The power of one? -The long-term increase in one-person households in Sweden, 1900-2017

Project description

Background and aim

Today, more people than ever before are living in one-person households. Political concerns about the consequences of increased single living, such as loneliness and vulnerability, are being raised in several Western societies (Edlien 2018; Starbrink 2018). Sweden exhibits one of the highest proportions of single living in the world, reaching almost 40% of all households (Eurostat 2019; Statistics Sweden 2019). The proportion of the population not living with someone else has increased from every fifth to every fourth person aged 40-65 since the early 1990s (Statistics Sweden 1992, 2019).

In the discussion of increased levels of single living in contemporary societies, Scandinavians are often described as being distinctly individualistic. They prioritize values of personal independence over mutual obligations between family members, which in turn promotes high levels of one-person households (see e.g. Klinenberg 2012; Lesthaeghe 2010; Reher 1998; Esping-Andersen 1990). However, in an historical perspective, one-person households is not a new phenomenon. Already during the first decades of the twentieth century, every fifth household in Sweden consisted of only one person according to the censuses (Statistics Sweden 1969). Surprisingly little is known about this long historical background and the processes that have led to today's high proportion of one-person households. In fact, why single living was relatively common more than one hundred years ago is not researched at all. Also, our knowledge is very fragmented about the socioeconomic and health related consequences of single living in the past.

The aim of this project is to analyze the demographic and socioeconomic (SES) characteristics of one-person households in Sweden, from 1900 to 2017. By applying a life course perspective, with recently available and unique large-scale microdata, we will answer the following questions:

Q1: How has the demographic characteristics of the one-person households changed during the twentieth century?

Q2: How has socio economic inequalities between individuals in one-person households and those cohabiting developed since the early twentieth century up until today?

Q3: How has inequalities in health between individuals living in one-person households and those that cohabit developed since the early twentieth century until today?

This project will challenge the notion that single living is an intrinsically modern phenomenon. It will bring new knowledge on why this form of living arrangement has become so common today and how it has developed over a period of more than 100 years. It will examine how SES, health inequalities and living arrangements have changed during a period when Sweden progressed from a largely agrarian to an industrial welfare state.

The project will also provide important insights into how this association might develop in the coming decades. Investigating the characteristics of those who have contributed to the increased single living is essential for our ability to assess the shifts we can expect in the future, and its societal consequences.

Survey of the field

Studying the long-term growth of one-person households in Sweden is important for several reasons. Sweden, together with the other Nordic countries, are repeatedly identified as forerunners in a trend towards weaker family ties and more individualistic behaviors (Höjer 2014; Iacovou and Skew 2011; Popenoe 1987). Several characteristics of contemporary Nordic societies are claimed to explain the large share of one-person households. It is suggested that Nordic countries are more de-familialized (Esping-Andersen 1990), that they are typical examples of weak family systems (Reher 1998) and people are putting more emphasis on individualistic values that prioritize self-actualization over collective solidarity (Lesthaeghe 2010). But, this characterization of Sweden as a “forerunner” in the growth of one-person households, is almost entirely based on evidence from the last few decades of the twentieth century, which is problematic for several reasons.

Our early censuses show that the existence of one-person households is not something that has emerged in the latter part of the twentieth century. In fact, already in 1900 one-person households counted as much as 20 % of the entire population. Therefore, the role of “individualization” needs to be challenged in an historical context. And, at present, there is no research describing the long-term trends in the incidence and characteristics of one-person households and its relationship with the economic, institutional and cultural restructuring of Sweden.

An influential work regarding the development of new family behaviors in the Nordic countries is Gøsta Esping-Andersens (1990; 1999) theory on welfare state regimes. He underlines the role of changing institutions when trying to explain different demographic behaviors. He argues that Nordic countries have become more de-familialized than other developed societies. Nordic citizens can rely on tax financed universalistic social security systems. They are therefore more independent of family ties for social and economic security.

David Reher´s important work on different family systems identifies “weak family” systems in Northern Europe and “strong family” systems in Southern Europe (Reher, 1998). Archetypical examples of the former are the Nordic countries characterized by early home leaving, late and non-universal marriages, more symmetrical power relations between parents and children and neolocal household formations. In comparison, “strong-family” systems, where archetypical examples are Spain Italy and Greece, are characterized by the opposite: Leaving home late, nearly universal marriages, strong influence of parents over adult children’s family formation and multigenerational households.

A key precondition for the reduced dependence on the family is the commodification of women’s labor. This moves reproductive services out from the family and into the market. Weak family countries like Sweden tend to have had earlier and much more extensive expansion of female labor force participation in comparison with strong family countries (Reher 1998; Hajnal 1982). Esping-Andersen argues that a specific trait of the Nordic welfare states is that policies actively worked for the integration of women into the labor market. That strongly contributed to a reduced interdependence between family members.

Mass employment among unmarried women came quite early in Sweden, already during the period 1920-1940 (Stanfors 2003). These changes in the economic roles of women were accompanied by significant changes in family behavior, with the first major upshift in divorce rates. Divorces spread from the upper social strata to the lower (Sandström 2012). Fertility patterns changed as family formations and fertility increased sharply among tertiary educated women who, up until the 1940s and 50s, mostly had remained childless (Bavel et al. 2018; Sandström and Marklund 2018). Divorce rates amplified even more in the 1960s, at the same time as marriages were increasingly replaced by cohabitation as the primary form of union (Sandström 2012; Trost 1975).

Apart from amplified divorce rates and changing fertility patterns, the censuses in the 1940s and 1950s recognized that one-person households were increasing distinctly during this period (Statistics Sweden 1950, 1955). In the latter census it was also stated that women were disproportionately driving the increase of one-person households (Statistics Sweden 1955). However, we do not know what characterized these women, like their age, occupation, income and vulnerability.  Therefore, it is difficult to assess how this trend was related to societal changes, such as industrialization, urbanization, labor force participation and the expansion of the Swedish welfare state.

Arguably, the growth of one-person households is a key expression of the increased family complexity that is developing in our modern societies. It represents a fundamental societal transformation related to changes in the gender regime and in the structures and functions of the family. Family complexity has in recent decades given raise to new concepts such as “rainbow families”, “yours, mine and our children”, “single parenthood”, “living apart together”, etc. Recent qualitative studies, based on interviews in Sweden, also talks about an increased single culture were living without a partner is actively chosen and preferred as a way of safeguarding personal freedom (Henriksson 2014). The growing proportion of one-person households can thus be seen as one of the most important indicators of how the social, economic and cultural functions of the family have changed over time. Consequently, single living has sparked a growing interest among scholars trying to understand the recent transformations in family dynamics (Reher, 2018; Klinenberg 2012; Lesthaeghe 2010; Fokkema and Liefbroer 2008; Demey et al. 2013; Falkingham et al. 2012; Umberson and Montez 2010).

The SES structure of one-person households is also changing. At least since the 1980s we know that living in a one-person household in adult working age is more common among individuals with low SES (Sandstedt 1991). But there are indications that this tendency, with a negative association between single living and SES, is now being even further reinforced. Becoming a parent, getting married and not experiencing divorce is more and more associated with high levels of education, stable employment and high income (Boschini and Sundström 2018; Esping-Andersen 2016; Esping-Andersen and Billari 2015; Goldscheider et al. 2015). The extent to which single living was more common among individuals with a low SES position prior to the 1980s is, however, an open question.

Apart from the theoretical reasons that stress Sweden's status as a “forerunner” towards increased family complexity, it is important to study one-person households due to its connection with increased vulnerability. It has a profound impact on health, with higher mortality and higher disease load (Fritzell et al. 2007; Ringbäck Weitoft et al. 2004; Weitoft et al. 2003; Whitehead et al. 2000). Access to social support networks and personal relationships are essential to both subjective and objective wellbeing of adult individuals (for reviews see House et al. 1988; Umberson and Montez 2010). Those living in a one-person households in general suffer from poorer health and have more extensive needs of care (Stenflo 2006). They also run higher risks of substance abuse and mental health problems and are known to have a higher risk of cardiovascular and alcohol-related causes of death (see e.g. Koskinen et al. 2007; Pulkki-Råback et al. 2012; Scott et al. 2010).

Compared to men, women are often found to have less negative health effects of living in one-person households. Better access to social support and social networks are suggested as vital explanations (Hanson et al. 1989; Koskinen et al. 2007). However, these findings on health inequalities between cohabiting and one-person households are exclusively based on research on contemporary societies. Again, we know very little whether health disparities between these groups are a recent phenomenon, or how such inequalities may have developed in a longer time perspective.

Research design

We will do an in-depth statistical analysis of the demographic and socioeconomic characteristics of the individuals which have contributed to the increase in single living over time. The association between these personal characteristics and the probability to be a one-person household will give us indications on the kind of societal changes that have contributed to the increased share of the population living in one person households over time. The project will analyze how living as a one-person household is associated with economic and physical wellbeing in a long-term perspective. Identifying these changes over time allow us to better understand the changing role of the family for individual wellbeing and how this is linked to economic, cultural and institutional modernization.

Q1: How has the demographic characteristics of the one-person households changed during the twentieth century?

By utilizing the censuses between the years 1900-1990 in combination with register data for the period up until 2017, we will provide a detailed descriptive account of the development of the demographic and socioeconomic characteristics of one-person households. These qualities will be contrasted against individuals living in other forms of living arrangements (nuclear families with or without children, single parents and other more complex households). The most important dimensions for comparison across time will be differences according to age, income, education, occupation and labor market attachment, sex/gender, civil status, fertility history, household dynamics, and urban-rural context. The descriptive account is methodologically straightforward given the new microdata available to us through the SWEDPOP-project. Determining the development of the demographic characteristics of one-person households relative to those cohabiting is a necessary first step. It provides us with the key background information for our analyses of socioeconomic and health differentials. Potential changes over time will provide important input for answering the question why the share of one-person households have grown at certain points in time.

Q2: How has socio economic inequalities between individuals living in one-person households and those cohabiting with others developed since the early twentieth century up until today?

The changes in SES characteristics of one-person households compared to those in different forms of cohabitation are important as it provide information on how economic changes have contributed to changing living arrangements. But, information on SES differentials also gives us information on the living conditions and relative vulnerability of one-person households in different time periods. The most important indicators available for us are labor market attachment, occupation, income and education for one-person households, relative to individuals in other living arrangements. The information on occupation is the most consistently recorded information already from 1900 and can be used to stratify individuals into socioeconomic groups. However, in the census of 1930 we also have access to income, wealth and education. From 1960 this information is present in all the censuses as well as in the annual register data from Statistics Sweden. A detailed account of how we plan to use this information is found in the Empirical Strategy section below. Apart from conventional descriptive methods the analysis will utilize generalized linear probability models to estimate how the probability to be a one-person household have been associated with socioeconomic resources for the entire period 1900-2017.

Q3: How has health inequality between individuals living in one-person households and those that cohabit developed since the early twentieth century until today?

In order to address the question of how health inequalities between one-person households and individuals living in multi-person living arrangements, we will primarily use the straightforward outcome of differences in mortality risks. With the combined information on living arrangements taken from the censuses and longitudinal mortality data from the Swedish Death Index (Sveriges Dödbok 7, 1901-2017), we can use survival analysis to determine differences in the hazard of death for one-person households at different points in time. With this analysis, we can determine how living arrangements have been associated with health differentials, net of other influences such as period effects and the demographic and socioeconomic characteristics described in Q1.

Empirical strategy

There have been very limited possibilities to study one-person households prior to 1960, as computerized population registers did not exist for the censuses before 1960. The only available information was limited to aggregate statistics on the share of one-person households, making a project like this almost impossible. However, the empirical landscape has very recently changed. Historical censuses are now available as computerized microdata. The SWEDPOP consortium — ­­The Demographic Data Base (Umeå University), The Scanian Economic Demographic Database (Lund University), SweCens (National Archives, Stockholm), The Roteman archives (City archive of Stockholm) and the Gothenburg Population Panel (Gothenburg University) — links historical censuses to other individual level longitudinal datasets (https://www.swedpop.se). These new and unique set of sources enable us to study if and consequently how and why single living has transformed over time and also how it has influenced people's material and physical wellbeing.

National Censuses 1900-1990 - Demographic, socioeconomic and household information
The population and housing censuses that was conducted with 5-10-year intervals will be used to determine the prevalence of one-person households. We can also establish changes in the demographic and socioeconomic characteristics of one-person households compared to those that cohabited. As individuals are linked across censuses, we can determine changes in living arrangements, as long as the person is not lost to death or out-migration from the country. The censuses are also linked to other historical population data sources including the Swedish Death Index, which means that we know when a person died, even if it happened between censuses. It will allow us to analyze changes in the composition of households and understand which kind of individuals that lived alone during the entire twentieth century.

The 1880-1910 censuses are available in the SweCens-database as well as through the North Atlantic Population Project (https://www.nappdata.org/napp/). The 1930 census is particularly useful as it also includes income, wealth and education for the entire population. The complete registration of this census is going to be finished in 2019. The 1940 census public release is scheduled for the spring of 2020. SWEDPOP will also disseminate the 1950 census in late 2019, already digitized by Arkiv Digital.

The Swedish Death Index 1901-2017- Longitudinal information on mortality
We will use all-cause mortality from the Swedish Death Index as the main health outcome for one-person versus different types of cohabiting households. It contains the names, marital statuses, place and dates of births and deaths of all deceased persons in Sweden between 1860 and 2016. The Death Index has already been successfully linked to the 1880-1910 censuses, both in terms of linkage rates and quality (Bengtsson et al. 2018). The SWEDPOP project is now replicating this process for the 1930, 1940 and 1950 censuses and integrating the information into the infrastructure.

Longitudinal population databases c. 1900-1960 (POPLINK / SEDD / Rotemansarkivet)
Censuses with 10-year intervals are not fully longitudinal, even if they are linked to each other and combined with death registers. In the last stage of our project we will use fully longitudinal demographic sources to be able to, in much more detail, take household dynamics into consideration. Such data only exists for certain geographical areas in Sweden. Three important longitudinal data sources will be the POPLINK-database covering the Skellefteå region and Umeå region for the period 1900-1960 (https://www.umu.se/enheten-for-demografi-och-aldrandeforskning/databaser-sokverktyg/kyrkoboksdatabaser/) and the Scanian Economic Demography Database (SEDD) with nine rural parishes and the city of Landskrona, 1900-2011 (https://www.ed.lu.se/databases/sedd). The Roteman database at Stockholm City Archives offers linked individual-level data on the Stockholm population from 1878-1926 (https://stadsarkivet.stockholm.se/hitta-i-arkiven/i-arkiven/rotemansarkivet/). Besides capturing household dynamics, these sources allow us to do detailed analyses of differences in causes of deaths, net of other socioeconomic and demographic characteristics.

Census and register data from Statistics Sweden 1960-2017, The Aging well database
For the period 1960-1990 the censuses are obtainable in computerized form from Statistics Sweden. This data is already accessible to us through our own Aging Well database. For the period after 1960 we are therefore no longer limited by having to rely on mortality as a proxy for health. From that year information on morbidity (cause specific hospital care) and cause-specific mortality allows for a more in-depth analysis of the health disparities between cohabiting and one-person households. Since 1990, Statistics Sweden no longer issues traditional censuses. Instead they publish longitudinal and annual register-based population data. The Dwelling register, which provides detailed and continuously updated information on housing and apartment arrangements (Statistics Sweden 2013), enables even more detailed analyses of household dynamics than what was previously possible. This information is also already available for us in the Aging Well database.

Methods

Demographic and socioeconomic determinants of single living
Our exposure is straightforward in terms of the individual not cohabitating with anyone but living separately in his/her household. This state is contrasted against other forms of living arrangements: nuclear families with or without children, single parents and other more complex households. Given that our data consists of both cross-sectional and longitudinal population data our analytical approach will employ a range of quantitative demographic techniques. First, standard descriptive methods will be used to ascertain how individuals living in one-person households have differed from other types of multi-person households in terms of demographic and socioeconomic characteristics, from 1900-2017. Second, we will use generalized linear probability models (i.e. logistic regression) to estimate the relative importance of these demographic and socioeconomic determinants for the probability to be a one-person household.

It is important to note that there are some changes in the definition of a household in the censuses during this long time period. The most important change is the shift from defining the household as “the individuals living and sharing the main meal of the day” employed until 1945 and the modern definition of a “dwelling household” (Statistics Sweden 1950, p. 21, 1955, p. 58). Tenants or servants living in the same dwelling was counted as separate households up until 1945, although they lived in the same building/dwelling as the other household units. In the modern definition from 1960, they are included in the household if they reside in the same dwelling. This subtle difference in the definition needs to be kept in mind when interpreting changes in the composition of one-person households. Methodologically the change in the classifications does not pose any serious obstacle for our research agenda for two reasons. First, the overlap of the two definitions in the census of 1950 makes it straightforward to assess the potential influence it might have on estimates and descriptives. Second, the practice to have tenants and servants living in the same dwelling was relatively common only in the prewar era. It rapidly became a fringe phenomenon as living standards and housing supply improved after World War II. This was also the motivation for Statistics Sweden to definitively switch to the modern definition in the census of 1960.

The main socio-economic information available in the censuses prior to 1960 is the occupation of all household members. This information will be used to classify individuals into socio-economic groups by coding the occupational information according to the HISCO classification system for historical occupations (Van Leeuwen et al. 2004). These harmonized occupational codes will then be used to assign the individual a social stratification indicator (SES) according to the Social Power coding scheme (SOCPO) developed by Vande Putte & Miles (2005). These methods for comparison of socioeconomic factors have already been successfully used by the main applicant in studying changes in fertility and nuptiality in Sweden and other European countries (Reher et al. 2017; Sandström 2017; Sandström and Marklund 2018). In addition to these occupations, we will use the extended information in the census of 1930 which includes income and education for all household members. This allows for an in-depth analysis of socioeconomic differentials between individuals in one-person households and those in other types of households. This can then be compared with the situation in 1960, when this information is available once again, in order to identify potential changes in the association between education and income since the 1930s. For the period 1970-2017 we can use income and education based on the register-based information available in the Aging Well database and we do not need to rely on occupations as a basis for stratification.

Health and living arrangements
The data provided by SWEDPOP allows us to determine the living arrangements of individuals with a 10-year interval from 1900 and onwards. With the linked information from the Swedish Death Index we can analyze how their living arrangements are associated with survival, net of the demographic and socioeconomic factors. As we have the date of death, we will use standard survival analysis techniques such as Kaplan-Meier survival estimates and Cox-regression (Cox 1972). To the extent that we encounter non-proportional effects that cannot be easily addressed with e.g. stratified analysis, flexible parametric Royston-Parmar models will be utilized to describe effects that vary with analysis time (Royston and Lambert 2011).

Time plan

2020: Getting ethical permissions and all necessary datasets. Commencing descriptive analyses of the censuses 1900-1950 and the Swedish Death Index.

2021: Continued analyses of the census data and register data, now including 1960-2017. Longitudinal analyses of censuses for the full period. Writing key conference papers.

2022: Longitudinal analyses data from the POPLINK-database covering Skellefteå and Umeå for the period 1900-1960, the SEDD database, 1900-2011 and the Roteman database at Stockholm City Archives, 1878-1926. These databases offer individual demographic, socio-economic and cause-specific mortality data that add details not available in the censuses alone. Writing and submitting articles to key international journals.

2023: Final analyses of census data and individual longitudinal data. Writing and submitting final articles. Publishing our anonymized data in open access.

Target journals for publications are Population Studies, Demography, Demographic Research, Population and Development Review, The History of the Family and Historical Life Course Studies. We plan to submit three conference papers 2020 and at least two articles per year, 2021-2023.

Research team

Project leader, Associate Professor Glenn Sandström, is an historian and a demographer. Sandström has extensive experience and expertise in quantitative research on the history of family behavior, focusing especially the development of divorce, nuptiality and fertility in Sweden and Europe from the late 19th century up until the present. Sandström has also published several studies on living arrangements in contemporary Europe and Sweden (Sandström et al. 2019; Sandström & Karlsson, 2019; Padyab et al. 2019) and is well established within the field of family demography.

Anders Brändström is full professor in historical demography and has done extensive research in the secular development of mortality and public health in Sweden. Brändström has played a key role in the establishment and expansion of the Demographic Data Base at Umeå University, as its director 1998-2015. He also founded and led the Center for Population Studies until 2015.

Therefore, both applicants have well-documented expertise in working with large and complex datasets, as well as big relational databases. Both applicants are also well experienced in the quantitative methods required to properly leverage the unique data and the analyses required to complete this project within its time frame.

References

Publications in bold are authored/co-authored by the main applicant

Van Bavel, et al. (2018). Seeding the Gender Revolution: Women’s Education and Cohort Fertility among the Baby Boom Generations. Population Studies 72(3): 283–304.

Bengtsson, T., Dribe, M., & Eriksson, B. (2018). Social Class and Excess Mortality in Sweden During the 1918 Influenza Pandemic. American Journal of Epidemiology, 187(12), 2568–2576.

Cox, D. R. (1972). Regression Models and Life-Tables. Journal of the Royal Statistical Society. Series B (Methodological), 34(2), 187–220.

Demey, D., et al. (2013). Pathways into living alone in mid-life: Diversity and policy implications. Advances in Life Course Research, 18(3), 161–174.

Edliden, A. (2018, March 29). Ingen ensamhetsminister i Sverige. https://www.svt.se/nyheter/inrikes/annu-inga-politiska-forslag-for-att-bryta-ensamhet.

Esping-Andersen, Gøsta. (1999). Social foundations of postindustrial economies. New York: Oxford University Press; (2016). Families in the 21st century. Stockholm: SNS Förlag; Esping-Andersen, Gøsta, & Billari, F. C. (2015). Re-theorizing Family Demographics. Population and Development Review, 41(1), 1–31.

Eurostat. (2019). Distribution of households by household type from 2003 onwards - EU-SILC survey. Income and living conditions (ilc). http://appsso.eurostat.ec.europa.eu

Falkingham, J., et al. (2012). The Demography of Living Alone in Mid-Life: A Typology of Solo-Living in the United Kingdom. The 2012 European Population Conference, Stockholm.

Fokkema, T., & Liefbroer, A. C. (2008). Trends in living arrangements in Europe: Convergence or divergence? Demographic Research, 19, 1351–1418.

Fritzell, S. et.al. (2007). From macro to micro: The health of Swedish lone mothers during changing economic and social circumstances. Social Science & Medicine, 65(12), 2474–2488.

Goldscheider, F., Bernhardt, E., & Lappegård, T. (2015). The Gender Revolution: A Framework for Understanding Changing Family and Demographic Behavior. Population and Development Review, 41(2), 207–239.

Hajnal, J. (1982). Two Kinds of Preindustrial Household Formation System. Population and Development Review, 8(3), 449–494.

Hanson, B., et.al. (1989). Social network and social support influence mortality in elderly men. The prospective population study of “Men born in 1914,” Malmö, Sweden. American Journal of Epidemiology, 130(1), 100–111.

Henriksson, A. (2014). Organising intimacy : exploring heterosexual singledoms at Swedish singles activities. Karlstad: Karlstads universitet.

House, J., et.al. (1988). Social relationships and health. Science, 241(4865), 540–545.  

Höjer, H. (2014). Ensamboendet ökar i hela världen och Sverige ligger i topp. Forskning och framsteg, (10), 42–45.

Iacovou, M., & Skew, A. J. (2011). Household composition across the new Europe: Where do the new Member States fit in? Demographic Research, 25, 465–490.

Klinenberg, E. (2012). Going Solo: The Extraordinary Rise and Surprising Appeal of Living Alone. Penguin Press.

Koskinen, S., et.al. (2007). Mortality differences according to living arrangements. International Journal of Epidemiology, 36(6), 1255–1264.

Lesthaeghe, R. (2010). The Unfolding Story of the Second Demographic Transition. Population and Development Review, 36(2), 211–251.

Padyab, M., Reher, D., Requena, M., & Sandström, G. (2019). Going It Alone in Later Life: A Comparative Analysis of Elderly Women in Sweden and Spain. Journal of Family Issues, online.

Popenoe, D. (1987). Beyond the Nuclear Family: A Statistical Portrait of the Changing Family in Sweden. Journal of Marriage and Family, 49(1), 173–183.

Pulkki-Råback, L., et al. (2012). Living alone and antidepressant medication use: a prospective study in a working-age population. BMC Public Health, 12, 236.

Putte, B. V. D., & Miles, A. (2005). A Social Classification Scheme for Historical Occupational Data. Historical Methods: A Journal of Quantitative and Interdisciplinary History, 38(2), 61–94.

Reher, D. (1998). Family ties in Western Europe: persistent contrasts. Population and development review, 203–234.

Reher, D., et al. (2017). Agency in Fertility Decisions in Western Europe During the Demographic Transition: A Comparative Perspective. Demography, 54(1), 3–22.

Reher, D., & Requena, M. (2018). Living Alone in Later Life: A Global Perspective. Population and Development Review, 44(3), 427–454.

Ringbäck Weitoft, G., Burström, B., & Rosén, M. (2004). Premature mortality among lone fathers and childless men. Social Science & Medicine, 59(7), 1449–1459.

Royston, P., & Lambert, P. C. (2011). Flexible parametric survival analysis using Stata : beyond the Cox model. College Station, TX: Stata Press.

Sandstedt, E. (1991). Att bo ensam: om enboendeliv i Sverige. Stockholm : Statens råd för byggnadsforskning.

Sandström, G. (2012). Ready, Willing and Able: The Divorce Transition in Sweden 1915-1974 (dissertation). Umeå University, Umeå.

Sandström, G. (2017). A reversal of the socioeconomic gradient of nuptiality during the Swedish mid-20th-century baby boom. Demographic Research, 37(50), 1625–1658.

Sandström, G., & Marklund, E. (2018). A prelude to the dual provider family – The changing role of female labor force participation and occupational field on fertility outcomes during the baby boom in Sweden 1900–60. The History of the Family, online.

Sandström, G., et al. (2019). Persistent High Levels of Single Living Among Adults with Disabilities in Sweden, 1993–2011. Presented at the Annual Meeting of the Population Association of America, Austin Texas.

Sandström, G., & Karlsson, L. (2019). The Educational Gradient of Living Alone: A Comparison Among the Working-Age Population in Europe. Presented at the Annual Meeting of the Population Association of America, Austin Texas.

Scott, K. M., et al. (2010). Gender and the relationship between marital status and first onset of mood, anxiety and substance use disorders. Psychological Medicine, 40(9), 1495–1505.

Stanfors, M. (2003). Education, labor force participation and changing fertility patterns : a study of women and socioeconomic change in twentieth century Sweden. Stockholm: Almqvist & Wiksell.

Starbrink, A. (2018, October 9). Sverige behöver tillsätta en ensamhetsminister. svd.se.

Statistics Sweden. (1950). Folkräkningen den 31 december 1945. Del 7; (1955). Folkräkningen den 31 december 1950. Del 6; (1969). Historisk statistik för Sverige. Del 1; (1992). Population and Housing Census 1990; (2013). Lägenhetsregistret- Efterlängtad statistik. Källa (3), 10–13; (2019), SCB Stockholm; Antal personer efter hushållsställning, personernas bakgrund och kön. År 2011 - 2017.  http://www.statistikdatabasen.scb.se.

Stenflo, G. A. (2006). Äldres omsorgsbehov och närhet till anhöriga. Stockholm: SCB.

Trost, J. (1975). Married and Unmarried Cohabitation: The Case of Sweden, with Some Comparisons. Journal of Marriage and the Family, 37(3), 677–682.

Umberson, D., & Montez, J. K. (2010). Social Relationships and Health A Flashpoint for Health Policy. Journal of Health and Social Behavior, 51(1), 54–66.

Van Leeuwen, M. H. D., Maas, I., & Miles, A. (2004). Creating a Historical International Standard Classification of Occupations. An Exercise in Multinational Interdisciplinary Cooperation. Historical Methods: A Journal of Quantitative and Interdisciplinary History, 37(4), 186–197.

Weitoft, G. et.al. (2003). Mortality, severe morbidity, and injury in children living with single parents in Sweden: a population-based study. The Lancet, 361(9354), 289–295.

Whitehead, M., Burström, B., & Diderichsen, F. (2000). Social policies and the pathways to inequalities in health: a comparative analysis of lone mothers in Britain and Sweden. Social Science & Medicine, 50(2), 255–270.