Final Report Summary - MIGFAM (International migrant families)
Demographers agree on three forces that drive the development of any population, namely births, deaths and migration. While fertility and mortality are extensively studied, migration and subsequent demographic behaviours of international migrants have received less attention in contemporary Europe. This is surprising given that some European Union (EU) member states, e.g. Germany and France, estimate that about 20 % of their current populations are immigrants or were born to immigrant parents. Migration movements are causing a growing trend toward ethnic, cultural and religious diversity. This cultural diversification is also likely to lead to different demographic patterns of sub-groups within the population of one country and this heterogeneity needs to be considered in population research and public policy making.
The project 'International migrant families' aimed at studying the demographic behaviour of international migrants in greater detail, focussing on Germany. International migration is associated with a rapid change in the migrants' environment and is often associated with social downward mobility. These changes usually take place within a much shorter time span than societies alter as a whole and immigrants must cope with these changes. Therefore, the study of the demographic behaviour of migrants gives us insights into the patterns and speed of the demographic responses of individuals or vulnerable groups when exposed to sudden changes in their physical and social environment. On the individual level, the life-course approach allows us to analyse the sequencing of several events, thereby studying both short-term as well as long-term effects of migration on a person's life.
The aims of the project were:
1. to address the heterogeneity within immigrant populations
2. to study the impact of the country of origin
3. to identify migrant-specific indicators.
Heterogeneity within immigrant populations
Migrant populations are heterogeneous in several aspects. The types of migrants vary over time and between countries. Western Europe experienced large inflows of labour migrants from the 1950s through the beginning of the 1970s, followed mainly by family migration. Since the 1980s there has been an increasing number of refugees and asylum seekers. In addition, Germany has received large numbers of ethnic Germans from eastern European countries. Each of these groups differs from others in their demographic composition as well as in the legal conditions under which they migrated. Hence, we must also distinguish between the immigrant generations as well as different countries of origin and cultural traditions.
Carnein, Milewski, Doblhammer and Nusselder present new aspects of health expectancies among elderly Turkish immigrants living in Germany, relative to the health expectancies of the German host population. The study has two objectives. First to research the differences of health expectancies between German and Turkish foreigners and second to examine potential factors which affect the health status of a population. The remaining years of life at age 50 are divided into years spent in good and in bad health by considering the age-specific prevalences of disability. This study uses population data from the German Office for National Statistics (Destatis) and the Central Register of Foreigners (AZR), as well as data on age-specific prevalences of disability from the German Generations and Gender Survey (GGS). The measurement of health expectancy is used to highlight health inequalities between Turkish migrants living in Germany and Germans. The decomposition method is employed in order to explore the impact of mortality and disability on differences in health expectancy. The results on health expectancy show significant differences by sex and nationality in the proportion of remaining life expectancy spent in good and in bad health. Within the Turkish population, women in particular show greater advantages in life expectancy, but also greater disadvantages in health relative to German non-migrants.
Milewski and Doblhammer analyse mental health among immigrants in Germany. Several hypotheses have been discussed, such as the healthy migrant effect and the salmon bias, focussing mainly on outcomes of physical health. The statuses of mental health have not been studied for the German context, however. Since immigrant status, socioeconomic disadvantages and discrimination are highly correlated, an analysis of the mental health of immigrants is necessary. Furthermore, a worse mental health may be a predecessor of physical health impairments or may be interrelated with them and may therefore serve as a predictor for health risks and inequalities in later life. We use data of the German Socio-Economic Panel (SOEP), which contains sub-samples on older immigrant groups, i.e. immigrant workers from the Mediterranean regions and new immigrant groups, such as from Eastern Europe. We use data from the latest available wave 2010. Mental health was included in the questionnaire as SF12, a set of questions that address both physical and psychological health. The latter one was included mainly with items that refer to symptoms of depression. The study investigates the risk of bad mental health by applying logistic regression analysis. Our sample contains about 17 000 cases (aged 18 to 80), where about one third are immigrants. We control for standard socio-demographic variables, such as sex, age and education, as well as migrant-specific indicators, such as duration of residence, social networks and discrimination. We find that immigrant status is not generally associated with a higher prevalence of bad mental health. By contrast, we find large differences between immigrant groups. Whereas immigrants from Eastern Europe have better mental health than non-migrants, the health of immigrant workers is worse. These differences can partly be traced back to the lower socio-economic status of the latter group.
Impact of the country of origin
In previous studies, attention has usually been given to comparing several immigrant groups in one country of destination. In order to address the question of why there are differences between migrants and non-migrants, one must not only consider individual characteristics and the background of the country of origin, however, but also the influence of the receiving/host society. This can be done properly only by comparing emigrants from the same country who leave for different countries of destination and who must thus adapt to alternate cultures and societal frameworks.
Milewski (2011) examines the fertility behaviour of descendants of Turkish immigrants in Western Europe. We used data from the project The Integration of the European Second Generation (TIES), which was carried out in Germany, France, Austria, Switzerland, the Netherlands and Sweden from 2006 to 2008. Each country sample includes about 250 women who were born in these countries and who have one or two parents born in Turkey, as well as approximately 250 non-migrant women. The respondents were 18 to 35 years old. We applied event-history techniques to the transition to a first birth. Descendants of Turkish immigrants were found to have lower first-birth transition rates in Germany and in Switzerland than in Sweden, the Netherlands and France. These differences cannot be explained in full by compositional differences of the Turkish second generation. This finding supports the hypothesis that immigrant descendants adapt to the host society's fertility behaviour.
Milewski and Otto study religiosity of young Turkish immigrants living in Germany. Whereas previous research concentrated on comparisons of immigrants and Germans, the focus of this study is a within-group comparison of Turks who are either descendants of immigrants or members of the first generation. We pay special attention to the role of the family, in particular the country background of the spouse. We use data from the Generations and Gender Survey (2005 and 2006). The sample contains about 3800 Turkish citizens aged 18 to 81, most of whom are Muslim. We apply logistic regression analysis to four indicators of Muslim religiosity: the importance of religion for weddings and for funerals, the importance of religion in childrearing and rates of attendance at services. Apart from the area of funerals, we find declining religiosity among descendants of Turkish immigrants relative to the first generation. This can be traced back to higher educational attainment among immigrant descendants. By contrast, immigrants in transnational unions show higher religiosity. These findings suggest that the Muslim community is undergoing two contradictory processes: on the one hand, they are being secularised through education; while on the other, their religiosity is strengthened when they are married to a partner who comes from Turkey.
Migrant-specific indicators
Since previous studies on fertility and mortality have usually not specifically addressed immigrants, these studies generally use indicators that apply to the indigenous populations and therefore rarely utilise migrant-specific factors. An exception can be seen with the duration of stay, now included in demographic studies with a growing attention to the life-course approach. One of the challenges for demographic research on international migrants is to single out relevant migrant-specific factors because different demographic patterns between the sub-populations cannot fully be explained by the standard indicators used thus far.
Milewski and Peters demonstrate that the study if immigrant health needs to consider more indicators than used so far, using the example of birthweight. The paper analyses birth outcomes among immigrant women in Germany relative to those of non-migrant mothers. In Germany, about one-third of all newborns are born to immigrant mothers. Since immigrant status and socioeconomic disadvantages are highly correlated, the health of immigrant children and their mothers had received an increasing amount of attention. When investigating perinatal outcomes, the evidence on the effect of the immigrant status of the mother on the health of her child, measured as birthweight, is contradictory. Some authors even claim a protective effect of the migrant status on low birthweight. We use the newborn sample of the German Socio-Economic Panel (SOEP), which contains pre-, peri- and postnatal variables that allow us to analyse the determinants of adverse birth outcomes. The data are on approximately 1500 births that occurred between 2001 and 2009. Our study investigates the risks of low and of high birthweights by applying logistic regression analyses to four birthweigh indicators. We find that immigrant status is not associated with a higher prevalence of low birthweight (LBW) and of newborns who are small for gestational age (SGA). By contrast, immigrant status is associated with a higher prevalence of high birthweight (HBW) and of increased prenatal growth (LGA). Control variables of the mother - i.e. age, parity, height, body mass index (BMI) - cannot explain the birthweight differences between immigrants and non-migrants. Thus, the findings contradict recent assertions in the literature that the disadvantage in birthweight among newborns of immigrant mothers has been levelling off. Instead our results suggest that the birth outcomes of immigrant mothers are doubly disadvantaged: their share of high birthweight outcomes is larger than among non-migrants and higher birthweight correlates with preterm birth. Since high birthweight is a risk factor for overweight and obesity, efforts to prevent these problems from developing should be undertaken during pregnancy and in early childhood.
Milewski and Kulu investigate the effect of native-immigrant intermarriage on divorce. International literature often uses the rate of mixed marriages as indicator for a (successful) integration of immigrants. We ask about the stabilty of these unions. We use a rich longitudinal data-set from Germany (SOEP) and apply event-history techniques to examine the risk of divorce among 'labour migrants' in Germany. The analysis of the divorce risk of 5 648 marriages shows that immigrant couples have lower divorce risks than non-migrants. However, marriages between German-born individuals and immigrants have a higher likelihood of divorce than marriages between German-born individuals or between immigrants from the same country supporting the heterogamy hypothesis. The socio-demographic composition of mixed couples only partially explains their high divorce levels. We discuss the role of various factors in explaining the high divorce risk of mixed marriages including the differences in values and preferences between the spouses, reduced support from social networks of the spouses and potential discrimination they experience in everyday lives.
These examples of research stress that the population of Germany is getting more heterogeneous because of immigration streams and different demographic behaviours of the sub-groups. The growing demographic and cultural heterogeneity causes on one hand challenges to research: migrant-specific indicators need to be included in studies and data collections should include new and smaller immigrant groups as well. On the other hand, the examples show that - arising from the heterogeneity of the population - immigrant women, men and children as vulnerable groups should receive more attention as targets of public-health measures.
The project 'International migrant families' aimed at studying the demographic behaviour of international migrants in greater detail, focussing on Germany. International migration is associated with a rapid change in the migrants' environment and is often associated with social downward mobility. These changes usually take place within a much shorter time span than societies alter as a whole and immigrants must cope with these changes. Therefore, the study of the demographic behaviour of migrants gives us insights into the patterns and speed of the demographic responses of individuals or vulnerable groups when exposed to sudden changes in their physical and social environment. On the individual level, the life-course approach allows us to analyse the sequencing of several events, thereby studying both short-term as well as long-term effects of migration on a person's life.
The aims of the project were:
1. to address the heterogeneity within immigrant populations
2. to study the impact of the country of origin
3. to identify migrant-specific indicators.
Heterogeneity within immigrant populations
Migrant populations are heterogeneous in several aspects. The types of migrants vary over time and between countries. Western Europe experienced large inflows of labour migrants from the 1950s through the beginning of the 1970s, followed mainly by family migration. Since the 1980s there has been an increasing number of refugees and asylum seekers. In addition, Germany has received large numbers of ethnic Germans from eastern European countries. Each of these groups differs from others in their demographic composition as well as in the legal conditions under which they migrated. Hence, we must also distinguish between the immigrant generations as well as different countries of origin and cultural traditions.
Carnein, Milewski, Doblhammer and Nusselder present new aspects of health expectancies among elderly Turkish immigrants living in Germany, relative to the health expectancies of the German host population. The study has two objectives. First to research the differences of health expectancies between German and Turkish foreigners and second to examine potential factors which affect the health status of a population. The remaining years of life at age 50 are divided into years spent in good and in bad health by considering the age-specific prevalences of disability. This study uses population data from the German Office for National Statistics (Destatis) and the Central Register of Foreigners (AZR), as well as data on age-specific prevalences of disability from the German Generations and Gender Survey (GGS). The measurement of health expectancy is used to highlight health inequalities between Turkish migrants living in Germany and Germans. The decomposition method is employed in order to explore the impact of mortality and disability on differences in health expectancy. The results on health expectancy show significant differences by sex and nationality in the proportion of remaining life expectancy spent in good and in bad health. Within the Turkish population, women in particular show greater advantages in life expectancy, but also greater disadvantages in health relative to German non-migrants.
Milewski and Doblhammer analyse mental health among immigrants in Germany. Several hypotheses have been discussed, such as the healthy migrant effect and the salmon bias, focussing mainly on outcomes of physical health. The statuses of mental health have not been studied for the German context, however. Since immigrant status, socioeconomic disadvantages and discrimination are highly correlated, an analysis of the mental health of immigrants is necessary. Furthermore, a worse mental health may be a predecessor of physical health impairments or may be interrelated with them and may therefore serve as a predictor for health risks and inequalities in later life. We use data of the German Socio-Economic Panel (SOEP), which contains sub-samples on older immigrant groups, i.e. immigrant workers from the Mediterranean regions and new immigrant groups, such as from Eastern Europe. We use data from the latest available wave 2010. Mental health was included in the questionnaire as SF12, a set of questions that address both physical and psychological health. The latter one was included mainly with items that refer to symptoms of depression. The study investigates the risk of bad mental health by applying logistic regression analysis. Our sample contains about 17 000 cases (aged 18 to 80), where about one third are immigrants. We control for standard socio-demographic variables, such as sex, age and education, as well as migrant-specific indicators, such as duration of residence, social networks and discrimination. We find that immigrant status is not generally associated with a higher prevalence of bad mental health. By contrast, we find large differences between immigrant groups. Whereas immigrants from Eastern Europe have better mental health than non-migrants, the health of immigrant workers is worse. These differences can partly be traced back to the lower socio-economic status of the latter group.
Impact of the country of origin
In previous studies, attention has usually been given to comparing several immigrant groups in one country of destination. In order to address the question of why there are differences between migrants and non-migrants, one must not only consider individual characteristics and the background of the country of origin, however, but also the influence of the receiving/host society. This can be done properly only by comparing emigrants from the same country who leave for different countries of destination and who must thus adapt to alternate cultures and societal frameworks.
Milewski (2011) examines the fertility behaviour of descendants of Turkish immigrants in Western Europe. We used data from the project The Integration of the European Second Generation (TIES), which was carried out in Germany, France, Austria, Switzerland, the Netherlands and Sweden from 2006 to 2008. Each country sample includes about 250 women who were born in these countries and who have one or two parents born in Turkey, as well as approximately 250 non-migrant women. The respondents were 18 to 35 years old. We applied event-history techniques to the transition to a first birth. Descendants of Turkish immigrants were found to have lower first-birth transition rates in Germany and in Switzerland than in Sweden, the Netherlands and France. These differences cannot be explained in full by compositional differences of the Turkish second generation. This finding supports the hypothesis that immigrant descendants adapt to the host society's fertility behaviour.
Milewski and Otto study religiosity of young Turkish immigrants living in Germany. Whereas previous research concentrated on comparisons of immigrants and Germans, the focus of this study is a within-group comparison of Turks who are either descendants of immigrants or members of the first generation. We pay special attention to the role of the family, in particular the country background of the spouse. We use data from the Generations and Gender Survey (2005 and 2006). The sample contains about 3800 Turkish citizens aged 18 to 81, most of whom are Muslim. We apply logistic regression analysis to four indicators of Muslim religiosity: the importance of religion for weddings and for funerals, the importance of religion in childrearing and rates of attendance at services. Apart from the area of funerals, we find declining religiosity among descendants of Turkish immigrants relative to the first generation. This can be traced back to higher educational attainment among immigrant descendants. By contrast, immigrants in transnational unions show higher religiosity. These findings suggest that the Muslim community is undergoing two contradictory processes: on the one hand, they are being secularised through education; while on the other, their religiosity is strengthened when they are married to a partner who comes from Turkey.
Migrant-specific indicators
Since previous studies on fertility and mortality have usually not specifically addressed immigrants, these studies generally use indicators that apply to the indigenous populations and therefore rarely utilise migrant-specific factors. An exception can be seen with the duration of stay, now included in demographic studies with a growing attention to the life-course approach. One of the challenges for demographic research on international migrants is to single out relevant migrant-specific factors because different demographic patterns between the sub-populations cannot fully be explained by the standard indicators used thus far.
Milewski and Peters demonstrate that the study if immigrant health needs to consider more indicators than used so far, using the example of birthweight. The paper analyses birth outcomes among immigrant women in Germany relative to those of non-migrant mothers. In Germany, about one-third of all newborns are born to immigrant mothers. Since immigrant status and socioeconomic disadvantages are highly correlated, the health of immigrant children and their mothers had received an increasing amount of attention. When investigating perinatal outcomes, the evidence on the effect of the immigrant status of the mother on the health of her child, measured as birthweight, is contradictory. Some authors even claim a protective effect of the migrant status on low birthweight. We use the newborn sample of the German Socio-Economic Panel (SOEP), which contains pre-, peri- and postnatal variables that allow us to analyse the determinants of adverse birth outcomes. The data are on approximately 1500 births that occurred between 2001 and 2009. Our study investigates the risks of low and of high birthweights by applying logistic regression analyses to four birthweigh indicators. We find that immigrant status is not associated with a higher prevalence of low birthweight (LBW) and of newborns who are small for gestational age (SGA). By contrast, immigrant status is associated with a higher prevalence of high birthweight (HBW) and of increased prenatal growth (LGA). Control variables of the mother - i.e. age, parity, height, body mass index (BMI) - cannot explain the birthweight differences between immigrants and non-migrants. Thus, the findings contradict recent assertions in the literature that the disadvantage in birthweight among newborns of immigrant mothers has been levelling off. Instead our results suggest that the birth outcomes of immigrant mothers are doubly disadvantaged: their share of high birthweight outcomes is larger than among non-migrants and higher birthweight correlates with preterm birth. Since high birthweight is a risk factor for overweight and obesity, efforts to prevent these problems from developing should be undertaken during pregnancy and in early childhood.
Milewski and Kulu investigate the effect of native-immigrant intermarriage on divorce. International literature often uses the rate of mixed marriages as indicator for a (successful) integration of immigrants. We ask about the stabilty of these unions. We use a rich longitudinal data-set from Germany (SOEP) and apply event-history techniques to examine the risk of divorce among 'labour migrants' in Germany. The analysis of the divorce risk of 5 648 marriages shows that immigrant couples have lower divorce risks than non-migrants. However, marriages between German-born individuals and immigrants have a higher likelihood of divorce than marriages between German-born individuals or between immigrants from the same country supporting the heterogamy hypothesis. The socio-demographic composition of mixed couples only partially explains their high divorce levels. We discuss the role of various factors in explaining the high divorce risk of mixed marriages including the differences in values and preferences between the spouses, reduced support from social networks of the spouses and potential discrimination they experience in everyday lives.
These examples of research stress that the population of Germany is getting more heterogeneous because of immigration streams and different demographic behaviours of the sub-groups. The growing demographic and cultural heterogeneity causes on one hand challenges to research: migrant-specific indicators need to be included in studies and data collections should include new and smaller immigrant groups as well. On the other hand, the examples show that - arising from the heterogeneity of the population - immigrant women, men and children as vulnerable groups should receive more attention as targets of public-health measures.