Two centuries of disability disadvantages in Swedish partnerships

Abstract

Partnership signifies a key transition for social recognition in society. This study identifies long-term trends of disability and partnership in Sweden evidenced by the chances to marry or cohabit during two centuries (1800s–2010s). We compare results from studies within one comprehensive disability project, making use of quantitative life-course analysis and population records. Our findings uncover a remarkably persistent trend from the 1800s until the 2010s. Disability impeded both men and women’s partnership chances significantly (by about 60%), with some variations across disability types, genders, and periods. That disabled people did not enjoy greater access to a partner relative to others while Sweden moved from a poor country to a wealthy welfare state, suggests that disability persistently affords fewer possibilities to participate in social life and society. Our study is exceptional by combining disability with partnership and comparing recent results with the past.

Points of interest

• Partnership is a milestone for becoming an adult and for social recognition in society.

• In this article, partnership refers to marriages between a man and a woman, or when living together as a couple, both among people with disabilities and without using Swedish population records from the 1800s until 2000s.

• The most important result of the article is that people with disabilities kept being about half as likely to experiencing partnership compared to others no matter of time period studied.

• This result suggests that negative attitudes from the past not recognizing disabled people as potential partners persisted over time when Sweden moved from a poor country in the 1800s to becoming internationally known for its welfare in the 1900s creating more equal opportunities for all citizens.

• This result reveals inequality in that people with disabilities still today do not experience partnership on an equal basis with others.

Keywords:

• Disability
• inequality
• life course
• marriage
• partnership
• Sweden

1. Introduction: background and aims of the study

According to Article 23 of the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD, 2006), disabled people are not to be discriminated against in matters relating to partnership and family formation. Similar to education and work, these matters act as milestones in bringing individuals from childhood to adulthood, and play a significant role for social wellbeing and recognition in society. The relationship between disability and partnership has been far less examined in comparison with how disability leads to disadvantages in health and access to education or employment (EC Eurostat 2017; Lengnick-Hall, Gaunt, and Kulkarni 2008; Kavanagh et al. 2015; Tideman et al. 2017; Turner, Bohata, and Thompson 2017; Shandra 2018; Schur, Kruse, and Blanck 2013; UN Factsheet 2020; Vornholt et al. 2018; WHO 2011, 2018). The small bulk of research on disability and partnership demonstrates that impairments make people more likely to remain single and less likely to head families or households (e.g. Clarke and McKay 2014; Fekete et al. 2020; Fujiura 2014, cf. section 2). One plausible reason for this paucity of research is that disabled people have not been expected or afforded the rights to occupy a social role as a partner on an equal basis with others (Gill 1996; Kallianes and Rubenfeld 1997; Shah 2008, 2017; Van Trigt 2020; Vikström, Shah, et al. 2020). Studies further show that academic as well as popular discourses tend to portray disabled people as asexual beings or infantilized subjects (Bahner 2012, 2015, 2020; Crawford and Ostrove 2003; Finkelstein 1980; Kudlick 2008; McRuer and Mollow 2012; Robillard and Fichten 1983; Schalk 2016; Wälivaara and Ljuslinder 2020). This dismissive portrayal might have far-reaching effects on the partner prospects of the people concerned, which historical perceptions can contribute to (Sigurjónsdóttir and Rice 2020). Our study moves back in time to trace long-term trends in how disability has shaped the transition to marriage or cohabitation up until the 2000s. This transition has been under-researched in the field of disability studies, while the literature on partnership in general populations has ignored disability issues. The topic and scope of our investigation are exceptional, as it combines partnership with disability and compares recent patterns with the past, which seldom come together in one single study.

Following the social model of disability (Barnes and Mercer 2004; Oliver 1990), we primarily view disability as the outcome of various dysfunctions caused by people’s impairments relative to circumstances in their specific time-space setting such as socio-economic structures, welfare provisions, and attitudes to (dis)ability. In recognizing the surrounding society as the major source of disablement, the social model has come to challenge or even replace the medical or individual model according to which disability stays within the person’s body/mind in need of some health intervention to be possibly restored (Oliver and Barnes 2012; Retief and Letšosa 2018). However, the social model has received criticism, as it tends to ignore how both the body/mind and society can bring disability to people (Barnartt and Altman 2001; Kudlick 2003, 2018; Norberg 2019; Oliver 2013, 2017; Shakespeare and Watson 2001; Thomas 2007). This notion makes us apply the holistic framework of the life-course concept (cf. section 4.1), which recognizes circumstances within both societal structures and individuals (their characteristics and lives) to understand the issue under study. Further, it stresses the roles of social ties between individuals. Marriages/cohabitations exemplify close ties, which we examine to increase the understanding of how disability shapes human life and the consequences of it. Examining the partnership chances this way would further suggest how people and society perceive disability, as expressed in the partner pool.

The major aim of our study is to identify how disability has affected partnership chances in Swedish society from the 1800s until the 2010s. We look for variations in these chances across time with regard to gender and even disability type when the data allows. Partnership includes relationships represented by marriage or cohabitation in the varieties of population data we use, which also reports information on disability or impairments; these two concepts we use interchangeably (cf. section 4.2). Although our study is mainly an empirical contribution on how disability has affected partnership for about two centuries in Sweden, we conceptualize partnership as indicative of people’s participation in social life as well as in wider society. Viewing the actual access, or possibly the right, to a partner this way, makes our results reflect disabled people’s opportunities in society from inequality perspectives (cf. Fekete et al. 2020; Mithen et al. 2015; Shandra 2018). Today, Sweden is regarded as one of the wealthier and more equal societies worldwide (Esping-Andersen 1990, 1999; Ginsburg 1992; Gunnarsson 2013), providing more comprehensive welfare programs for citizens with disabilities compared to many other nations. About 150–200 years ago, it was one of Europe’s poorest countries. Whether such a dramatic societal development brought about any shifts in the relationship between disability and partnership has until date not been investigated, primarily due to a lack of data covering extended periods and populations over time.

The next two sections account for a selection of literature of importance to our study and for the Swedish context. Then, we clarify our use of the life-course concept and present the data and methods, followed by a discursive presentation of the results. In the concluding section, we highlight (dis)similarities in how disability has shaped people’s partnerships across time and stress some of the strengths and limits with our study. We also discuss the findings from aspects addressing social inequality and social capital, and propose a few avenues for future research.

2. Disability and partnership: previous results and explanations

Individuals with disabilities are the largest minority group worldwide (15% or 1 billion, UN Factsheet 2020). In the EU countries, disability affects 65–80 million (10–12%) (EC 2017). Despite this sizeable proportion, there are relatively few studies on how disability influences humans’ life opportunities in relation to partnership. One recurrent result is that disability tends to decrease the chances of cohabiting/marrying and make adults more likely to reside with their parents (Fujiura 2014; Namatovu, Häggström Lundevaller, and Vikström, Shah, et al. 2020; Savage and McConnell 2016; Scott-Marshall et al. 2013; Tumin 2016; Vikström, Shah, and Janssens 2020). Moreover, when disabled people experience partnership, they do it later in life than others (Franklin 1977; Janus 2009; MacInnes 2011; Osgood et al. 2005). A couple of studies echoes these results and is of special interest to us in following disabled and non-disabled people over their lifetime. In her thesis of deaf people in East Flanders from 1750–1950, Sofie De Veirman (2015) finds that only 15–19% married while 65–82% of their hearing siblings did. According to the British study of Clarke and McKay (2014) from c. 1990–2005, about 50% of the non-disabled people experienced partnership while this only held for 35% of the disabled individuals. These two studies concern different disabilities and time-space contexts but show similar results. In their study of 22 countries across all continents 2017–2019, Feteke et al. (2020) find that the partnership status of people with spinal cord injury (SCI) depended on socio-demographic and SCI characteristics while the country context was insignificant.

Some of the above studies and other works indicate the negative social and psychological effects disability can have on partnerships (MacInnes 2011; Osgood et al. 2005; Robillard and Fichten 1983; Queirós, Wehby, and Halpern 2015). Research by Shah (2017) and by Wiegerink et al. (2006, 2008) show that even though adolescents with physical impairments do go out on dates, they are not as prone to be involved in romantic relationships as their non-disabled counterparts are. The latter are also less prone to encounter disapproval from others than their disabled peers are. Wiegerink et al. (2010, 2011) suggest that barriers are rising for young disabled people entering into adulthood because the support and services they enjoyed during adolescence then decrease or even stop. This makes them dependent on their parents, thus cutting off their possibilities to develop an autonomous life as well as romantic relationships.

A few new studies on disability and partnership from both past and contemporary times are collected in a special journal issue of The History of the Family (Vikström, Shah, and Janssens 2020). This collection highlights disabled people who marry/cohabit and couples who survive in their relationship when one of the spouses has an impairment (Kim and Mee Kim 2020; Shembri Lia and Abela 2020; Vikström, Haage, and Häggström Lundevaller 2020). However, this collection also contends that people with disabilities face significant social and societal barriers when aiming at partnership, which limits their opportunities to participate in the community and social life. Many of them wish to find a partner despite being consistently reminded that they are not worthy a life like others because of their disability (Amin, Shaari, and Khairuddin 2020; Shah 2008). Cultural studies show an increasing interest in the ‘impossible’ combination of disability and sexual-romantic relationships in order to understand why impairments make people rarely recognized as subjects or objects of love and partnership (McRuer and Mollow 2012; Schalk 2016; Wälivaara and Ljuslinder 2020). According to these studies, media representations (e.g. films or TV series in traditional or social media) rarely portray romantic relationships as viable options for disabled people. As mainstream media can reinforce disabling attitudes in society, disability and partnership risk being regarded as an impossible combination.

Since disability studies in general conclude that impairments limit people’s participation in society at large, it might not come as a surprise if this conclusion holds for the partner market as well. However, there are more advanced explanations. One strand of theories links low partnership to scarce socio-economic resources (Colella and Varma 1999; Jahoda 1981; Oppenheimer 1994, 1997) and impairments can reduce the opportunities to work, generate income and have a career (e.g. Shah 2008; Tideman et al. 2017). Studies suggest that adults with disabilities make up one significant underemployed group in the global workforce (Colella and Bruyère 2011; Lengnick-Hall, Gaunt, and Kulkarni 2008; Vornholt et al. 2018; WHO 2011). Research has found a strong negative association between unemployment and marriage/cohabitation (González-Val and Marcén 2018; Lester 1996), since establishing a household or family with a partner usually comes with costs. Hence, insufficient material resources and low social status inhibit people’s appeal in the partner pool rather than their impairments. Workforce participation also offers opportunities to form social relationships that unemployed persons might miss. Studies report that close relationships to a partner and/or family benefit humans’ health and social wellbeing (Dafoe and Colella 2016; Kalmijn 2017; Koball et al. 2010; Mithen et al. 2015; Robles 2014; Ross, Mirowsky, and Goldsteen 1990; Sandström et al. 2021). Insufficient access to close social ties can thus have detrimental effects on life in the shorter as well as the longer run.

The theory of homogamy suggests that similarities attract, implying that people select a partner resembling their own values, socio-economic status, or cultural background (Kalmijn 1998; Vikström, Haage, and Häggström Lundevaller 2020). Such assortative partner preferences might manifest ‘disablist views’ among non-disabled people, not recognizing disabled persons as potential partners (Crawford and Ostrove 2003; Helmius 1999; O’Toole 2002; Robillard and Fichten 1983). This notion resembles the concept of ‘courtesy stigma’, which sociologist Erving Goffman (1963) coined about sixty years ago and has been extensively in use ever since (Corrigan and Miller 2004; Phillips et al. 2012; Scambler 2004). This stigma is associative in moving beyond the individual afflicted to include his/her friends or relatives, and this can make potential partners avoid disabled people. The fear of becoming trapped in a caregiving role could be yet another factor to avoid partnership with someone having a disability (Fiduccia 2000; Gill 1996; Savage and McConnell 2016). Studies show that negative attitudes from the surroundings, especially when coupled with sexual relationships, can be transmitted to those subjected to these attitudes (Amin, Shaari, and Khairuddin 2020; Gill 1996; Schembri Lia and Abela 2020). Not seeing themselves in the role of having a spouse, they become more likely to stay single. This adds an explanation to the low partnership level linked to disability and suggests that environmental views and circumstances matter in line with the social model of disability.

3. The Swedish context: from a poor country to a wealthy welfare state and beyond

As our study shows Swedish findings from the past two centuries, we outline some of the societal structures and living conditions below. Three overlapping periods help inform us about the socio-demographic and economic developments, and about welfare measures and norms linked to partnership and health issues like impairments. There are no nationwide figures about the prevalence of impairments during the whole study period because records are not available or comparable over time. Disability definitions can vary considerably depending on the time-space context, as perceptions and the documentation of impairments are not consistent, neither is the collection of data (Grönvik 2009; Kaplan 1999; Oliver 2017). Present-day estimations suggest that between 1.3 and 1.8 million of the total Swedish population (10.3 million) has some type of disabilities (http://www.funka.com/design-for-alla/tillganglighet/statistik/.) In the early 1800s, only about 2.5 million made up the population. In 1900, this number had doubled to 5 million.

In pre-welfare Sweden (c. 1800–1930), the governmental resources to aid people in need due to impairments were scant. Welfare support depended on kin, local parishes (poor relief) and charity, especially during the 1800s. In an economic structure based on household production, marriage was key to finding subsistence and cohabitating as couples. Spouses and especially the ‘breadwinning male’ were to gain the income to afford a household. Legislation prohibited marriage if the spouses were too young or could not provide for themselves or their new family. Until the 1840s, men should be 21 and women 15 years to marry while the eligible age for the latter was set to 17 in 1892, and to 18 in 1915 (Gaunt 1983; Lundh 1997). Certain diseases (e.g. epilepsy) and mental disorders (e.g. ‘insanity’, ‘idiocy’) were regarded impediments for marriage and continued to be well into the 20th century (Hafström 1974). Urban-industrial processes went on at a slow pace during the 19th century and did not alter the labor market and living conditions more rapidly until the 1870s. Factory work replaced agricultural production and handicraft, and provided new employment. However, the 19th-century large-scale processes led to income inequalities and made about 1 million Swedes leave the country for North America before WW1 (Runblom and Norman 1976).

During the 1930–1990 period, the reputation of the Swedish welfare state spread internationally due to its comprehensive welfare programs and economic progress (Esping-Andersen 1990, 1999; Ginsburg 1992; Hirdman 1989; Lundberg and Åmark 2001). The first phase (c. 1930–60) took place within the hegemony of the social democratic party and confidence in the knowledge of experts and authorities was profound. They acted as ‘normative advisors’ regarding citizens’ lifestyle and health, and regarding what citizens the state should afford welfare to and what the care should consist of. Confinement was a highly regarded ‘treatment’ and institutions were built rapidly, many of which were directed towards people with impairments. Influenced by eugenic ideas, more than 60,000 Swedes were subject to forced sterilization programs from 1935–1975 due to ‘undesirable’ hereditary conditions, such as impairments (Runcis 1998; Tydén 2002). Although eventually abolished, these invasive measures reflect the dual character of the Swedish welfare state in relation to disabled people. Yet, this period showed improvements in health care, economy, education, and social services that increased the material and medical conditions of average citizens and decreased the inequality between them. In the 1950s and 1960s, working-age persons with long-term limitations due to handicap or ill health could access some income support through benefits. Women entered the workforce in large numbers and it became less necessary for them to find subsistence through a husband. During the 1960s, couples increasingly cohabited and bore children without marrying and the divorce ratio rose (Sandström 2012). These population developments exemplify the process of de-familiarization associated with the second demographic transition (Lesthaeghe 2010; Van de Kaa 1987), and manifest a move away from traditional and religious norms about that couples must marry to cohabitate. Economic booms and de-institutionalization, the latter of which aimed to integrate disabled people in society, characterized the 1980s. This goal goes back to the 1960s when the ‘normalization principle’ became a key concept in Swedish disability policy to enable for disabled citizens a more ‘normal’ life regarding education, work, and inclusion in the community (Nirje 1969; Tøssebro 2016). Universal suffrage did not come to include disabled people until 1989, while others had enjoyed the right to vote since 1921.

Sweden’s most recent era, which we call the post-welfare period, starting approximately around 1990, shows a continuation of the de-familiarization process and pluralities regarding partner choices (Therborn 2004). Swedish disability policy moved further away from large-scale and centralized public services and put an emphasis on citizens’ freedom, diversity and individual choice (Lindqvist 2000). Today, living a life without a partner is possible and more accepted in contrast to 100–150 years ago. In 1995, the right to register a homosexual partnership was sanctioned, and marriages from 2009. In 2008, Sweden ratified the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD 2006), according to which states are to ensure access to a range of support services to prevent segregation. Recurrent economic recessions and ideas of New Public Management have, however, come to limit the public costs in Sweden overall since the 1990s, partially governed by debates about ‘welfare fraud’ and austerity (Altermark 2017; Goodley, Lawthom, and Runswick-Cole 2014; Norberg 2019). One law of special interest is the Act for Support and Service (1994) for Persons with Certain Functional Impairments (Lagen om Stöd och Service till vissa funktions­hindrade, LSS). It provides disabled citizens the right to apply for a personal assistant to facilitate participation in society on more equal terms with others (Askheim, Bengtsson, and Richter Bjelke 2014). Since Swedish disability policy favors active citizenship through redistributive and regulatory measures, and especially through employment, citizens must negotiate their social rights to attain or maintain security (Christensen, Guldvik, and Larsson 2014; Halvorsen et al. 2017; Lindqvist 2000). As not all enjoy the capability nor have the socio-economic resources to claim their rights accordingly while public benefits are shrinking, citizens have increasingly been denied support and disabled people have not received service assistance upon stricter assessments of their needs. Consequently, inequalities exist in Sweden as research and reports show (Bahner 2012, 2015, 2020; Bylund 2022; National Board of Health and Welfare 2009, 2010; Norberg 2019; Sépulchre 2018; Tideman et al. 2017). This has made citizens in need of support more dependent on their relatives, like partners, parents or children, similar to the past.

4. Research design: theoretical framework, data and methods

4.1. Disability and partnership within the life-course concept and partners as social resources

Scholars conceive the life course in terms of trajectories (Elder 1985; Giele and Elder 1998) that reflect people’s living conditions, constraints, and possibilities in society. Individuals follow a line of development with phases such as childhood, adolescence, and adulthood, which in turn include transitions to education, work, partnership, and parenthood, for example, before entering into old age. These phases and transitions affect people’s status, behavior, identity, social activities, and rights in society. Finding a job or a spouse, or acquiring disability, can influence social wellbeing and subsequent course of life. In disability studies, the life-course approach has been increasingly used and debated, but primarily concerning its qualitative application (Heller and Harris 2011; Priestley 2003; Shah and Priestley 2011; Tarvainen 2021). Peter Siminski argues this is ‘because large-scale disability surveys have usually been framed around the medical model, which posits disability as a consequence of disease and limits analysis of disability as a social product’ (Siminski 2003, 717). We seek to avoid such a framing by addressing the social disability model (above) and social capital (below).

This study examines the outcomes of disability on partnership quantitatively by considering four life-course principles (Kok 2007; Mortimer and Shanahan 2003), one of which is the life-span development. It proposes that past experiences shape present circumstances and even future developments. We apply this notion examining whether impairments since birth or young adulthood lead to disadvantages in life represented by low partnership chances. Another principle recognizes humans as agents and that the life course mirrors their choices and actions taken, in this case to enter partnership. Third, the principle of time and setting stresses that the environment influences people’s choices and actions. Making temporal comparisons, we come across different normative views in Swedish society concerning gender, (dis)ability, and partnership. Sweden exhibited shifting structures over time, providing different opportunities for less fortunate citizens regarding access to a job or a partner or to welfare, which can decrease or increase inequalities between people. Research has called for an increasing recognition of disability as an analytical category on par with gender, age, and class to comprehend how human characteristics intersect to define people’s (dis)advantaged position in society (Barnartt and Altman 2013; Grönvik and Söder 2008; Kavanagh et al. 2015; Kudlick 2003; McCall 2005). Our multivariate statistical analysis helps respond to this call, as we can estimate the impact of disability on partnership relative to other individual-level characteristics in different temporal contexts of Swedish society.

Fourth, the principle of linked lives is central to us, as it states that familial and social ties mark the life course. People’s possibilities depend on the assistance and appreciation of others, which we gain data on through partnerships like cohabitation/marriage. A partner/spouse can provide socio-economic and emotional resources, just as oneself can act as a resource to him/her. In this sense, partners/spouses constitute ‘social capital’, as the French sociologist Pierre Bourdieu (1986) conceived it. According to him and his followers, social capital is a resource that flows between members of social networks and usually is beneficial for social wellbeing and health among those attached to these networks (Karhina et al. 2016; Kawachi, Subramanian, and Kim 2008; Mithen et al. 2015). Although scholars can differ in their conception of social capital, they usually regard it as a major determinant of health-related outcomes, which is rarely equally distributed between people because class or economic and socio-cultural resources define social capital. Economically disadvantaged groups, like people with disabilities, risk further marginalization since they often lack networks that can promote social capital to help improve their position in society. Examining lives linked together through marriages/cohabitation makes social capital of interest to consider. Formal partnership means close access to social capital through the partner/spouse, at least potentially. Irrespective of disabilities, people not experiencing partnership might lose some of the advantages associated with social capital. Further, if potential partners think that disabled individuals have insufficient resources to bring in terms of social capital, the latter would be less likely to enter partnership.

4.2. Data, methods and definitions

In this study, ‘partnership’ refers to relationships represented by marriage or cohabitation. We use the two concepts of ‘disability’ (funktionshinder) and ‘impairment’ (funktionsnedsättning) interchangeably as the Swedish language does not have any exact equivalent for ‘disability’ and its distinction from ‘impairment’, as Norberg has explained (2019, 20–38). Disability/impairment refers to various limits in people’s functionalities as described in the data, which helps us identify individuals with disabilities, also termed ‘disabled people’ and ‘disabled groups’ in this study. To enable temporal comparisons within the life course framework (section 4.1), we bring together studies within the DISLIFE Liveable Disabilities project (https://www.umu.se/en/research/projects/dislife-liveable-disabilities-life-courses-and-­opportunity-structures-across-time) on how disability has affected partnership in past and recent Sweden (Haage, Vikström, and Häggström Lundevaller 2017; Namatovu, Häggström Lundevaller, and Vikström 2020; Sandström et al. 2021; Vikström, Haage, and Häggström Lundevaller 2020). Each of these studies and the preprint of this paper (Vikström, Junkka, and Karhina 2021) further details the research design and statistical results presented in the next section. Our results cover the two most recent centuries (except the 1960–1990 period) and are based on micro-level data from nationwide records on Swedish population or samples of it (c. 1990–2011), or from parish registers for selected regions (c. 1800–1960) (Table 1). The use of different datasets from different periods has implications for the definition of disability, as explained below and detailed in the individual studies. All the studies and use of datasets comply with ethics issues upon approvals from the National Ethics Review Authority.

Table 1. The databases on Swedish populations and the periodic time studied.

Database/study population	Time period	No. of persons	Swedish area	Type of data
1. HISTORICAL POPUM Database Demographic Data Base (DDB), Umeå University, Sweden	c. 1800–1900	c. 28,500 (Disabled: 468)	Sundsvall region (14 parishes)	Parish records showing continual micro-level data on birth, occupation, marriage, migration, death, impairments
2. HISTORICAL POPLINK Database Demographic Data Base (DDB), Umeå University, Sweden	c. 1900–1960	c. 116,000 (Disabled: 3,035)	Västerbotten region (11 parishes)	Parish records (same type of data as above)
3. SIMSAM Database Umeå SIMSAM-Lab, Umeå University, Sweden	1990–2010	c. 550,000 (Disabled: 12,636)	Nationwide	Extensive longitudinal socio-economic & demographic micro-level data, in-patient & health data at birth, drug prescription records, hospitalization, disability benefits
4. SILC/ULF Database (National Survey of Individuals’ Living Conditions, Statistics Sweden), Dept of Sociology, Umeå University, Sweden	1993–2011	c. 16,500 (Disabled: 553)	National sample	Rich micro-level data on living conditions, indicators of impairments, cross-sectional panel waves, disability benefits, self-reported disabilities & accounts of quality of life

The results are obtained through statistical life-course methods. Logistic regression or Cox regression models estimate the chances of partnership (marriage/cohabitation vs. singlehood) and how impairments (congenital or acquired at a young age) affect the probability of experiencing partnership in different periods of Swedish society shown as hazards ratios (HR) or odds ratios (OR), which are comparable (Vikström, Junkka, and Karhina 2021). As the data is time-stamped, we know whether and when in life impairment occurred and similarly marriage/cohabitation, in addition to socio-demographic characteristics of the individuals and their biological sex as the data report this. Adjusting for the impact of other variables, our key interest is the effects disability has on people’s partnership chances and variations over time by gender and even type of disability when the data allows.

Although the statistical methods are similar and the datasets resemble each other in showing micro-level socio-demographic data and impairments, the regressions were run separately for each dataset. This is because the analyses took place in different labs at different times and the datasets cannot be merged all together due to ethics issues. Further, the data does not always provide the identical type of information. While marriage is the only evidence of partnership in the historical data, the recent datasets also show cohabitation and parenthood. Moreover, the disability documentation varies since the datasets comprise records maintained in different periods for shifting purposes. As for the historical datasets (1800s and 1900–1960, Table 1), ministers reported impairments among parishioners to assist national authorities with collecting data on the population and its health status. Such recording would have been hard to collect without the ministers, as Sweden was a vast and sparsely populated country back then. Historians have found the Swedish parish registers to be highly credible, having resulted in numerous scientific works (Edvinsson and Engberg 2020; Westberg, Engberg, and Edvinsson 2016). These registers indicate abilities and behaviors (lyten) that the ministers regarded as disabling or abnormal at the time and in accordance with authorities’ guidelines, in turn governed by medical expertise and perceptions of (dis)ability. This type of reporting enables us to distinguish between ‘sensory’, ‘physical’, and ‘mental’ impairments. While many of the concepts we come across in the historical data are offensive today (e.g. ‘crippled’, ‘idiot’ or ‘insane’), especially concerning the group of ‘mental’ impairments, we have no intention to offend anyone when using these concepts; this is only to avoid translation issues both across time and language. Parishioners not showing these impairments constitute a group of references we term ‘non-disabled’ and whose partnership chances we compare with the disabled group (Vikström, Junkka, and Karhina 2021; Haage, Vikström, and Häggström Lundevaller 2017; Namatovu, Häggström Lundevaller, and Vikström 2020; Sandström et al. 2021). Due to a higher number of cases and a more detailed reporting of disabilities in the 1900–1960 period, we can distinguish between ‘auditory’ and ‘visual’ impairments in the sensory group, and in the group of mental impairments between acquired (‘psychiatric’) or congenital (‘intellectual’) disabilities (Junkka, Sandström, and Vikström 2020).

The data on recent populations allow comparisons accounting for impairments in two ways (cf. Table 1). First, in the SIMSAM dataset, there is data on the benefits entitled to citizens regarded unable or unfit to work on medical grounds ranging from physical to psychological conditions, though not specified further (Namatovu, Häggström Lundevaller, and Vikström 2020). This is a less differentiated definition of disability compared to past parish records, but it is based on diagnoses in the SIMSAM dataset which is also nationwide. Second, the SILC/ULF sample provides data on both disability benefits and self-reported impairments among citizens in recent Sweden. In this study, we only account for mobility disabilities (Sandström et al. 2021) while the self-reported data also informs about how people experience their health and social wellbeing given their current life situation.

5. Results: long-term trends in disability and partnership

Our results show that disability made the partnership chances differ among Swedish men and women across the centuries studied, as disability impeded these chances with about 60% or even more. This difference is substantial and statistically significant. From the 1800s until 1960, disability affected women’s partnership chances slightly more negatively (between 71–79%) than among men (between 63–75%) (Vikström, Junkka, and Karhina 2021; Haage, Vikström, and Häggström Lundevaller 2017). In the 1990–2010 period, this gender difference diminished as both men and women with disabilities had 60% lower partnership chances than non-disabled groups (Vikström, Junkka, and Karhina 2021; Namatovu, Häggström Lundevaller, and Vikström 2020). While we do not access data for the 1960–1990 period, another Swedish study covers the major part of it and men born in Stockholm in 1953. Using a similar statistical life-course approach, Grotta and Högnäs (2019) find that disability reduced men’s marital chances with about 50%.

As for the 1993–2011 period, our results concern the risk of ending up living alone, thus not experiencing partnership, when having mobility disabilities (Vikström, Junkka, and Karhina 2021; Sandström et al. 2021). Compared to non-disabled women, women with mobility disabilities had 94% higher risks to live alone while the corresponding risk among their male counterparts was 41%. From the self-reported data, it further becomes clear that both men and women with mobility difficulties rated their quality of life lower when living alone than others did (Vikström, Junkka, and Karhina 2021; Sandström et al. 2021). This result is interesting for two reasons. First, living alone and low life satisfaction is associated with adverse health outcomes that disability might promote if it implies singlehood (Collins, Glei, and Goldman 2009; Wiest et al. 2011). Second, the high dissatisfaction from living alone suggests that people with disabilities represent normative expectations in wanting to experience partnerships to the same level as non-disabled (Amin, Shaari, and Khairuddin 2020; Arnold and Chapman 1992; Tarvainen 2021).

Regarding how the marital chances have varied up until 1960 depending on disability type, particularly mental disabilities limited the probability to marry (Vikström, Junkka, and Karhina 2021). Compared to non-disabled groups, the marital chances were 79% lower among both men and women with mental disabilities in the 1800s, and these chances stayed similarly low (between 70–93%) in the 1900–1960 period (Vikström, Junkka, and Karhina 2021). This indicates that mental disabilities were associated with poor access to the partner pool, probably due to high levels of stigmatization and/or institutionalization in comparison with other impairments. As for sensory and physical disabilities, there were some gender variations over time although the results for women are less statistically significant. During the 1800s (Vikström, Junkka, and Karhina 2021), sensory disabilities affected men’s marriage more negatively (67%) than physical disabilities did (44%), while the latter disability type reduced women’s marriage to a higher extent (65%) than sensory disabilities did (56%). In the 1900–1960 period (Vikström, Junkka, and Karhina 2021), sensory as well as physical disabilities continued to decrease women’s marital chances profoundly (c. 70% and 78% respectively) and even more than in the 1800s. Sensory disabilities continued to limit the marriage of men as well (visual 28%, auditory 54%), but to a lower extent than the century before. While physical disabilities decreased men’s marital chances (51%) in the 1900s, women’s chances were considerably more negatively affected (78%) by this disability type (Vikström, Junkka, and Karhina 2021). One plausible reason is that disabled women’s bodies in particular have been asexually objectified by medical and legal discourses that have made them perceived as undesirable partners (Kallianes and Rubenfeld 1997; McFarlane 2004; Shah 2008; Shembri Lia and Abela 2020; Smith and Hutchison 2004).

The differences in marital chances between men and women over time reflect gendered expectations linked to work and normative views concerning (dis)ability (Vikström, Junkka, and Karhina 2021; Haage, Vikström, and Häggström Lundevaller 2017; Vikström, Häggström Lundevaller, and Haage 2017). During the period before public welfare provisions, a job and income was key to affording an independent life and preparing oneself for marrying, which required the capability to perform work. In 19th-century Sweden, which largely depended on agricultural production, impairments might not have entirely hindered women from doing domestic work associated with their gender role (e.g. maidservants, seamstresses). They might have kept some of their appeal in the marriage market, as women were to take care of the household tasks and family members. Our results suggest that sensory disabilities did not modify their feminine appearance and domestic work capability as much as physical impairments did, as the latter made women less prone to become wives. Men were expected to be ‘breadwinners’ and provide for themselves, their wife and family by taking up rough work as farmers, farmworkers or craftsmen (e.g. tailors, shoemakers) or in factories (e.g. skilled or unskilled laborers). Such manual occupations were likely more difficult to conduct when having sensory disabilities than if being physically impaired and help explain the gender variation in marriage.

The job opportunities and gendered expectations changed in the 20th century as economic modernization and the expansion of the welfare state came to differentiate the labor market. People found more diverse occupations in industry and in the service and public sectors and women increasingly entered the workforce. Possibly, labor market developments modified the gendered marital chances that people with sensory and physical disabilities experienced in the 1800s. Sensory disabilities became more limiting for women during the 20th century than back then, while the opposite trend held for men. Physical impairments continued to impede women’s marital chances substantially in the 1900–1960 period, while similar impairments did not inhibit men’s marriage to similar high extent as in the 1800s (Vikström, Junkka, and Karhina 2021). Physical disabilities kept having adverse effects especially on women’s partnership, as the most recent decades show that mobility difficulties made them more prone to live alone than men (Vikström, Junkka, and Karhina 2021). On the one hand, this might reflect women’s independence, or that they are more inclined to live alone or to cope with this situation than are men. In Sweden, comprehensive welfare programs have long supported the idea of independent life without having to lean on a partner or relatives, which has particularly benefitted women, allowing them to stay single if they like (Gunnarsson 2013). On the other hand, that mobility difficulties decrease the partnering chances more for women than men suggests that bodily dysfunctions or defects still make women less regarded as potential partners because female bodies continue to be more subject to gendered ideals than are male bodies.

6. Concluding discussion

Historically, marriage was the aim of most young people. Even today, partnership continues to be one key transition to adulthood and social recognition in society. The aim of this study was to identify how disability has affected people’s partnerships across time in Sweden using population records from the past to recent times. This section sums up the findings and discusses them from social inequality perspectives, as we conceive partnership as one indicator of participation in social life and society that has largely passed unnoticed in research. We also address some strengths and limits with our life-course approach and suggest avenues for disability research to take.

Our major contribution are the long-term trends we can uncover using longitudinal data and statistical life-course methods. To our knowledge, no other study has been able to trace such trends. First, they show evidence that disability kept weakening individuals’ position in the partner pool to a similarly profound extent in recent times as two centuries ago. No matter the period or gender, disability cut partnership chances by some 60% or even more. This trend stays remarkably strong despite the different structural conditions that Swedish society has witnessed since the 1800s. Second, another profound but less surprising trend is that mental disabilities implied the ever-lowest partnership chances for both men and women from the 1800s until 1960. This result resonates with disability studies, past and present, which generally find that this type of disability brings difficulties in domains concerning education, labor, and health, for example, and often due to social interaction issues. Third, during the study period we came across some partnership variations depending on whether men or women had sensory or physical impairments, probably shaped by current perceptions regarding gender roles and (dis)abled bodies. These results would not have come to the fore without distinguishing the effect of different disability types in vast populations across extended times. Fourth, our trends suggest that the overall disability effect on partnership became more similar between the genders over time. Up until 1960, disability made women’s partnership chances slightly lower, while this gender gap had narrowed by 1990. This mirrors Sweden’s move toward a more gender-equal society. During the preceding decades, welfare reforms like parental leave had worked to encourage gender equality in Swedish society and women’s incomes improved through their increasing participation in the labor force. These developments likely balanced the genders’ chances in disabled people’s partnerships, although their chances relatively speaking remained less than half of others’ chances across the whole period.

Although the explanations are not clear-cut, disability tends to make human life differ from the typical life course regarding partnership. As this difference can both indicate and lead to disadvantages, it is relevant to apply social inequality perspectives on the long-term trends we find. Impairments would be less disabling in the partner pool and likely in society as well, if people, irrespective of their capability or attributes, could enjoy various socio-economic resources and attitudes enabling their interaction with peers in the community, which in turn can result in partnership. This makes partnership and disability of the utmost concern to scrutinize simultaneously. Despite Sweden’s development of welfare systems and targeted policies for creating and sustaining an equal society for all, including citizens with disabilities, the negative association between disability and partnership has persisted over time. It was slightly more prominent in the first half of the 20th century due to the dominant ideology that promoted institutionalization and largely inhibited access to partners among those confined. Yet, the recent relationship between disability and partnership does not differ much from then. The Act for Support and Service for Persons with Certain Functional Impairments (1994) to assist the participation of disabled citizens in wider society in accordance with the UNCRPD (2006) has not made any significant impact on their partnership chances. Developments in the 1900s such as the de-familiarization process and differentiation of labor markets have not helped to narrow the disability disadvantage in partnership, neither have more tolerant attitudes towards partner preferences. Disability keeps on distributing the partnership chances unequally between people.

This bias in partnership might not only indicate the level of social inequality in human life and Swedish society across time due to disability. It can have wider implications in making the group with disabilities less equal in other domains as well, since many advantages come with living in partnership in terms of social capital. Partnership tends to benefit one’s ties to firm social bonds and to emotional support, which benefits social wellbeing and health. Two partners’ pooling of socio-economic resources can improve their living conditions and increase quality of life. Consequently, small chances to experience partnership might have detrimental outcomes in the longer run and risk reinforcing existing inequalities between people through the accumulation of disadvantages across the life course. In turn, this suggests that there would be many gains with implementing policies and programs promoting social capital among people in general including those with disabilities, for example, by increasing their access to the partner pool possibly to enjoy positive effects coupled with social capital. However, such policy measures hold some normative notions that are not without drawbacks as discussed below.

While the life-course concept implies many strengths that our results have benefitted from, it also carries limitations. One limit is that it has normative implications of how life is to be lived in taking the life course of able-bodiedness/mindedness for granted. This expects people to attain an education, take up work, or engage with a partner, which has influenced the research design of our study, for example the comparisons between disabled and non-disabled groups. The normative pathways imposed by the life-course concept have yielded constructive criticism among scholars who have positioned disability within a life-course frame (Priestley 2003; Shah and Priestley 2011) and among scholars who call for rethinking human life and its courses through the lens of ableism (Campbell 2012; Cherney 2011; Halberstam 2005; Ljuslinder, Ellis, and Vikström 2020). We welcome these avenues of research and think our long-term results will encourage a further debate on the use of the life-course approach in disability research.

Another limit with our study worthwhile to recall is that partnership is not something that all are aspiring to, nor does it always bring joy in life. Studies suggest that people with disabilities are much more likely to become victims of domestic violence than others are (Shah, Tsitsou, and Woodin 2016; Thiara, Hague, and Mullender 2011). They are also more prone to split up from partners (Clarke and McKay 2014; Osgood et al. 2005; Singleton 2012). These issues urge a further investigation of disability in close relationships. There is also a dearth of research treating the couple as the major unit for analysis when disability is involved through one partner or both, which would yield knowledge about how the couple’s partnership is established and survives across its lifetime. Moreover, people with disabilities might live in partnership but without communicating it to authorities. This is one means to avoid the risk of losing some of the assistance or benefits, which is key to not depend on their partner (Bylund 2022; Norberg 2019). In escaping public recording, this type of partnership is not included in our study and exemplifies the need to know more about the complex relationship between disability and partnership. Romantic partnerships must not only involve one man and one woman, which is the only type of partnership our data enables us to examine. This under-recording further suggests that the actual partnership levels of disabled people is higher than the levels this study uncovers, at least for the recent decades. During these decades, technological advancements such as the Internet and cell phones have also come to link people increasingly together as partners, but still little is known about to what extent this includes people with disabilities and the consequences (Molin, Sorbring, and Löfgren-Mårtenson 2017; Shah, Bradbury-Jones, and Taylor 2020).

Despite dissimilarities between the datasets and in their definition of disability, there are many strengths with our long-term study on how disability affects partnership. The statistical results show the same outcome measured similarly, although the reverse side of partnerships is concerned in the sub study on the link between disability and living alone (Sandström et al. 2021). Cox regressions and logistic regressions are joint for all sub studies we base our long-term results on and allow temporal comparisons. Furthermore, individuals not having disabilities reported in the datasets serve as references/controls that help to discern partnership between groups of people by disability and gender across an extended time. In all, this has enabled us to identify two centuries of disability disadvantages in partnership.

Our study results are exceptional yet disappointing. Sweden has long held a reputation internationally for its generous welfare policy providing public services and social security to limit citizens’ dependence on support and care from relatives. However, about 150–200 years ago, Sweden was one of Europe’s poorest countries. Since then, it has witnessed large-scale transformations that have shaped peoples’ opportunity structures such as urban-industrial changes, shifts in gender regimes, democratization, and advancements in medical treatment, as well as the advent of New Public Management promoting a dismantling of public welfare in more recent decades. Among people and scholars across the globe, Sweden might still hold its reputation of being especially good to marginalized groups like women, children, immigrants, ill citizens, or those without employment or to people with disabilities. Similar to new Swedish disability research (Altermark 2017; Bahner 2020; Bylund 2022; Norberg 2019; Sépulchre 2018), our partnership findings cannot support this reputation or that all the policy efforts implemented over time have made disabled people participate in wider society on a more equal basis with others. Why this disadvantage in partnership persists is difficult to pinpoint. It might be an outcome of social marginalization having prevailed even though Swedish society has changed fundamentally. Disablist attitudes not recognizing disabled people as potential partners have likely yielded the persistence of the low partnership trends we find and such attitudes can be more important than material or physical factors in determining the access to a partner or finding one. These low trends do not only indicate high social inequality levels but also the weight history can have in shaping circumstances to which past negative values of disability contribute to compromise people’s opportunities in present-day society.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme, Grant Agreement No. 647125. DISLIFE Liveable Disabilities: Life courses and opportunity structures across time’, headed by Lotta Vikström.