Migration abstracts

Strand organiser: Professor Hill Kulu, University of Liverpool

Internal migration: Monday 7th September, 1:30pm

Union dissolution and migration
Thomas J. Cooke1, Clara Mulder2, Michael Thomas2. 1 Department of Geography, University of Connecticut 2 Department of Demography, University of Groningen

While there is a limited body of research regarding residential mobility following union dissolution, there are no studies of the effect of union dissolution on long-distance migration. This research isolates and identifies the processes that influence migration in the period immediately following the dissolution of a marital or long-term cohabiting union. Multi-level logit models of the probability of inter-state migration following the dissolution of marital and long-term cohabiting unions are estimated using data drawn from the 1975 through 2011 US Panel Study of Income Dynamics (PSID). The results provide support for the gendered model of family migration, indicate that separated parents are less likely to migrate than ex-partners without children, and that the migration decisions of former partners may remain linked through their children even after the end of their union. These findings imply that growing family complexity may be a factor in the long-term decline in aggregate US migration rates.  

Email: thomas.cooke@uconn.edu

Overcoming data issues to project interregional migration flows amongst Australia’s Indigenous population
James Raymer, Nicholas Biddle, Paul Campbell, Research School of Social Sciences, Australian National University

In this paper, a model is developed to project the interregional migration flows for the Aboriginal and Torres Strait Islander (Indigenous) populations in Australia at the state and territory level by age and sex. Recent migration flow data, obtained from the two most recent Australian quinary censuses, are first assessed and analysed in comparison with the patterns of the corresponding non-Indigenous population. Log-linear models are used to identify the key structures and patterns over time. A model is then developed to project the migration flows by origin, destination, age and sex forward in five-year increments to 2021. This includes incorporating techniques to overcome the small number cell issues associated with the very small population size of the Indigenous population. The results of this research provide (i) insights into the different migration patterns of an important but disadvantaged minority population in Australia and (ii) inputs for a dynamic multiregional model of Indigenous population change. This research is funded by the Australian Research Council as part of the Linkage Project on ‘Improved Indigenous Population Projections for Policy and Planning.’

Email: james.raymer@anu.edu.au

Long-term trends in spatial mobility: An order-specific analysis of migration of young adults in Sweden
Hill Kulu1, Emma Lundholm2, Gunnar Malmberg2, 1 University of Liverpool, 2 Umea University

The aim of this study is to investigate spatial mobility over time. Most studies on internal migration focus on spatial redistribution of population and determinants of inter-regional migration flows; surprisingly little research has investigated dynamics of spatial mobility in industrialised societies over time. By using Swedish register data we will calculate annual age standardised migration rates to investigate spatial mobility of young adults (aged 18 to 29) over the last three decades (from 1986 to 2009). We will then disaggregate mobility rates by calculating order-specific migration rates. We will next standardise order-specific mobility rates for educational enrolment and level and for family status to determine how much changes in various life domains of individuals explain the change in mobility levels over time. The analysis shows that migration rates for young adults significantly increased in the 1990s; while all order-specific migration rates increased, the first migration rates grew the most. Changes in population composition, particularly increased enrolment in higher education accounted for much of elevated spatial mobility levels in the 1990s.

Email: hill.kulu@liverpool.ac.uk

Internal migration in Austria along the rural-urban continuum and across the life course, 1996–2014
Ramon Bauer1, Nikola Sander2, 1 Vienna Institute of Demography (Austrian Academy of Sciences) – Wittgenstein Centre for Demography and Global Human Capital & University of Vienna, Department of Geography and Regional Research, 2 Vienna Institute of Demography (Austrian Academy of Sciences) – Wittgenstein Centre for Demography and Global Human Capital

Research on the demography of Austria has focused largely on trends in fertility, ageing and international migration. Surprisingly little attention has been paid to spatial patterns and trends in internal migration, despite its strong influence on population change at the regional and local scales. Understanding how much, where, and why cities growing and rural areas are shrinking is essential for the provision of services of general interest, but Austria’s research community has been reluctant to grapple with the complexities of origin-destination-specific migration flows at finer spatial scales. Earlier work has focused mostly on net migration at the state level or on case studies for particular regions or municipalities. The recent shift from sub-urbanisation to re-urbanisation that has been noted in other Western countries has not yet been studied comprehensively in the Austrian context. This paper adopts a spatially explicit approach to the study of migration flows between Austria’s cities, hinterlands and rural areas, how they changed over time, and how they differ across the life course. We draw on annual migration flows at the municipality-level to create a series of alternative custom geographies that allow for a more effective classification of settlements along the rural-urban continuum than traditional state- or NUTS-based analyses. We demonstrate how the observed migration patterns differ depending on the underlying geography and propose a typology of Austria’s municipalities tailored to the study of migration and its effect on population redistribution. We conclude by comparing our results to those obtained for the US, Germany and the UK.

Email: ramon.bauer@oeaw.ac.at

Immigrant selection and integration – Tuesday 8th September, 11am

Selective sorting and the stayers left behind: what does this mean for changing ethnic health gradients in England?
Fran Darlington 1, Paul Norman 1, Dimitris Ballas 2, 1 University of Leeds, 2 University of Sheffield

A number of studies explore the contribution of selective migration to health gradients, particularly in the UK. Using longitudinal data, these studies find that through migration or residential mobility (however defined), a sorting process can differently distribute (un)healthy groups of people between areas, and that over time this can widen health gradients. The mobility of these differently healthy groups is therefore of central importance. However, the immobile groups and their possible contribution to changing health gradients receive less specific coverage in the literature. Further, no work has investigated ethnic variations in either the mobile or immobile groups, or their influence on changing ethnic health gradients. This paper asks whether ‘stayers’ influence health gradients, and whether the relationship between ‘staying’ and health varies by ethnic group. It is likely that propensities for immobility may be as important as propensities for mobility in respect of the influence on changing health gradients. Starting with the premise that selective sorting between area-types through migration can widen health gradients, as various work has demonstrated, this paper will explore how socio-demographic attributes, health and experience of deprivation help explain propensity to migrate for different ethnic groups in England. Using data from the Office for National Statistics Longitudinal Study in 1991, 2001 and 2011, binary logistic regression models will be used to illustrate the increased likelihood of immobility for certain population subgroups which may exacerbate health gradients already widened by selective migration.

Email: gyfd@leeds.ac.uk

Union dissolution and residential mobility among Dutch and non-Western immigrant women in the Netherlands
Ilse Rooyackers1, Marjolijn Das2, Helga de Valk1,3, 1 Netherlands Interdisciplinary Demographic Institute, The Hague/KNAW/UoG, The Netherlands, 2 Statistics Netherlands, 3 Vrije Universiteit Brussel, Brussels

Separating from a partner usually sets into motion other important life-events, such as moving out of the shared home. The focus of this paper is on rates of union dissolution and residential mobility among Dutch and immigrant women of various origin in the Netherlands. We questioned to what extent diversity in dissolution and mobility between origin groups could be explained by the socio-demographic characteristics of women. Unique administrative population data were used (System of Social Statistical Datasets, Statistics Netherlands), containing information on all Dutch, Turkish, Moroccan, Surinamese, Antillean, other non-Western and Western immigrant women in the Netherlands in 2008 (N=868,831). Using logistic regression models, we analyzed to what extent separation rates were accounted for by income, home ownership, type of union, number of children at home, living area and age. Next, focusing on separated women only (N=30,707), we examined how these features explained residential mobility. Our results affirmed that levels of union dissolution differed across origin groups. The multivariate analyses indicated that while the lower rates of separation among Turkish and Moroccan women could be completely accounted for by their likelihood to have more children and marry rather than cohabit, none of the socio-demographics had explanatory value for the higher rates of separation among Surinamese and Antillean women. In all four migrant groups, women were found to be less likely to move compared to the Dutch. This was partly explained by the fact that they less often owned their house, lived in larger cities and had more children.

Email: rooyackers@nidi.nl

Economic and policy determinants of immigrants’ interstate migration
Barry Edmonston, Sharon M. Lee, Population Research Group and Department of Sociology, University of Victoria, Canada

Studying immigrants’ interstate migration is important in countries with high levels of immigration because it affects population distribution and growth, and related social, economic, and political processes. There are two research questions. First, what are the effects of economic factors (including employment growth, wage rates) on immigrants’ interstate migration? Using data from the 2000, 2005, 2010, and 2013 American Community Surveys and other sources, we found that geographical mobility is relatively common for both U.S.-born and foreign-born residents, and the effects of economic factors on interstate migration are fairly similar for both groups: people move from states with low employment growth and wages to states with high employment growth and wages. However, there are important differences, for example, the U.S.-born tend to move from states with a higher proportion of immigrants, while immigrants are more likely to stay in states with a higher proportion of foreign-born co-ethnics. The second research question is: What are the effects of state immigration-related policies (for example, laws requiring employers to verify legal status of workers and schools to determine students’ immigration status) on immigrants’ interstate migration? To date, there is little empirical research on the effects of state policies that might encourage out-migration and discourage in-migration. This study makes a new contribution by studying the effects of state immigration-related laws, enacted from 2005 to present, on interstate migration of immigrants. Results show that states with unfavourable immigration policies have higher rates of immigrant departures and are less likely to attract immigrants who move interstate.

Email: be@avic.ca

Labour market characteristics and conditions of first generation migrants: Is the Southern European model of migration still valid today?
Anna Di Bartolomeo1, Giuseppe Gabrielli2, Salvatore Strozza2, 1 European University Institute, Robert Schuman Centre for Advanced Studies, 2 University of Naples “Federico II”, Department of Political Sciences

The paper aims at analysing the differential demographic, social and labour market characteristics of immigrants in Italy, Spain and the UK, three of the most important destination countries of Europe during the last fifteen years. The focus is on employment patterns and over-qualification levels addressing a number of research questions. Which is the recent evolution of migration patterns in the three countries? How do migrants perform compared with natives in terms of employment rates and over-qualification levels? How do migrants differ in terms of demographic, social and labour market characteristics? Is the Southern European model of migration still valid today in terms of labour market performances? What is the role of migrant characteristics by sex and origin dynamics? Data come from the 2008 Ad-Hoc Module of the European Union Labour Force Survey which has a specific focus on the integration of migrants and their descendants in the labour market. We analyse different labour market indicators including over-qualification rates of natives and migrants according to several demographic characteristics. In order to disentangle the determinants of the differences in over-qualification by country of residence, we apply the two-step Heckman procedure that fits a probit model with sample selection. Preliminary results confirmed the largest growth of migration during recent years. East-West migration flows have become more important than the South-North flows. Compared to the UK situation, Italy and Spain share several common traits suggesting that the Southern European Model of migration is still valid today, also with respect to over-qualification dynamics.

Email: Anna.DiBartolomeo@EUI.eu 

Family dynamics among the descendants of immigrants – Wednesday 9 September, 9am

Union formation among the children of immigrants in Sweden
Ognjen Obucina, Stockholm University

The aim of this paper is to analyse the patterns of the first union formation among the children of immigrants in Sweden. The study is based on Swedish register data and covers the time period between 1990 and 2012. The main comparison is made between Swedish-born children with two foreign-born parents and children of intermarriage, here defined as Swedish-born individuals with one native parent. Union formation is analysed with respect to partner’s origin and the type of living arrangement. Descriptive analysis shows that children of intermarriage are much more likely to start the first union with a native person as compared to the children with two foreign-born parents. They are also more likely to form childbearing non-marital cohabitations. Multivariate analysis is modelled as event history analysis with six competing risks: 1) cohabitation with native, 2) marriage with native, 3) cohabitation with partner of the same origin, 4) marriage with partner of the same origin, 5) cohabitation with person of other immigrant background, and 6) marriage with person of other immigrant background. The results show that higher education implies a higher likelihood of union with a native person or a person of other immigrant background, regardless of the type of living arrangement. The role of education is less pronounced when it comes to the formation of cohabitation or marriage with a partner of the same origin. Having one native parent has a stronger impact on union formation if the immigrant parent originates from a country culturally less similar to Sweden.

Email: ognjen.obucina@sociology.su.se

Socio-economic effects on first union formation among descendants of immigrants in Belgium
Layla Van den Berg, University of Antwerp

Belgium has seen a steady growth of immigrants since the 1960s which has also resulted in an increasing group of children of immigrants. While the association between socio-economic position and union formation has mainly been studied among majority populations, these patterns of union formation have not been studied to the same extent among the descendants of immigrants. The objective of this paper is to examine to what extent socio-economic theories of union formation mainly developed for the majority population are also applicable to the second generation migrant women in Belgium. We make use of longitudinal micro-data from the Belgian Administrative Socio-Demographic Panel to examine patterns of union formation among women of Southern-European, Eastern-European, Turkish and Moroccan origin in Belgium. Using event history models of union formation, conditional logistic regression for union type and multinomial logit models of partner choice and socio-economic position of the household we reach two main conclusions. First, a disadvantaged socio-economic position increases the odds of union and marriage among women Turkish and Moroccan origin while the effect is inverse among native Belgians. Second, marrying a first generation migrant significantly increases the odds of starting a household with a weak socio-economic position for the descendants of migrants. Because Belgian-born women of Turkish and Moroccan origin   often form a union with an immigrant of the same origin this implies a disadvantaged starting point for the newly formed household.

Email: layla.vandenberg@uantwerpen.be

Fertility among the descendants of immigrants in eight European countries
Hill Kulu1, Tina Hannemann1, Ariane Pailhé2, Karel Neels3, Leen Rahnu4,  Allan Puur4,  Sandra Krapf5,  Amparo González-Ferrer6,  Teresa Castro-Martin6,  Elisabeth Kraus6, Laura Bernardi7,  Andrés Guarin7, University of Lausanne Gunnar Andersson8,  Lotta Persson8; 1University of Liverpool, 2Institut national d’études démographiques, 3University of Antwerp, 4University of Tallin, 5University of Cologne, 6Centro de Ciencias Humanas y Sociales, 7University of Lausanne, 8Stockholm University

This study investigates the childbearing patterns of the descendants of immigrants in selected European countries, with a focus on ethnic minority women whose parents arrived in Europe from high-fertility countries. While the fertility levels of immigrants to Europe have been examined in the recent literature, the childbearing patterns among their descendants have received little attention. In the last decades, the share of individuals with at least one foreign-born parent has increased in many European countries. Using longitudinal data from eight European countries (UK, France, Germany, Belgium, Switzerland, Sweden, Spain and Estonia) and applying Poisson regression models, the study shows that many descendants of immigrants exhibit first-birth levels that are similar to the ‘native’ population in their respective countries; however, first-birth levels are elevated among women of Pakistani and Bangladeshi origin in the UK and for those of Turkish descent in France and Belgium. Transition rates to a second child vary less across ethnic groups. Most ethnic minority women in the UK, France and Belgium show significantly higher third-birth levels than ‘natives’ in those countries. The inclusion of women’s level of education in the analysis has little effect on fertility differences across the ethnic groups, whereas country context matters. Overall, the childbearing behaviour of the descendants of immigrants falls in between the fertility pathways experienced by their parents’ generation and the respective ‘native’ populations. The analysis supports the idea that both the mainstream society and the minority subculture shape the childbearing patterns of the descendants of immigrants in Europe.

Email: tina.hannemann@liverpool.ac.uk

Does partner choice affect the transition to higher order births? A Belgian study analysing the influence of the partner choice on the transition to second and higher order births for second-generation women of Turkish and Moroccan origin
Lisa Van Landschoot1, Jan Van Bavel2, Helga A.G. de Valk3, 1 Vrije Universiteit Brussel, Interface Demography, 2 Vrije Universiteit Brussel/University of Leuven, Centre for Sociological Research, 3 Vrije Universiteit Brussel/Netherlands Interdisciplinary Demographic Institute/ Koninklijke Nederlandse Akademie van Wetenschappen/University of Groningen

Belgian research analysing the influence of the partner choice on the transition to parenthood revealed for second-generation women of Turkish and Moroccan origin differentiated transition rates to a first birth. Second-generation women of both origin groups in union with a partner of the same origin (endogamous unions) record higher transition rates to parenthood compared to unions where the partner is a native Belgian man (exogamous unions). However, contrary to expectation, the generation of the endogamous partner did not differently affect the probability of entering parenthood. Whether the partner is of first- or second-generation, the transition rates to parenthood are approximately the same. Therefore, the aim of this study is to analyse whether the distinction of having a first- or second-generation partner of same origin emerges when the transition to second and higher order births is examined. In order to do so, we use event history techniques applied on data from the 2001 Belgian Census linked with data from the National Population Register (2006). These data include the full population legally present in the country and provide a wide range of individual-level indicators. By analysing the transition to second and subsequent births, we expect that the  contrast between endogamous and exogamous unions will diminish. Moreover, we expect to find the highest rates of higher order births when the partner immigrated to Belgium (first-generation), followed by endogamous unions with a second-generation partner and the lowest when the partner is a native Belgian man.

Email: Lisa.Van.Landschoot@vub.ac.be

Transition to adulthood among descendants of immigrants in France: Cultural versus structural determinants
Giulia Ferrari, Ariane Pailhé, INED 

Patterns of transition to adulthood have changed substantially in most Western countries from the late 1960s onwards. The emerging trend reveals more complex and longer trajectories due to changes in norms, greater labour market insecurity and the increased cost of living. Socio-economic stratification plays a crucial role in this process. This study investigates if such changes have also taken place among descendants of immigrants in France. Focusing on the interplay between cultural and structural factors, it investigates whether individual behaviours are affected by tradition and open to individual choice and/or shaped by structural factors, especially economic resources. Based on event-history data collected in the Trajectories and Origins survey (Ined, Insee 2008), we construct individual sequences of states occurred during the transition to adulthood, paying particular attention to residential, occupational, conjugal and parenthood dimensions. Using optimal matching sequence analysis, we identify similar multidimensional trajectories to adult roles and by means of a hierarchical cluster analysis we classify them into five transition typologies, namely Married parents (37%), Single and childless (32%), Cohabiting (15%), Latest nest leavers (12%), Unemployed in union (3%). To investigate whether different adult behaviours depend on individual and background socioeconomic characteristics, as well as on structural aspects, we apply a multinomial logistic regression model on the likelihood of belonging to one of the five selected clusters. Our findings highlight the significant role played by individual characteristics (namely, gender and origin) and structural factors (namely, level of education) in explaining differences in the trajectories to adulthood. Finally, we will extend the multivariate analysis by testing whether these differences persist once we account for interactions with gender and other important individual characteristics. 

Email: giulia.ferrari@ined.fr 

Contextual effects and population change – Wednesday 9 September, 11.30am

Contextual influence on mate selection: Is there an effect of adolescent neighborhood characteristics on future partner choice?
Karen Haandrikman, Bo Malmberg, Department of Human Geography, Stockholm University 

Extensive research on mate selection shows that similarities in terms of social status, education, religion and geographical origin are important factors in partner choice. In this study we will expand on this research by considering an additional factor that can increase the probability of a match: coming from a neighborhood with shared socio-demographic characteristics. In the neighborhood effects literature, the importance of neighborhood context for the development of social norms, behavioral patterns and demographic outcomes has repeatedly been stressed. The paper will address the following research question: Is partner choice affected by neighborhood context in youth? The study employs a longitudinal design, using register data for the 1980 Swedish birth cohort, measuring neighborhood context at age 15 and partner selection from age 18 to 32. A conditional logit model is estimated using the actual partner and a number of alternative partners, including individual and neighborhood level variables. Neighborhood context is measured using aggregate statistics for the population of individualized neighborhoods. Earlier studies using the same cohort have shown that neighborhood context has significant effects on educational achievement, early income career, and the timing of family formation. This suggests that coming from neighborhoods with shared characteristics can have a positive influence on the likelihood that two individuals become partners. In addition, this study may help to clarify whether partner choice patterns vary between individuals brought up in different types of neighborhoods. 

Email: karen.haandrikman@humangeo.su.se 

Geographical inequalities and population change in Britain, 1971-2011
Chris Lloyd, University of Liverpool 

This paper details a study of small area population change in Britain between 1971 and 2011. The analysis is made possible by a project which is reallocating available population counts from Census geographies for 1971, 1981, 1991, 2001 and 2011 onto a common gridded system. Given this resource, it is possible to explore change in small areas for any variables which are available for two or more Census years. The main focus of the paper is on the analysis of change in total population and selected variables including employment status and housing tenure, as represented using 1km by 1km grid squares for the whole of Britain. The geography of change in these variables is explored, in addition to identifying during which periods neighbourhoods experienced the greatest change (e.g., 2011-2001 or 1991-1981) are produced. The paper also seeks to explore how geographical inequalities have changed with respect to the selected variables. As an example, the question of how far the distribution of unemployed persons has become more or less uneven, or geographically concentrated, into particular regions or neighbourhoods, will be explored. It is demonstrated that the new grid square Census resource will enable the answering of questions which could not be answered using larger geographical aggregations of counts and that this resource has the potential to open up a major programme of new research on population change and its determinants. 

Email: c.d.lloyd@liverpool.ac.uk 

What is left behind? Gender-specific effects of commuting and relocation on social relationships and social integration
Natascha Nisic, Stefanie Kley, University of Hamburg 

This study analyses gender-specific effects of long-distance commuting and residential relocation on the social integration of couples. Research on geographical mobility in the partnership context focused very much on gender differences in determinants and outcomes of regional mobility with respect to labour market outcomes and economic success (Cooke 2008). The effects of regional mobility and their possibly gendered impact on social networks and the social integration of households have been little studied. This might present a shortcoming in many ways. In the literature on social capital and social networks there is a large body of evidence indicating that social support is important for individual well-being and that personal relationships are vital for many life domains. Drawing on exchange theory, one would expect that regional mobility, like residential relocation or long-distance commuting, imposes severe temporal and geographical restrictions on the formation and upholding of social relationships. Yet, literature on domestic and care work suggests that the management and maintenance of social relationships constitute a crucial aspect of household labour and thus falls in the domain of women. Against this background we analyse the effects of internal migration and job-related commuting on the amount and quality of social relationships male and female partners maintain. The analysis is based on the BHPS waves 1992-2008 and the panel fixed-effects models are applied. Results indicate that while women’s role as kinkeepers increases the quantity of social contacts in particular after relocation, these contacts are not perceived as individually satisfying. Among men social ties remain less affected by regional mobility. 

Email: natascha.nisic@wiso.uni-hamburg.de 

The importance of forced migration as a driver of population change during war: A case study of Sierra Leone, 1991-2002
Amie Kamanda, Jakub Bijak and Sabu Padmadas University of Southampton 

Forced migration is a human rights violation prohibited under the Geneva Convention (1949). However, since World War II, the forced displacement of civilians has been documented in most conflicts. During the conflict in Sierra Leone (1991-2002), an estimated 2.5 million people were forcibly displaced. Approximately 400,000 of those displaced were refugees. In total, 70% of the country’s pre-war population of 3.5 million people were forcibly displaced by the conflict. This paper aims to reconstruct the conflict population in order to assess how the three demographic components – mortality, migration and fertility – contributed to population change during this turbulent period in Sierra Leone’s history. The cohort component method is applied to mortality and fertility rates reconstructed from the 2008 and 2013 Sierra Leone Demographic and Health Survey. Furthermore, indirect estimation methods are applied to a 10% sample of the 2004 census of Sierra Leone to estimate post-conflict fertility and mortality. Information on net migration is extracted from Statistical Yearbooks published by the United Nations High Commissioner for Refugees. According to the results of the reconstruction, the total population loss during the war was 446,198. The difference in the reconstructed population attributed to migration is -350,000 while the difference attributed to mortality is -310,000. Hence, the analysis reveals that the main drivers of population change during the war were migration and mortality. Net emigration of more than 400,000 people from 1991 to 2000 slowed down the rate of population growth which resulted in a lower total population size, compared to the counterfactual projections. 

Email: ak904@soton.ac.uk