Youth labour migration and its family setting, the Netherlands 1850-1940
The province of Utrecht
Data and methods
The extent of youth migration
Family background and family strategies behind the first migration
Youth mobility and individual career
Acknowledgements & Notes
The rapid succession of jobs is a recurrent theme in the memories of adolescence of working men and women, born in the nineteenth century. The shift from one workplace to another often meant moving to a different community and sometimes even to another area (cf. Maynes 1995, Weber 1985, Burnett 1974, Giele 1979). High youth mobility, both in the occupational and the geographical sense, was an important feature of pre-industrial society. The meaning of this phenomenon can be specified on the levels of society, family and individual.
Youth mobility had an essential role in the functioning of either agrarian production and the pre-industrial demographic régime. The labour exchange between agrarian households was directly geared to the cycles of individual families. For instance, a shortcoming of labour in a family with young children could be met with a surplus of adolescent children in another household (Kussmaul 1981b, p. 26, Lundh 1995, pp. 57-62). Adolescent workers moved regularly, because they were generally hired for one year only. The yearly labour contracts of servants reflected the fact that on non-mechanized farms many tasks were age-specific and therefore required a rapid succession of farm-hands. By living and working in other households, adolescents were able to save and wait for the moment they could start their own household (Schlumbohm 1996, p. 84). Postponement of marriage until economic independence was reached was the quintessential feature of the reproductive system of the European pre-industrial society. On the level of the family, attention is momentarily being focussed on the more or less conscious strategies concerning the moving or staying of the adolescent children. Whereas one child might be kept at home, to take care of the parents in old age, others will be sent away. Their moves may take on many forms. Some will be sent away simply to spare the parents an extra mouth to feed. Some will return regularly, because they had joined the army of seasonal workers. Their earnings will be pooled at home. Others will live more or less permanently outside the parental home, but will still send the parents part of their income. The experiences of present-day migrant workers testify to the strength of these familial bonds (Baud 1994). Finally, children are directed to relatives in other places, who may be in need of their help or who can take care of them. The functioning of family networks in this respect has been amply described by Hareven in her study of French-Canadian textile workers in the United States (Hareven 1982). Obviously, many youths will leave on their own accord and a number of them will have severed the economic ties with their parents. Some demographers even speak persistently of 'emancipation' when discussing adolescent mobility (Blanchet and Kessler 1992, p. 354). On the level of the individual, youth migration has been recognized as an important life course transition (Moch 1992, p. 34). Moving youths ended the life stage of direct dependency on the parents and their experiences in other households had an important bearing on their later life course. They acquired skills and contacts which for many of them may have been essential for starting their own workshop or farm. Furthermore, in their new surroundings they came into contact with potential spouses. Nor surprisingly, marriage markets coincided geographically with the labour markets for adolescent workers.
Because the traditional sources of historical demography (e.g. censuses, vital registers) are extremely reticent on (youth) migration, early studies on this subject were by necessity limited to particular villages or small areas over a relatively short time-range (Anderson 1971, Schofield 1970, Wall 1978, Langholm 1975). The theme was explored more fully in countries like Sweden, Belgium and Italy where the sources permit a reconstruction of individual migration paths (Martinius 1977, Eriksson and Rogers 1978, Alter 1988, Gribaudi 1987, Kertzer and Hogan 1989). Fortunately, in several countries large-scale projects are presently combining data from different sources into large databases with family reconstructions and individual life histories. The study of youth migration is receiving a strong impetus from this work as well (Bouchard 1991, Bideau, Foroni and Brunet 1995, Rosenthal 1990).
In this article an exploration of the phenomenon is endeavoured for a rather large area in The Netherlands over a period of ninety years (1850-1940). The following questions will be adressed. How many persons actually participated in youth migration? How frequent did they move and what distances did they cover? The centre of interest will be the family setting of youth migration. How often did migrating youths return to their parents? How important was the socio-economic position of parents for their childrens' tendency to migrate? I will also analyze the effect of other aspects of the family background, e.g. literacy of the father, the age of the parents, their own migratory experience and their decease. In combining the results, I will try to answer the question: to what extent were youth migrations influenced by family strategies? Throughout the analysis, I will make distinctions by sex, type of birthplace and birth-period. Men and women participated in different labour markets, and these markets differed also between the countryside and the cities. We can therefore expect migration patterns to be differentiated by sex and birthplace. In this period, Dutch society underwent transformations that all had impact on the incidence and extent of youth labour migration. Living-in farmhands were replaced by labourers who had households of their own. In the countryside, the number of domestic women servants dimished as well. Employment in agriculture gave way to the rapidly growing industrial and service sectors. Developments in transport made it possible to stay at the parental home and work elsewhere. In what ways are these changes reflected in the migratory patterns of the youths?
This broad approach of the subject has been made possible by the recent construction of a large database containing the life courses of 3300 persons born between 1812 and 1912 in the province of Utrecht (The Netherlands). In the next section a short description of the area is given. In the third section I elaborate on the possibilities and limitations of the database for the study of (labour) migration. Also, the methods employed in analyzing the data are described. In the fourth and fifth sections I go into the extent of the phenomenon of youth migration and describe the patterns found. In the sixth section I probe into the family backgrounds in trying to detect family strategies behind mobility patterns. Finally, I look at the impact of adolescent migration experience on the occupational position held at a later stage in the life course. In addition to the database, I use several nineteenth- and early twentieth-century inquiries on labour relations and living conditions in the area.
The province of Utrecht
The province of Utrecht is, with respect to its surface, the smallest of all Dutch provinces. It is centrally located in the mid-western part of the country. According to the census of 1849, the province numbered 149.380 inhabitants. The size of the population more than tripled to 489.608 in 1940. Population growth accelerated particularly after 1875 when mortality dropped much faster than fertility. Until after the Second World War, net immigration did not contribute significantly to the population increase. The province of Utrecht has only two 'real' cities: the capital, also called Utrecht, and Amersfoort. Of these, Utrecht is the most important with, in 1849, 47781 inhabitants. In that year Amersfoort counted 12377 inhabitants. The cities have always been demographically important. In 1800 forty percent of the Utrecht population lived in one of the cities, in 1920 fifty percent. Initially, the two cities did not take a very active part in the Dutch industrialization. From the 1880's onwards, however, Utrecht became an important centre of metal and machine industries. These industries were drawn to the city because it is Holland's junction of most railway lines. In addition, a large number of persons in Utrecht, and in Amersfoort as well, found work in one of the railroad companies (Knippenberg 1995).
In the countryside, agriculture was the most important means of existence. Because of the variety of soil conditions, three of the most important agricultural systems of The Netherlands can be found here. On the peaty soil of the western part of Utrecht, cattle raising and dairy farming was dominant, whereas in the north-western communities bordering on North-Holland peat was cut, mostly by migrant workers. The soil of southern Utrecht consists of sediments of river clay, allowing for mixed farming with grain cultivation. Finally, on the poor sandy soils of eastern Utrecht, mixed farming with a three-field system was common. Only once or twice in every three years rye could be grown on fields which had to be restored with the cultivation of buckwheat.
The first three-quarters of the nineteenth century was a prosperous period for Dutch farmers, particularly because of the growing exports of dairy and meat to England and Germany. These conditions were especially favourable to western Utrecht, where farmers concentrated on cheese making (Gouda). During the nineteenth century the acreage of grazing land was increased, as well as the stock of cattle and pigs. On the other hand, the acreage of farmland diminished strongly. In the areas with mixed farming, dairy farming became of central importance. In the sandy regions in eastern Utrecht, agricultural productivity was very low in the early nineteenth century. A gradual improvement took place when the construction of new roads stimulated the import of dung to fertilize the fields. Production was also enhanced by large-scale reclamations. Reclamation was accompanied by the splitting up of the last remaining commons in the area, part of which were used for commercial forestry. The international crisis in agriculture struck The Netherlands between 1878 and 1895. Eventually, the crisis accelerated the development towards a modern, mechanized agriculture. However, Utrecht farming was not in the forefront of this modernization.
Finally, the service sector was of importance in the rural labour market, and not only on the farms. From the late eighteenth century onwards, the well-to-do built country houses on the range in south eastern Utrecht, as they had done in earlier times along the river Vecht in the northwestern part of the province. These houses offered employment for many servants, especially female. In the cities, the service sector increased strongly in the period 1860-1930. Not only railway companies, but banks, insurance companies and educational institutions as well opened national headquarters in Utrecht, which meant an enormous boost to the commercial service sector of the city. Amersfoort became an important centre of services as well. In 1930 both in Utrecht and Amersfoort 30 percent of the working population was employed in the commercial services (trade and transport).
What were the effects of these structural economic changes on the migration patterns of unmarried youths? Throughout the period studied, rural labour became gradually less important, particularly for women. In Utrecht, domestic service on farms started to decline already in the middle of the nineteenth century. Female wage labour (as day labourers) on farms was of little importance in the area. On the other hand, employment for female servants grew strong in the urban middle classes from the 1860s onwards (Van Zanden 1985, pp. 69-70). These developments probably increased, after the 1860s, the tendency of unmarried rural women to migrate to cities and towns. In the early twentieth century, the attraction of domestic service in the cities gave way to the expanding opportunities in industry and, in particular, the service sector (Henkes and Oosterhof, 1985, p. 14, Van der Woude 1989, p. 173). For young men, the growth of building and trading activities in the larger cities of Holland (Amsterdam, Rotterdam, The Hague), particularly in the period 1870-1885, will have formed an impetus to live and work outside their Utrecht area, at least for a certain period. After this period, employment in the Utrecht cities expanded, which may have reduced mobility, at least for urban boys.
In the first half of the twentieth century, working and living tended increasingly to be separated. In relative terms, the number of young women that entered service declined. In principle, women who worked outside domestic service could remain in their parental homes. Staying at home was stimulated by the enhancement of public (railroads, trams) and private transport. In the first decades of the twentieth century, the bicycle became a very popular means of transport (Knippenberg and De Pater 1988, pp. 47-48). We may hypothesize that these developments decreased the mobility of adolescents in the first half of the twentieth century.
Data and methods
The database on the Utrecht population actually forms the first part of a much larger one, which is named the Historical Sample of The Netherlands. In this project a sample of 0,5% is taken from all the birth certificates in the period 1812-1922. All information on people in the sample that is found in vital registers (certificates of birth, marriage and death) is stored in the computer. In the future, this database will consist of about 70 000 individuals whose family backgrounds, occupations, civil status and residence at birth, marriage and death are known (Mandemakers 1994). Moreover, geographical moves between these vital events can be reconstructed with the use of the Dutch population registers. Obviously, the database will be an invaluable asset for the study of migration. This is not to say that the information on migration is in all respects perfect. The population registers started to function in 1850. However, some municipalities stalled registration for another decade. Temporary moves were underregistrated in the period 1850-1861, because the officials were only recording persons who legally resided within their municipalities. In 1861 the definition of inhabitant was changed to any person actually living in the municipality. Changes of adress within a town or village were recorded, but generally not dated. Therefore, the registers in most places do not allow us to establish exactly the age at leaving home. A serious problem is the underrecording of servant mobility. Registrars often failed to note their residence, because of their extreme mobility and because, since they did not pay taxes, they did not warrant special attention. Finally, even when they had been recorded, it might simply be impossible to trace them. This is because either the indexes are inadequate, or because (parts of) the registers are missing. Particularly in the Second World War several town halls were destroyed by fire, or even raided by resistance fighters aiming to destroy the population registers.
Generally, the registrars meticulously recorded all basic characteristics and vital events pertaining to the members of households. Birthplace and date, sex, religion, relation to the head of the household, muncipalities of provenance and destination are given for every person. Occupations, however, are mostly mentioned only for the heads of the household and for persons entering the household (e.g. servants). It was not obligatory to report a change of occupation. For the young, this means that we are better informed on their work in other households, than in their parental homes.
Migration is defined as the crossing of municipal borders. This definition allows for comparison with other studies, although we have to bear in mind that municipalities in different areas can differ strongly in surface. Also, the number of very small communities has diminished over time, because of the tendency to consolidate them into larger municipalities. In analyzing migration we employ some of the techniques associated with event history analysis (Alter 1988, pp. 25-62). To begin with, the information on all sample persons is converted into observed 'person-years'. Age-specific migration rates can be calculated by taking the number of the events in question (migration) per each age-group as the numerator, and the total amount of observed person-years at risk for each age group as the denominator. In life tables basically the same method is used in calculating per time interval the probability of an event experienced by a hypothetical cohort. When the end of observation is not caused by the event one is analyzing, the data are called right censored. In itself, right censoring is not a problem, provided that the censoring event does not interfere in one way or the other with the dependent variable. In our case, right censoring caused by death or by the closing of a register is not related with migration. When we want to relate the probabilities of the event (also called hazard rates) to certain individual characteristics, we have to move on to multivariate analysis (Mayer and Tuma 1990, Blossfeld 1989). Proportional hazards models measure the effect of independent variables on the 'baseline' hazard of experiencing an event like migration. The assumption is that this baseline hazard is related only to age. The effect of the independent variable is presented in the form of relative risks. For each variable a subgroup functions as a reference group. For instance, we may assume that the risk of youth migration is related to the occupation of the father. For the subgroup 'artisans' we set the risks to 1,00 and we calculate the effect of the other subgroups relative to this reference group (see table 3). The method is called 'proportional', because it is assumed that the relative effect of the independent variables on the risks remains the same at each age. Obviously, the broader the age-range studied, the less tenable this assumption becomes.
The extent of youth migration
In my introduction I boldly stated that unmarried adolescents had the highest propensity for migration of all groups in society. Just how strong this tendency was is illustrated in Figure 1. Here migration rates are given for all persons in our sample, by sex and civil status (for the calculating method, see above). Marriage has been treated as a censoring event and observed episodes of marrying persons have been split in two, one that ends and one that starts with the date of marriage.
Migration rates of youths started to rise from age 12 onwards, when most of them left school. After age 23 migration rates for single men started to drop, for single women this occurred only after age 25. The high migration rates for young married women reflect the fact that women frequently left their hometown within a week or two after marriage. It is interesting to see that single persons, particularly women, remained more mobile in their thirties and forties than married ones. Because of population growth, the birth cohorts after 1870 are overrepresented in this figure. Also, the experiences of urban and rural persons are roughly lumped together. The figure is therefore only useful as a first impression. In our following tables we further differentiate by birth-period and birthplace. Further, we will concentrate on the unmarried adolescents born between 1833 and 1912 and aged from 12 to 26. We cannot use the earlier birth cohorts, since our reconstruction of individual itineraries starts in 1850 with the beginning of continuous registration. The end of the age range corresponds more or less with the mean age at first marriage for the sample population (27,5 for men (N=547) and 26,0 for women (N=539), birth cohorts 1833-1912).
In table 1 we look at the relative number of persons participating in youth migration. Youth migration is defined as any registered move between municipalities. Here, we only consider 'independent' migration, that is a move made without (one of) the parents. The total number of observed persons (N=1533) is lower than the original sample from the birth certificates (N=2718). Persons who have died before age twelve are obviously not included. Also we lack all information on a small number of persons, namely 2.5% for the birth cohorts 1833-1902. We were unable to trace these persons in the population registers, probably because their parents were officially residing somewhere else and their birth in a particular municipality was therefore, to use the official phrase, 'accidental'. For the birth cohort 1903-1912 this percentage is much higher (33.6) because it is more difficult to find information on persons who are still alive.
About 30-55% of all single persons in this age-group migrated at least once independently. These percentages are slightly higher than those found in the French region of Haut-Bugey in the same period (Bideau, Foroni and Brunet 1995, p. 132). Rural girls were the most prone to migrate before their marriage or before their 27th birthday. This finding agrees with results from other studies, in The Netherlands as well as elsewhere (De Vries 1971, p. 104, Moch 1992, p. 39, Bideau, Foroni and Brunet 1995, p. 133). On the other hand, urban girls had the lowest tendency to migrate. Probably, their native city offered them enough employment opportunities. The percentage of migrating rural girls increased rather strongly in the youngest cohort. Contemporary reports testify to the ongoing exodus of rural girls to the cities, leaving the farmers with a strong shortage of female labour (Verslagen 1908, p. 11, Lijst IV approx. 1930, pp. 133, 140). Apparently, improved transportation did not induce them to stay at home. The differences between urban and rural boys are not very large. It is, however, rather surprising that urban boys born between 1833-1872 were more prone to migrate than rural boys.
In this section, I will look more closely at various characteristics of youth migration, such as mean age at first migration and distances covered. However, it is important to note that the individual variation in migratory behaviour was enormous. To illustrate this point, I compare the migration histories of two girls in our sample. Both were born in the small town of Zeist in 1887. Moreover, these girls (Geertruida S. and Geertruida van R.) were daughters of workers who lived in the same quarter and who were themselves not natives of Zeist, and neither were their wives. At this point, however, the similarities end. Geertruida S. left her parents at age fifteen in 1902 for a short spell in the provincial capital. After four months she already returned to her parental home. In 1903 she migrated once again to the city of Utrecht and in 1905 to the village of Driebergen, each move followed by a return to her parents in Zeist. Finally, in 1908 she went to the village of Doorn, where she married a worker. Even though she moved quite frequently, she was never farther than eleven kilometers away from her native village. Geertruida van R., on the other hand, stayed in Zeist with her parents and, after their death, she lived with her brothers and sisters until 1910 when she married a gardener in the same town.
In table 2 I try to capture both the general patterns of youth migration and the amount of individual variations in them. Firstly, I discuss the age at first migration from the native town or village, or, to be more precise, from the municipality where the parents lived at the time. The mean age at which migrants left was rather similar for all groups. The small differences found can be combined with the results of table 1. Apparently, a stronger tendency to migrate generally went parallel with a lower age of the migrants. Rural women, for instance, left frequently and at a relatively early age. However, the similarity of the means may be misleading. In fact, the standard deviations show that there was no set age at which youths migrated for the first time (Wall 1978, p. 191). Life course studies have revealed that in the twentieth century, until about the 1970s, transitions such as starting to work, leaving home, getting married and having a child, tended to take place more and more within narrow age-limits. This is because individuals were more likely to conform to societal expectations on the proper age and the proper sequence of education, work, marriage and parenthood.
In earlier times, the absence of fixed age patterns reflected the fact that individuals conformed more to family expectations regarding their behaviour (Hareven 1982, p. 167, 188, Modell and Hareven, 1978, p. 267, Mulder 1993, p. 123.) Leaving or staying depended on the labour requirements of the parental home and on the person's position in the sibling order. Among farmers, for instance, the inheriting son generally remained at home. Among labourers, the parents counted strongly on the income of the eldest children, thus leaving room for the younger children to be educated (Mitterauer 1992, pp. 91, 98). The choise between leaving and staying was also strongly influenced by demographic events, in particular the death of (one of) the parents (Maynes 1995, pp. 69-70, Mitterauer 1992, p. 94).
In the fourth column the mean number of migrations (without parents) is given. The figures can be used solely as a minimal indication. A number of persons were missing (in an administrative sense) for a couple of years, because their stay in a municipality was not recorded properly. As we discussed in the second section, these lapses occurred particularly in this stage of life, because adolescents were not particularly interesting to the registrars. Throughout the period, rural women appear to have migrated the most and urban women the least frequent, compared with men.
It is likely that occupation is more relevant to migration patterns than either sex, birth place or birth cohort. Unfortunately, occupational titles are often not provided. Only for a small number of persons can we be sure about the nature of their work away from home. Farmhands appear as a rather mobile group (mean number of moves 3,6, s.d.=2, N=30), compared with, for instance young craftsmen (mean 2,3, s.d.=1,5, N=30). Female domestic servants were relatively mobile as well (mean 3,4, s.d.=2,1, N=154). Obviously, the mean number of migrations hides a much larger number of intra-municipal moves from one job to another. The relatively strong mobility of living-in servants (either male of female), who were hired by the year, can be explained by the fact that it was beneficial to both parties. Servants could leave because they were displeased with their master, and because a change of job meant some amount of free-time. More importantly, a new job might entail new contacts, skills, responsibilities and higher wages. Kussmaul has suggested that service on farms can be considered '(...) an informal type of apprenticeship to farming, and servants might have moved to experience a variety of farming practices' (Kussmaul 1981a, p. 226). For female servants, moving on to another mistress was the normal way to get ahead, that is, to acquire a better paid and higher status position (Poelstra 1981, p. 62). Employers, on the other hand, had several motives to send servants away after the contract had expired. Firstly, the longer the servants remained on the farm, the more they expected their wages to grow. But for many tasks, farmers needed cheap labour, more than they needed experienced labour. Secondly, a swift rotation of workers might lower the risk that they would eventually become a burden on the local poor relief (Kussmaul 1981a, p. 226-227). The Dutch Poor Law of 1870 stipulated that the municipality of actual residence was solely responsible for the care of paupers (Cottaar 1996, pp. 138-141). Finally, employers needed to pay less or no personal taxes for young servants. For different age-groups, different tariffs applied and therefore one's age was a reason to be sent away (Drieling 1853, p. 408). The employment of servants younger than 15 was exempt from the personal taxes. In 1843 and again in 1869 the taxes on servants were lowered and these measures appeared to have strongly increased (female) employment in service.
Table 2 also gives the median of distances covered by independently migrating youths. In the entire period, rural girls made the shortest moves. Rural boys moved on average five kilometer farther than girls. We can safely assume that, as in other rural areas in Europe, most migrating boys and girls kept well within a radius of 10-15 kilometers from their parental home (Langholm 1975, p. 50, Schofield 1970, p. 271). Apparently, the change in employment opportunities we described in section two, hardly affected the migratory behaviour of rural youths. Urban youths moved no less than three to four times farther than their rural counterparts. Their destinations were often other cities, like nearby Amsterdam, The Hague and Rotterdam. It is interesting that young urban women, although they migrated somewhat less (and less frequent) than urban boys, took also part in medium and long-distance travels. This finding does not necessarily contradict Moch's statement that '(...) young women rarely searched from city to city for work' (Moch 1992, p. 129). Migrating urban girls probably already had a job and moved because they were assured of a better position elsewhere. Domestic servants responded to ads or used the arbitration of a hiring agency (Henkes and Oosterhof 1985, pp. 41-44). Also, several urban girls worked in various parts of the country as nuns, nurses or teachers. In the twentieth century, urban women (birth cohort 1873-1912) seemed to have moved over shorter distances than in the second half of the nineteenth century. This may be acccounted for by the growth of employment in the Utrecht cities themselves.
Finally, table 2 shows the importance of return migration, defined as moves directed to the parental household as a percentage of all moves. As we have seen in the example of Geertruida S., the home of the parents functioned as an intermediate station between successive employments. It shows that migrating youths were not 'emancipated' but, on the contrary, entertained close contacts with their parents. The tendency to return was generally stronger in the rural areas than in the cities. However, the tendency was increasing among those born between 1873-1912 in the cities, particularly among women. They ventured on a rapidly expanding labour market, where fixed labour contracts became scarce. In this context the possibility to fall back on familial support and housing became even more important (Janssens 1993, pp. 86, 92-93, 96).
Family background and family strategies behind the first migration
In this section I will evaluate the influence of the family setting on the migration of adolescent children. Furthermore, I try to find out to what extent family strategies were involved. The detection of strategies is not a simple task. Since oral history methods cannot be applied, the actual motives of the individuals remain obscure. Only by combining quantitative and qualitative information some insight in the processes of decision making can be gained. In my quantitative analysis, I have in mind Louise Tilly's definition of family strategies: '(...) principles that lie behind predictable, interdependent behaviors in which one outcome is regularly favored over another' (Moch 1987, p. 124). But statistics cannot instantaneously unearth these underlying principles. It goes without saying that the keeping at home or the sending away of Utrecht youths was related to regional property and labour relations, inheritance practices and customs relating to the training of the young. The contemporary reports from the area enable us to reconstruct a range of potential strategies. The behavioral tendencies that result from our statistical exercises, can be linked to the strategies we expect to have been followed by specific socio-economic groups.
A hazard model is used to see which characteristics of the family have an impact on youth migration. The variables uses in the model are the age and eventual decease of the parents, the father's literacy and occupation, and the migratory history of the parents. These variables certainly do not represent the family situation completely. In fact, information is lacked on some of the probably most important factors. The composition of the household, in terms of the number of children or the labour force present, may have been crucial in determining who should go or who should stay (Alter 1988, pp. 154-160, Kurosu 1996). For a better understanding of the socio-economic position of the household, information on income or property would have been most welcome. Notwithstanding these shortcomings, the statistical exercise proved instructing. Table 3 presents separate hazards models for respectively rural men, rural women, urban men and urban women. In this analysis, the dependent variable is the time that passes until an event takes place. In other words, subgroups where migration of youths occurs relatively frequent and at a relatively young age, will stand out with a risk higher than 1.00. The impact of each variable is calculated independent of the other variables. The overall significance of the models is satisfying, particularly for rural women.
Cohort effects were, on the whole, of little importance. However, the risk of migration of rural boys, born in 1873-1892, was twice as high as the risk of boys born in 1843-1852. Most of their migrations will have taken place in the period 1890-1915. In explaining this, the strong growth of urban employment in the province was probably more important than the push-effect of the agrarian depression (1878-1895). Inversely, the migration risks of urban men were lower in each birth cohort.
The municipalities of birth of the parents are compared with the municipality in which our 12-year old sample persons were residing. This allows us to make a distinction between sedentary parents, who both still lived in their native town or village, and parents of whom one or even both have immigrated. The migratory history of the parents had a strong impact on their children, which was significant for rural men and urban girls. The latter even had a two-hundred percent higher risk of leaving the city (on their own) when one or both of her parents had their origins elsewhere. The most likely explanation for this is that these girls disposed of useful contacts in other places. A longitudinal cohort study of a village in the northern part of the Netherlands has shown that young migrants often moved to the birthplace of (one of) their parents (Clement 1993, p. 177). Also, the presence of relatives was an important motive to move to the textile city of Tilburg, in the south of The Netherlands (Janssens 1993, pp. 184-185, 191).
I assume that there was a pressure on persons with elderly parents to stay at home and provide for them. The large-scale government inquiries on the conditions of farmers and agricultural labourers (respectively 1890 and 1908) indicate that it was customary for one of the children to be designated to stay at home. Often, a contract was drafted whereby the child was promised the farm or, in the case of workers the house and the furniture. This was meant as a compensation for the boarding of the parents and the agreement was made with the consent of the siblings. The reports indicate that the designated person was either the eldest son or the last marrying child, often the youngest (Uitkomsten 1890, part II, XLIV Cothen, p. 15 and part IV, pp. 265-268; Algemeen overzicht 1908, pp. 288, 291). This custom was known at the time as 'bringing the parents to their end'. In the table, the age of the parents is given at the time of birth of the sample person. Unfortunately, we do not know the exact rank-number of the individuals. The figures suggests that boys with relatively old parents tended, particularly in the countryside, to stay at home more than other boys (Wall 1978, p. 198). For girls this relation is not visible. It is, however, interesting that urban girls with relatively young parents migrated frequently and early. Apparently their labour was not needed at home.
It may be hypothesized that literate parents had a relative advantage in obtaining information on opportunities for work and housing elsewhere (Boonstra 1993, p. 237). Therefore they might be better able to find for their older children a job, an apprenticeship or a boardinghouse. As predicted, children of illiterate parents had lower risks of migration. The results, however, are not significant.
The decease of one of the parents proves very important in explaining migration hazards. In her synopsis of workers' autobiographies, Maynes suggests that widowed fathers were even less able to keep the family together than widowed mothers (Maynes 1995, p. 69). We find this reflected in our table. Remarriage of the surviving parent may have hastened the departure of their adolescent children. The arrival of stepparents frequently caused tensions in the family, which made children decide to enter into service (Mitterauer 1992, p. 95, Schlumbohm 1996, p. 85). We may conclude that when the (biological) parents were both alive, they were best able to keep their children at home, apparently because they helped in the farm or workshop, brought in wages, or simply provided for the parents. In Alters study of the life courses of the women of Verviers (1846-1880), a similar relation between the death of (one of) the parents and the chances of marriage of their resident daughters is found. In Alters words: 'The important point is that adult daughters acted as if their parents had first claims on their labor and their actions were subject to parent approval' (Alter 1988, p. 161).
Finally, I turn to the variable assumed to be the most important in predicting the migratory behaviour of adolescents: the socio-economic position of the household. Indeed, some of the outcomes are very significant, but, rather unexpectedly, there are no significant results in the group of rural men. In the cities, youths whose fathers were skilled workers were most prone to migrate. These fathers form a broad group consisting of white collar workers of low and middle rank, skilled manual workers and factory workers. They possessed the means (contacts, information and some money) to get their children an interesting job or apprenticeship in another town. Besides, there was not much their sons and daughters could do at home, at least not in the sense of production. The same tendency to migrate is found among sons of merchants, and even stronger among the sons of the elite (ministers, lawyers, high-ranking officers etcetera). The latter frequently sent their boys to a boarding school.
Artisans have been used as the reference category and it therefore not easy to evaluate their relative risks. However, boys from several other occupational groups had higher risks of moving than the children of artisans. Also, the sons of artisans who did migrate, had about the average number of migrations. Clearly, the local movements of Dutch apprentices and young artisans are insignificant when compared with the journeys of their colleagues in Central Europe. Here, artisanal traditions of migrations appeared remarkably resilient during industrialization (Ehmer 1994, pp. 101-129). In fact, in Dutch history there was no equivalent to the German Wanderjahr, the French tour de France or the British tramping (Lucassen 1995, p. 397). An enquiry into the circumstances of the self-employed in the province of Utrecht (1909) indicates that most artisans had worked with and succeeded their fathers. Others had worked for a period of 10 to 15 years with an employer in a neighbouring community, before they started for themselves. In fact, the report attributes the low quality of the crafts to the self-employed' unwillingness to travel and take note of innovations (Middenstandsenquête 1918, pp. 7, 15, 20, 55, 75, 91, 96.)
Since the number of urban 'farmers' (mostly market gardeners) is very small, we will only discuss the countryside. Here the children, in particular the daughters, tended to stay on the parental farm where they were expected to assist the parents. According to the reports, living-in adolescent daughters and sons spared the costs of female servants and workers (Uitkomsten 1890, part II. XLIV Cothen, p. 12,17, LII Alphen aan den Rijn, p. 14 ). However, the urge to keep them at home was not very strong. Farmers' sons frequently worked elsewhere as farmhands until they could, with the help of their fathers, rent a farm for themselves. Their fathers could help them because they purposefully held a large reserve of liquid assets, such as stocks. In 1890 more than half of all homesteads in the area was leased. There are some interesting differences with the Italian tenants studied by Kertzer and Hogan. In Italy, the short lease of one year required the concentration of as many family members as possible in order to maximize production. This led to relatively low migration rates of tenant sons (Kertzer and Hogan 1990, p. 500). In Utrecht, the leases were minimally for a term of six years and they were almost always extended, so that tenants were succeeded by one of their sons. Tenant farmers had the same strategy as owners of trying to place as many children as possible on their own farm (Uitkomsten 1890, part II, XLV Benschop, p. 6, XLIV Cothen, p. 8).
Daughters of unskilled workers in the countryside (mostly agricultural labourers) were very prone to leave home and migrate early. Table 3 shows that their migration risk is more than three times higher than the risk of daughters of farmers. Predominantly, labourers' daughters swelled the ranks of the highly mobile group of young domestic servants. According to the 1890 report, at age 13 '(...) the girls start serving, unless the mother cannot miss them at housekeeping' (Uitkomsten 1890, part II, LII Alphen aan den Rijn, p. 15, Verslagen 1908, part II, p. 133 ). How can we explain that workers' sons had much lower migration hazards than their sisters? Firstly, they could often find employment on the same farm as their fathers (Verslagen 1908, part II 133). Secondly, they could also work on their father's own plot of land. It was very common for agricultural workers to own or rent some land or live stock. Finally, parents could claim the wages of their resident sons and daughters more effectively than of those who lived elsewhere. According to contemporary enquiries, resident children handed over their entire salaries to their parents. A mid-nineteenth-century observer was highly critical of this custom. It was held responsible for the lack of able farmhands in Utrecht: 'The cause of this lies herein, that the workers prefer to keep their children at home, in order to find in their earnings a compensation for the costs of the family and [therefore] they are not properly trained in agricultural work. And yet already after a stay of one year on a farm, children of day labourers show the favourable effect of a better and broader education on their development' (Drieling 1853, p. 407). The report depicts the parents as selfish and short-sighted, depriving the children of any motive to remain at home: '(...) what advantage would [the young worker] at home gain by diligence, indeed none, because he would be dressed and fed anyhow' (Drieling 1853, p. 436). According to the reports, the strategy of farmers and labourers alike was directed towards keeping the children in the vicinity. From the observers' point of view, therefore, leaving home and migration were individual acts of liberation from a suffocating parental control, or, at least, a sign of domestic discord (Drieling 1853, p. 436, Verslagen 1908, p. 133).
In the late-twentieth century, the intergenerational wealth-flows would be completely reversed (De Regt 1993, pp. 52-74). This reversal was a gradual process, beginning in our period. Around 1900, parents at least gave their adolescent children pocket money in return for their salaries. In this period also, the practice of paying board and lodgings to one's parents was spreading, especially in the Northern provinces. (Verslagen 1908, part II, p. 13, 132-133; Algemeen overzicht 1908, pp. 285-291). Obviously, girls working as domestic servants could not stay at home. But they, like migrated boys, sent their parents at least part of their wages. An interesting observation is that 'The farther the children live away from their parents, the smaller is their support' (Algemeen overzicht 1908, p. 291).
The hazards models suggest a rather strong parental influence on the (first) migration of their adolescent children. By combining the behavioral patterns with information from contemporary observers, we can infer two types of family strategies. Firstly, parents who needed labour or assistance at home (e.g. farmers, the elderly and the poor) would refrain in particular their sons from migrating. Daughters from these groups migrated more often, simply because becoming a domestic servant was one of their few options. Secondly, some groups of parents, in particular urban ones with relatively 'better' occupations, stimulated their sons and daughters to migrate, presumably because this was beneficial to their career. The eventual connection between migrations in adolescnece and the career in later stages of the life course will be discussed in the next section.
Youth mobility and individual career
Adolescent youths from the cities may have migrated because they could find formal education, an apprenticeship or a better-paid job in another town. Their parents may have helped them either with money or with contacts. Perhaps they expected that their investment in their children's higher education or career opportunities would be returned in the form of larger financial support. A suggestive example of the above is the life course of Nicolaas de R. Born in 1854 in Utrecht as a carpenter's son, he moved in 1876 (22 years old) to the provincial capital of Den Bosch, in the southern part of The Netherlands. Here he worked as a draughtsman. After a few weeks, however, he headed for Arnhem, the provincial capital of Gelderland. In 1879 he lived and worked, still as a draughtsman, in the city of Hengelo, close to the German border. In 1880 he travelled to the northern city of Groningen, where he got a fine position as inspector of the public railroads. In 1887 he married in this city.
In this final section we will take a more systematic look at the effect of youth migration on one's future career. Did youth migration enhance one's opportunities, compared with non-migrating persons? If so, this may have been the result of long-term planning by the parents. To answer the question whether youth migration stimulated upward intergenerational mobility, we have to look at occupations held in a later stage of the life course. Unfortunately, most occupational groups are too small to include in a table of occupational mobility. In table 4 we only compare the occupations of sons of artisans and unskilled workers who had and who had not migrated during adolescence. I have taken the occupation of the sample persons at first marriage, or, in the case of permanently single men, their occupation most regularly mentioned in the population registers between age 28 and 40. The occupations of these men are compared with the ones of their fathers as stated on the birth certificates.
Although the absolute numbers are still small, we may conclude that, at least for the artisans, youth migration was not conducive to upward mobility. Quite the contrary: throughout the entire period sons of artisans who migrated had higher chances of becoming an unskilled worker than those who did not migrate. Conversely, their chances of pursuing one of the occupations in the 'other' category (higher occupations, merchants, shopkeepers and farmers) were lower. We may combine this finding with the comments on the Utrecht artisans in the previous section. It appears that inheriting sons tended to stay at home, and that non-inheriting, and therefore downwardly mobile, sons left their hometown.
In the first period, migrating sons of unskilled workers became more often an unskilled worker themselves than sons who remained in their home town. Staying in their home town and obtaining a job that involved some skill (preferably in an artisan's workshop) was the best way to escape a life of casual and unskilled labour (Gillis 1981, pp. 127-128). In the later period, however, extension of the service sector created many new jobs, most of them headed under 'skilled workers'. Workers' sons born in 1873-1912 could take advantage of these developments more readily by migrating.
In the period between 1850 and 1940, no less than 30 to 55% of all youths (of the ages 12-26) left their parental homes to live in another municipality. The mean age of first migration was 18-19 years. Generally, migrated adolescents dwelled in several other households, more or less as a member of the family, until they set up a household of their own. For the overwhelming majority of these youths, work was the motive to move. In this respect, women did not differ greatly from men. Migrating youths were employed mostly as farmhands, domestic servants and apprentices. The propensity to migrate was not diminished by the gradual disappearance of living-in personnel in the countryside, the shift towards employment in the industrial and (non-domestic) service sectors, and the improvement of transport facilities.
Leaving home and moving to another community has been rightly recognized as a major step in the individual life course. However, it was by no means an act of individual emancipation. Rural youths rarely travelled more than ten kilometers and they remained therefore well within the reach of parental influence. According to commentaries, daughters were expected to help their mothers with housekeeping. Sons were either trained to succeed their father on his farm or workshop, or they were encouraged to find work near home and hand over all, or most, of their earnings to their parents. Often, one of the children was appointed to 'bring the parents to their end'. In fact, migration hazards for sons with elderly parents were markedly lower than for other sons. To keep children at home, therefore, seems to have been more in the parents' interest than to send them away, at least in most social groups. Therefore, many sons and daughters who migrated, did so because they could not be employed at home. Even when working somewhere else, for instance as a domestic servant, the parents expected them to bring home at least part of their income. They probably mediated actively in finding work and housing for their children. Parental contacts in other areas were a powerful stimulus for adolescent children to move. They often returned to their parents when necessary. The hold of parents on their offspring seemed to have loosened up when one of them died. For the youths themselves migration, at least in the case of sons of artisans and unskilled workers, was not beneficial to their individual career. Clearly, for adolescents from urban and relatively better-off social groups, the above picture is too bleak. They were sent to a another town to enter a boarding school or a university, to become an apprenticeship or trainee, or to live with relatives who helped them to find a (better) job.
For a more profound insight in the complex relationship between individual moves and family interests, we need to know more about the composition of the family, in terms of sibling-order and labour force. Detailed information is required on the economic situation of the family, e.g. its access to land and its income position. Further research on the subject should include moves within communities and the families that received migrating youths.
Youth migration was a prominent phenomenon in Western society, strongly embedded in its demographic and economic structures. However, without a detailed investigation of family background the mechanism and the implications of youth mobility cannot be properly understood.
Earlier versions of this article were presented as papers to the European Social Science History Conference (Noordwijkerhouter, May 9th, 1996) and the conference of the Social Science History Association in New Orleans, October 13th, 1996. I would like to thank Jan Lucassen, Theo Engelen, Daniel Scott Smith and Marco van Leeuwen for their useful comments. Henk Delger has been most helpful with the occupational classifications.
1. Migration distance is defined as the distance between the geographical centre of the municipality of origin with the centre of the municipality of destination.
2. The birth cohort 1833-1842 has not been used in this analysis, because the twelve-year old had to be followed in the continuous records starting in 1850. When one of both of the parents died when the sample persons where aged 12-26, the episode of residence has been split. As an indication for literacy we have used the ability of the father to sign the birth certificate of his child. Occupations have been grouped according to a classification described in: Marco van Leeuwen and Ineke Maas, 'Social Mobility in a Dutch Province 1850-1900', Social History, March 1997. The relative risks are Cox partial likelihood proportional hazards coefficients, calculated in SPSS 6.1 for Windows.
- Algemeen overzicht. 1908. Algemeen overzicht van den oeconomischen toestand der landarbeiders in Nederland.'s-Gravenhage.
- Alter, George. 1988. Family and the female life course. The women of Verviers, Belgium, 1849-1880. Madison: The University of Wisconsin Press.
- Anderson, Michael. 1971. Family structure in nineteenth-century Lancashire. Cambridge: Cambridge University Press.
- Baud, Michiel. 1994. 'Families and migration: Towards an historical analysis of family networks.' Economic and social history in The Netherlands 6: 83-107.
- Bideau, Alain, Fabrice Foroni and Guy Brunet. 1995. 'Migration and the life course: mobility in Haut Bugey (France) during the nineteenth and twentieth centuries'. Journal of Family History 20:127-138.
- Blanchet, D., and D. Kessler. 1992. 'La mobilité géographique de la naissance au mariage'. Pp 343-377 in La société française au XIXe siècle. Tradition, transition, transformations, edited by Jacques Dupâquier and D. Kessler. Paris: Fayard.
- Blossfeld, Hans-Peter, Alfred Hammerle and Karl Ulrich Mayer. 1989. Event history analysis. Statistical theory and application in the social sciences. New Jersey: Lawrence Erlbaum Associates.
- Boonstra, O.W.A. 1993. De waardij van eene vroege opleiding. Een onderzoek naar de implicaties van het alfabetisme op het leven van inwoners van Eindhoven en omliggende gemeenten, 1800-1920. Wageningen: Landbouwuniversiteit.
- Bouchard, Gerard. 1991. 'Mobile populations, stable communities: Saguenay 1840-1911.' Continuity and Change 6: 59-86.
- Burnett, John. 1974. Useful toil: autobiographies of working people from the 1820s to the 1920s. London: Allen Lane.
- Clement, Marcel. 1993. 'Demografisch gedrag, leefsituatie en mobiliteit. Een analyse van vier generaties'. Pp. 160-201 in: Dorp naast een stad. Hoogkerk 1770-1914, edited by Pim Kooij. Assen: Van Gorcum.
- Cottaar, Annemarie. 1996. Kooplui, kermisklanten en andere woonwagenbewoners. Groepsvorming en beleid 1870-1945. Amsterdam: Het Spinhuis.
- Drieling, F.H.C. 1853. 'Verslag over den toestand der arbeidende klasse, vooral ten platten lande, in de provincie Utrecht, en over de middelen tot verbetering van hunne zedelijke en stoffelijke welvaart', Tijdschrift voor Staathuishoudkunde en Statistiek 8:405-452. Reprinted in Jan Luiten van Zanden. 1991. 'Den zedelijken en materiëlen toestand der arbeidende bevolking ten platten lande'. Een reeks rapporten uit 1851. Groningen: Nederlands Agronomisch-Historisch Instituut.
- Eriksson, I., and J. Rogers. 1978. Rural labor and population change. Social and demographic developments in East-central Sweden during the nineteenth-century. Stockholm: Almqvist and Wiksell International.
- Giele, Jacques. 1979. Arbeiderslevens in Nederland 1850-1914. Nijmegen: Socialistiese Uitgeverij Nijmegen.
- Gillis, John R. 1981. Youth and history. Tradition and Change in European Age Relations, 1770-Present. New York: Academic Press.
- Gribaudi, Maurizio.1987. Itinéraires ouvriers. Espaces et groupes sociaux à Turin au début du XXe siècle. Paris: Editions de l'E.H.E.S.S.
- Hareven, Tamara Kern. 1982. Family time and industrial time. The relationship between the family and work in a New England industrial community. Cambridge: Cambridge University Press.
- Henkes, Barbara and Hanneke Oosterhof. 1985. Kaatje, ben je boven? Leven en werken van Nederlandse dienstbodes 1900-1940. Nijmegen: Socialistiese Uitgeverij Nijmegen.
- Janssens, Angélique. 1993. Family and social change. The household as a process in an industrializing community. Cambridge: Cambridge University Press.
- Kertzer, David I., and Dennis P. Hogan. 1989. Family, political and demographic change, the transformation of life in Casalecchio, Italy, 1861-1921. Madison: The University of Wisconsin Press.
- Kertzer, David, and Dennis Hogan. 1990. 'Household organization and migration in nineteenth-century Italy.' Social science history 14: 483-505.
- Knippenberg, Hans en Ben de Pater. 1988. De eenwording van Nederland. Schaalvergroting en integratie sinds 1800. Nijmegen: Socialistiese Uitgeverij Nijmegen.
- Knippenberg, Hans. 1995. 'Het hart van Nederland. De provincie Utrecht in de periode 1800-1940.' Pp. 7-36 in: De levensloop van de Utrechtse bevolking in de 19e eeuw, edited by Kees Mandemakers and Onno Boonstra. Assen: Van Gorcum.
- Kurosu, Satomi. 1996. 'Leaving home in a stem family system: Departures of Heirs and Non-Heirs in Pre-Industrial Japan', History of the Family. An International Quarterly 1: 329-352.
- Kussmaul, Ann. 1981a. 'The ambiguous mobility of farm servants'. Economic history review 34:222-235.
- Kussmaul, Ann. 1981b. Servants in husbandry in early modern England. Cambridge: Cambridge University Press.
- Langholm, Sivert. 1975. 'Short-distance migration, circles and flows: movement to and from Ullensaker according to the population census lists of 1865', Scandinavian Economic History Review 23: 36-62.
- Lucassen, Jan. 1995. 'Labour and early modern economic development.' Pp 367-409 in: A Miracle Mirrored. The Dutch Republic in European Perspective, edited by Karel Davids and Jan Lucassen. Cambridge: Cambridge University Press.
- Lundh, Christer. 1995. 'Households and families in pre-industrial Sweden.' Continuity and change 10:33-68.
- Lijst IV. Approximately 1930. Lijst IV. Werktijden, vrouwen- en kinderarbeid, dienstverband, arbeidsovereenkomsten, huisvesting, inwonend personeel.
- Mandemakers, Kees. 1994. 'Historical sample of the Population of the Netherlands (HSN)'. Backgrounds, Objectives and International Context'. Pp. 174-181 in: Yesterday. Proceedings from the 6th International Conference. Association of History and Computing, edited by Hans Joergen Marker and Kirsten Pagh. Odense: Odense University Press.
- Martinius, Sture. 1977. Peasant destinies: the history of 552 Swedes born 1812-12. Stockholm: Almqvist and Wiksell International.
- Mayer, Karl-Ulrich, and N.B. Tuma, eds. 1990. Event history analysis and life course research. Madison: University of Wisconsin Press.
- Maynes, Mary Jo. 1995. Taking the hard road. Life course in French and German workers' autobiographies in the era of industrialization. Chapel Hill and London: The University of North Carolina Press.
- Middenstandsenquête. 1918. Verslagen van het onderzoek naar den toestand van den handeldrijvenden en industrieelen middenstand in de provincie Utrecht. Uitgegeven door de Commissie voor de middenstandsenquête ingesteld bij Koninklijk Besluit van 15 september 1908. Amsterdam.
- Mitterauer, Michael. 1992. A History of Youth. Oxford: Basil Blackwell.
- Moch, Leslie Page et al. 1987. 'Family strategy: a dialogue'. Historical Methods 20: 113-125.
- Moch, Leslie Page. 1992. Moving Europeans. Migration in Western Europe since 1650. Bloomington and Indianapolis: Indiana University Press.
- Modell, John and Tamara Kern Hareven. 1978. 'Transitions: patterns of timing.' Pp 245-269 in: Transitions: the family and the life course in historical perspective, edited by Tamara K. Hareven. New York: Academic Press.
- Mulder, Clara H. 1993. Migration dynamics: a life course approach. Amsterdam: Thesis Publishers.
- Poelstra, Jannie. 1981. 'Dienstboden tussen arbeiders en burgerstand.' Tijdschrift voor vrouwenstudies. 2:45-68.
- Regt, Ali de. 1993. Geld en gezin. Financiële en emotionele relaties tussen gezinsleden. Amsterdam: Boom.
- Rosenthal, Paul-André. 1990. 'Maintien/rupture: un nouveau couple pour l'analyse des migrations.' Annales ESC 1403-1431.
- Schlumbohm, Jürgen. 1996. 'Micro-history and the macro-models of the European demographic system in pre-industrial times: life course patterns in the parish of Belm (Northwest Germany), seventeenth to the nineteenth centuries', The History of the Family. An International Quarterly. 1:81-95.
- Schofield, Roger. 1970. 'Age-specific mobility in an eighteenth-century rural English parish.' Annales de Démographie Historique: 261-274
- Uitkomsten. 1890. Uitkomsten van het onderzoek naar den toestand van den landbouw in Nederland ingesteld door de landbouwcommissie, benoemd bij Koninklijk Besluit van 18 september 1886. 's-Gravenhage.
- Verslagen. 1908. Verslagen betreffende den oeconomischen toestand der landarbeiders in Nederland. 's Gravenhage.
- Vries, Hille de. 1971. Landbouw en bevolking tijdens de agrarische depressie in Friesland (1878-1895). Wageningen: Veenman.
- Wall, Richard. 1978. 'The age of leaving home', Journal of Family History 3:181-202.
- Weber, Therese ed. 1985. Mägde. Lebenserinnerinungen an die Dienstbotenzeit bei Bauern. Vienna: Böhlau Verlag.
- Woude, Ad van der. 1989. 'De trek van alleenlopenden naar Eindhoven tussen 1865 en 1920'. Pp. 167-183 in: Werk, kerk en bed in Brabant. Demografische ontwikkelingen in oostelijk Noord-Brabant 1700-1920, edited by Gabriel J.M. van den Brink, A.M.D. van der Veen and A.M. van der Woude. 's Hertogenbosch: Het Noordbrabants Genootschap.