+91-8290613225 How to get love marriage success | Remedies to remove obstacles in love marriage
Partner Choice and Marriage Migrants - · PDF [email protected] Koen Matthijs...
Transcript of Partner Choice and Marriage Migrants - · PDF [email protected] Koen Matthijs...
Partner Choice and Marriage Among
Migrants The Acculturation Process of Migrants in
the Port‐City of Antwerp from a Life
Course Perspective (1846‐1920)
Paul Puschmann ‐ Nina Van den Driessche –
Koen Matthijs ‐ Bart Van de Putte
Working paper
WOG/HD/2012‐5
D/2012/1192/10
© Centrum voor Sociologisch Onderzoek (CeSO)
Parkstraat 45 – bus 3601
B – 3000 Leuven
All rights reserved. Except in those cases expressly determined by law,
no part of this publication may be multiplied, saved in an automated
datafile or made public in any way whatsoever without the express prior
written consent of the author
Alle rechten voorbehouden. Behoudens de uitdrukkelijk bij wet bepaalde
uitzonderingen mag niets van deze uitgave worden verveelvoudigd,
opgeslagen in een geautomatiseerd gegevensbestand of openbaar
gemaakt, door middel van druk, fotokopie, microfilm, of op welke
andere wijze ook, zonder voorafgaande schriftelijke toestemming van de
uitgever
Partner Choice and Marriage Among Migrants The Acculturation Process of Migrants in the Port‐City
of Antwerp from a Life Course Perspective (1846‐1920)i.
Paul Puschmann Family and Population Studies
Centre for Sociological Research
Parkstraat 45 – 3601
KU Leuven, Belgium
Room 02.205
+32 (0)16 32 34 72
Nina Van den Driessche Health and Demographic Research
Ghent University
Korte Meer 5
9000 Gent
+32 (09) 264 84 53
Koen Matthijs Family and Population Studies
Centre for Sociological Research
Parkstraat 45 – 3601
KU Leuven, Belgium
Room 02.207
+32 (0)16 32 31 73
Bart Van de Putte Health and Demographic Research
Ghent University
Korte Meer 5
9000 Gent
+32 (0) 9 264 68 00
1
Table of contents
Abstract 3
1. Introduction 4
2. Acculturation in the nineteenth and twentieth century 6
3. Theoretical framework: partner choice and marriage among migrants 7
4. Research Questions and Hypotheses 10
5. Historical Context 11
6. Data 14
7. Competing Risks Event History Analysis 19
8. Results 21
9. Conclusion 29
End notes 32
References 33
2
List of tables
Table 1: Migration in Antwerp, 1806 – 1890 (annual averages) 11
Table 2: Origin of Antwerp’s Population in 1900 According to 13
Birth Municipality (%)
Table 3: Example Person – Period File 16
Table 4: Descriptive statistics of the variables 21
Table 5: Subhazard Ratios for the Competing Risks of Getting Married: 27
Endogamous, Exogamous(migrant), Exogamous(local).
List of figures
Figure 1: Acculturation Attitudes of Immigrant Groups (Berry, 1997) 9
Figure 2: Antwerp District 15
Figure 3: Conceptual Model 20
Figure 4: Nelson‐ Aalen Cumulative Hazard Estimates for the 24
Competing Events
3
Abstract
This article deals with the question to which degree migrants married in Antwerp
(1846‐1920). It is assumed that marriage was a way of escaping marginalization. With
marginalization we mean that migrants were not able to build up a social network. For
those migrants who did marry, it is investigated with whom they married. We
analyzed which characteristics of the migrants increased the risk of entering
matrimony. We discerned four acculturation processes which were linked to outcomes
related to marriage and partner selection: marginalization (staying single), segregation
(endogamous marriage with a migrant from the same area), integration (exogamous
marriage with a migrant from another area) and assimilation (marriage with a native).
A competing risk regression was carried out on a dataset that consisted of all migrants
from the COR* database who settled as singles at marriageable ages in Antwerp or one
of its suburbs. The results show that males, literates, and short distance migrants were
at an increased risk of an endogamous marriage (separation). Migrants who settled at
young ages, and migrants who were born before 1845 had a higher incidence of
marrying to a migrant with a different geographic background (integration). Being
literate and arriving young increased one’s risk of marrying to a native (assimilation).
Migrants who were born in a city had a higher incidence of marrying to a native. We
conclude that many migrants did not marry at the place of residence and that making a
distinction among marriage types increases the insight in the acculturation processes.
Key words: marriage, acculturation, urban, immigration, 19th century, competing‐risk
regression
4
1. Introduction
Marriage is an important institution in society and a core transition in the life course.
Marriage has far‐reaching consequences for individuals (emotional, financial, legal,
etc.) and society at large (social, demographic, economic, cultural, ethic, religious, etc.).
As marriage reveals a lot about how society is organized, it is not surprising that
sociologists and historians use partner choice and characteristics of marriage partners
as indicators of various types of socio‐cultural processes. Social heterogamy, for
example, is used as an indicator of societal openness (Kalmijn 1998; Van de Putte et al.
2005; Van Leeuwen & Maas 2005; Zijdeman 2010). Age homogamy is treated as a proxy
of egalitarian relationships (Atkinson & Glass 1985; Bozon 1991; Van de Putte, et al
2009) Religious heterogamy functions as a measure of pillarization and the power of
the church over the people (Beekink et al 1998). Geographic heterogamy points
towards the decreasing importance of local contexts and mentality changes which
result from technological innovation and improved infrastructure in local communities
(Ekamper et al. 2011). Finally a transition towards ethnic heterogamy shows that
minorities are accepted and that integration into mainstream society takes place
(Schrover 2004).
In this article, acculturation processes are studied of various groups of migrants who
settled in the latter half of the nineteenth or the early twentieth century in the city of
Antwerp. An attempt is made to explain these acculturation processes on the basis of
the characteristics of the migrants. Research on mixed marriages regarding ethnicity
and race originates from countries with a long immigration tradition, notably the
United States of America. The aim of those studies is usually to investigate to which
extent migrants are assimilated (Alba & Gordon 1986; Kennedy 1994; Pagini & Morgan
1990).
Contrary to recent waves of immigrants and the trans‐Atlantic movement, nineteenth
and early twentieth century migration within and towards Western Europe is not part
of the collective memory (Lucassen 2005b). That is why historians, sociologists and
historical demographers only recently – inspired by the contemporary debate on
integration – have become interested in acculturation processes of migrants who
settled in Western Europe before WWII. Migration is both in Western Europe and
elsewhere in the world, a phenomenon of all times, and like today acculturation in the
past was not always an easy‐going process (Bade et al 2011; Hoerder 2002).
Most historical studies on the structural adaptation process of migrants focus on mixed
marriages between natives and migrants. Such marriages are often viewed as the
ultimate assimilation test, the last successful step in a series of adaptations which proof
that a minority has become part of mainstream society (Gordon 1964; Alba & Nee 2003;
Lucassen 2005a). Mixed marriages show that intensive and intimate relationships
between natives and migrants exist. That is because social and cultural differences
between both social groups have grown smaller and form no longer barriers for the
collective social life. According to classic assimilation theorists, like Warner & Srole
(1945) and Gordon (1964) barriers between natives and migrants grow smaller, because
migrants manage to adapt to the native population. In this classic assimilation
approach adaptation is perceived as a unilateral and linear process and it is assumed
that migrants eventually will resemble more and more to natives (Gans 1973; Alba &
5
Nee 2003). These studies depart from the idea that the (behavior of) the native
population is not at all influenced by the arrival of migrants.
In recent migration and acculturation studies the classic assimilation paradigm has
been criticized in several ways. Some of the critics have put forward that this
perception of assimilation presupposes that the native population is superior to
newcomers. In the eyes of Warner & Srole (1945), for example, migrants had to unlearn
their former way of life and they had to adapt to the lifestyle of the native population.
In this way assimilation is perceived as a natural process with a single possible
outcome: adaptation of newcomers to the receiving society. This angle does in the long
run not leave any room for the existence of minorities and beforehand it is ruled out
that the behavior of the native population might change as a consequence of
immigration (Alba & Nee 2003). The adherents of the segmented assimilation theory
have heavily criticized this perception of assimilation. They rather believe that
assimilation is a segmented process, which means that the process of adaptation results
in different outcomes for various groups of migrants (Barth & Noel 1972; Portes &
Zhou 1993). The idea that assimilation is not a one‐sided process and can lead to
different results, is today acknowledged by most scholars in the field. However,
contrary to most assimilation theorists, we conclude that it is not only interesting to
investigate mixed marriages between natives and migrants, but to look also at
marriages among migrants and to analyze who in the migrant population stays single.
We think, that not only mixed marriages can learn us something about acculturation.
Following, de Graaf & Kalmijn (2003) we assume that a certain level of social
integration is a necessary condition for mating and marrying. That is why, in our view,
all migrants who settle as singles in a society can be included in an analysis of
acculturation processes. If migrants manage to set up a long‐term relationship with no
matter what kind of partner in the receiving society, it shows that a certain level of
social acceptation has been reached. However, at the same time we think that it is
telling with whom a migrant married. We put forward that migrants who find their
marriage partner from within the same group undergo different acculturation process
compared to those migrants who selected their partner outside the own group.
In this article, we first of all examine whether migrants were isolated at the Antwerp
marriage market during the latter half of the nineteenth and the early twentieth
century. In a next step, we analyze ‐ if they married ‐ with whom migrants married and
which personal features were important in escaping marginalization through marriage.
We distinguish four possible outcomes regarding acculturation which we link to
results concerning partner choice and marriage: marginalization (staying single),
segregation (endogamous marriage with a migrant with the same geographic roots),
integration (exogamous marriage with a migrant from a different area of origin) and
assimilation (marriage with a native). This typology is based on Berry’s (1997)
acculturation boxes. Berry departs from the idea that two basic preferences among
migrants – ( 1) whether they desire to have intensive contacts with people outside
their own group and (2) whether they strive to maintain their own culture and identity
– result in four different acculturation strategies. Migrants can choose to give up their
own culture and identity and to mingle intensively with natives. This strategy is
known as assimilation. If, however, migrants wish to preserve their own cultural
identity and avoid interaction with natives and other migrant groups, they strive for
6
separation. Integration takes place if migrants wish to keep their own culture, but do
have intensive contacts with members of other social groups. Finally, marginalization
occurs if migrants do not wish to preserve their own culture, but at the same time no
interaction with other groups takes place. We try to avoid the exclusive agency
approach inherent to Berry’s scheme ‐ in this scheme acculturation is solely viewed as
the outcome of the purposeful behavior of migrants – by adding structural elements
like social class and literacy to the analysis. In that sense we rather view Berry’s
typology as a continuum of acculturation processes in which group boundaries become
wider in one direction, while they grow narrower in the other direction.
Eventually we want to know which individual characteristics help to release migrants
out of a marginalized situation. We are interested in the following variables: age at
arrival, social class, gender, country of origin, distance, rural‐urban background,
literacy and birth cohort. The data we use for this research is retrieved from the
Antwerp COR* database, a letter sample from the Antwerp population registers of the
period 1846‐1920 (Matthijs & Moreels 2010). The dataset we used for our analysis
consists of the life‐course of all those migrants who arrived as singles in Antwerp. We
carried out a competing risk regression.
2. Acculturation in the nineteenth and twentieth century
Acculturation is not to be mixed up with enculturation. Scholars who study theories of
adaptation treat both process as two orthogonal phenomena (Kim & Omizo 2006),
which means that both are necessarily related to each other, but describe each
separately something else. Enculturation is the process in which subjects obtain their
primary culture on the basis of processes of socialization. Acculturation is the process
of adaptation which starts once somebody crosses a cultural border.
The nineteenth century can be viewed as a century of progress. Industrialization, the
construction of dense networks of rail‐ and tramway connections, the lay‐out of water
pipes and sewers and the gradual coming into being of social security systems,
improved in the long run the standard of living and the health situation of the Western
European population. Extended life‐expectancy was another positive outcome of these
developments. However, the nineteenth century was also an age in which people
worried about processes of transformation. Migration, especially rural‐to‐urban
migration, was such a process which scared contemporary observers. Sociologists like
Ferdinand Tönnies, Georg Simmel and Max Weber all paid attention to massive rural‐
to‐urban migration. They perceived growing cities as places, where contrary to
villages, anonymity, chaos, loneliness and confusion were prevalent (Jackson 2012). It
did not take long before the link between migration and all kind of urban evils was
made. Migrants were, for example, at an increased risk of deviant behavior like
criminality, alcohol abuse, prostitution, out of wedlock fertility, infanticide and suicide
(Bouman & Bouman 1955; Chevalier 1984; Moch 2003; Winter 2009). Moreover,
migrants clustered in poor and overcrowded city quarters (Lawton & Lee 2002). The
adherents of the Chicago School of Sociology thought that the migrants’ lack of a social
network was one of the main sources of the evil. Many urban in‐migrants had grown
up in the countryside and their movement to the city had a disrupting effect upon
them. The marginal man was born (Park 1928).
7
The image of the acculturation process of nineteenth and early twentieth century
migrants has been re‐adjusted during the last decades. Problems of adaption are now
believed to have been smaller among long‐distance migrants and among stayers
(Sewell 1985; Lucassen 2006c; Moch 2003). This has to do with processes of selection.
Long‐distance migrants and stayers had more human capital at their disposal. Thanks
to their schooling and work experience, these migrants performed well at the labor
market. Lucassen (2004) reached, for example, the conclusion that German stayers in
Rotterdam reached even higher social positions than the average native inhabitant of
that Dutch port city. That study and comparable research (Sewell 1985; Jackson 1997)
have made necessary adaptations. However, this more positive picture of acculturation
can also be put into question. Above all, stayers were a small minority among the
waves of migrants. Five out of six migrants left Rotterdam within a year and other port
cities experienced equally high turnovers of migrants (Lawton & Lee 2002; Lesger,
Lucassen & Schrover 2002). Like stayers, long distance migrants were also a minority.
In the nineteenth and early twentieth century the majority of newcomers in the city
originated from the hinterland (Moch 2003). This implies that the more rosy picture of
acculturation concerns only a small minority of all migrants.
3. Theoretical framework: partner choice and marriage among migrants
At this moment few studies exist in which the acculturation process of different types
of migrants during the nineteenth century is analyzed from a life‐course perspective
(Wingens et al. 2011). This has to do with a shortage of longitudinal sources on the
micro‐level. First of all, only a few countries disposed of passive registration systems in
the form of population registers. Belgium was the first country in which the keeping
up‐to‐date of a nation‐wide population register became compulsory by the
government. During the nineteenth century only Italy (Kertzer & Hogan 1985) and the
Netherlands introduced a comparable nation‐wide system (Van Baelen 2007).
Elsewhere in Europe, some longitudinal registration systems for smaller geographic
entities were established, like the Roteman’s registration system in Stockholm
(Geschwind & Fogelvik 2002). Migration research elsewhere in Europe is based on less
appropriate sources, like censuses and the vital registration of births, marriages and
death. The problem of the cross‐sectional sources is, that only at a restricted number of
moments during the life‐course information of migrants is obtained. As a result, many
migrants, especially temporary migrants, cannot be taken into the analysis, simply
because they did not leave any traces in the cross‐sectional sources. They arrived after
a census was taken and left before a new‐one was carried out. Their risk of ending up
in the vital registration was equally considerably smaller.
Many historical studies show that migrants used to marry later and less than natives
(Lee 1999; Lynch 1999; Kok 2006; Moreels & Matthijs 2011; Oris 2000). This is often
ascribed to the problems migrants encountered upon arrival in an alien environment.
Language problems and the absence of a social network could hinder migrants from
finding a suitable marriage partner (Van Poppel 1992). Furthermore, one had to
dispose of sufficient financial means in order to set up an independent household after
the wedding. It took couples‐to‐be often years to save enough money and to find
appropriate accommodation. Again, the situation for migrants was more complicated.
Moreover, a considerable proportion of the migrants arrived at an advanced age in the
8
city, which meant that they had already a time‐lag. Recently, Moreels and Matthijs
(2011) showed that due to heavy urban in‐migration, the mean age at first marriage
increased in Antwerp in the course of the nineteenth century. This happened at a time
when elsewhere in Flanders mean ages at first marriage were in decline (gradual
disappearance of the Western European Pattern).
Next to marriage timing, partner choice is used as an indicator of acculturation
(Gordon 1964; Blau 1977; Lieberson & Waters 1988; Hwang et al. 1997). In classic
migration studies, the principle question regarding partner choice was to which degree
migrants marry natives (Gordon 1964; Blau 1977; Hooghiemstra 2001; Hwang et al.
1997; Schrover 2005; Van de Putte 2003). Mixed marriages were considered a result and
a further stimulus for the acculturation process (Schrover 2005). Mixed marriages were
seen as proof that migrants were assimilated to a certain degree. After all, high
percentages of exogamous marriages assume that important intimate relations exist
between migrants and natives and that both groups perceive each other as social
equals (Kalmijn 1998). Marrying a native stimulates the acculturation process of the
migrant even further as this, amongst other things, leads to an extension of the social
network of natives. Moreover, it is likely that a marriage with a native improves the
language skills of the migrant and the children of the couple.
Although a considerable proportion of immigrants in European cities married, the
number of marriages between migrants and natives was limited (Rutten 2005; Schrover
2002; Van de Putte 2003). At the beginning of the nineteenth century, for example, only
21% of German women in Amsterdam married to a native husband from Amsterdam.
Amsterdam‐born men highly preferred native Amsterdam women (De Vries 2007). It
seems that mixed marriages were rather exceptional, because natives and migrants
tried to evade each other in daily life. Both social groups competed with each other on
the labor and housing markets, which caused hostility. Moreover, most of the
migrants had a rural background and city dwellers were prejudiced against rural
dwellers and vice versa. Furthermore, it was not uncommon that migrants practiced
another religion and marrying someone with a different faith was taboo (Van de Putte
2003). Language and other cultural differences could likewise act as obstacles for
natives and migrants to enter intimate relationships. Finally, in the nineteenth century
most people married within their own social class. Differences in social class between
migrants and natives could therefore also have given rise to a segregated marriage
market. Industrialization and modernization did not terminate geographical
homogamy on the marriage market. However, the degree to which there was
homogamy differed among social groups. In the Flemish cities of Ghent, Leuven and
Aalst rural migrants and migrants from the lower social classes, had lower chances of
marrying with a native bride compared to urban migrants and migrants from the
higher social classes (Van de Putte 2005).
In recent migration studies, the exclusive focus on intermarriage between migrants and
natives has been contested, because, amongst other things, such an approach assumes
a unilateral adaptation process. The classic assimilation paradigm is founded on the
assumption that all groups of migrants will start to behave more and more like the
native population (Gans 1973; Alba & Nee 1997). According to the adherents of the
segmented assimilation theory, the process of adaptation is rather divided into
9
segments, and leads for different groups of migrants to diverging results. Moreover,
the mere fact that a migrant settles as a single and marries in the receiving society can
be used as an indicator of acculturation process. We are therefore of the opinion that it
is informative to evaluate to which degree single migrants upon arrival in a nineteenth
century Western European city married and ‐ if they did so ‐ to study different
marriage types on the basis of their partner choice.
Berry (1997) has developed a useful scheme to study these different outcomes. He
discerns four types of acculturation strategies, namely assimilation, integration,
separation and marginalization which follow from two fundamental preferences
among migrants: (1) whether they aspire cultural maintenance or not, and (2) whether
they are willing to mingle with other cultural groups, especially natives (figure 1).
Migrants can give up their own cultural identity and mingle with natives. This strategy
is known as assimilation. If migrants, by contrast, wish to preserve their own culture,
and avoid interaction with other groups, we can talk about separation. Integration occurs
if there is an interest in cultural maintenance, but exchange with other groups takes
place. Finally marginalization, occurs when migrants do not strive for cultural
maintenance, but at the same time do not mingle with other groups.
Figure 1. Acculturation Attitudes of Immigrant Groups (Berry, 1997).
Source: Berry, J. (1997). ‘Immigration, Acculturation and Adaption’, Applied Psychology: An
International Review 46(1), p.10
We believe that partner and marriage choices reflect to a certain degree these
acculturation strategies. We adapted the acculturation scheme to fit our purposes.
Whereas Berry gives in his scheme an active role to the migrants by discerning four
acculturation strategies, we consider these four dimensions as acculturation processes.
By doing so, we avoid this agency‐approach inherent to Berry’s scheme. This way, we
try to overcome recent criticism on the universalist approach of cross‐cultural
psychologists. Most empirical studies in the field treat acculturation strategies as
individual preferences that guide people’s actions and ignore that immigrants and
their offspring are part of social structures that are sources of both opportunities and
constraints. This approach denies the influence of the historical context and hence fails
to explain differences in the process of acculturation (Ali 2008; Ngo 2008; Ward 2008).
By considering different marriage outcomes as a reflection of acculturation processes,
10
both active and passive components regarding the human behavior are taken into
account.
We assume that the different marriage types reflect the most important social ties
migrants have in society. We treat the marriage types as ideal types. Together they
form a continuum in which group boundaries in one direction grow wider, whereas in
the other direction they grow narrower. It can be argued that migrants who stay
(against their will) single over longer periods of time, experience marginalization. Little
interest in the own cultural background or unattractive characteristics at the marriage
market may keep these migrants away from members of their own group, while they
do not manage to set up intensive relations with natives and other groups of migrants,
most likely because of exclusion and discrimination. Migrants who marry migrants
with the same geographic and cultural background go through a process of separation.
They have their most important and intimate relations with people within their own
group and maintain their cultural heritage. This can be both the result of a strategy or a
consequence of less favorable treats on the marriage market. Migrants who marry to
migrants with another cultural and geographic background experience integration.
These people do have intensive contact with people outside their own group, but
cultural differences or less favorable personal features might hinder them from
marrying natives. At the same time, the fact that they have intensive contacts with
people from other groups, decreases their likelihood of marrying another migrant with
the same cultural background. Finally, we interpret that migrants who marry natives
go through a process of assimilation. After all, the fact that they marry someone outside
their own group makes it very assumable that they are open to mingle with natives
and that cultural maintenance is not their highest aim. Furthermore, the migrants
probably have characteristics that make them attractive prospective partners, since
natives are also willing to interact with them, even when it comes to the most intimate
aspects of life.
4. Research Questions and Hypotheses
In this paper, we aim to study the acculturation process of migrants by looking at
partner choice outcomes from a life course perspective. To boot, we consider factors
that can facilitate or hamper these outcomes. To this end, we make a distinction
between several forms of acculturation, based on Berry’s acculturation scheme (1997)
which we linked to the different marriage outcomes. In other words, we do not only
look at exogamous marriages in which migrants marry locals (assimilation), but we are
also interested in migrants that marry endogamous with migrants that have the same
geographical and cultural background (separation), as well as in migrants that marry
exogamously with migrants from a different origin (integration). Censored individuals
are those that remain single (marginalization). From this point of view, we expect not
only that the process of acculturation is segmented among different migrant groups,
but that marriage in itself (irrespective of the origin of the partner) is seen as evidence
of social acceptance and a prerequisite to meeting a partner (Kalmijn & Graaf 2003).
Based on these ideas, we formulated the following main research questions. (1) To
which degree migrants managed to marry? (2) Which individual characteristics
increased the risk of a marriage and decreased the risk of marginalization? (3) Had
certain groups of migrants a higher incidence of getting married and were there
11
difference in risks among the different marriage types? We checked whether social
class, age at migration, gender, country of origin, distance to place of origin, rural‐
urban differences, literacy and period played a role. We wonder which effect these
covariates had on the incidence of an endogamous marriage vs. staying single in the
presence of the exogamous marriages (with a native or with another migrant) as
competing events and vice versa. How these covariates are associated with the
independent variables is discussed in the paragraph that deals with the data.
5. Historical Context
During the nineteenth century, Antwerp transformed from a medium‐sized textile
center into a world port. In a time when Belgium’s industrialization gained
momentum, Antwerp’s textile industry vanished completely. This was the direct result
of a lack of investments, through which Antwerp technically started to lag behind
other textile cities, like Ghent (Lis 1986). However, the fall of the textile industry did
not cause a long‐term economic crisis, since port‐related jobs were growing swiftly.
Under the ideals of the French revolution, Antwerp’s port, which had been kept closed
for more than two centuries by the Northern Netherlands, was re‐opened at the end of
the eighteenth century. In the following decades Antwerp developed into the most
important port‐city of the Southern Netherlands and under Belgian rule a competitive
rivalry with the less than a hundred kilometer north situated Dutch port city of
Rotterdam started. The expansion of Antwerp’s port was impressive. Already in 1840,
less than half a century after its re‐opening, Antwerp had become the twelfth largest
port in the world measured by incoming cargo (Greefs 2008).
Table 1. Migration in Antwerp, 1806‐1890 (annual averages).
Source: Anne Winter (2009). Migrants and Urban Change: Newcomers to Antwerp, 1760‐1860.
London: Picekering & Chatto, p.60
Economic and demographic expansion went hand in hand. Due to declining mortality,
Western Europe experienced during the nineteenth century the largest population
expansion in its history. In Antwerp natural population increase was reinforced by
strong urban in‐migration. Especially in the latter part of the nineteenth century, urban
Period Immigration Net‐Migration Gross Mobility
N ‰ N ‰ N ‰
1806‐1815 2,160 38.6 ‐77 ‐1.4 4,397 78.6
1816‐1820 1,614 30 445 8.3 2,783 51.7
1821‐1829 2,285 36.1 1,296 20.5 3,274 51.8
1830‐1846 1,830 23.7 254 3.3 3,406 44.2
1847‐1856 4,445 48.6 737 7.7 8,153 85.5
1857‐1866 6,300 57.9 1,505 13.4 11,096 98.5
1866‐1880 8,480 60.2 2,076 14.1 14,885 101.4
1880‐1890 12,841 66 2,408 12.1 23,273 117.4
12
in‐migration increased massively. Whereas on average between 1830 and 1846 – the
period after the collapse of the textile industry – yearly some 1,830 migrants arrived in
Antwerp, this figure rose to 12,841 in the period 1880‐1890 (see table 1). Rising urban
in‐migration was on the one hand a result of crops failure and agricultural crises in the
Flemish countryside, on the other it was a reaction to the growing demand for laborers
in the city on the river Scheldt (Winter 2009). As a result of these developments,
Antwerp became in the course of the nineteenth century the largest city of Belgium. Its
total population grew from 55,925 inhabitants in 1800 to 272,831 in 1900. According to
Jaap Kruithof (1964), between 57% and 72% of the population growth can be ascribed
to urban in‐migration.
Antwerp’s 19th century urban in‐migration can be characterized by the following
features. First, most urban in‐migrants were short distance migrants who originated
from the city’s direct hinterland: the province of Antwerp. Foreign migrants originated
mostly from the Netherlands and Germany (De Munck, Greefs & Winter 2010). Not
only the number of urban in‐migrants grew in the course of the nineteenth century,
but also the area of recruitment was extended. Between 1796 and 1855 the average
distance between Antwerp and the migrant’s place of birth increased from 61 to 131
kilometers (Winter 2009 107). Striking is also the fact that short and long distance
migrants settled separately. Migrants made up a considerable part of the population of
Antwerp and its suburbs. However, only a minority of the migrants were of foreign
descent (about 10%). It is remarkable that the proportion of international migrants in
the city of Antwerp was considerable higher that in the city’s suburbs, while the
proportion of internal migrants was lower in the city than in its suburbs. This makes it
assumable that international migrants had a profile which differed completely from
that of the internal migrants. Whereas the majority of internal migrants originated from
the countryside, international migrants were probably mostly recruited from the urban
environment. Their preference for the city of Antwerp above semi‐rural municipalities
like Hoboken and Deurne strengthens this idea. Moreover, this settlement pattern
presumes, that international migrants were wealthier than internal migrants. After all,
in the city of Antwerp housing was many times more expansive than in the suburbs.
This settlement pattern is in line with earlier findings. Lis (1986) demonstrated for
example that the majority of rural migrants in Antwerp lived in misery, while Greefs
(2008), on the other hand, showed that the re‐opening of the port went hand in hand
with the settlement of a new wealthy trade elite, largely of foreign descent.
13
Table 2. Origin of Antwerp’s Population in 1900 According to Birth Municipality (%)
% born in the same
municipality
% born in another
Belgian municipality
% born abroad
Antwerpen 56.7 33.1 10.2
Hoboken 51.3 45.9 2.8
Wilryck 59.7 39.2 1.1
Berchem 37.9 57.1 5.0
Borgerhout 45.1 51.1 3.7
Deurne 50.5 47.4 2.1
Merxem 47.8 47.2 5.0
Eeckeren 62.1 34.6 3.3
Source: Récensement Général du 31 décembre 1900
Another important feature of the migration pattern of Antwerp is the high total
mobility of migrants. Similar to other European and American cities, the largest part of
the Antwerp in‐migrants left the city within three months (Darroch 1981; Moch 2003;
Thernstrom 1973; Winter 2009). There are several reasons for this. First of all, Antwerp
was a place where many migrants from Central Europe took a ship to America. These
migrants had not the intention to stay any longer in the city on the Scheldt. The same
goes for most of the life cycle servants and apprentices. They only settled for a limited
amount of time in Antwerp in order to save money and acquire skills. Subsequently,
they moved back to where they came from or tried their luck elsewhere. A
considerable part of the agricultural migrants also returned after having saved some
money in the industry or the port. Upon return these peasants invested their savings in
agriculture. They stayed on the countryside until they encountered financial problems
once again (Klep 1981; Hochstadt 2002). High mobility reflected also trouble on the
urban labor market. Only a minority of the labor force was from January till December
employed in the port (Winter 2009). Unemployment is therefore another reason why
many urban in‐migrants left Antwerp after a few weeks or months of settlement.
Problems at the housing market were still another reason for the high mobility. Due to
high natural population growth and high rates of urban in‐migration, the pressure on
the housing market increased in the course of the nineteenth century. The fact that
Antwerp was a military stronghold, caused also trouble on the housing market. For
that reason the city of Antwerp could not be extended outside its ramparts until 1860.
As a consequence population pressure reached a maximum during the 1850s.
Most of the migrants arrived early in their life (60% of the males and 62% of the
females arrived between age 16 and 30) and most of them were single (Winter 2009).
Many of them were life cycle servants and apprentices, who wanted to learn a trade.
Antwerp attracted both males and females. Males dominated long‐distance migration,
while females were over‐represented among the short‐distance migration from
Antwerp’s hinterland (De Munck, Greefs & Winter 2010).
14
6. Data
To analyze the research questions we have put forward, we used data from the
Antwerp COR*‐database (Van Baelen 2007; Matthijs & Moreels 2010). The data has
been collected by employees of the research group Family and Population Studies
(FaPOS) at the Centre for Sociological Research of the KU Leuven between 2003 and
2010. The COR*‐database contains a representative letter‐sample of the population
living in the district of Antwerp between 1846 and 1920 and is based on the population
registers and the vital registration of births, marriages and deaths. Geographically, the
database encompasses both the city of Antwerp, as well as the surrounding semi‐rural
and rural municipalities of the entire Antwerp district. Personal, demographic and
social‐economic details were gathered from those population registers and vital
registration of births, marriages and deaths for individuals whose last name starts with
the letter‐combination COR*. In addition to this, the same information has been
gathered for live‐in family members with a different surname (partners, uncles, aunts,
nephews, nieces, etc.). After the cleaning and linking of the dataset, the dataset
comprises of 33,583 life courses, of which 17.2% (N=5777) are completed. In other
words, we have the ability to follow roughly 6000 individuals from cradle to grave.
The high proportions of incomplete life courses is due to migration and the large
amount of people that were still alive at the opening of the new population register in
1920. No data has been collected after that date for practical reasons concerning
privacy legislation. The choice for a letter sample based on individuals whose name
starts with COR* is well‐considered. For more information about the database, choices
made and motives, see Matthijs & Moreels (2010).
15
Figure 2. Antwerp District
Source: Matthijs K. & Moreels S. (2010). The Antwerp Cor* Database: A Unique Flemish Source
for Historical Demographic Data. The History of the Family p. 111.
The dataset we have used in order to answer our research questions, contains both
internal and international migrants that lived in the city of Antwerp and its suburbs
Hoboken, Wilryck, Berchem, Borgerhout, Deurne, Merksemii and Eeckeren. Migrants
are those individuals that are not born in Antwerp or the suburbs mentioned before,
but have settled in the area between 1846 and 1920. Of all these migrants, we selected
those migrants that were single at the time of arrival in Antwerp. Subsequently,
following events and accompanying details about these migrants have been sorted out
of the sources: birth, immigration, first marriage, emigration and death. Next, we
extracted the dates where the migrant was respectively 16 and 50 years old. Those
dates were added as events, on the condition that the migrants reached these ages and
were still living in Antwerp at the time.
Using this information, we created a person‐period file (PPF) for a total of 1912
individuals, of whom 904 are men and 939 are women. A person‐period file
reproduces the life course of an individual, distinguishing between the well‐known
phases in the life course and starts and ends with an event. In this case we differentiate
between 16th birthday, age of arrival in Antwerp, marriage, 50th birthday, emigration
16
and death. When the migrant got married during his stay in Antwerp, the same
information was added for the partner. We intended to exclude migrants that married
within one year after arrival, because those individuals probably met their partner
before moving to Antwerp and would not be reflecting any acculturation processes in
Antwerp and would lead to possible biased results. However, none on the migrants
married within one year after arrival. From the migrants that established themselves in
Antwerp, 671 (35.1%) arrived as a child (i.e. younger than 16). In all probability, these
youngsters accompanied their parents. A group of 922 (48.2%) migrants moved to
Antwerp between the ages of 16 and 30, whilst 222 (11.6%) migrated to Antwerp after
their 30s. For the remaining 32 migrants (7.1%) the exact age at migration is unknown.
392 migrants (20.5%) are foreign and 1455 (76.1%) are born on Belgian territory.
Among the international migrants, men are in the majority (224 men, 57.1% vs. 167
women, 42.6%). For the national migrants, on the other hand, women are more well‐
presented (680 men, 46,7% vs. 772 women, 49,6%). In other words, both men and
women are well represented. This is a striking observation since most research on the
Antwerp case observes a skewed gender ratio (Lis 1986; Moreels & Matthijs 2011;
Winter 2009). Those studies put forwards that the skewed gender ratio is the result of
an enormous demand for longshoremen, which need to be physically strong with a lot
of stamina. Our data does not support this line of reasoning. For each year of analysis,
we observe a similar number of male and female migrants arriving in Antwerp. The
discrepancy in findings might be due to the fact that we study a very specific group of
migrants, namely only those migrants that were singe upon arrival in Antwerp.
Table 3. Example Person‐Period File
ID Event 1 Date
Event 1
Event 2 Date
Event 2
Marriage
Type
Covariate
1
Covariate
2
1 16th
birthday
14‐09‐1899 Arrival in
Antwerp
05‐06‐
1905
…. …
1 Arrival in
Antwerp
05‐06‐1905 Marriage 07‐08‐
1907
Exogamo
us
marriage
(local)
… …
2 Arrival in
Antwerp
12‐05‐1870 16th
birthday
22‐10‐
1871
2 16th
birthday
22‐10‐1871 Emigratio
n
07‐02‐
1872
Single
The selected information was used to create following variables that are used as
covariates in our analysis.
Age at arrival: the age at which migrants arrive at their new place of residence
influences the chances of getting married. In this regard, assimilation theory
states that the degree of assimilation is higher for migrants who have migrated
at an early age (Gordon 1964, Hwang, Saenz & Aguirre 1997). In terms of effects,
we expect a negative effect of age of arrival on both types of exogamous marriages, at
17
which the effect is larger for those migrants that married a native than for migrants that
married another migrant from a different geographical background. For the endogamous
marriages we expect no effect. The variable was constructed as a categorical
variable with three categories: migration < 16 years old, migration between 16
and 30 years old and migration > 30 years old.
Social class: a higher social class is associated to higher chances of getting
married (Elder 1969; Kalmijn 1994; Mare 1991), because a higher social position
goes hand in hand with bargaining power. Partner choice is in this case
determined economically and fits with the idea that the marriage market is
characterized by competition for scarce resources. Moreover, the occupational
structure of the higher class occupation enables migrants to have more contact
with the local population, which could lead to an increase in intermarriage
between migrants and locals. We hypothesize for the higher social classes to have a
higher incidence of marrying exogamous with a local than the lower classes. For the
other types of outcome, we expect no significant class differences. The social class
variable, based on occupational titles, was transformed from standardized
HISCO‐codes (Van Leeuwen et al., 2002) into a meaningful class‐scheme, the
SOCPO‐classification, which is based on the concept of ‘social power’ (Van de
Putte & Miles 2005). Due to the small sample size, this class scheme was
dichotomized into a variable that distinguishes between the lower classes
(unschooled, semi‐schooled and schooled workers) and the higher classes
(middle class and elite).
Gender: meeting opportunities are considered an important structural factor in
the search for a possible spouse (Kalmijn 1998). It is possible that men have a
higher possibility to engage in any kind of marriage because they have,
compared to women, less isolated occupations. Women’s occupations were
often confined to occupations such as household servants. These isolated
occupations, in which there was few opportunity to interact with others, could
make it difficult for women to build up a social network and find a potential
spouse. There are indications that this was especially the case by the end of the
nineteenth century, due to the centralization and specialization of factory labor
and with the rise of the male breadwinner model (Van Hautte 2002). Regarding
gender, we hypothesize that male migrants were more likely to get married, especially
exogenously.
Origin migrant (distance and country): as stated before, long‐distance
migration in the second part of the nineteenth century was characterized by a
higher social class since they had to travel further to find a suitable job
(Lucassen 2005b). Following the reasoning that long‐distance migrants have
more human capital at their disposal due to selection‐effects, we assume long‐
distance migrants to have a higher probability to marry in general and that the effect is
stronger regarding both types of exogamous marriages. To test the hypothesis, we
constructed a variable which measures long‐distance migration. According to
Moch (2003) a distance of 20 kilometer can be considered short‐distance
migration. Migrations exceeding those 20 kilometers, on the other hand, were
considered long‐distance migration.
18
Furthermore, we distinguish between national migrants (originating from
Belgium) and international migrants (originating from other countries than
Belgium, especially the Netherlands, Germany and France). These international
migrants miss country‐specific human capital. International migrants may have
a harder time communicating with the local population, due to language
barriers. We hypothesize that foreign migrants were less likely to marry, and that the
effect was stronger for marrying exogamous, especially with a local.
By controlling for both country of origin and the distance of migration, we
should be able to get a more refined image. Moreover, by using this approach,
we also take into account the overrepresentation of long‐distance migrants,
which could possible lead to an overestimation of the effect of the country of
origin (Moch 2003).
Rural‐urban differences: following the ideas of the adherents of the Chicago
School of Sociology, reiterated by Bouman & Bouman (1955) for Rotterdam, we
will check whether migrants who came from the countryside to the Antwerp
area had lower odds of marrying with a local. A lack of a social network is
considered the largest cause. However, other forces, such as prejudices, can be
at work as well. Effect‐wise, we expect urban migrants to have a higher chances of
marrying exogamous with a local. We expect no difference in the effect between
endogamous and exogamous marriages between a migrant with another migrant from a
different geographical background.
Literacy: rising levels of literacy and geographical mobility seem to have been
interrelated in the past (Heffernan 1989). We include literacy because it can be
seen as an indicator of standard‐of‐living, i.e. as an example of the possibility to
invest in long‐term determinants of well‐being (Nicholas & Nicholas 1992).
Besides this, literacy is also related to schooling and points to human capital.
Moreover, literacy might also be a necessary condition of entry or to acceptance
within a certain social milieu (Bonneil & Rosental 1999). We assume that literacy
is a favorable trait on the marriage market. Therefore, we expect literate migrants to
have higher odds of getting married in general. We expect the largest effect in the group
that marries exogamous with a local, followed by the other type of exogamous marriage.
Period: the period under investigation is 1846 until 1922. Antwerp was
characterized by steady growth, both with regard to population growth and
immigration, as well as with the expansion of the harbor. A few important
breaking points during that period are the fall of ramparts in 1860 and WWI
(1914‐1918). Given the fact that Antwerp was a military stronghold in the
period before 1960, building outside the ramparts was prohibited. The fall of
the ramparts is therefore important because it decreased the pressure and
competition on the housing market (Lis 1977), whereby negative feelings
towards migrants possibly declined. We expect the migrants who are living in
Antwerp after the fall of the ramparts to be more at risk at marrying exogamous with a
local resident. Because we only have dates of birth at our disposal, we created
birth cohorts based on the tipping points when the subjects of research reached
marital age (16 years old). The variable that was constructed consisted
originally out of three categories (1803‐1844; 1845‐1903; 1904‐1922). Because this
19
led to too small number within the categories, we dichotomized the variable in
a period before 1845 and a period after 1845.
7. Competing Risks Event History Analysis
To test the hypotheses, we employ event history (or survival) analysis, which refers to
a broad range of statistical techniques that examine both the patterns and correlates of
the occurrence of events (Yamaguchi 1991). More specifically, we use competing risks
regression because our research questions presuppose a situation where the individual
is exposed to more than one event, in this case multiple marriage types.
For each event type a type‐specific hazard function can be constructed, incorporating
time and effects of explanatory variables (Allison 1986). A type‐specific hazard is then
defined as ‘the instantaneous risk of failure from a specified cause given that no failure
from any cause has yet occurred’ (Cleves et al. 2010: 366). For T equal to the time to
first failure from any cause, the type‐specific hazard for cause i at time t is (Cleves et al.
2010: 366):
When stating that an event is ‘competitive’, we refer to the fact that only one event can
occur first. The estimation of our postulated model via well‐known and popular Cox
regression (Cox 1972) is not sufficient in the presence of competing events because the
cause‐specific Cox models considers the competing risk we are interested in as
censored observations. In the case of dependent competing risks, the result is that the
cause‐specific hazard function offers no direct interpretation in terms of survival
probability (Kim 2007). An adaption of the Cox models is possible, but is very time
consuming and demands a specially adapted interpretation. Fine and Gray (1999) have
developed a semi‐parametric method that estimates cumulative incidence functions,
based on the proportional hazards models for cause non‐specific models (i.e. multiple
events). The cumulative incidence function (CIF) is defined as the cumulative
probability that the event of interest has occurred in the presence of alternative events.
The use of these functions explicitly enables us to study a competing risks situation,
but is interpretatively closely related to the hazard and survivor functions (= 1 –
probability of event). The use of this method of analysis allows us to study the different
paths of acculturation and the factors that contribute in a way that is both easy to
interpret and that permits the inclusion of various covariates in the analyses (Cleves et
al. 2010).
In this research paper, we estimate three different models, using the Fine and Gray
method for estimating cumulative incidence functions (Fine & Gray 1999). Model A is a
competing risks regression in which the dependent variable is the incidence of getting
married endogenously in a given year vs. staying single. Model B is a competing risks
regression in which the dependent variable is the incidence of getting married
exogenously with a migrant from a different geographical background in a given year
vs. staying single. Model C is a competing risks regression in which the dependent
variable is the incidence of getting married exogenously with a local in a given year vs.
t
tTicausefromfailurettTtP
tht
i
,
lim)(0
20
staying single. For all three dependent variables, the other types of partnering are
treated as competing risks (Allison 1982).
Figure 3 shows the conceptual model that we use in order to answer our research
questions. In a survival analysis, time or duration until event‐occurrence is referred to
as survival time, an episode, interval or risk period (Mills 2011). We will use the term
risk period, which is defined in years. The beginning of the risk period is the moment
when everybody in the population is still at risk to experience the event and occupy the
same initial state (Singer & Willett 2003). In this research project, the observation time
starts at age 16. In other words, we assume that migrants will not marry before the age
of 16. The end of the risk period is defined either by the moment when the event occurs
(i.e. age at marriage – either endogamous or exogamous with a migrant or a local) or
by the moment the individual is censored for cases that do not experience the event
during the observation period (Yamaguchi 1991). We assume that migrants who are
still single at the age of 50, will remain single for the remainder of their days
(permanent celibacy). Migrants who died before or emigrated before age 50, were also
censored. This type of censoring is called right‐censoring and is quite common
(Yamaguchi 1991; Mills 2011). This type of censoring is no disadvantage, since event
history analysis can adequately deal with it, which is not the case for some standard
statistical methods such as linear regression or logistic regression (Yamaguchi 1991).
Figure 3. Conceptual Model
21
8. Results
Table 4 shows the descriptive results of the explanatory variables for each type of
migrant marriages. A first striking observation based on these descriptive results is
that the bulk of the migrants remain single throughout their stay, namely 80.8%. Of the
migrants that married (N = 367), 6.8% married a migrant with the same geographical
background, whilst 35.4% of the migrants married another migrant ‐ from a different
place of origin ‐ and 24.3% of the migrants married a local. For the remaining 33.5% of
the migrants, we know they married, but have no information regarding their spouse.
Reasons for this high percentage of unmarried migrants are the high mobility of
migrants and the fact that many migrants married in the place of origin. Which
consequences this finding has for the interpretation of our results will be discussed
later on in this section. At first glance, this finding points to large barriers for migrants
to marry and is in line with earlier research (Lucassen 2005a). The non‐parametric
descriptive Nelson‐Aalen estimates (see figure 4) confirm this first impression.
Furthermore, figure 4 also gives us an idea about the timing of marriage, for each of the
subgroups. We chose Nelson‐Aalen estimates because it estimates the cumulative
hazard, which suits the study of marriages intuitively better than the survivor function.
Moreover, Nelson‐Aalen estimates are proven superior when estimating the
cumulative hazard function in small samples (Cleves et al. 2010). All these subgroups
have a relatively low risk of getting married. The cumulative hazard of marriage was
the highest for migrants marrying exogamous with another migrant from another birth
village. From our perspective, these migrants could be considered as integrated. These
people have contact with other groups, but cultural differences and lack of contact
might hinder them from marrying natives. These migrants also seem to get married
somewhat earlier and later than those that marry endogamous or with a local,
although the differences in timing remain small.
Table 4. Descriptive statistics of the variables
Single
Endogamous
Marriage
Exogamous
Marriage (migrant)
Exogamous
Marriage (local) Other Total
N Valid % N Valid % N Valid % N Valid % N Valid % N Valid %
Age at arrival
<16 years old 597 39,3 2 6,7 17 18,3 30 40,5 25 21,2 671 37,0
16‐30 years old 728 48,0 9 75,0 64 68,8 39 52,7 82 69,5 922 50,8
>30 years old 193 12,7 1 8,3 12 12,9 5 6,8 11 9,3 222 12,2
Total 1518 12 93 74 118 1815
Missing 28 2 2 0 1 33
Social class
Lower classes 633 73,9 9 30 55 82,1 37 77,1 65 77,4 799 75
Middle class + elite 223 26,1 1 10 12 17,9 11 22,9 19 22,6 266 25
Total 856 10 67 48 84 1065
Missing 690 4 28 26 35 783
Gender
Male 762 49,4 7 50 45 47,4 38 51,4 52 43,7 904 49,1
Female 779 50,6 7 50 50 52,6 36 48,6 67 56,3 939 50,9
Total 1541 14 95 74 119 1843
Missing 5 0 0 0 0 5
Literacy
Illiterate 1445 93,5 9 64,3 42 44,2 28 37,8 111 93,3 1635 88,5
Literate 100 6,5 5 35,7 53 55,8 46 62,2 8 6,7 212 11,5
Total 1545 14 95 74 119 1847
Missing 1 0 0 0 0 1
22
Country of origin
International 358 23,2 0 0 8 8,4 8 10,8 18 15,1 392 21,2
National 1187 76,8 14 100 87 91,6 66 89,2 101 84,9 1455 78,8
Total 1545 14 95 74 119 1847
Missing 1 0 0 0 0 1
Distance
Short distance migration 326 21,1 6 42,9 31 32,6 27 36,5 35 24,9 425 23
Long distance migration 1220 78,9 8 57,1 64 67,4 47 63,5 84 70,6 1423 77
Total 1546 14 95 74 119 1848
Missing 0 0 0 0 0 0
Rural‐urban differences
Rural 876 57,2 7 50 70 73,7 42 56,8 79 66,4 1074 58,6
Urban 656 42,8 7 50 25 26,3 32 43,2 40 33,6 760 41,4
Total 1532 14 95 74 119 1834
Missing 14 0 0 0 0 14
Birth cohort
1801‐1845 173 11,2 0 0 6 6,3 5 6,8 4 3,4 188 10,2
1846‐1922 1372 88,8 14 100 89 93,7 69 93,2 115 96,6 1659 89,8
Total 1545 14 95 74 119 1847
Missing 1 0 0 0 0 1
23
24
Figure 4. Nelson‐ Aalen Cumulative Hazard Estimates for the Competing Events
0.0
02
.00
4.0
06
.00
8.0
01
0.00
0 10 20 30 40 50Age
Endogamous marriage
Exogamous marriage - other migrantExogamous marriage - local
No information
Cululative Hazard According to Marriage Type
In reality, however, we are probably under‐estimating the prevalence and over‐
estimating the time at risk of migrants ever marrying in Antwerp. Two reasons are
found for this possible bias. An explanation for the under‐estimation of the prevalence
of the marriages lies in the fact that many migrants married elsewhere. An explanation
for the over‐estimation of the time at risk is to be found in the under‐registration of
out‐migration.
Marriages were almost exclusively registered in the vital registration of the
municipality where a marriage was contracted, because population registers were
based upon the de jure principle. The Register Office, on the other hand, was founded
on the de facto principle (Gutmann & Van de Walle 1978). For individuals that married
in a different municipality than the one they lived, this gives rise to a registration
problem. It is probable to assume that this is the reason for the low number of
marriages in the group of endogamous marriages. It is probable that these migrants
did not marry in Antwerp, but in the place of origin. In order to improve our dataset,
we could try to retrieve those marriages that were contracted elsewhere. Guttman &
Van de Walle (1978) made a successful attempt to do so, by supplementing marriages
from the population register of La Hulpe with the help of the so‐called bound register
of that municipality. Every person who intended to marry had to make note of his
intention to do so at the city hall in the municipality where he or she was domiciled.
All the bans, which intended to prevent clandestine marriages, were bundled in the
bound register. By comparing the marriages of the population registers with the entries
in the bound registers of Antwerp, we can possibly retrieve some marriages that were
contracted elsewhere, but never entered the population register of the larger Antwerp
area. Even though this under‐registration could be the cause of serious bias, especially
with regard to the endogamous marriages, we think this problem should in the light of
25
our research questions not be over‐estimated. 42,6% of the migrants officially registers
emigration out of Antwerp before they are 50. From that percentage, 80,8% emigrates
out of Antwerp before they are 25. For these migrants it is very likely that they married
somewhere else. Seeing that we are only interested in acculturation processes within
Antwerp, we are only interested in single migrants marrying in Antwerp. Therefore,
we believe this type of bias does not hamper the validity of our research findings.
A second limitation results from the under‐registration of out‐migration which leads to
an over‐estimation of the time the individual is at risk (Alter 1988; Gutmann & Van de
Walle 1978; Alter, Devos & Kvetko 2009). Reporting moves in and out of the
municipality was compulsory by law, but like today not all migrations were registered,
as people intentionally or unintentionally neglected to inform the authorities about
their move. There are several reasons to assume that nineteenth century migration was
more under‐registered than births, marriages and deaths. First of all, contrary to births,
marriages and deaths, the registration of migration had only recently become
compulsory. Secondly, migration was not coupled with any kind of rites of passage
(Gutmann & Van de Walle 1978). Consequently, some people might not have been
aware that they had to follow a legal procedure, which required them to sign out at the
municipality of origin and to sign in at the municipality of arrival. If they neglected
this procedure, they were not entered in the population register of the municipality
where they moved to and not crossed out in the population register of last residenceiii.
Other migrants might have consciously not registered their move initially, as they
wanted to avoid losing (the right of) any kind of benefits (e.g. poor relief) from the
municipality where they had lived before.
In the case of unregistered out‐migration, the time at risk of getting married is over‐
estimated. This type of bias is problematic because both the CIF’s, the coefficients and
the significance levels are influences. We tackled this problem by comparing
information from the previous population register with information from a following
population register based on a new census. After all, the population register was based
on the de jure population from the census updated with events related to migration and
the vital registration of births, marriages and deaths. If a migrant moved out of the
municipality in population register t (with or without having reported so), we will not
retrieve him or her in register t+1 (unless he or she returned). Consequently, the person
must have moved out between the last known moment of presence and the opening of
the new register. The exact moment when the person actually left, stays unknown, but
the range in which the migrants stopped being at risk reduces considerably. This way
we avoid over‐estimating the time at risk of this specific group of migrants. We created
two subsamples, where we censored migrants that where not present in the following
census either at time t or at time t+1. Since no significant differences in the estimation
of our models were found between the two subsamples, we found it unnecessary to
turn to imputation techniques. The results in the paper thus reflect the most
conservative approach, where we based the censoring upon t+1.
In the final part of the analysis, we turn to the multivariate analysis of the data. By
using competing risks models, where the CIF’s are calculated, we are able to study the
marital behavior of different types of marriages (endogamous, exogamous with a
migrant and exogamous with a local) vs. staying single in the presence of the
26
alternative events as competing events. The interpretation of the CIF’s is equivalent to
that of the survivor function (= 1 – probability of event). The analysis provides us with
subhazard ratio’s (SHR) or the exponentiated regression coefficients. The SHR is the
ratio of the hazards associated with the CIF (Cleves et al. 2010).
A required assumption of both Cox regression, as well as competing risks regression, is
the proportionality assumption, which presumes that the (sub)hazard function for each
group of a time‐contact predictor is proportional to the baseline (sub)hazard function
(Singer & Willet 2003). A test of the proportionality assumption will be performed for
each covariate using the time‐dependent covariate option. If the subhazard ratio
changes over time for any given covariate, then this covariate is said to interact with
time, so adding a X*time interaction term, making them time‐varying covariates (TVC),
provides a better fitting model and fixes the problem of non‐proportionality (Cleves, et
al. 2010). When these time‐varying covariates are not significant and thus do not
violate the proportionality assumption, they are dropped from the model for the sake
of a more parsimonious model. However, we did not turn to a more parsimonious
model that only includes predictors that are statistically significant at the p<0,05 or
even stricter criteria, because the potential for residual confounding in such models is
significant (Vittinghof et al. 2005).
Table 5 shows the results for the competing risks regression. In model A, we
investigate how the characteristics of the migrant influence the incidence of marrying
endogamous versus staying single, which we define as a migrant marrying a migrant
from the same background (birth place), which we consider as indicative of a
separation process. Age at arrival, country of origin, birth cohort and social class are
taken out of the analysis as a consequence of a lack of variation. All the migrants who
married endogenously are between 15 and 30 years old, originated from within
Belgium and were born after 1846 and all but three are situated into the lower social
strata. Hence, analysis with these covariates was impossible. The small N (=19 failures)
might pose a problem for overfitting, which can lead to poor predictive performance of
the model, as it can exaggerate fluctuations in the data. However, competing risk
regression is able to deal with small sample sizes. Only four variables are added to our
analysis, which is borderline acceptable with the rule of thumb for model inclusion
(N/5). For this model no violations against the proportionality assumption were
ascertained, so no time‐varying covariates with interaction terms are added to the
model. The results indicate that for migrants that married individuals from their own
village, literacy and distance to origin seem to be of importance. In line with our
expectations, literacy is positively related to the incidence of an endogamous marriage.
The incidence is almost 3 times higher for individuals that are literate in comparison
with individuals that are not literate. Long distance migration is negatively related to
endogamous marriages. The incidence of marrying endogamous vs. staying single is
70% higher for short‐distance migrants, which is not that surprising given the fact that
contact between short‐distance migrants and the home town was a lot easier.
27
Table 5. Subhazard Ratios for the Competing Risks of Getting Married:
Endogamous, Exogamous(migrant), Exogamous(local).
A. Endogamous
Marriage
B. Exogamous
Marriage (migrant)
C. Exogamous
Marriage (local)
Covariates SHR S.E. Sig. SHR S.E. Sig. SHR S.E. Sig.
Age at arrival (ref: < 16
yo)
16‐30 years old 0.017 0.031 * 0.007 0.146 **
>30 years old 0.002 0.005 ** 0.001 0.007 ***
Social class (ref: Lower
class)
Middle class + elite 0.783 0.192 1.217 0.388
Gender (ref: Male)
Female 0.515 0.224 0.830 0.183 0.856 0.243
Literacy (ref: Illiterate)
Literate 2.928 1.218 * 4.877 1.100 *** 5.692 1.707 ***
Country of origin (ref: Int.)
National 1.649 0.544 1.183 0.443
Distance (ref: Short dist.)
Long distance migration 0.303 0.122 ** 0.946 0.237 0.818 0.238
Rural‐urban diff. (ref:
Rural)
Urban 1.621 0.661 0.890 0.205 1.672 0.500 +
Birth cohort (ref: 1801‐45)
1846‐1922 3.947 1.547 *** 2.132 1.081
Time‐varying covariates
Age at arrival (ref: < 16
yo)
16‐30 years old 1.174 0.820 ** 1.169 0.088 *
>30 years old 1.224 0.109 ** 1.247 0.103 **
Birth cohort (ref: 1801‐45)
1846‐1922 1.890 0.490 **
N of events 19 93 57
Log pseudolikelihood ‐155.677 ‐525.30 ‐312.69
Wald Chi² ‐ test *** *** ***
Significance Level:+ < 0,1 * < 0,05 ** < 0,01 *** 0,001
Note: time‐varying covariates equation interacted with analysis time (_t)
28
Model B investigates the link between individual characteristics and the incidence of
marrying exogamous with a migrant from a different town vs. staying single. We
consider these marriages to be a result of an integration process. To account for the
violation of the proportionality assumption for age at arrival and birth cohort, time‐
varying covariates that interact with analysis time were added for those variables. In
this model age at arrival seemed to be associated with the incidence of getting married
to a migrant from a different background vs. staying single. Compared to migrants that
moved to Antwerp between the ages of 16 and 30, migrants that moved to the city as a
child, have almost twice the incidence (1/0,017) of getting married to a migrant (vs.
staying single) from a different background. For migrants that arrived at older ages
(<30) the difference in incidence is five‐fold (1/0,002). These results are in line with
reasoning that marriage in itself is evidence of social acceptance (de Graaf & Kalmijn
2003). The longer a migrant has been living in Antwerp, the more likely he is to have
built up a social network. Next to age, plays birth cohort also a role in this model.
Migrants that became at risk to be married after the fall of the ramparts, had a much
higher incidence – almost 4 times higher) of marrying a migrant (vs. staying single)
from a different geographical background than individuals that were at a marital age
before 1846.
Age at arrival, literacy and rural‐urban differences seem to be two important variables
in explaining the incidence of getting married to a local vs. staying single. The results
are shown in model C. We consider migrants that marry a local an outcome that
reflects the assimilation process. Time‐varying covariates were added for the two
categories of age at arrival to account for the non‐proportionality of the variable. We
find the same age effect for migrants marrying other migrants as for migrants
marrying locals, but more pronounced. Migrants that arrive at a younger age have a
85% higher incidence of marrying a local (vs. staying single) than migrants arriving
between the ages of 16 and 30 and a ten‐fold higher incidence in comparison with
migrants that arrived after age 30. This is in line with the assimilation thesis that argues
that assimilation is more likely for those individuals that migrated at an earlier age
and have stayed in Antwerp for a longer time. As expected, the effect of literacy on
marrying a local is greater than for the other marriage outcomes. Individuals that are
literate have an incidence that is 5,5 times higher of marrying a local (vs. staying single)
than for illiterates. This is consistent with the contact theory or exchange theory
approach. Finally, rural‐urban distance is also explanatory in explaining the incidence
of marrying a local (vs staying single). Immigrants with urban background have a 70%
higher incidence than immigrants from a rural background. A possible explanation for
this are the mutual prejudices rural and urban dwellers have against each other.
Model D (not included in the output) models the results for the migrants that marry
but for whom we do not have any information on the spouse. The reason we included
them separately was to take them into account as competing events rather than to treat
them as censored. Since we do not have any information regarding the spouse,
interpretation of these results is meaningless for our purposes.
29
9. Conclusion
The goal of this paper was to study different types of acculturation processes migrants
undergo upon arrival in a city and to explore to what extend these differences can be
explained by individual characteristics. This study adds to the existing body of
research on nineteenth century immigration in three ways. First of all, instead of only
focusing on intermarriage between migrants and locals which reflects the assimilation
paradigm, we see the process of change as segmented. Hence, following the basic
tenets of the segmented assimilation paradigm, we distinguish between four different
partner choice outcomes and investigate how these differences can be explained by
individual characteristics. Second, we made an adaption of the commonly used
acculturation scheme of John Berry (1997) by taking the social and historical realities of
acculturation processes into account and surpassing the agency‐oriented view on
acculturation processes often employed by cross‐cultural psychologists. Thus, we
consider Berry’s typology (1997) more as a continuum, whereby group boundaries
become more narrow in one direction, while they become more open in the other.
Third, we study the different marriage outcomes, which reflect acculturation process as
competing events. This way, we integrate various perspectives on acculturation in one
comprehensive framework.
Antwerp lends itself perfectly to study the research questions we have put forward. On
the one hand because this port city was characterized by fast urbanization, exponential
population growth and a large influx of different groups of migrants throughout the
entire 19th century. On the other hand, we have for this city the COR*‐database at our
disposal, comprising of socio‐demographic information of migrants, which allows us
to study migration history in detail. Analyzing data on single immigrant arriving in
Antwerp for the period 1946‐1920 allows us to draw several conclusions.
A first general conclusion is that only 19,2% of our single migrants married in Antwerp
or one of the surrounding municipalities, which points to marginalization and large
barriers for migrants on the marriage market within the Antwerp context. However,
this conclusion does not need to be exaggerated. One possible explanation for the low
percentage of migrants who marry in general lies in the fact that the majority of
migrants left Antwerp within the year. A second explanation is that is that marriages
were only registered in the vital registration of the municipality where a marriage was
contracted. As a consequence, the chance that a marriage that was conducted
elsewhere was not entered into the population register in Antwerp was considerable.
This would especially be the case among short‐distance migrants who married a
migrant from the same place of origin. Therefore, one should be wary for possible bias
when interpreting the results for the endogamous marriages. Both these matters are
problematic for those who are interested in studying the complete life‐course of
migrants. However, these issues are far less pressing within the framework of this
paper, because we are purely interested in acculturation processes within the Antwerp
context. The problem of under‐registration of the emigration could prove more
problematic, but since the majority of migrants officially emigrated at a young age, we
do not have to exaggerate the issue. Seeing that many migrants emigrated officially
before age 25 and that the majority of migrants left within the year, we believe this
does not have any consequences for the generalizability of our results because these
30
migrants did not undergo any structural acculturation in Antwerp. Furthermore, since
a large proportion of migrants left and probably married somewhere else, this can be
seen as an indication of the marginalized position of the migrant in Antwerp.
We can conclude that age at arrival and literacy are the most important individual
characteristics that can help a migrant to escape marginalization. In regard to age at
migration, it seems that for both types of exogamous marriages ‐ which reflect an
integration and assimilation process ‐ age at migration influences the risk at marriage
(vs. staying single). Migrants that have moved to Antwerp before the age of 16 have
almost doubled their chances of getting married, irrespective of the origin of the
partner, compared to those migrants that migrated between the ages of 16 en 30. This
effect is even more pronounced after the age of 30. In line with our expectation, the
effects are stronger for those migrants who married a local than for the migrants who
married a migrant from a different geographical background. Unfortunately, a lack of
variation does not allow us to test the effect of age at arrival on migrants that married
endogamous.
For literacy it was shown, in line with the expectations, that being literate had an
overall positive effect on the incidence of getting married (vs. staying single). Also in
line with our expectations, the incidence for literate versus non‐literate is strongest for
exogamous marriages in which migrants marry locals. Endogamous marriages, on the
other hand, had the smallest impact of literacy.
Our results also corroborate findings of earlier research on the place of origin of
migrants (Lucassen 2005b; Moch 2003). The origin of migrants seems to be an
important explanatory factor in explaining differences in marriage outcomes. Migrants
that originated less than 20 kilometers from their place of birth had higher incidences
of getting married endogamously (vs. staying single), which is not that surprising
given the fact that it was a lot easier to keep in touch with their home town. Contrary
to what we expected, no positive effect was found for long‐distance migration and
marrying exogenously with a local, which would have pointed to selection‐effects.
Rural‐urban differences only turn out to be relevant for exogamous marriages with a
local, which can point to mutual prejudices urban and rural dwellers have against each
other.
Finally, we tested for the effect of historical context. We found that migrants who are at
a marital age after the fall of the ramparts, had an incidence almost 4 times more of
marrying a migrant from a different geographical background (vs. staying single)
compared to those that were at risk whilst there was a lot of pressure on the housing
market. This could point to the fact that negative feelings between migrants and
natives were diminished due to a declining pressure and competition on the housing
market after the fall of the ramparts. However, this did not lead to a decrease in
negative feelings between migrants and locals. If this was the case, we also would have
found an effect of birth cohort on those exogamous marriages in which a migrant
marries a local.
This article shows that it was not easy for migrants to marry in a modernizing context.
A first question in research on partner selection for migrants should therefore be if
migrants got married at all or if they stayed marginalized. Related to this, the second
question one should pose is which characteristics are important in ‘escaping’
31
marginalization. Besides, research that considers marriage and partner selection as
indicators of acculturation processes, benefit from a differentiation according to
marriage type (endogamous, exogamous with a migrant, exogamous with a local)
because marriage in itself can be seen as prove of social integration (de Graaf &
Kalmijn 2003). The process of adjustment is not linear and, therefore, is characterized
by different outcomes. By looking at marginalization and marriage in this manner, we
are able to shed more light on the complexity of nineteenth century acculturation
processes.
32
End Notes i This research in this article was supported by the Research Foundation Flanders (FWO). The authors acknowledge the helpful comments of Jan Kok and Hilde Bras on earlier versions of this paper. ii Migrants who settled in Merksem are poorly represented. This has to do with the annihilation of the population registers of this municipality. iii These individuals were incorporated into the population registers only upon the opening of a new register. In most cases, however, we do not have the exact date of arrival at our disposal.
33
References
Alba, R. & Golden, R. (1986). ‘Patterns of Interethnic Marriage in the United States’,
Social Forces 65, 203‐223.
Alba, R. & Nee, V. (2003). Remaking the American Mainstream: Assimilation and
Contemporary Immigration. Cambridge: Harvard University Press.
Ali, S. (2008). ‘Understanding Acculturation among Second‐Generation South Asian
Muslims in the United States’, Contributions to Indian Sociology 42, 383‐411.
Allison, P. D. (1984). Event History Analysis: Regression for Longitudinal Event Data.
Beverly Hills: Sage Publications.
Alter, G. (1988). Family and the Female Life Course. Madison: University of Wisconsin
Press.
Alter, G., Devos, I. & Kvetko, A. (2009). ‘Completing Life Histories with Imputed Exit
Dates A Method for Historical Data from Passive Registration Systems’. Population
64(2), 293‐318.
Atkinson, M., & Glass, B. (1985). ‘Marital Age Heterogamy and Homogamy’, 1900‐
1980, Journal of Marriage and Family 47(3), 685‐691.
Bade, K. J. (2003). Migration in European History. Malden: Blackwell.
Bade, K. J., Emmer, P.C., Lucassen, L., & Oltmer J. (eds.) (2011). The Encyclopedia of
European Migration and Minorities. Cambridge: Cambridge University Press.
Barth, E., & Noel, D. (1972). ‘Conceptual Frameworks for the Analysis of Race
Relations: An Evaluation’, Social Forces 50(3), 333‐348.
Berry, J. (1997). ‘Immigration, Acculturation, and Adaption’, Applied Psychology: An
International Review 46(1), 5‐68.
Blau, P. (1997). Inequality and Heterogeneity. New York: The Free Press.
Boneuil, N. & Rosental, P. (1999). ‘Changing Social Mobility in Nineteenth‐century
France’, Historical Methods 32, 53‐74.
Bouman, P., & Bouman, W. (1955). De groei van de grote Werkstad. Een studie over de
bevolking van Rotterdam. Assen: Van Gorcum.
Bozon, M. (1991). ‘Les femmes et l’écart d’âge entre conjoints. Une domination
consentie’, Population 2, 327‐360.
Bozon, M. (1991). ‘Les femmes et l’écart d’âge entre conjoints. Une domination
consentie’, Population 2, 565‐602.
Chevallier, L. (1984). Classes laborieuses et classes dangereuses à Paris, pendant la première
moitié du XIXe siècle. Paris: Hachette.
Cleves, M., Gould, W., Guttierez, R. G., & Marchenko, Y. V. (2010). An Introduction to
Survival Analysis Using Stata. Second Edition. Texas: Stata Press.
Cox, D. R. (1972). ‘Regression Models and Life‐tables (with Discussion)’, Journal of the
Royal Statistical Society Series B 34, 187‐220.
34
Darroch, G. (1981). ‘Migrants in the Nineteenth Century: Fugitives or Families in
Motion?’, Journal of Family History 6, 257‐277.
De Graaf, P. M. & Kalmijn, M. (2003). ‘Alternative Routes in the Remarriage Market:
Competing‐Risk Analyses of Union Formation after Divorce’, Social Forces 81(4): 1459‐
1498.
De Munck, B., Greefs, H. & Winter, A. (2010). ‘Poorten en papieren. diversiteit en
integratie in historische perspectief’, In: I. Bertels, B. de Mucnck, and H. van Goethem
(eds.). Antwerpen. Biografie van een stad (pp. 211‐243). Antwerpen: Meulenhoff/
Manteau.
De Vries, J. (2007). European Urbanization 1500‐1800. London & New York: Routledge.
Ekamper, P., Van Poppel, F. & Mandemakers, K. (2011). ‘Widening Horizons? The
Geography of the Marriage Market in Nineteenth and Early‐Twentieth Century
Netherlands’, International Studies in Population 9, 115‐160.
Elder, G. H. (1969). ‘Occupational Mobility, Life Patterns, and Personality’, Journal of
Health and Social Behavior 10(4), 308‐323.
Fine, J. P. & Gray, R. J. (1999). ‘A Proportional Hazards Model for the Subdistribution
of a Competing Risk’, Journal of the American Statistical Association 94, 496‐509.
Geschwind, A. & Fogelvik, S. (2000). ‘Sweden: The Stockholm Historical Database’, In:
P. K. Hall, R. McKaa, & G. Thorvaldsen (eds.) Handbook of International Microdata for
Population Research (pp. 207‐231). Minneapolis: IMAG.
Gordon, M. (1964). Assimilation in American Life. New York: Oxford University Press.
Greefs, H. (2008). ‘Exploiting International Webs of Relations: Immigrants and the
Reopening of the Harbour of Antwerp on the Eve of the Nineteenth Century’, In: A.
Jarvis, and R. Lee (eds.). Trade, Migration and Urban Networks in Port Cities. 1640‐1940
(pp. 81‐107). St. John’s: International Maritime Economic History Association.
Gutmann, M.P. & Van de Walle, E. (1978). ‘New Sources for Social and Demographic
History: The Belgian Population Registers’, Social Science History 2(2), 121‐143.
Heffernan, M. J. (1989). ‘Literacy and Geographical Mobility in Nineteenth Century
Provincial France : Some Evidence from the Department of Ille‐et‐Vilaine’, Local
Population Studies 42, 32‐42.
Hoerder, D. (2002). Cultures in Contact. World Migration in the Second Millenium.
Durham: Duke university press.
Hwang, S., Saenz, R. & Aguirre, B. (1997). ‘Structural and Assimilationist Explanations
of Asian American Intermarriage’, Journal of Marriage and Family 59(3), 758‐772.
Jackson, J. (1997). Migration and Urbanization in the Ruhr Valley, 1821‐1914. Boston:
Humanities Press.
Kalmijn, M. (1994). ‘Assortative Mating by Cultural and Economic Occupational
Status’, American Journal of Sociology 100, 422‐452.
35
Kalmijn, M. (1998). ‘Intermarriage and Homogamy: Causes, Patterns, Trends’, Annual
Review of Sociolgy 24, 395‐421.
Kennedy, R. (1944). Single or Triple Melting Pot? Intermarriage Trends in New Haven,
1870‐1940. American Journal of Sociology 49, 331‐339.
Kertzer, D. & Hogan D. (1985). ‘On the Move: Migration in an Italian Community,
1865‐1921’, Social Science History 9(1), 1‐23.
Kim, B. & Omizo, M. (2006). ‘Behavioral Acculturation and Enculturation and
Psychological Functioning among Asian American College Students’, Cultural Diversity
and Ethnic Minority Psychology 12(2), 245‐258.
Kim, H. T. (2007), ‘Cumulative Incidence in Competing Risks Data and Competing
Risks Regression Analysis’, Clinical Cancer Research 13, 559‐565.
Klep, P. (1981). Bevolking en arbeid in transformatie. Een onderzoek in Brabant 1700‐1900.
Nijmegen: Socialistische Uitgeverij Nijmegen.
Kok, J. (2006). ‘Eigen baas zijn da’s maar alles.’ Huwelijksmotieven van Rotterdammers
uit de tweede helft van de negentiende eeuw. In: P. van de Laar, L. Lucassen, and K.
Mandemakers (eds.) Immigratie en levensloop in Rotterdam vanaf het einde van de
negentiende eeuw (pp. 99‐118). Amsterdam: Aksant.
Kruithof, J. (1964). ‘De Demografische ontwikkeling in de XIXde eeuw. In:
Genootschap voor Antwerpse Geschiedenis’ (Ed.). Bouwstoffen voor de geschiedenis van
Antwerpen in de XIXde eeuw (pp. 508‐543). Antwerp: Anversois.
Lawton, R. & Lee, R. (2002). Population and Society in Western European Port Cities c.1650‐
1939. Liverpool: Liverpool University Press.
Lee, R. (1999). ‘Labor Markets, In‐Migration, and Demographic Growth: Bremen, 1815‐
1914’, Journal of Interdisciplinary History 30(3), 437‐473.
Lee, R. & Lawton, R. (2002). ‘Port Development and the Demographic Dynamics of
European Urbanization’, In: R. Lawton, and R. Lee (eds.). Population and Society in
Western European Port‐cities c.1650‐1939 (pp. 1‐36). Liverpool: Liverpool University
Press.
Lesger, C., Lucassen, L. & Schrover, M. (2002). ‘Is there Life Outside the Migrant
Network? German Immigrants in 19th Century Netherlands and the Need for a More
Balanced Migration Typology’, Annales de Démographie Historique 2, 29‐50.
Lieberson, S., & Waters, M. (1998). From Many Strands. Ethnic and Racial Groups in
Contemporary America. New York: Russell Sage Foundation.
Lis, C. (1986). Social Change and the Labouring Poor: Antwerp, 1770‐1860. New Haven:
Yale University Press.
Lucassen, L. (2004). ‘De selectiviteit van blijvers. Een reconstructive van de sociale
positie van Duitse migranten in Rotterdam (1870‐1885)’, Tijdschrift voor sociale en
economische geschiedenis 1(2), 92‐115.
Lucassen, L. (2005a). ‘Huwelijken van Duitse migranten in Nederland (1860‐1940)’,
Tijdschrift voor sociale en economische geschiedenis 2(2), 54‐80.
36
Lucassen, L. (2005b). The Immigrant Threat: the Integration of Old and New Migrants in
Western Europe since 1850. Urbana: University of Illinois Press.
Lucassen, L., Feldman, D., & Oltmer. J. (eds.) (2006). Paths of Integration. Migrants in
Western Europe (1880‐2004). Amsterdam: Amsterdam University Press.
Lucassen, L. (2006a). ‘Poles and Turks in the German Ruhr Area: Similarities and
Differences’, In: L. Lucassen, D. Feldman, and J. Oltmer (2006). Paths of Integration.
Migrants in Western Europe (1880‐2004) (pp.27‐45). Amsterdam: Amsterdam University
Press.
Lucassen, L. (2006b). Gelijkheid en onbehagen. De wortels van het integratiedebat in West‐
Europa. Leiden: Inaugurele rede Universiteit Leiden
Lucassen, L. & Lucassen, J. (2011). Winnaars en verliezers. Een nuchtere balans van
vijfhonderd jaar immigratie. Amsterdam: Bakker.
Mare, R .D. (1991). ‘Five Decades of Educational Assortative Mating’, American
Sociological Review 56(1), 15‐32.
Matthijs, K., & Moreels, S. (2010), ‘The Antwerp COR*‐database: A Unique Flemish
Source for Historical‐demographic Research’, History of the Family 15(1), 109‐115.
Mills, M. (2011). Introducing Survival and Event History Analysis. London: Sage
Publication.
Moch, L. (2003). Moving Europeans. Migration in Western Europe since 1650. Bloomington
& Indianapolis: Indiana University Press.
Moreels, S. & Matthijs, K. (2011).’Marrying in the City in Times of Rapid Urbanization’,
Journal of Family History 36, 72‐91.
Nicholas, S. & Nicholas, J. (1992). ‘Male Literacy, ‘Deskilling’ and the Industrial
Revolution’, Journal of Interdisciplinary History 23, 1‐18.
Ngo, H. (2008). ‘A Critical Examination of Acculturation Theories’, Critical Social Work
9(1), 1‐6.
Oris, M. (2000). ‘The Age at Marriage of Migrants during the Industrial Revolution in
the Region of Liège’, The History of the Family 5(4), 391‐413.
Pagini, D. & Morgan, P. (1990). ‘Intermarriage and Social Distance among U.S.
Immigrants at the Turn of the Century’, American Journal of Sociology 96, 405‐432.
Park, R. (1928). Human Migration and the Marginal Man. Human Migration and the
Marginal Man 33(6), 881‐893.
Rutten, W. (2005). Twee nationaliteiten op één kussen. Huwelijk en partnerkeuze in de
Zuid‐Limburgse grensstreek in de negentiende en twintigste eeuw. In: J. Kok, and M.
H. D. van Leeuwen (eds.). Genegenheid en gelegenheid. Twee eeuwen partnerkeuze en
huwelijk (pp. 159‐180). Amsterdam: Aksant.
Sewell, W. (1985). Structure and Mobility. The Men and Women of Marseille, 1820‐1870.
Cambridge: Cambridge University Press.
37
Singer, J. D. & Willet, J. B. (2003). Applied Longitudinal Data Analysis: Modelling Change
and Event Occurrence. Oxford: Oxford University Press.
Schrover, M. (2002). Een kolonie van Duitsers: groepsvorming onder Duitse immigranten in
Utrecht in de negentiende eeuw. Amsterdam: Aksant.
Schrover, M. (2004). ‘De Duitscher is immers tegenwoordig ook zeer galant voor
dames’. Huwelijksgedrag van migranten in Nederland in de 19e eeuw. Bevolking en
Gezin 33, 103‐125.
Schrover, M. (2005). Huwelijk, gender, migratie en integratie. Partnerkeuze van
Duitsers in Utrecht in de negentiende eeuw. In: Kok, J., and Van Leeuwen. M. H. D.
(eds.). Twee eeuwen partnerkeuze en huwelijk (pp.135‐158). Amsterdam: Aksant.
Thernstrom, S. (1973). The Other Bostians. Poverty and Progress in the American Metropolis,
1880‐1970. Cambridge: Harvard University Press.
Van Baelen, H. (2007). Constructie van een historisch‐demografische longitudinale database.
Methodologie van de Demographica Flandria Selecta. Leuven: Centrum voor sociologisch
onderzoek.
Van de Putte, B. (2003). ‘Homogamy by Geographical Origin: Segregation in
Nineteenth‐Century Flemish Cities (Gent, Leuven and Aalst)’, Journal of Family History
28, 364‐389.
Van de Putte, B. (2005). Partnerkeuze in de 19de eeuw: klasse, geografische afkomst,
romantiek en de vorming van sociale groepen op de huwelijksmarkt. Leuven: Leuven
University Press.
Van de Putte, B. & Miles, A. (2005). ‘A Class Scheme for Historical Occupational Data.
The Analysis of Marital Mobility in Industrial Cities in 19th Century Flanders and
England’, Historical Methods 2, 61‐92.
Vanhaute, E. (2002). ‘Breadwinner Models and Historical Models. Transitions in
Labour Relations and Labour Markets in Belgium’, in: H. Jensen (ed.). The Welfare State.
Past, Present and Future (pp. 59‐77). Pisa: Edizioni Plus.
Van Leeuwen, M., Maas, I., & Miles, A. (2002). HISCO. Historical International Standard
Classification of Occupations. Leuven: Leuven University Press.
Van Leeuwen, M., & Maas, I. (2005). ‘Endogamy and Social Class in History: An
Overview’, International Review of Social History 50, 1‐23.
Van Poppel, F. (1992). Trouwen in Nederland. Een historisch‐demografische studie van de
19de en vroeg‐20e eeuw. Wageningen: Landbouwuniversiteit Wageningen.
Van Poppel, F., Liefbroer, A., Vermunt, J. & Smeenk, W. (2001). ‘Love, Necessity and
Opportunity: Changing Patterns of Marital Age Homogamy in the Netherlands, 1850‐
1993’, Population Studies 55(1), 1‐13.
Vittinghoff, E., Glidden, D. V., Shiboski, S. C. & McCulloch, C. E. (2005). Regression
Methods in Biostatistics: Linear, Locgistic, Survival, and Repeated Measures Models. New
York: Springer.
38
Ward, C. (2008). ‘Thinking outside the Berry Boxes: New Perspectives on Identity,
Acculturation and Intercultural Relations’, International Journal of Intercultural Relations
32(2): 105‐114.
Warner, L. & Srole, L. (1945). The Social Systems of American Ethnic Groups. New Haven:
Yale University Press.
Wingens, M., Windzio, M., De Valk, H., & Aybek, C. (2011). The Sociological Life
Course Approach and Research on Migration and Integration. In: M. Wingens, M.
Windzio, H. de Valk & C. Aybek (eds.). A Life Course Perspective on Migration and
Integration (pp. 1‐26). Dordrecht, Heidelberg, London, New York: Springer.
Winter, A. (2009). Migrants and Urban Change: Newcomers to Antwerp, 1760‐1860.
London: Pickering & Chatto.
Yamaguchi, K. (1991). Event History Analysis. Newbury Park: Sage Publications.
Zijdeman, R. (2010). Status Attainment in the Netherlands, 1811‐1941: Spatial and Temporal
Variation Before and During Industrialization. Ede: Ponsen & Looijn.