1.1 BACKGROUND

Whether in relation to the past or present-day society, there are few studies investigating how individuals' lives develop with regard to work and family if disability was or is present in life. This analysis contributes historical knowledge on this issue by uncovering the life trajectories of thousands of people living in Sweden some 150–200 years ago. Recent studies of the contemporary situation have found that disabilities make individuals more disadvantaged in the labour market than the "able" majority (Arvidsson, Widén & Tideman, 2015; Barnes & Mercer, 2005; Kavanagh et. al, 2013; National Board of Health and Welfare Report, 2009; Reine, Palmer & Sonnander, 2016; Tøssebro & Kittelsaa, 2004). In turn, this disadvantage may explain why disabilities tend to make people more likely to live without a partner or a family on their own (Franklin, 1977; Sainsbury & Lloyd-Evans, 1986; Schur et al., 2013), as uncertain resources make individuals less attractive in the partner market. Disability scholars argue that the impairment itself cannot explain these patterns of inequality. Environmental factors must be considered and not only in terms of welfare provisions, as prevailing perceptions about normalcy in the surrounding society tend to promote marginalization that further impedes the possibilities of people with disabilities finding a job or a partner (Barnes & Mercer, 2005, 2010; Oliver, 1996; Priestley, 2003; Siminski, 2003). As for past societies, information on the working and social lives of people with disabilities is scarce due to deficient or mediated documentation in historical sources. Scholars in disability history have primarily analysed records from institutions to which some people labelled disabled were admitted, or poor relief and crime records in which some disabled people crop up as a result of the economic difficulties they faced (Engberg, 2005; Vikström, 2006; Turner, 2012). Another, smaller group of disabled individuals have left better primary source materials because they led exceptional lives or enjoyed public careers that were sufficiently notable to have been documented (Nielsen, 2001). Such problems with the primary source material jeopardise historians' chances of constructing histories that recognise and account for the diverse experiences of people with disabilities in the past.

Finding a job, getting married and having children were expected and common events as young people in the past transitioned into adult life. While it is difficult to know the extent to which individual people's lives approximated this idealised pathway generally, it is an even greater challenge ascertaining whether people with disabilities did so. Two individuals with disabilities according to the parish registers for the Sundsvall region, Sweden, in the nineteenth century exemplify that a life with disabilities could develop in quite different directions. Both were born in 1820. When Nils Jonas was four years old, the parish registers reported him as being weak-sighted, while they recorded Anna Brita to be deaf when she was eleven years old. In 1837, she was noted as a 17-year old maidservant (piga). In 1846, and at the age of 26, Anna Brita married a crofter, and in the following year she gave birth to their first baby, a girl. While her hearing impairment did not seem to stop her from finding work and forming a family, Nils Jonas did not experience all these events during the time we follow him (i.e. until the age of 33). We know he was a farmhand from the age of 18, but, by the time he was 33 years old, he was still recorded as an unmarried farmhand.

1.2 AIMS

With the different experiences of Anna Brita and Nils Jonas in mind, our aim is to clarify how disabilities affected the life course of people using Swedish parish registers. We compare a series of essential events in their lives with a group of non-disabled individuals, all of whom resided in the nineteenth-century Sundsvall region of Sweden. The latter cohort work as a control group that helps make clear whether and to what extent getting a job, marrying a spouse and becoming a parent were less common events among individuals with disabilities. The acquisition of a job is of primary interest to us, while we examine the other two events to assess the role of having an occupation for marriage and family formation. To differentiate the individuals' life courses, we investigate if the occurrence and timing of these events varied by gender and disability, and also with regard to different types of disabilities.

2.1 THEORETICAL NOTES ASSOCIATED WITH THE LIFE COURSE

The life course approach has some theoretical implications of importance to our study and it has governed the choice of method as explained below (Section 3). This approach has attracted an extensive cross-disciplinary interest (Giele & Elder, 1998; Kok, 2007), in addition to attention from disability scholars (Jeppsson-Grassman & Whitaker, 2013; Priestley, 2003; Sandvin, 2003). This is because it conceptualises the course of life as shaped not only by the identities of the individuals and their decision-making but also by the opportunities, values and norms of the society that work to influence the choices people make across their lifetimes. As people's life course is the outcome of both their individual agency and structural circumstances, it brings each of the two models debated within disability research to the fore. The social model has come to move the focus from the person or the impairment per se (i.e. the individual or medical model) to gain a social understanding of disability, focusing on the limitations or barriers placed in the way of disabled people's full participation in society (Oliver, 1996; Longmore & Umansky, 2001).

According to its proponents, the life course reflects humans' living conditions from cradle to grave, their constraints and possibilities in society (Giele & Elder, 1998; Elder, Kirkpatrick Johnson & Crosnoe, 2004). It involves a line of development that includes phases, such as childhood, adolescence, adulthood and parenthood, and old age, that impact on individuals' status, behaviour, identity, social activities and rights in society. Getting a job, marrying or having a child exemplify life events that affect this line of development. However, having disabilities may jeopardise one's chances to take the course of life one would otherwise have taken or may postpone certain life events, which makes some scholars distinguish between "pathways" and "trajectories" (Kok, 2007). The former prescribes the "appropriate" life course, for instance, the age or timing to take up a job and marry, and may suggest the preferable order of certain events, such as that children should be born within marriage, and not before or beyond marital unions. Pathways thus refer to standardised patterns of human behaviours perceived as "normal" by society at a given time, whereas trajectories show the actual life courses or behaviour of individuals. To gain information on how disabilities affected human life historically, our study investigates life trajectories making use of demographic data on individuals with and without disabilities from the parish registers. We do that by focusing on a transitional phase of young adulthood (between 15 and 33 years of age) when certain crucial demographic events, such as getting the first job, marrying for the first time and giving birth to the first child, tend to occur. The age of 33 was chosen to ensure comparability across the whole period and to ensure the selection of a sufficiently large number of cases to make comparison meaningful.

In both disability studies and disability history, life course analyses are few in number, and even more rare with regard to the quantitative examination of disability. Even though we conduct such an examination to ascertain how life developed in the past among a comparatively large quantity of people, we cannot tell whether their trajectories were primarily the outcome of the individual impairments or rather were shaped by physical obstacles or intolerant attitudes associated with the specific environment. Further, life course scholars argue that there are more aspects to consider in relation the life-span development of individuals, the timing of events and the social ties that shape people's lives (Kok, 2007). Our approach to the life course involves the study of individuals with disabilities from adolescent years (age 15 or before) to see how disabilities influenced the three events of job, marriage and parenthood, and their occurrence and timing up to the age of 33. Our results also reflect how their life courses interlinked with others by their involvement in the labour market (i.e. employers) and in the marriage market (i.e. spouses). Theoretically, we conceive of access to the labour market and family formation as indicative of people's social inclusion.

2.2 THE HISTORICAL CONTEXT AND DISABILITY HISTORY

Disabled or not, young individuals in eighteenth- and nineteenth-century Western Europe, particularly in countries such as Sweden where the processes of urbanization and industrialization moved at a slower pace than in other countries, were required to spend a few years of their young lives as servants before progressing to other occupations. This meant that the majority of adolescents, who originated from farming families or the lower socio-economic strata, sought employment as maidservants, farmhands, servants or apprentices in households other than their parents' (Bras, Liefbroer & Elzinga, 2010; Dribe, 2000; Hajnal, 1962; Lundh, 1999, 2003). In Sweden, they usually did so after having passed confirmation around the age of 15 or 16 years (Jacobsson, 2000). The servant system had a profound influence on the marriage patterns resulting in rather high ages at first marriage, as it took time for the future couple to gather the material resources through work to afford an independent household (Hajnal, 1965; Whittle, 2005). Men in nineteenth-century Sweden married at the average age of about 27 and women at 25 (Harnesk, 1990; Gaunt 1983; Lundh, 1997; Vikström, 2003).

Disability may have constituted a direct obstacle for people if the impairment restricted their ability to perform work or establish social relationships with employers or a spouse. This would cut their opportunities in both the labour and marriage markets, especially for men since their gender role posited them as "breadwinners" expected to support themselves and their families (Horrell & Humphries, 1997; Janssens, 1998). However, historical research suggests that disabled people confronted obstacles in life indirectly, not as an immediate consequence of their impairments but more due to discriminatory attitudes towards their disabilities (Borsay, 2005; Jaeger & Bowman, 2005; Kudlick, 2003, 2008; Turner & Stagg, 2006). Yet, impairment did not allow disabled people to escape the social norms and expectations in relation to labour and self-subsistence. In relation to nineteenth-century Scotland, when entitlement to any economic support from society was meagre, Hutchison (2007) concludes that a job and an income was not impossible for disabled persons to find and key for their chances of finding a spouse. On the other hand, other studies indicate that mechanization during industrialization promoted the exclusion of disabled people from labour and society, since factory jobs replaced handicraft and agricultural labour, and disabled people were considered less able to cope with these new industrial tasks (Barnes & Mercer, 2010; Oliver, 1990). Correspondingly, even though the establishment of institutions increased in the nineteenth century (Foucault, 1965), not all persons with disabilities were admitted to them nor could authorities afford to admit them into such institutions. The latter were aware that a job would make disabled persons become productive citizens and thus less of a burden to society (Bengtsson, 2012; Borsay, 2005; Förhammar, 1991). The establishment of special schools for children with sensory or physical impairments was yet another means to help them become productive citizens in adult life through work. However, such educational institutions were few in number in nineteenth-century Sweden (Förhammar & Nelson, 2004; Olsson, 2010).

Apart from this broader context and these particular historical developments, Swedish law also worked to enforce people's moral obligation to work by regulating the relationship between the employed person and his/her employee. The Servant and Master Act (tjänstehjonsstadgan) was in effect from 1664 until 1926 and stipulated that citizens who were not self-employed or did not live off some property or gained support from relatives, were required to find subsistence through employment to enjoy their civil rights in society. Historians argue that the intention with this legislation was to supply employers' call for a workforce and make people work to cut local authorities' costs in terms of poor relief (Harnesk, 1990; Petersson, 1983). While many disabled people were required by this piece of legislation to work to support themselves, some did nevertheless succeed in gaining relief and they have been subject to historical research which provides some insights into their living conditions (Engberg, 2005; Vikström, 2006; Turner, 2012).

The various issues that arise with primary source material when undertaking research on individuals with disabilities in history suggest why studies incorporating life course perspectives are rare, but a few historians have attempted to overcome these difficulties. Olsson (1999) makes use of parish registers to follow some 400 individuals with disabilities over their lifetimes in the Swedish town of Linköping in the nineteenth century. Men primarily worked as unskilled labourers or craftsmen or farmhands, and women as maidservants or seamstresses. Olsson further finds that about 25% of them married, and all these individuals held jobs. She recognises work as key to marriage, showing that they could support their spouse and future children without rendering any cost to society. However, as Olsson does not conduct comparisons with non-disabled people, it is difficult to assess the extent to which disability limited individuals' opportunities to work and to start a family. Such a comparison is present in the thesis of Sofie De Veirman (2015), who makes quantitative life course analyses of people with auditory impairment and their hearing siblings in Flanders before and during industrialization. This comparison leads De Veirman to conclude that hearing impairments impeded the employment and marriage patterns of hearing impaired people, particularly among the men. Similarly, in his thesis on Revolutionary War veterans in the USA, Daniel Blackie (2010) examines the experiences of some 300 men, of whom about 150 had disabilities. He contends there were similarities between these men regarding their post-war working and family lives. Our own life course studies show that disability decreased both individuals' survival chances and marital prospects, as a possible consequence of disadvantaged opportunities in the labour market (Haage, 2017; Haage, Häggström Lundevaller & Vikström, 2016; Haage, Vikström & Häggström Lundevaller, 2017; De Veirman, Haage & Vikström, 2016).

3.1 THE AREA AND DIGITIZED PARISH REGISTERS

The Sundsvall region (Figure 1) is representative of the demographic and economic structures found elsewhere in nineteenth-century Sweden and north-western Europe. The majority of population depended on agricultural production until the middle of the century but, in four parishes (Alnö, Skön, Njurunda and Timrå), the demographic and socio-economic structure came to change from the 1860s, as economic activity and the labour market became increasingly based on the sawmill industry. The stimulus given to business and commerce were most evident in the town of Sundsvall, the only significant urban community in the region. While the mortality decline of the nineteenth century served to bring about a natural increase in population, the sawmill industry also exerted a pull on migrants which added to the population growth (Bergman, 2010; Edvinsson, 1992; Vikström, 2003). While there were 18,793 inhabitants in the Sundsvall region in 1840, this number had increased to 46,418 in 1880 (Alm Stenflo, 1994). This particular setting, and the existence of digitized parish registers, make the Sundsvall region a unique "historical laboratory" for our analysis.

Apart our own recent studies, this region has not yet been subject to disability history research but has been the object of demographic studies on socio-economic vulnerability in industrial society. A few of these studies consider how norms about labour and marriage in contemporary literature and legislation were reflected in people's lives as recorded in the parish registers. For instance, the increase of migrants and seasonal labour in the sawmills, which was more difficult to control for authorities and more insecure for those employed accordingly, worried social and political elites during the latter part of the century (Bergman, 2010; Vikström, 2003), as did illegitimacy and crime (Brändström, 1999; Vikström, 2008, 2011). These phenomena were viewed as moral threats to the social order, and labour and marriage were highly valued to safeguard the reproduction of families and communities, partly in the hope that such values would decrease the costs incurred because of unemployment, crime and illegitimacy. Studies based on the parish registers for the Sundsvall region provide a complex picture in which a range of attitudes and experiences are evident. On the one hand, some individuals were stigmatized by continued unemployment or recidivism or by never marrying, or, in the case of women, through giving birth to illegitimate children. On the other hand, surprisingly many of these groups of people took up work and married later in life and did not seem to suffer an untimely death due to poor or unhealthy living conditions. The outcome of the analysis below will suggest whether such tolerant attitudes included people who may have looked or behaved differently because of their impairments.

The Demographic Data Base (DDB) at Umeå University, Sweden, has digitized the parish registers of the Sundsvall region. They consist of original records for parishioners' births, baptisms, marriages, out- or in-migration, deaths and burials, the catechetical examination registers, and provide notes on occupations and impairments. Being linked on an individual level, the DDB's parish registers yield a demographic description of each person and the timing of essential events in their lives, such as the first employment, marriage, parenthood, migration or death (Vikström, Brändström & Edvinsson, 2006). Providing data across individuals' lives makes the DDB's parish registers excellent for conducting life course research.

The catechetical examination records (husförhörslängder) explain why Sweden's parish registers are exceptionally rich. Since the seventeenth century, the ministers had to keep these records on an annual basis to document parishioners' knowledge of the catechism and their reading skills (Nilsdotter Jeub, 1993). In these records, the ministers also made notes about their individual impairments (lytesmarkeringar), which indicate disabilities (Drugge, 1988; Haage, 2017; Rogers & Nelson, 2003). Parishioners whom the ministers labelled as disabled have been manually categorised as such for the purposes of this study, since the DDB has not consistently coded this type of information. Due to translation issues, both across time and between languages, we group similar types of disabilities as shown in Table 1 and only account for relatively evident impairments, such as auditory and visual defects and another few types of physical or mental dysfunctions. This facilitates comparisons both within the group of people with disabilities and with non-disabled parishioners in the Sundsvall region selected from the parish registers. Consequently, individuals who were not reported to be blind, deaf mute, crippled or mentally disabled and so forth (Tables 1 and 2) we recognise as having no disabilities. The latter people represent typical life trajectories of the population living in the Sundsvall region.

Table 1. Disability types based on the impairments reported in the parish registers for the Sundsvall region.

Blind	Visual defects, including weak-sighted, short-sighted and blind
Deaf mute	Hearing or communication dysfunctions, ranging from bad hearing to deafness, and from difficulties speaking to stammering to mute
Crippled	Physical dysfunctions, e.g. lame, limping, walking with crutches, missing body parts, harelipped, small in size
Mental disabilities	Mental dysfunctions, e.g. foolish, silly, less cognisant (mindre vetande), insane, feeble-minded or crazy
Multiple disabilities	Combination of two or more of the above disabilities

Source: Digitized parish registers, the Sundsvall region, Demographic Data Base, Umeå University, Sweden

3.2 DATASET AND METHODS

In order to focus upon young people who were in the beginning of the transition to adulthood, the dataset comprises 8,874 unique 15-year-old persons born between 1820 and 1860, of whom 117 had different forms of impairment (Table 2). At the start of observation (age 15 years), none of them were married and none had any children, and only a few had any occupation reported. From then onwards, we observe them for a maximum of 18 years (to age 33 years) to investigate how disabilities influenced their life courses, focusing on the three events of first employment, first marriage and first occasion of parenthood.

Table 2. Population under analysis in the Sundsvall region by number according to gender, disability and type of disability.

Disability category	Men N	Women N	Total N
Blind	9	5	14
Deaf mute	23	13	36
Crippled	19	5	24
Mental disabilities	20	15	35
Multiple disabilities	4	4	8
Sum with disabilities	75	42	117
Sum without disabilities	4385	4372	8757
Total	4460	4414	8874

Source: Digitized parish registers, the Sundsvall region, Demographic Data Base, Umeå University

Having access to longitudinal individual-level data on people enables the method of sequence analysis (Abbott, 1990, 1995; Aisenbrey and Fasang, 2010; Oris & Ritschard, 2014), which we explain and employ more extensively in another study (Vikström, Häggström Lundevaller & Haage, 2017). Broadly conceived, it helps to examine a series of events during an extended period of the life course to identify how disability, for example, interfered with life. A sequence is the chronological states that one person holds over his or her lifetime, here from 15 to 33 years. This age interval constitutes our "observation window" of the trajectory and the events of occupation, marriage and parenthood we look for. As will be evident in the results section, the most common state at the start of observation (age 15) was 0. This refers to "zero" events and means that the individuals had no occupation and were not married and had no children. This state changes if one event in the series of events being analysed occurs. Figure 2 demonstrates the different states and transitions we observe in our sequence analysis. To illustrate the method further, Table 3 outlines the life sequence of the woman and man referred to in the Introduction. While it shows three different states of Anna Brita – who got her first job as maidservant when she was 17 years old, married at the age of 26 and gave birth to her first child in the following year – the life sequence of Nils Jonas only has the event of occupation. At the age of 18, he began working as a farmhand, a job he held throughout the whole observation time, but did not marry or become a father.

Table 3. Image displaying all states by age in the life sequence of one woman and man in the dataset during observation time.

The life sequence of Anna Brita (deaf-mute)
Age		15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33
State		0	0	1	1	1	1	1	1	1	1	1	101	111	111	111	111	111	111	111
The life sequence of Nils Jonas (weak-sighted)
Age		15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33
State		0	0	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1

As for the first job, this appears from the first occupational notes recorded in the parish register, which we regard as indicative of people's employment and transition to the labour market. However, if there is flawed documentation of occupation in the parish registers, or indeed if it is absent, this does not necessary imply that individuals did not work, nor does the recognition of them as being unskilled labourers, farmworkers or maidservants, for instance, confirm that they were actually employed in these occupations. Historians have noted that ministers' registration of people's occupations in this type of population record is not always consistent, and particularly so regarding women and especially if they were married (Leeuwen, Maas & Miles, 2002; Putte & Miles, 2005); Vikström, 2010). Similar to the practice adopted in those studies, we conclude that if there are occupational notes about individuals, which is frequently the case in the Swedish parish registers and particularly for men, it is most likely that they performed the type of work reported and it indicates their occupational status with some degree of accuracy. The two events of people's first marriage and first child are less difficult to determine, as the ministers documented these types of events and their exact dates more systematically than occupational changes. The same is true for the other two events we make some account of, death or migration from the parish before the age of 33, since they were also recorded carefully in the registers. All our analyses are performed in the statistical environment R, complemented with the package and toolbox TraMineR (Life Trajectory Miner for R) (Gabadinho, Ritschard, Müller & Studer, 2011; Gabadinho, Studer, Müller, Buergin & Ritschard, 2015).