretation of the term.
Second, it is important to position our intellectual curiosity relative to other social science research on 'mobilities' (Cresswell, 2012). Recent scholarship makes very clear that movement is not the same as mobility (or migration), and that if we consider every physical move within the term mobility then the term becomes meaningless (Adey, 2006). In this paper the focus is therefore on understanding migration as a 2 relocation of residence for diverse periods ranging from months to years across a person's lifecourse. By locating our interest in migration relative to the lifecourse, this necessitates a consideration of the relationship between moves motivated by very different forces (including labour migration, family-related moves and student mobility) as well as examination of the possible inter-relationships between these moves across an individual's lifespan. At the same time we exclude from our paper many of the fascinating new research findings emerging from the so-called 'mobility turn' (Urry, 2007) relating to moves that are part of a person's daily routine or involving limited duration relocation such as vacationing. Third, what we take from recent thinking about mobilities (Killick, 2012) is the important conception of 'mobility as a relationship through which the world is lived and understood' (Adey, 2010, 270), in the case of this paper in relation to migration and the lifecourse. The overall motivation therefore is to conceptualise how this type of thinking enlightens our understanding of the changing nature and meanings of migration across the lifecourse. Fourth, we acknowledge from the outset that the nature of the CPC's funding, means that inevitably our field of interest is particularly to advance understanding of migration in, to and from the UK in comparison with the experience of other western societies.
The paper commences by mapping some key changes in perspective that have been adopted by researchers studying migration. It then considers the opportunities offered in population studies to advance the understanding of the links between migration and the lifecourse as a result of the availability of rich new datasets, as well as new insights made possible by longitudinal research methods. Discussion then turns to introducing a simple schema for analysing migration within population studies.

CHANGING PERSPECTIVES ON MIGRATION AND MOBILITIES
Traditional demographic analysis of a population recognises fertility, mortality and migration as the three key drivers of population change. Migration as the third driver was somewhat of a poor relation that demographers often wished to ignore (for 3 example, in making population projections, the starting assumption has usually been that net migration is zero). Some demographers have even tried to define demographic regions in such a fashion as to minimise migration. This involved seeking to zone areas in such a way as to minimize inter-regional migration, thus reducing the complicating effect of migration on demographic change, and allowing researchers to focus more easily on the dominant roles of fertility and mortality (Hollingsworth, 1969). The last fifty years, however, has seen a shift in perspective, with recognition now being given by demographers to the key role of migration in driving population change.
Figure 1 suggests that over time there have been several profound switches of perspective in migration and mobilities research. Use of the term 'mobilities' signals the emergence of a 'movement driven' social science (Urry, 2007, 43) in which a new significance is found in studying interdependent mobilities, potential mobilities (motility) and virtual mobilities. For demographic studies what is arguably most important about this is that migration and other mobilities have become understood as relational -relational between groups of people (linked lives) who move together, relational between movers and non-movers (those left behind and those in the social groups at a migrant's destination), relational between migrants and those with power and resources in the housing and labour markets impacted by their departure/arrival, relational between the individual mover and the institutions to which she/he is linked (the transnational employer, the international banking system, the pension scheme) and relational between the migrant and those in the state who govern mobility. In all these contexts uneven power relations and knowledges are implicated (Bailey, 2009).

Figure 1: Changing Perspectives on Migration/Mobilities and Demographic Change
A second dimension illustrated in Figure 1 is the switch in perspective, from analysing migration as an isolated measurable event taking place at one point in time and involving the crossing of a physical boundary (such as an administrative census area or international frontier), to analysing migration as a set of economic and social practices. We conjecture that Denzin's idea of an epiphany is helpful in conceptualising migration. For Denzin epiphanies are interactional moments or a series of moments which leave marks in the individual's life (see Denzin 1989). The epiphany concept has been successfully deployed in empirical work, for example in an analysis of ageing in a former mining town in Northern England (Humphrey 1993).
Employing different language Turner (1986) adopts the term 'liminal phases of experience' to describe epiphanies in the lifecourse.
The exemplars listed in the previous paragraph capture just a few of the relational dimensions of migration that involve social practices whose meanings encode the ways in which economic and cultural forces can be understood to shape migration. Thus, for example, within phase one of ESRC's Centre for Population Change, researchers Shubin et al, 2014) began to explore the nature of the social practices involved in the recruitment of Latvian labour migrants to the UK. While not ignoring the significance either of the positioning of Latvia and the 5 UK within a European Union labour market that facilitated migration between national labour markets with very different wage levels, or the uneven effects of the Great Recession on the UK and Latvia , analysis of the social practices adopted by recruiters in the migrant selection process helped to extend understanding of how people are 'produced' as 'migrants'. Furthermore the social nature of the relationship was shown to result in migrants engaging in selfdisciplining practices in order to perform in ways that they considered to be compatible with images of what employers would expect of a 'good migrant' (Scott, 2013 presented the possibility of researching some of the social practices underpinning 'intimate mobilities' (Holdsworth, 2013). For the social scientist, researching mobility as an outcome of relational living has therefore become a very exciting possibility. This is true, not only in the context of residential mobilities and the family, but also in 6 relation to other types of linkages between the individual and the employer or the individual and the institution. The wider reframing of the nature of research has involved a change from the position in the 1970s and 1980s where research praxis assumed that migrants were 'out there' (Law, 2004) waiting to be discovered if only researchers could ask the right questions and use the most powerful analytic tools to elicit an understanding of the economic and social regularities shaping migration, to a contemporary emphasis by many researchers (for example in sociology and anthropology) on how the migrant is produced as subject. Consider for example the contrast between the research agenda on skilled international migration drawn up in the late 1980s by Findlay and Gould (1989) and the nature of contemporary research that analyses how the expatriate is performed as subject (Cranston, 2014). Researchers have therefore begun to study how culturally regulated behaviour amongst senior staff in transnational companies is produced through a range of actors (Faulconbridge and Beaverstock, 2010;Walsh, 2014). Moreover, the application of 'research' within disciplinary sub-fields such as human resource management illustrates how researchers have a role in co-producing the very world that we imagine we are trying to understand (Barnes, 2008). 7

Longitudinal Analysis
Longitudinal analysis has been identified in Figure 1 as one of the key shifts in research interest over recent decades, and we now focus on this new departure in analysing large scale survey data resources to illuminate our understanding of migration.
The UK has an unparalleled portfolio of large-scale longitudinal social science datasets, and these studies have been referred to as 'jewels in the crown' of the research infrastructure (Diamond, 2008). have also been coordinated efforts to make some of these cross-national data comparable (see Frick et al, 2007).
Following the explosion of longitudinal data there has also been a steady increase in the availability of suitable statistical methods which are appropriate to the analysis of both repeated measures and duration data (see Blossfeld, 2001;Diggle et al, 2002;Singer and Willett, 2003;Baltagi, 2008;Wooldridge, 2010). These techniques are now more readily incorporated in data analysis software packages and 8 open up the possibilities for social science research (see Singer and Willett, 2003;Gelman and Hill, 2006;Rabe-Hesketh and Skrondal, 2008;Long, 2011;Blossfeld et al, 2012;Heck et al, 2013).
Longitudinal surveys allow detailed consideration of the timing of migration in relation to demographic characteristics such as age, cohort and period effects. ways. First, it suggests that migration may not be a one off 'event', but may be oscillatory (as in some transnational households moving back and forth between households that are spread across countries), or in some complex inter-regional movements by couples living in long distance relationships (Reuschke, 2010). Other complex movements include the multiple residential mobilities of students in the early stages of entering the labour market, which Sage et al (2013) have shown to involve multiple temporary relocations between place of study, parental home and residence close to a new place of employment. Equally one can imagine skilled transients with a partner recorded as living in one global city, but with short-term relocations to multiple other global cities for work assignments of varying durations. Second, Figure 2 signals the potential of researching not only the effects of demographic and economic events (such as divorce, widowhood, redundancy) on migration (Feijten and Mulder, 2002;Feijten and van Ham, 2013), but also using longitudinal data to explore emotions (such as happiness) and social practices in relation to mobility (Nowok et al, 2013). Our purpose is not to rehearse once again the diverse and emerging research field arising from work on the timing of migration in relation to demographic concepts (e.g. the value of longitudinal datasets for studying mobility amongst older people (Malmberg et al, 2010)). However, it is useful to illustrate a few exemplars of work that have begun to show the significance of longitudinal research on age, cohort and period effects on mobility (although in practice these effects are often more complicated to tease out even with high quality longitudinal data). The propensity of members of a population to migrate has long been recognised to change with age as captured in model age migration schedules (Rogers and Castro, 1981). Cohort effects capture the commonalities of experience of similar people, for example those born at the same time and experiencing the same structural effects on their mobility. Period effects on mobility by contrast recognise that people of any age living in a particular time may exhibit distinctive mobility responses to the structural forces operating at that time (such as the impact of periods of economic prosperity or recession). For example, in relation to cohort effects in the Netherlands, Feijten and Mulder (2002) note the effect on residential mobility of a range of events including leaving the parental home, marriage and birth of the first child. They then use longitudinal methods to explore leads and lags in mobility in relation to specific cohorts. In a similar vein, Lundholm (2007), using Swedish data, examines the changing cohort effects of having children in inhibiting residential migration, with longer-distance commuting becoming more likely over time.
Period effects, have been suggested by Dunford and Fielding (1997) to account for variations in migration rates within the UK, with downturns in the housing market associated with periods of higher in-migration to the South East, while during boom periods Fielding (1993)

A framework for analysing demographically-linked drivers of migration
The academic literature has several existing frameworks for analysing mobility trends. One recent example based on an extensive study of migration in Britain (Fielding, 2012) makes a threefold distinction between 'conjuncture', 'restructuring' and 'deep structure'. While any framework is limited by its ontological and epistemological basis, Fielding's schema has the distinctive merit of highlighting the parallel, simultaneous influences of short and longer-term economic effects. His schema challenges researchers to seek to disentangle cyclical effects on migration from structural forces such as changing spatial divisions of labour and deeper and longer run cultural-economic trends.

MOBILITY AND THE LIFECOURSE
The lifecourse has been defined as 'an age graded sequence of socially defined roles and events that are enacted over historical time and place' (Elder et al, 2003, 15).
Much sociological research assumes that in the mid-twentieth century the lifecourse 12 in western society was organised in a remarkably linear fashion with individuals passing through a series of life stages (Rossi, 1955) starting at birth, spending childhood and the adolescent years in the parental home, before leaving home for marriage or proximity to employment. It anticipated that early married life would be followed by the birth of children triggering the likelihood of further residential mobility to find accommodation appropriate to family size. The subsequent departure of these children and retirement from work represented later stages in the life cycle.
An extensive literature exists that refines these ideas in relation to residential mobility (Brown and Moore, 1970). Others have critiqued the deterministic nature of a linear life stage model, recognising both the complexity of household decisions on mobility/immobility and also the diversity of mobility drivers (including education, the labour market and social and cultural forces relating to household formation and dissolution) that trigger mobility (Elder, 1994;Clark et al. 2006;Mulder and Hooimeijer, 1999 and linked lives) that explore the relation between mobility and the multiple transition in roles (and sometime reversals) that occur across the lifecourse. We choose to discuss each of the principles in more detail in the text below for two reasons. First, Elder et al (2003) gave surprisingly little attention to the mobility dimensions of the lifecourse, and second we stress these principles because they underpin fundamental demographic research issues that need to be evaluated with a stronger evidence base.
We illustrate below some migration and mobility questions that arise from thinking about each of the principles.
Rising life expectancy, in lengthening the lifespan, has not only introduced more years in the latter part of the lifecourse for residential relocations to take place in relation to people's leisure desires, as well as moves associated with adjustments to the vulnerabilities of the ageing body, but it has also led to questions about the role of migration and other mobilities in maintaining links between the generations across the extended lifespan (Lundholm and Malmberg, 2009 to explore differences in the physical and emotional well-being of migrants and nonmigrants, but much remains to be done in fully developing an understanding of the role of migration in the uneven 'accumulation of experiences, resources and vulnerabilities' (Bailey, 2009, 411) across the lifecourse.
In terms of human agency, Elder et al (2003, 11) point to the idea that 'individuals construct their own life course'. This has been applied within some areas of sociology (for example the sociology of youth) in the form of the concept of choice biography, which has its genesis in the work of the luminary Ulrich Beck (see Beck, 1992). In this approach there is a distinction made between normal and choice biographies. Normal biographies refer to the linear, and relatively predictable, move through the lifecourse (for example from youth to adulthood) which was seen to have The relative significance of human agency in migration is a matter that has been debated to some extent relative to ideas about structuration (Halfacree and Boyle, 1993;Halfacree, 1995), but the availability of rich new longitudinal datasets, as discussed above, does now open the prospect of evaluating more fully than even before the determining influence of the neighbourhoods within which people live on their subsequent migrations and life trajectories, as opposed the effects of human agency within the 'structuration engines'  of residential mobility or career migration.
Since time and place is a principle addressed directly in Figure 3 in terms of how the lifecourse is embedded historically and geographically, we do not discuss this further here, but turn to the principle of 'timing'. The most obvious significance of the second demographic transition (van de Kaa, 2004) for the study of migration has been that it has changed the timing and nature of mobility in relation to transitions from one household state to another. Longitudinal data now presents the opportunity to discover whether the changing meanings of, for example, home, family, pregnancy or parenthood relative to migration produce different outcomes later in the lifecourse.
Moreover, researchers have recognised that it is not just the timing of migration in relation to lifecourse transitions that matters but also the sequencing, and that a change in sequencing fundamentally affects the meaning of a particular migration move. This is an important and unresolved gap in migration research.
Explanation of the decision of whether or not to migrate has involved analysis of linked lives (Bailey, 2009). This has focused attention on how people within a household negotiate from different age, gender and class positions the relative desirability and feasibility of, for example, residential migration .

Coulter et al (2013) provide an extensive literature review of the implications of this
for researching residential mobility and this ground is not rehearsed again here.
Others have looked at the effects of the spatial mismatch between where people live and where jobs are available and have shown how neo-classical labour economics has had to be adapted to explain which individuals (in terms of age, gender, marital status and family composition) are likely to move to enter the labour market or for later career advancement within a national labour market (Van Ham, 2001;. Moreover, in parallel with the shift from the assumptions of the traditional household with a single male wage-earner to households with multiple earners and complex multiple labour market links, it has been increasingly been recognised how complicated decisions are about household relocation following opportunities for career progression for the highest earning household member (Raghuram, 2004). Over recent decades mobilities linked to the labour market have also increasingly disrupted the lifecourse as a result of a reduction in the security of employment (Clark and Withers, 2002) although the experience of entering unemployment may reduce rather than stimulate mobility. There has also been the rise of part-time and flexible working and the complexities of occupational mobility within company labour markets that are not only multi-site, but trans-national (Beaverstock, 2005).
Not only are decisions about labour market motivated-mobility negotiated between the linked lives of household members, but they are also structured by the decisions of significant actors such as employers and recruitment agencies. At the level of links to the employer and labour market, Scott (2013) has observed how an employer-led perspective on labour migration has helped to deepen understanding of the significance of sectoral specialisation and the intensification of production (Rogaly, 2008). Others have explored the link between international labour migration and the lifecourse.

MIGRATION AND STRUCTURAL CHANGE
Mobile lives are linked not only to increasingly fluid lifecourses, but they are also associated through work, study and residential behaviour with other socio-economic contexts that are driving changes in mobility (Figure 3). This fluidity of lifecourses has been accentuated by national and structural changes that have allowed free movements of individuals across certain state boundaries (for example within the EU). Given the tendency to privilege residential migration in the examples chosen in the previous section of this paper, we focus here on migration linked to temporal 16 changes in the labour market. Also excluded from our discussion is the very significant reshaping of international student flows in response to the internationalisation of higher education (King and Ruiz-Gelices, 2003;Brooks and Waters, 2011). It should be noted however that much of the student migration literature parallels international labour migration research in charting the effects on student flows of the transnational hierarchical differentiation of the higher education sector. Moreover the meanings given to cross-border student moves have also been argued to be part of longer-term lifecourse mobility strategies held by some students ).
An important starting point in analysing how migration may evolve in relation to labour market change is the observation that in most western societies, there is not a single integrated labour market but many sub-markets 'hierarchically structured by location and by class, gender, ethnicity and age' (Fielding, 2012, 98). Turning to the issue of labour market change over time, the period effect of short run boom and bust periods within the business cycle can now be observed using longitudinal data to assess the effects on labour mobility (see above). Within neoclassical economics it has been hypothesized that during an upturn in the economy, the increase in investment produced by rising demand for labour will result in greater opportunities for labour mobility (including international migration) with the inverse effect during recessionary times (Fielding, 1993). The reality of segmented hierarchically structured labour markets was more complex in terms of the trends observed in internal and international migration (Fix et al, 2009;  The wider economic restructuring of employment from a regional sectoral division of labour to a new spatial international division of labour (Massey, 1984) has had huge effects in producing the patterning both of internal and transnational migrants observed in the modern world. Once again our purpose here is not to rehearse debates that have been the subject of many journal papers describing the shift from mass migration for an era of Fordist mass production (Skeldon, 2012;Castles, 2010) to the new mobilities that were to follow. These emerged on the one hand from the globalization of production involving the planned transnational separation of labour tasks, and on the other the expansion and deepening of new political-economic blocks such as the European Union that were to enshrine the right of freedom of movement of labour between member states (King, 2002 Recent socio-economic restructuring has therefore resulted in many new mobilities (in some cases hyper-mobilities) especially amongst younger cohorts of the population (Fielding, 2012;Favell, 2009). At the same time as facilitating significant flows of labour within economic blocks such as the EU, international political groupings have also added to the efforts of nation states to regulate (often to severely restrict) legal labour immigration from other destinations (such as from the majority world of less wealthy nations). Some would argue that these efforts have done little more than to produce flows of illegal migration (Anderson and Ruhs, 2010) and to distort migration flows by encouraging those motivated to move to enter western democracies by other channels (such as asylum, study and so on).
Fielding (2012) recognises a deeper and longer run level of cultural and economic change that also impacts on migration over the long run. He points for example to the slow decline of the west with the end of empire and the decline of the west's economic dominance of transnational trade and global production systems (Held and McGrew, 2007). In place of traditional expatriates, there is the emergence of the new capitalist class (Sklair, 2001).  Elder et al's (2003, 12) third principle was defined as 'the principle of time and place'. They noted that the 'lifecourse of individuals is embedded and shaped by the historical times and places they experience over their lifetime' (Elder et al, 2003, 12).

SPACE-TIME CONTEXTS
Given the prominence given to place with their schema, it is surprising to find that impacting on any given lifespan (level 2), but that it is relational relative to geographical understandings of space (Bailey, 2005). Thus space needs to be considered as the active context rather than a passive property associated with mobilities across the lifecourse. Not only have the space-time contexts in which mobilities are studied been transformed by processes of time-space convergence, time-space compression and distanciation, but the interpretation of mobilities need to be read through a relational understanding of the uneven meanings of mobility reported by different actors and in different places. Moreover researchers engaging in studies of the space-time contexts of new mobilities need to recognise their roles in the co-construction of the societies and spaces (Little, 2014) that they are studying.
A few examples serve to illustrate the significance of this third level of engagement. Even if space were understood to be a passive force, with geographical 20 boundaries being no more than defining lines around the empty containers of the spatial dimension of society, it would still be worth undertaking a comparative analysis of population change in general and migration in particular. This would highlight for example how differences between states in the way that resources are distributed and in the way that society is governed influence demographic events and processes. One would expect to find significant contrasts for example between the frequency and nature of population mobility in countries such as the UK, Sweden and the Netherlands simply because of the distinctive ways that these countries have constituted the relations between the family, the welfare state and the economy.
Indeed the skill type of a migrant is affected not only by immigration laws but also by the generosity of the welfare state (Razin and Wahba, 2014). A major challenge to undertaking even this most basic comparative type of research remains, and this is the very different basis on which movement data are collected in each country.
Most researchers would recognise that space is not passive, but active.
Population movements are embedded in a spatial context that is constantly changing and doing so unevenly. For example, improved travel technologies have continued to change the threshold defining the break point between residential migration and commuting opportunities (Lundholm, 2007), thus redefining the desirability of residential migration relative to other mobilities. This change has not however happened evenly across space with greater investment in improved transportation in large urban centres and in core economic regions, with rural and peripheral areas being less favoured. Over time electronic media and the world wide web have also enabled new virtual mobilities to substitute for some physical moves (Verne, 2014), but once again these changes have transformed certain more economically favoured places in western societies more rapidly than less connected parts of the majority world. Even within the favoured locations that are best served by telecommunications, access has been uneven with a digital divide emerging in many Western societies (with the elderly and the less well-off falling on the wrong side of this divide). It is in this context that hypotheses such as the idea that secular society is increasingly rooted (Cooke, 2011) needs to be explored, since not all members of society have the same opportunities to fulfil their mobility and immobility desires. Simultaneously the redefining of social roles (for example in relation to gender and parenting, or ageing and inter-generational care relationships) within the close and extended family have 21 changed the timing and significance of mobility-lifecourse relationships, but they have done so unevenly between social groups and between different communities. In summary, space has not only been an active force in shaping mobility-lifecourse relationships simply because society is spatially constituted, but it has also been the arena through which the asymmetries of power relations have been played out (in terms of the structuring influence of gender, class and race).
The asymmetry of power between ethnic groups and their host society has been recognised by migration researchers as one of the reasons contributing to ethnic segregation (Finney and Simpson, 2009). Much remains to be researched about how the mobility of ethnic groups has changed over time in relation to their positioning in the housing and labour markets of host societies (Finney, 2011). A key question remains how powerful are residential neighbourhoods in shaping people's life opportunities, and consequently the importance of neighbourhood in determining the likelihood and destinations of moves in the housing and labour markets . Changing scale, similar questions might be asked of the role of interregional migration in enhancing social mobility (and inversely of immobility in hampering opportunities for self-improvement). While the so-called escalator region effect has been noted as significant both in the UK and Sweden in relation to inmigration to their capital cities in accounting for regional differences in upward occupational mobility (Fielding, 1993;Fischer and Malmberg, 2001;Van Ham et al, 2012) work remains to be done to identify whether similar spatial effects can be identified for metropolitan centres at lower levels in the urban hierarchy (Champion and Shuttleworth, 2014;Champion et al, 2014) and to establish the effect of interactions between international and internal migration in relation to this effect.
Power asymmetries are perhaps nowhere as evident as in the spatial context of the global city. Spatial analysis of ethnic niches in the labour markets of global cities have been re-theorised as part of a new 'migrant division of labour' (May et al, 2007;Wills et al, 2010). In Wills et al's landmark study of London, the gendered, ethnic and class dimensions of this are recognised, but research on the 'relational lives' of the migrants remains limited, and this is true both relative to the other parts of the world from which the migrants have come and also relative to mobilities of the host population served by the migrants. If research on mobility in relation to the power asymmetries evident in the global city provides a large canvas for future research (Walsh, 2014), even more remains to be done on topics such as transnational localism and transnational ruralism as key dimensions in understanding the place-based nature of the meanings given to mobility.

DISCUSSION AND CONCLUSIONS
In the absence of a conceptual framework it is all too easy for research effort to dissipate and for individual studies of the new mobilities produced by, and producing, population change in western societies to end up being no more than a few extra case studies scattered across the burgeoning academic literature on the topic. Some may even question if there can be any strength in a research canvas that is as broad as the schema presented in this paper. Nevertheless, it is hoped the current paper offers an opportunity for some coordinated research progress to be achieved in the course of the Centre for Population Change second five-year phase of funding.
This paper has been based on two over-riding assumptions. First, it has been assumed from lifecourse theory that changes in the way people organise their lifecourses will have a strong impact on mobility patterns (both locally and internationally). Second, it has been assumed that the economic, social and spatial contexts of people's lives affect the timing, frequency and meanings associated with mobility. Testing the veracity of these assumptions will be a first step to advancing new understandings of migration and mobility practices.
The conceptual framework proposed in this paper leads to a range of specific research questions as outlined below. The list of six questions is not intended to be a comprehensive one. Instead the questions are illustrative of how key ideas reported in In parallel, is it possible to advance longitudinal methods seeking to compare population mobilities over time?

24
There will of course be many research challenges in the attempt to address these and other questions. There will be a need to avoid creating new false binaries between the migrant and non-migrant communities under study. And there will need to be vigilance to avoid reifying the constructs we choose to privilege in our research if we are not simply to end up advocating some new totalising discourse in place of the work that has gone before. These challenges should not however be an excuse for retreating to an indulgent engagement with small questions that can be answered from the data feast that faces us. Instead, this is a moment of opportunity to make our research count by seeking to advance the conceptual understanding of the new mobilities evident in contemporary Western society using a multi-disciplinary approach.