Moving towards the centre or the exit? Migration in population studies and in Population Studies 1996–2021

This paper examines the position of migration in population studies, focusing on the period 1996–2021. It considers the reasons why migration remains problematic for demographers, but also how approaches to migration have changed over the last 25 years. While it has arguably become more important to both demography and population studies because of the transition to low fertility and mortality, migration has metamorphosed into a complex field in its own right, almost independently from changes in demography. Both internal and international migration form the subject of this examination and four main themes are pursued: data and measurement; theories and approaches; migration and development; and migration and political demography. The papers published in the journal Population Studies are used to provide a mirror through which to view these changes over the last 25 years. This paper concludes by looking at likely future directions in migration studies, demography, and population studies.


Introduction: Migration in demography
That the population of any area can only grow or decline through a change in one or more of three processes-fertility, mortality, and migration-is an axiom that needs little justification. Demography is, however, more than this. Donald Bogue, a distinguished scholar of population studies in the United States (US) after the Second World War, expanded the definition to incorporate five processes: the three just listed, but adding marriage and social mobility before going on to say that demography's 'long-run goal is to develop a body of theory to explain the events that it charts and compares' (Bogue 1969, pp. 1-2). Whether demography or even the broader subject area of population studies have fulfilled this mandate lies beyond the scope of this paper, which will focus primarily on just one of the processes-migration -and how demographers and those who specialize in population have tended to conceptualize and analyse the process. Despite migration being one of the key pillars, demographers have appeared ambivalent towards it. As a reviewer of an influential book on population has put it, 'Demographers have tended to regard migration as a nuisance' (Laxton 2020, p. 184).
As will be seen from the papers published in the journal Population Studies, migration has been a lesser concern in demography, while paradoxically becoming a more important component of population change as fertility and mortality have moved to low levels. At the same time, migration studies has blossomed and moved towards being a separate field in its own right. This paper will not just review work on migration published in the journal over the previous 25 years, but will also examine broader trends in the field of migration studies in general over the same period, to identify common synergies and to emphasize that they are part of a common and integrated field of enquiry. Migration illustrates many of the issues that appear to challenge demographers, should they wish to move beyond the focus on fertility and mortality that has typified much of the discipline and towards the broader view advocated by Bogue. In the early and, some would argue, even in later approaches in demography, migration was on the margins compared with the other two principal processes, fertility and mortality. Although the study of mortality was an early and consistent topic for demographers, after the 1950s, in a world of rapidly expanding populations, fertility came to dominate the discipline (see, e.g. Greenhalgh 1996, particularly p. 30). Dudley Kirk, in his address to the Population Association of America in 1960, argued that migration was the stepchild of demography in that it was the poor relative among the key demographic processes (Kirk 1960). In the early 2020s, with low or even declining population growth across much of the world, the role of migration as a demographic variable has become more widely appreciated and studied. However, whether migration is moving closer to the centre of primary concern is much less certain. Demographers still have an ambivalent relationship with migration and, as will be argued, migration is still the stepchild of demography, even if it has become a richly endowed stepchild (van Dalen 2018). Nevertheless, studies of migration have grown in number and taken on an identity of their own as they moved away from their demographic roots.
The reason that migration has been and will remain a stepchild in demography lies in its very nature and the data required for its measurement. Unlike fertility and mortality, migration is not a unique event but a continuous process throughout an individual's life. Migration is also determined by space and time: that is, over what distance a person must move, and over what period of time, in order to be defined as a 'migrant'. Hence, the volume of migration identified depends entirely on the definitions applied in specific data-gathering instruments. For example, the smaller the areal unit across the boundary of which a person moves in order to be defined as a migrant, and the shorter the time period a person must reside at an address before they qualify as a migrant, the greater the number of migrants captured by the instrument. The objective of any survey, nevertheless, is not simply to capture a greater number of migrants. Short-distance and short-duration movers in the developing world, for example, tend to be different in terms of sex, education, and occupation when compared with longer-distance migrants of longer duration. Hence, the types of migrants captured through the methods of data collection vary. Multiple and return moves add yet further complications. In the absence of universal or common space-time categories, comparisons across countries and areas are problematic and essentially act to keep the measurement of migration in a different category of precision from the registration of births and deaths, even if linkages among the three processes clearly exist.
As we shall see later, measures are being taken to mitigate the effect of these issues of measurement, but demographers, attuned to dealing with quantitative models of clearly defined categories for births and deaths, become uneasy with some of the fuzziness of migration classification. The very precision of demographic methods has, however, led to an overly narrow focus on population in 'a field with an extremely high ratio of techniques to ideas' and approaches 'relatively untouched by conceptual developments' in cognate disciplines (Hauser, in Crimmins 1993, p. 586;Greenhalgh 1996, p. 27, respectively). The majority of studies in demography, as reviewed in the mid-1950s, were virtually dismissed as 'planlessly empirical and trivial in content' (Thomas in Crimmins 1993, p. 586). Thus, all three principal demographic processes were seen, fairly or unfairly, to have limited horizons and capabilities. The situation has changed since that time, as the papers in the 30th anniversary issue of Demography in 1993 and those in the 50th anniversary issue of Population Studies in 1996 imply; see also Greenhalgh (1996). Migration continued to have a low profile in demography during this time. The essays in this current 75th anniversary issue review the advances to date in all three processes, suggesting that migration is coming of age within demography, while at the same time setting a course of its own. Yet, as will be seen later, uncomfortable parallels with the influence of funding and the needs of policymakers have emerged in the respective trajectories of migration and fertility as the former has moved to greater prominence.

The background to changes in approaches to migration
The focus of this paper is on the direction of research on migration as indicated through publications during the period from 1996-the time of the 50th anniversary special issue of Population Studies-to the present. The paper will adopt two main perspectives: first, from the literature on migration and S28 Ronald Skeldon population in general within the context of the emergence of migration studies as a discipline in its own right; and, second, from the contribution that papers on migration published in Population Studies have made to the wider debate. These perspectives provide an insight into how migration studies has engaged with population and how population studies has contributed to and is perceived in the wider field of migration.
The impression that migration is a 'process apart' from the main concerns of students of demography has perhaps had unfortunate consequences. No chapter on migration appeared in the 1996 special issue of Population Studies or in the special issue, 'Population-The long View', published by the journal in 2015. In the 50th anniversary issue, Caldwell (1996), however, did cover the topic in his paper 'Demography and social science', in which he identified migration as one of four new themes that had emerged in demography as reflected in the papers in Population Studies from the 1950s, the other three being occupational (and social) mobility, marriage, and population theory. This suggested that the papers in Population Studies well reflected Bogue's 1969 definition of demography given earlier. In the 2020s, with the transition to low fertility and mortality virtually complete across so much of the world, with the exception of sub-Saharan Africa, migration has emerged as a, if not the, most important component of population change. The omission of a paper on migration in the 75th anniversary special issue would be inconceivable.
An opportunity to anchor migration more securely in both demography and population studies was perhaps lost in the 1980s and 1990s when the concept of innovation diffusion entered into ideas of fertility decline (Cleland and Wilson 1987;Greenhalgh 1996, pp. 56-9;Casterline 2001). The idea that cultural and societal transformation played important roles in influencing attitudes towards the number of births and the adoption of family planning programmes took on greater currency. Hence, policymakers and health officials could not expect such programmes simply to be adopted without a prior ideational shift in local populations, creating the small family as a norm that was both desirable and attainable. While networks were seen to play a part in reproductive health (Casterline 2001, p. 15), no references were ever made to their extensive application in migration studies at around the same time, particularly in the conceptual approaches of the new economics of labour migration; for reviews of this work, see Massey et al. (1993) and Skeldon (1990Skeldon ( , 1997. The vast circulation of people from villages to towns and cities during the process of urbanization across the so-called developing world transformed communities of origin, as returnees brought not just money but new ideas and ways of doing things in what came to be called social remittances (Levitt 1998;Vari-Lavoisier 2020). This is not to argue that social remittances might hold the key to an understanding of fertility decline: clearly, other institutional factors associated with the diffusion of education, the changing status of women, the construction of roads and other forms of communication, and so on were also important. However, the importance of return migrants lies in their position as agents of change 'from within', rather than change being imposed 'from without', and it seems regrettable that the application of similar diffusion approaches in two of the core demographic processes did not produce significant crossfertilization at the time.
Migration also impacts on origins and destinations in ways other than through return migrants and social remittances. The selectivity of migrants by age, sex, and education changes the population compositions of communities at both ends of the migration process. Given that one of the few virtually universal generalizations that can be made about migration is that the majority of those who move are young adults, destination populations tend to be youthful, with higher proportions in the labour force, while origin areas may experience the loss of the more dynamic elements of their populations. This can have profound consequences that are discussed later in the subsection on 'Migration and development'. The loss of large numbers of young men and women not only affects the productive capacity of origin populations, but their reproductive capability too. This demographic change is more obvious in the context of rural-tourban internal migration, which has led to rural depopulation across much of the developed world.
The volume of research on migration since the mid-1990s has increased exponentially. Some idea of this increase can be appreciated from a list of migration journals available in the English language from the International Organization for Migration (IOM), now the official United Nations (UN) organization that deals with migration, which gives some 95 journals (IOM 2020). While reflecting the range of topic areas pursued by scholars of migration, the titles of the majority of these journals suggest that they might be of limited interest to those researching population issues or migration as a demographic variable. Nevertheless, the Swiss Forum for Migration in population studies S29 Migration and Population (SFM) lists some 43 journal titles in English, French, German, and Italian on its website that, given the mandate of the institution, would be of direct interest to those studying migration and population issues (SFM 2020). Many of these journals have come into existence over the last few decades. The present paper attempts to sketch the main directions of enquiry taken in migration research (that are directly related to population issues) since approximately the mid-1990s and then examines the papers in Population Studies within this context.
The mid-1990s period forms a useful divide in research on migration because it was from around that time that studies of international migration came to dominate the field to the extent that the word 'migration' came to mean 'international migration'. Not that internal migration was entirely lost to the field, and perhaps, most significantly, it continued in demography, but it lost the prominence it had in the late 1960s and early 1970s, when the work had focused so much on less developed countries and on rural-to-urban migration in particular (Todaro 1969(Todaro , 1976. By the 1990s, internal migration had become 'population redistribution' while 'migration' referred to movements across international borders, giving rise to two very different traditions in the migration field (Skeldon 2006;King and Skeldon 2010).
The rise of international migration from the 1990s reflected growing economic disparities among countries, as well as disruptions in both the classification of migrants and real movements consequent on the break-up of the large spatial units of the Soviet Union and Yugoslavia. For example, what had been internal migrations within the Soviet Union became international migrations between the new republics of the former Soviet Union. Until the 1990s migration at the multilateral level had been dominated either by a focus on labour migration, essentially considered to be a temporary phenomenon, or on forced movements and refugees. Following the UN International Conference on Population and Development in Cairo in 1994, international migration became more comprehensively viewed, incorporating longer-term movements for settlement, skilled migrants, and students, as well as labour migrants and refugees, and all in the context of sustainable development. International migration was both a cause and a consequence of development.
Adding to the significance of the mid-1990s as a turning point was the publication of the first edition of the most successful and influential textbook on migration, The Age of Migration, written by Castles and Miller (1993). This volume focused only on international population movements and, now in its sixth edition, continues to be important in giving direction to the field and inspiring new generations of 'migrationists' (De Haas et al. 2020).
I also published a book on migration and development in the mid-1990s that attempted to review the work undertaken on both internal and international migration up to that time, concluding that migration was an integral part of development and it was impossible to imagine the one without the other (Skeldon 1997). However, from about this time, 'migration and development' entered the global policy discourse and the emphasis shifted to how the movement of people could be used to promote 'development' and how governments could better 'manage' migration to achieve this result. Thus, in a way not entirely dissimilar to the experience of research on fertility from the 1960s (see Greenhalgh 1996), research on migration came to be influenced by funding and, for policy-based research, driven by the needs of governments, international organizations, and non-governmental interests. This has led migration in directions away from its demographic roots, as described in more detail in the following section, where relationships with fertility and mortality are almost forgotten.
This path has also led to an uneasy and often unacknowledged division between migration and migrants: between the aggregate macro-level flows that impact populations, and their compositions and distributions, and the individual migrants and their rights, aspirations, capabilities, and life courses.
These different perspectives have been reinforced by the impact of the postmodern turn in the social sciences-which dated from the 1970s and 1980s, but came to prominence in certain approaches to migration from the 1990s-a trend which the more empirically oriented demography largely avoided. This is not to say that intellectual ferment did not exist within the discipline, as the views expressed in the first section of this paper surely attest, but that, unlike cognate fields, 'postmodernism has rarely been used within the field of demography' (Riley and McCarthy 2003, p. 3). Considering itself to be a science, demography largely eschewed approaches that might be considered to be subjective or lacking a robust empirical foundation. Studies in migration have not always been so exacting and have been profoundly influenced by postmodernism, but it is perhaps indicative of its marginal position in demography that in a book devoted to showing what that discipline could gain from postmodern approaches, only three pages are given to migration S30 Ronald Skeldon and the importance of the emergence of the ideas of transnationalism and diaspora (Riley and McCarthy 2003, pp. 145-8).
Postmodernist approaches, with their scepticism of metanarrative and theory, led to a cultural turn in migration studies (King 2012) that emphasized detail and avoided grand generalizations. Out of this intellectual ferment of the 1990s, several major lines of enquiry crystallized, perhaps the most intriguing of which was the emergence of 'diasporas' and 'diaspora studies', which essentially became virtual synonyms for migration and migration studies, always assuming that migration meant international migration. However, the word diaspora comes with 'baggage', implying exile and often victimhood, which provided subtleties of interpretation that those more interested in population studies and demography might well feel lies beyond their range of competence. The diaspora concept also provided scope for feminist interpretations of constantly shifting gender relations within multicultural and transnational migration fields; see, for example, Campt and Thomas (2008) or Vatsa (2016).
The literature on recent research into diasporas is vast; for summary overviews illustrating the range of topics considered, see Cohen (1997) and the essays in Cohen and Fischer (2019). Particular attention has been paid to two of the 'great' diasporas: the South Asian (e.g. Lal 2006;Oonk 2007;Jayaram 2012) and the Chinese (e.g. Pan 2006; Ma and Cartier 2003;Wang and Wang 1998). For the Scottish diaspora, see Devine (2011), and on trying to locate the English diaspora, see Bueltmann et al. (2012). Fully 14 of the 95 journals recorded in the IOM list referred to earlier have diaspora in their title, although many other journals have also published papers on this topic (IOM 2020). Perhaps significantly, only one of the 43 listed by the SFM had diaspora in the title, although again many of those listed have published papers on diasporas (SFM 2020).
While diaspora will be considered among the themes examined in more detail in the next section, its meaning there will be more specific and in the context of the migration and development debate. Perhaps most important to note, however, is that diaspora has brought a transnational perspective to the subject: that migration cannot be understood just in terms of origins, destinations, and integration, but that migrants live and operate across states and cultures. Questions of identity, citizenship, and belonging take on increasing importance, with these being generally flexible and dependent on context. Voting behaviour and participation in the political, economic, and social life of two or more localities emerge as themes in a globalizing world. Whether an individual can have two (or more) identities and can live in two (or more) countries remain intriguing questions in a world that is witnessing rising nationalism and increasingly seeks to manage or control migration. For a succinct review of migration in the transnational literature, see Vertovec (2009).

Research areas in migration of most relevance to demographers
Of the many strands of migration research to emerge since the mid-1990s, just four will be identified here because of their specific relevance to population studies and demography: data and methods; theory; migration and development; and political demography.

Data and methods
Given the importance of defining who a migrant is, outlined in the first part of this paper, much effort has been directed towards the improvement of migration-related data and the means through which to analyse them, both themes that are close to the core of demography. At the most basic level was the construction of a global origin-destination database for the analysis of global flows, based on the stock of migrants in a country at a particular time. Initially established as part of a United Kingdom (UK) government project supported by the Department for International Development (DfID) at the University of Sussex, it was later transferred to the World Bank and the UN Population Division for expansion and both updating and backdating (Parsons et al. 2007). Global matrices from 1990 to 2020 are available and currently updated on a biennial basis by the UN. The stocks are based on the latest population censuses from each country and very much depend on the reliability of those censuses plus on the fact that migration data are often collected only in a long form from a sample in the census. Given that migrants tend to be concentrated in particular areas, spurious results can be generated on scaling up, dependent on the nature of the sampling strategy.
Stock data are not ideal for the study of migration: they omit return and onward movements, which can be significant. Nevertheless, the global database has provided basic information on bilateral flows Migration in population studies S31 between countries, including much useful information about regional migrations, quite apart from those to the principal destination countries in the developed world. These data have been fundamental to global assessments of migration in a number of influential reports (e.g. UNDP 2009; IOM 2018). Comparisons of the matrix across time have generated indirect estimates of flows, globally and regionally, with the results illustrated in innovative ways (Abel and Sander 2014). The matrix has also been disaggregated to allow the flows to be generated by sex (Abel 2018). Where data-gathering systems are more robust, as in the countries of the Organisation for Economic Coordination and Development (OECD), much more detailed information on flows to and from those countries using continuous immigration registers generates information on migrants by sex, age, education, and employment on an annual basis, in a system that is being continuously enhanced and extended (OECD nd).
One of the curious myths in migration is that the data are better for international than internal migration. Yet, data on the internal movements of people from censuses, large-scale surveys, and registration systems exist in quantity, and some of the most innovative methods are being developed in this area. For example, the geographical problem of variation among countries in the size of the spatial units used to define 'a migration' makes international comparisons difficult. Building on Daniel Courgeau's work, Martin Bell and his colleagues have moved some way towards mitigating this problem through the application of a measure of 'migration intensity' that relates the number of moves to the population at risk of moving, which, for the first time, provides more robust comparative estimates of migration across countries (Courgeau et al. 2012;Bell et al. 2015aBell et al. , 2015bBell et al. , 2018Bell et al. , 2020. A more general acceptance that diverse forms of mobility-including regular daily and weekly commuting, as well as recreational and business travel -need to be considered together with longer-term migrations, return movements, and circulation between rural and urban areas is forcing researchers to seek new sources of data. This trend will almost certainly be accelerated following the Covid-19 pandemic. The use of digital sources, including mobile phone and internet records, together with administrative data in the context of the more traditional census, survey, and continuous registration systems, will provide a more comprehensive and ultimately more robust assessment of human interactions across space and through time (Recchi 2017;Favell and Recchi 2019). The 'footprint' left by the mobile phone and the existence of massive computing power for the subsequent analysis of big data provide a powerful tool to map each person's movements. However, critical issues remain: not just access to information held by private companies, but the tension between what researchers, and by extension the state, wish to know and the rights of the individual to privacy. Quite apart from these thorny future issues, an interesting convergence appears to be taking place, with those studying internal migration and those studying international migration now using the same kinds of databases, which implies moves towards the application of similar methods.

Theory
Demography, as noted earlier in this paper, has been viewed as a primarily empirical discipline with few ventures into theory. Perhaps the most influential of these ventures has been the theory of the demographic transition, although that theory, as originally proposed, incorporated only two of the processes of population change (fertility and mortality). Kirk (1996, pp. 383-4), in his paper for the 50th anniversary special issue of Population Studies, drew attention to the point stressed by Chesnais in his 1986 work on the demographic transition: that migration had been omitted from the model (see Chesnais 1992). Yet, in 1971, Wilbur Zelinsky had done precisely that in his hypothesis of the mobility transition (Zelinsky 1971), with my own modification of the model published just six years later (Skeldon 1977) and a more comprehensive assessment in 1990 (Skeldon 1990).
The failure to follow up on this conceptual development to bring migration into the theory of the demographic transition can be attributed to three factors. First was the sheer difficulty of linking changing patterns of migration with shifts in fertility and mortality that, given the measurement issues discussed earlier, meant the Zelinsky model was more intuitive than supported by hard data. Demographers were perhaps reluctant to move beyond the securities of such data to incorporate a variable with a perceived high level of imprecision, although a significant exception was the work of Dyson (2011), who considered the roles of both rural-tourban migration and fertility in the demographic transition during the process of urbanization. Second was the more general shift in the social sciences away from unilinear models, such as transitions that were associated with modernization S32 Ronald Skeldon theory. Third was what might be termed 'disciplinary blinkering': that despite population and migration studies being interdisciplinary, communication among the various disciplines was often not as effective as might have been anticipated and hence a number of disciplinary silos have emerged. Specialization within silos need not necessarily be a bad thing in some disciplines, but in demography (with its clear links in mortality and fertility to other research fields as outlined in the respective papers in this issue) and particularly in migration, the broader view seems essential. However, the exchange of ideas, even any mutual awareness, across disciplines, has become increasingly an aspiration rather than a reality, perhaps reflecting the very diversity of approaches that have come to characterize migration studies.
The result has been that while demography has gradually moved to incorporate migration in a limited way, primarily in terms of methods as noted in the previous subsection and will be further discussed in the following sections, migration studies itself has branched out in diverse ways using a number of prevailing approaches. While Greenhalgh (1996, p. 47) criticized demography for failing to engage with theoretical developments in cognate disciplines, migration studies went on to do exactly that in the areas she identified: political science (e.g. Portes and DeWind 2007; Phillips 2011) and feminism and gender studies (e.g. Chant 1992; Piper 2006; Bastia and Piper 2019). These approaches are just some from among the many to the extent that a return to a more unifying theoretical framework has been observed; not along the lines of a single unilinear model but more in terms of multiple pathways through a mobility transition; not in a simple reproduction of Zelinsky but certainly influenced by it (Skeldon 2012(Skeldon , 2019Cooke et al. 2018).

Migration and development
That migration is closely linked to development is clear, as discussed earlier. However, since the last years of the previous century, migration has been cast in a different way, as a variable to be managed in order to facilitate development rather than as an integral, natural part of any process of economic and social transformation itself. Just as interventions in fertility from the 1960s were seen to be necessary to bring down national and ultimately global rates of population growth, so international migration, if managed correctly, could bring about development in a win-win-win scenario: for countries of origin, for countries of destination, and for the migrants themselves (e.g. Koenig 2017). The jury is still out on whether this more applied approach to migration and development will be a passing phase (Skeldon 2008) or a sustained path in the future of the field, but it has certainly been a major theme in research on migration since the mid-1990s.
This shift came about with the realization that international migration appeared to be one of the most important international issues: a product of globalization to which no reference had appeared in the Millennium Development Goals. The result was an explosion of policy-relevant research coordinated through the UN, the World Bank, the IOM, the European Union, the OECD, governmental agencies (such as DfID in the UK and the Swiss Agency for Development and Cooperation), and civil society organizations (such as the Migration Policy Institute in Washington), together with many academic institutions and think tanks around the world. The evolution of the complex global architecture for migration governance is outlined in Betts and Kainz (2017) and McGregor (2020aMcGregor ( , 2020b. Reviews into the range of research topics and approaches in migration and development can be gathered from the essays in Lucas (2014) and in Bastia and Skeldon (2020), although see also Lucas (2005) and Skeldon (1997) for extended assessments. Amid all the complexity, three major subfields of enquiry can be identified: remittances (including social remittances), skilled migration, and diasporas. Into these subthemes is interwoven the thread of circular migration (circulation), to emphasize the issues of return and repeat migration and to show that migration is rarely a simple one-way movement. Here is not the place for a detailed exposition of the three subthemes, although suffice it to say that, in each, the linkages with development are contested and not always intuitive.
By far the major area of concern in the migration and development debate has revolved around the role of migrant workers, who, at some 164 million, represented 70 per cent of the global migrant population aged 15 years and over in 2017, two-fifths of these being women (ILO 2018). Assumed to be temporary, with few rights to long-term residence in destination countries, migrant workers are found in occupations that are low paid and often hazardous, and which local workers are unwilling or unavailable to undertake. While the potential for exploitation and the vulnerability of these workers appear to be the antithesis of development, it is what has brought the migrants into the system that has commanded attention: the need to earn more money than they could in their home countries, in order to support their families better. They choose to migrate in order to send money back in the form of remittances. It is the sheer volume of remittances sent back by international migrants to their home countries-estimated at some US$529 billion flowing to low-and middle-income countries in 2018 and the largest source of foreign exchange for these countries (World Bank 2019, p. 1)-that should have the most significant development potential. Certainly, remittances appear to improve the well-being of families that receive them but, as international migrants have been shown not to come from the poorest families or from the poorest areas, remittances may also reinforce existing wealth differences. Rather than providing a panacea for development funding, remittances reflect the broader exchanges between countries that are at the root of social transformation (see Carling 2020).
Research into skilled migration, the second subtheme, has examined equally ambivalent positions. The skilled tend to be highly mobile, often through the networks of transnational corporations, with rapid turnover as they move through corporate career ladders. Governments of most developed economies have policies to attract the skilled, not just because such workers bring key skills in critical areas such as health and education, but because they are highly educated and make 'good' potential immigrants. Some may stay and integrate, and they make up part of the transnational capitalist class (Sklair 2001), even if that elite is fractured into a number of different interest groups. The movement of skilled migrants from countries of origin may be seen to prejudice the development prospects of those countries in the classic 'brain drain' argument. Nevertheless, the evidence for such a scenario, outside conflict situations, seems elusive, even for relatively small states (Skeldon 2020).
As discussed earlier, diaspora, the third subtheme in migration and development, has been one of the most dynamic areas of enquiry in migration studies as a whole over recent decades, and its role in development is but part of a much wider research concern. In the migration and development literature, the diaspora (here, the community of migrants in destination areas) is seen to play an important role through its transnational links. It brings remittances and skilled migration together to the extent that members of the diaspora are also key investors in countries of origin and that the skilled in the diaspora return on both a longer-and shorter-term basis to support enterprise or to engage in training and education. Support can be provided through communal and institutional frameworks in what are known as 'home town associations', which not only link origins and destinations but are actively promoted by international organizations and national development assistance departments (Gamlen 2019). Home town associations can range in size and influence from national-level, elite diaspora institutions, through provincial and district organizations, to village or even sub-village groups. Factionalism can be an issue, with decisions on what kinds of activities to support, and where, being a constant source of tension.
The capability of diasporas to promote development is difficult to assess. Small-town projects, such as supplying simple computers to schools or skilled returnees participating in training programmes, can produce localized successes, but whether these cumulatively lead to long-term sustainable development is more questionable. Also, migrants are unlikely either to invest in or return to home areas with little potential, choosing instead to focus on urban areas where their investments make a much smaller relative impact. Equally, they may also invest in destination areas and migrants do integrate into their new home areas, some, such as the Scots who settled in diverse areas around the globe, more easily than others. Also, following on from the issue of factionalism, diasporas are not homogeneous groupings of altruistic individuals, but are made up of people with competing interests, who may not always wish to work constructively for their countries of origin: diasporas can have a 'Janus face' (Kapur 2007) that acts to undermine institutions in countries of origin. Thus, migration can have security implications for host countries. This role of the diaspora and of individuals in the diaspora to foment change in origin and destination countries brings us logically to the last of the recent research areas in migration and population studies: migration and political development.

Political demography
Political demography, an area 'under-attended by demographic science' until recently (Teitelbaum 2015), emerged as a consequence of two powerful demographic processes-the ageing of populations and the generation of age bulges-and how these might drive conservative political movements in the case of the former and more revolutionary movements such as the Arab Spring in the case of the latter (Cincotta and Doces 2012;Urdal 2012). The emphasis in population studies has again been on S34 Ronald Skeldon shifting patterns of fertility, although the 2015 flows of refugees to Europe turned attention towards immigration (see e.g. Kaufmann 2018).
In the 50th anniversary special issue, Keyfitz (1996, p. 342) argued that national borders should be closed to mass migration and that advances in technology and automation would eventually replace the need for many kinds of workers. However, the 25 years since the publication of that paper has seen the numbers of migrants increase to the point that they gave rise to Castles and Miller's (1993) book, The Age of Migration, an age that some see as a threat to the very existence of certain nations. Demographers such as Coleman (2012) have examined migration and ethnic change, although most work in the area of political change and migration is still dominated by political scientists.
The demographic shifts brought about by fertility decline have brought a different perspective on a long-standing dimension of research on migration: what happens to the migrants after arrival in destinations. As we have seen, many returned, but many others stayed to integrate with host and other migrant groups in a variety of ways. The issues of assimilation and integration have been one of the major areas of research into immigration in the great settler societies, those that have seen migration as a central part of nation-building, exemplified by Australia, Canada, and the US. Yet, from just before the start of the period under review, both the context of the immigrant groups and the approaches to their study have changed. Until the 1990s, immigrants were seen to integrate in line with one of two general models: to assimilate to the values of the host society; or to create separate communities in a process of multiculturalism. Both of these approaches proved problematic and, from the 1990s, with the realization of the importance of continued links with origins, a narrow nationalistic framework was seen as inadequate for the study of political involvement and identity (Basch et al. 1994;also Faist 2013). Nations became transnational and, as seen in my earlier discussions of diaspora, heavily influenced by their regional and global linkages, heterogeneous though these may be.
Over the decades under consideration, this transnational situation has become increasingly complex in destination states because of changing demography. Host populations feel challenged, not just in the face of continuing immigration but also because of their own declining rates of natural growth. Host populations, mainly white and once dominant, turn to populism, extremist politics, xenophobia, and anti-migrant sentiment as a response to a perception that they are becoming minorities in their own countries (Kaufmann 2018;also Polakow-Suransky 2017). Yet such a perception is more 'the great demographic illusion' than a necessary outcome grounded in empirical reality (Alba 2020). The ethnic or racial categories used to define majority/minority are not set in stone. The host white populations are also more heterogeneous than often assumed and class becomes an important marker. According to Alba (2020), it is the emergence of a more progressive and expanding 'American mainstream', the result of intermarriage between hosts and the various migrant and ethnic groups themselves, that will create the new identity. In the analysis of this process, the role of demographers will be crucial, not just in the matter of reinterpreting the categories used to define migrant and ethnic groups, but also in the more traditional areas of migrant/non-migrant fertility differences and trends, and the changing rates of interethnic marriage.
The role of migration in political change is, however, more than just the impact of immigration in developed and ageing societies; it covers state creation and political remittances that can transform origin countries too (Skeldon 2021). Migration impacts state formation in two paradoxical ways: through inclusion and exclusion. In terms of inclusion, migration to incorporate new lands through conquest or the more gradual extension of agricultural frontiers expands the state, but just as important are the internal circuits of movement that consolidate the state. Circuits of mobility essentially define the area under effective control of a state: for example, circuits of military recruitment and for education that bring people, initially males, into a common language and shared traditions; circuits of workers from places of reproduction to places of production; circuits of traders and pedlars filling the highways and byways, holding the population together in a constantly shifting collage. In the words of the Italian historian Arlacchi (1983, p. 200), 'The imposing circulation of men, manners and ideas has turned out to be the most powerful unifying element in the whole history of Italy. Mass internal migration has finally "made us into Italians".' These forces of inclusion are so often accompanied by forces of exclusion, the expulsion of those who do not meet the criteria of belonging as laid down by ruling classes, which introduces one of the most important and dynamic subsets of studies in migration: refugee studies. The refugee, Migration in population studies S35 defined by the UN High Commissioner for Refugees as a person forced to flee because of a well-founded fear of persecution, is primarily the result of state formation (Zolberg 1983) or disputes among elite groups over who controls legitimacy over the founding ideals of the state. Those who do not fit the ideals are forced to flee. However, flight need not necessarily be an immediate reaction to circumstance; it can be a more measured assessment of risk and opportunity. It may be more advantageous to flee as a student, for example, than to cross a border illegally and then claim asylum, but motives are varied and constantly change according to opportunities presented or taken away throughout the trajectory from origin to destination (Wissink 2019;Bonfiglio 2020). The difference between forced and 'voluntary' migrations is not as sharp as often assumed, and refugee movements need to be examined within the context of migration as a whole and not as an entirely separate category.
Not all forced migrations, however, are the result of violent or oppressive political forces. What might be called development-induced migration perhaps accounts for the movement of 10-15 million people per year (cited in Yan 2020, p. 373). The construction of dams displaces large numbers of people, as does the expansion of urban areas over prime agricultural land in rapidly developing economies. Land-grabbing, mainly in sub-Saharan Africa for the future cultivation of food and biofuels, also has the potential to force populations towards cities or more marginal lands (Sassen 2014). The great unknown is the number of those displaced through climate change, with estimates ranging up to the hundreds of millions, although these are almost certainly on the alarmist side. It could be argued that these migrants are displaced by political factors, or at least by the failure of political leadership. Although the topic is more a subset of the migration and development theme than of political demography as such, it will be one of the leading research areas as we move further through the twenty-first century and is likely to emerge as a major separate area in its own right. The most balanced assessments to date are still the composite reports by Foresight (2011) and Piguet et al. (2011), although see also Adger and Safra de Campos (2020) and Martin (2020).
Migrants in the diaspora have often been agents of political change, either revolutionary or reformist. Communities of migrants overseas can act as catalysts, exchanging ideas of how change can be brought to the homeland, and returnees have played significant roles in the subsequent political and economic transformations of their countries of origin. The overseas education and experiences of leaders such as Lee Kuan Yew (Singapore), Muhammad Ali Jinnah (Pakistan), Mahatma Gandhi and Jawaharlal Nehru (India), HồChí Minh (Vietnam), Sun Yat-sen, Deng Xiaoping, and Zhou Enlai (China), and Kwame Nkrumah (Ghana), as well as other pan-Africanists in the UK, France, and the US, were formative for their later political involvement in their homelands. The importance of the migrant experience and political development is well summarized in Moses (2011), and see also Goldstone (1999) and Skeldon (1987Skeldon ( , 2021. Migration, at both the aggregate level of flows and the individual level of students and potential leaders, has been a central pillar of political and state development.
This section has discussed some of the most significant issues in migration over recent decades that have direct implications for the study of human populations. The discussion now turns to how the journal Population Studies has engaged with such topics during the 25 years to its 75th anniversary.

The journal Population Studies and migration
Unsurprisingly, in Population Studies: A Journal of Demography, the major topics have revolved around fertility. Migration has been a minority interest, with a total of some 64 papers published on the topic during the first 50 years of its existence, 1947-96, giving an average of 1.3 papers a year. In the following quarter century to the 75th anniversary, from volume 51, issue 1, in 1997 to volume 74, issue 3, in 2020 (including supplements), 45 papers devoted to migration were published, of which 15 appeared in the last five years. The identification of 'migration-related' articles was by necessity somewhat rough and ready and was where the word 'migration' either appeared in the title or was the major thrust of the article. Papers that mention migration simply in passing or where it was secondary to the main theme of the paper were excluded, as were book reviews. Papers dealing with urbanization were included. The references to the migration-related papers in Population Studies that are given in this section appear in Annex 1, by theme, not in the main list of references.
To the 45 papers must be added an additional four papers on migration that appeared in the special issue, The Science of Choice, in 2017. The total of 49 migration-related papers over the 24-year period gives a crude average of just over two S36 Ronald Skeldon papers on migration per year between 1997 and 2020, an increase over the previous period certainly -particularly during the last five years, when about three papers per year on migration were published-but perhaps not enough to suggest that the journal would be foremost in the minds of those researching migration.
Yet, papers published in Population Studies did focus on areas of migration that might not have been covered to the same degree in other migration-related journals. The link between migration and fertility behaviour, either of migrants themselves or how migration could impact fertility, was covered in nine papers (see Annex 1). Migration and marriage, family formation, household composition, and the influence of family ties were covered in ten papers. More broadly, both marriage and migration were identified by Caldwell (1996) as being among the new themes in demography in the period after the Second World War. Both themes continued in the journal as minority concerns, particularly in relation to linkages between the two. Yet, with the partial exception of the 2012 paper by Hertrich and Lescling and, the spatial patterns of migration associated with marriage were rarely addressed. These circuits, dominated by women, have seen the emergence of new forms of migration, as some have moved away from predominantly local or extended kin networks to global and more commercial networks as gender relations changed in both origin and destination areas. Transnational marriage markets have emerged as potential partners proved difficult to find in more developed destination societies.
Only one paper in Population Studies during the period looked specifically at migration and mortality (see Annex 1). Although the majority of the 20 papers covered in the three aforementioned categories were based on detailed specific cases, several would become grist to the mill of research into links between the vital (fertility and mortality) and mobility (migration) transitions considered as central to the area of theory in migration research described earlier, even though that macro-level theory was unlikely to have been the primary objective of those specific studies.
More directly linked to the research areas in migration outlined earlier were an additional eleven papers on methods and the elaboration of techniques to improve the measurement of migration, plus six papers on migrant decisionmaking, including four papers in the 2017 special issue, which dealt with issues such as agent-based modelling (see Annex 1). Only two papers, both on remittances, specifically dealt with migration and development, although an additional two papers dealt with migration and health issues. The eight remaining miscellaneous papers covered general migration issues, either historically or relating to specific country contexts. With the partial exception of one of these papers (on migration and armed conflict in Nepal), not one paper dealt overtly with the last of the research areas identified: the political demography of migration.
It is perhaps invidious to force papers into the topic straightjackets suggested in the previous paragraph: for example, virtually all papers had a method and some kind of theoretical or conceptual framework. Nevertheless, the categorization does provide a general indication of the orientation of approaches to migration taken during the period. When compared with the first 50 years of the journal, the importance of papers dealing with the link to fertility has remained but, overall, the studies appear to have become more thematic and methodological. Country-specific discussions of historical immigration and emigration in more developed countries appear to have declined as the complexity of the migration process increased. Certainly, migration, if still not the primary demographic variable of concern to the journal, has moved a little closer towards the centre of demographic interest. However, at the same time, other developments were taking place in migration studies.

Migration studies: An emerging independent field
Over the last three decades, migration studies has emerged as a major field in its own right, drawing on researchers from anthropology, history, economics, law, sociology, and post-colonial studies into its teaching programmes, if we accept Wikipedia's definition of 'migration studies'. Demography and population studies are missing from the list, although perhaps the absence of the latter is understandable because it is itself interdisciplinary. So too is geography missing, a discipline that for long had provided a home for studies on migration, although more so in Europe and Australasia than in North America.
In the UK, the first centre to integrate migration as a major research concern was the Centre for Research in Ethnic Relations, founded in 1970 at the University of Bristol, with later moves to Aston University in 1978 and finally to the University of Warwick in 1984. Migration was thus approached more through the lens of anthropological and sociological interpretations of ethnicity rather than as a population variable in its own right. A more direct approach to the study of migration, albeit with an emphasis on forced movements, was the establishment of the Refugee Studies Centre at the University of Oxford in 1982. The first centre dedicated to research and teaching on migration as a subject in its own right was at the University of Sussex from the mid-1990s, with the first MA programme in Migration Studies launched in 1997, coordinated through the Sussex Centre for Migration Research (King 2011).
By the early 2020s, seven universities in the UK were offering MA programmes in migration studies, and these had been joined by universities in Sweden and the Netherlands in particular, but also in Denmark, Italy, Hungary, Israel, the US, New Zealand, and Ghana. While some of these programmes incorporated refugee and diaspora studies, these two subjects themselves are often expanded to programmes of their own and offered in yet more universities. The current concerns about the apparent 'crisis' in forced, refugee, and economic migrations have only accelerated this trend, with governments, international agencies, and civil society organizations seeking staff trained in these fields.
Thus, migration studies has developed an identity of its own, quite separate from that of population studies in general and demography in particular, an identity that has been reinforced through the emergence of new journals as outlets for the work of migration researchers. Until the 1960s, publications on migration tended to appear in journals that represented the disciplines of the researchers, a trend that has continued to this day for some, and particularly for economists. The requirements of academic promotion and assessment are still largely determined by disciplinary affiliation and woe betide those who deviate from the esprit de corps towards more nebulous 'centres' or interdisciplinary studies.
Nevertheless, in migration studies, specialized journals began to appear from the early 1960s with the International Migration Digest  Migration and Law (1999), the Journal of International Migration and Integration (2000), and the Journal on Migration and Human Security (2013), and so it went on to reach the numbers of journals listed in IOM (2020) and SFM (2020).
Several of these leading journals publish the kinds of papers seen in Population Studies, but most also include less quantitative, more qualitative studies and those that engage with the postmodernisms, postcolonialities, and feminist debates that reflect the times: precisely those areas that Greenhalgh (1996, pp. 27-8) accused demography of ignoring and which, too, seem to be largely missing in those recent papers on migration in Population Studies just discussed. Migration, as a demographic variable, while inching towards greater acceptance in the population field, has at the same time moved outwards to embark into the broader, more contested, environment of migration studies.

Conclusion
This paper has attempted to review the changing position of migration in demography and the broader field of population studies, essentially since the mid-1990s, even if an occasional sally into earlier years was required to clarify longer-term trends. With the decline in fertility, migration has moved from a quite marginal position in demographic studies towards a position of greater significance, even if it is still overshadowed by studies on fertility and mortality. Away from the focus on demography and even population studies, however, migration has emerged onto a platform of its own, the interdisciplinary migration studies, where a bewildering range of topics have been considered in a variety of often-contested ideological frameworks.
This trend reflects the increasing specialization of knowledge and has been mirrored in the other two demographic processes, with fertility studies moving in a strong medical and reproductive health direction, and the ageing of the human population leading to the emergence of centres specializing in gerontology. In migration studies, however, a disconnect has arisen between its demographic and statistical roots and its new and multifaceted forms. The underlying demographic context of fertility and mortality, and the methodological difficulties of measuring migration itself, have too often been sidestepped S38 Ronald Skeldon in so many of these studies. Central to future research is improved communication, even just awareness of each other, among those working in the 'migration in demography' tradition and those on that broader platform of migration studies. There is little that is new in this division: the bifurcation of approaches in migration reflects the broader divisions in knowledge between the scientists on the one hand and the humanists on the other, identified by C. P. Snow in his 'two cultures' over half a century ago (1959).
It is probably too idealistic to think that the schism can be unified in population studies over the next quarter century of Population Studies to the 100th anniversary issue in 2046. Perhaps it would be wrong even to seek it, as further fragmentation can perhaps be expected during this period: a fragmentation that is the result of new demands and realities. The division between macro-level approaches, based largely on the quantitative aggregate flows of migration, and the more micro-level focus on migrants is certainly likely to persist, with the former concentrated in disciplines such as economics and demography and the latter promoted in the expanding migration studies centres. While it is possible that the four themes identified in this paperdata and measurement, theory, migration and development, and political demography-will continue throughout the period, two additional interrelated themes guiding students of migration and demographers in the upcoming years can be foreseen: . environmental and specifically climate-changerelated movements, which were partially recognized by Keyfitz in the 1996 anniversary issue, but which have now taken on a pressing urgency; . following the Covid-19 pandemic, a growing interest in epidemiological research in which analysis of the spatial mobilities of populations will be a central concern.
To these might be added two general, cross-cutting issues: . the emergence of a slowing in migration, not just limited to the developed world or to internal migration, and an associated increase in different forms of spatial mobility (see Champion et al. 2018); . a shift from a focus on the national level to subnational levels, and particularly to the level of large cities. Migration and population studies need to drill down below the national to regional and local levels. Studies showing that migration occurred from China to the US or UK, for example, do not teach us very much. We need to know from which parts of China the migration originates and to which parts of the US or UK the migrants are going. We might expect much of the migration to be urban-to-urban.
These possible but largely unknowable future directions of migration within the discipline will occur against a knowable critical demographic shift: an increasing number of countries entering into a period of low, even declining, population growth. The challenges facing declining populations are likely to be very different from, but as great as, those that faced expanding populations in the previous century. Several of these challenges will involve migration: in terms of both the supply of potential migrants and the demand for workers. As already seen, one of the few robust generalizations about migration is that the majority of migrants are young adults. It follows that the volume of population movements will be impacted in some way by the size of the relevant cohorts of young adults, not in any simple or deterministic way, but variably, depending on other parameters. In an examination of global declines in the volume of internal migration, for example, about one-fifth can be attributed to ageing (Bell et al. 2018).
Shifting demand for migrants is likely to provide a greater challenge than the potential supply. Cohorts lost to fertility decline can never realistically be replaced by migration: the numbers required are simply too large in the case of most major developed economies (United Nations 2001). Nevertheless, migration can help to fill gaps in the labour market, but how it will interact with other policies to extend working life, increase the labour force participation of less active groups, and incorporate labour-saving technologies is not so clear. These interactions, however, are likely to be highly variable. What is clear is that migration, however defined, must become more central within this matrix of demographic change if the challenges provided by population decline are to be adequately addressed. In contexts where onefifth to one-quarter of births are to migrant women, the changing composition of sustainable populations poses critical ethical, social, and political questions that require sensitive and realistic resolution. Despite all the complexities, if this paper has a key message, it is that migration, irrespective of whether it remains a stepchild or not, has emerged as an important and necessary member of the population family, making it essential to integrate more fully into future demographic analyses and scenarios. So doing will help to bring the rigour of the demographic approach into the wider world of migration studies. And, in this way, migration may be upgraded from stepchild to the bonus child of demography.

Notes and acknowledgements
1 Please direct all correspondence to Ronald Skeldon, Kingston, Westbury Road, Nairn, Highland, IV12 4QF, United Kingdom; or by Email: R.Skeldon@sussex.ac.uk 2 I thank the editors of this special issue for their painstaking reading of previous drafts of this paper and the two anonymous referees for their detailed comments. The final version may not have entirely addressed all the points raised during the various rounds of discussion and assessment, but the paper has certainly benefited immensely from the process. A special thanks to Alice Reid for the felicitous term, 'bonus child in demography', apparently common in Denmark; it seemed appropriate to end the piece with this idea. Only the author, however, has the final responsibility for the opinions expressed and interpretations given.