Living arrangements can reveal much about a society, as they both reflect and determine wider societal trends. Throughout modern history, two broad forces have restructured the population composition of living arrangements: demographic change, which determines who is alive at the same time to potentially live with; and social change, which encompasses cultural, political and macroeconomic forces that shape preferences and needs for living apart or together.

Expected years of life by living arrangements is a useful measure for summarising historical change. Based on life expectancy at birth, it indicates the number of years a fictitious cohort would live if they experienced the same mortality and the same living arrangement profiles as those prevalent in the population in a given year. It provides a means to capture change over time and break it down into the contribution of different age groups. In our recent study (Floridi & Esteve, 2026), we summarise life expectancy at birth by living arrangement over 170 years with available census and register data (1850–2021) and complement this by looking at changes across cohorts of people born in the same decade. We focus on the United States, a country that has seen a doubling of life expectancy at birth since 1850 while undergoing radical economic, policy, and cultural shifts.

Three historical ages of living arrangements

Figure 1 shows life expectancy at birth for available census or population survey years between 1850 and 2021. Expected years of life are distributed across three broad categories and seven sub-categories: living alone (alone); living with primary kin (couple only, two-parent, and single-parent households); and living in an extended household (kin only, mixed, or nonkin only).

Based on this cross-sectional look and complementary analyses, we identify three historical ages of living arrangements in the United States.

The first is a large household system (1850–1940). During this phase, years spent living alone represent a negligible share of population life years, which are evenly split between primary kin and extended households. Primary kin living absorbs most of the increases in life expectancy over the period, which are largely attributable to improved infant survival. Years spent in extended households look stable over the period, but this masks substantial heterogeneity: extended kin households continue to expand until 1940, while mixed households (kin and nonkin) fall as a result of the decline in live-in servitude and lodgers. During this phase, household types where children are born into primary kin and extended kin households grow in line with life expectancy, with mortality decline among infants and children being the main driver of changes in living arrangements.

A second age of primary kin dominance follows (1940–1980). Living alone rises quickly, especially among older females, which can be attributed to women’s survival after adult children’s exit from the household and the death of a male partner. Years spent with primary kin increase faster than life expectancy, and for all age groups. Meanwhile, extended kin-only households decline for the first time. This period is marked by rapid social change, as all age groups shift from extended to primary-kin (or solo-living) arrangements, with changes outpacing mortality decline. It is interesting to note that while this represents a brief phase of rapid social change during the post-war period, it has led to so-called “nuclear” households being widely regarded as traditional and normative in the United States (Nicholson, 2016). 

1980 ushers in a new ongoing phase of diversification. We observe increases in living alone among younger age groups and males, with a stabilisation (and slight decline) in living alone for older females. Within the primary kin category, two-parent households with children decline in prevalence, while the share of couples without children, and single-parent households increases. Living in extended families returns to 1940s levels, with the rise concentrated among children and young adults. This has been attributed to racial/ethnic diversification of the population (Pilkauskas et al., 2020) and to economic constraints on economic independence (Ruggles, 2007). Again, it is worth emphasising that extended family households are common throughout the observation period, except for the relatively brief era of primary kin dominance.

Change across cohorts and future prospects

To complement our cross-sectional analysis, we select cohorts born in 1860–69, 1900–09, and 1940–49, and plot their living arrangements by age in Figure 2. 

The 1860s cohort is emblematic of the diversity in living arrangements that characterised the second part of the 19^th century. In the later cohorts, we observe a sharp decline in extended households, through the near disappearance of mixed households and the decline in extended family arrangements that characterised the early and mid-20^th century. The 1900s cohort (aged around 40 in 1940) displays the initial shift towards primary kin living, and a marked increase in living alone at older ages, especially among women, due to high male mortality in this cohort. The 1940s cohort reached age 20 during the “baby boom” of the 1960s. Life courses in this cohort are structured around primary kin living: most individuals go from living as a child with two parents (until age 20), to living as a parent with children (ages 20–45), followed by couple-only arrangements once children move out (45 onwards). Again, women are more likely to spend their older ages living alone than men, due to differential survival.

We cannot provide a picture for the birth cohorts who have not yet reached later life. However, our period analyses (Figure 1) suggest that things will look different for those born in the 1960s, 1980s, and beyond. As the United States enters a diversified era of living arrangements, successive cohorts will experience more years spent living alone and in extended arrangements, including with nonkin. Within primary kin households, couples without children will rise in prevalence and may become the dominant category, while single parenthood rates will also increase. More recent birth cohorts are thus likely to experience the decline in two-parent households that have misguidedly been represented as the historical norm in the United States.  

References

Floridi, G., & Esteve, A. (2026). 170 years of change in living arrangements in the United States using expected years of life: A research note. Demography. https://doi.org/10.1215/00703370-12546633 

Nicholson, L. (1997). The myth of the traditional family. In Lindemann, H. Feminism and families (pp. 27-41). Routledge. https://doi.org/10.4324/9780203760413 

Pilkauskas, N. V., Amorim, M., & Dunifon, R. E. (2020). Historical trends in children living in multigenerational households in the United States: 1870–2018. Demography, 57, 2269–2296. https://doi.org/10.1007/s13524-020-00920-5 

Ruggles, S. (2007). The decline of intergenerational coresidence in the United States, 1850 to 2000. American Sociological Review, 72, 964–989. https://doi.org/10.1177/000312240707200606 

The post Change is the only constant: living arrangements in the United States, 1850–2021 appeared first on N-IUSSP.

What are life expectancy and lifespan inequality?

Life expectancy at birth is a measure of the mean level of mortality. Simply put, it expresses the average number of years a newborn is expected to live given the mortality conditions at a point in time. Because it is an average, life expectancy does not show differences in length of life among individuals; these differences can be substantial, however. To fully understand mortality at the population level, life expectancy can be complemented by lifespan inequality, which measures variation in ages at death, capturing differences in lifespans. Lifespan inequality has been described as the most fundamental of all inequalities, because every other type of inequality is conditional upon being alive (van Raalte, Sasson, & Martikainen, 2018). Its level and evolution can tell us much about how equitably health improvements are allocated in a society.

At the individual level, lifespan inequality gauges the unpredictability of the timing of death. For example, using the standard deviation of ages at death (SD) as a measure of lifespan inequality, it was considerably more difficult to predict how long a newborn in Italy would live in 1920 (SD = 31.6 years) than it was in 2017 (SD = 12.4 years) because in 1920 deaths were more widely spread across different ages (Figure 1).

How does life expectancy relate to lifespan inequality?

In 1920, 13.3% of females born in Italy died before their first birthday and less than 10% of them were expected to reach the age of 82. In contrast, 99.7% of female babies born in 2017 survived to their first birthday and almost 70% of them are expected to be alive in 2100 (Figure 1). This exceptional progress in reducing mortality has raised life expectancy at birth from 46.4 years to almost 85 years over the last century. At the same time, life chances have become more equitable, i.e. lifespan inequality has decreased (Figure 2).

Higher levels of life expectancy usually correspond to lower levels of lifespan inequality (Figure 2 and e.g. Smits and Monden 2009). However, historically, this has not always been the case. In certain countries, lifespan inequality has been shown to stall or even increase during specific periods despite improvements in life expectancy, or vice versa. In Venezuela, for instance, between 1996 and 2013 life expectancy increased (from 69.4 to 71 years), but lifespan inequality also increased (SD from 20.7 to 21.7 years) because progress in reducing mortality at almost all ages, including infant mortality, was counteracted by a rise in violent deaths concentrated among young males (García & Aburto, 2019).

A recent article shows that changes in both life expectancy and lifespan inequality can be expressed as rates of progress in reducing mortality (Aburto et al 2020). For example, in 1920’s Italy, the biggest gains in life expectancy were achieved if progress was concentrated in infancy. By contrast, in today’s Italy, reducing infant death rates by 5% leads to the same gain in life expectancy as reducing mortality at age 79 by 5%. Lifespan inequality is reduced significantly if lives are saved at infancy but increases if lives are saved at older ages (Figure 3). This has potential implications for the relationship between life expectancy and lifespan inequality. In contemporary societies, where rising life expectancy is often attributable to improvements in old-age mortality, increases in lifespan inequality may become more frequent.

Societal implications of increased lifespan inequality

Lifespan inequality is important because people making major life decisions are influenced by the mortality experience of those around them. Subjective beliefs about one’s lifespan have been shown to affect decisions such as whether and when to invest in education, save, buy a house, or retire (e.g. Bloom et al., 2006).

A potential weakening or reversal of the negative relationship between lifespan inequality and life expectancy means that, while people live longer on average, individuals also face greater uncertainty about when they will die. Studies have shown that such increased uncertainty within a society is likely to be concentrated among those from disadvantaged socio-economic groups. For example, in Finland, between 1971 and 2014, lifespan inequality increased among subgroups with low educational attainment and low income but decreased among more advantaged groups (van Raalte et al., 2018). This represents a “double burden” for lower socio-economic groups experiencing not only shorter lifespans but also higher uncertainty compared to the more advantaged groups.

Diverging patterns in lifespan inequality across socioeconomic groups may have important consequences for the perpetuation of social inequalities. For instance, faced with greater uncertainty about their own lifespan, individuals from disadvantaged socio-economic backgrounds risk making unfavorable financial decisions that might further reduce their financial resources later in life. Lifespan inequality is also likely to create uncertainty about the timing of inheritances and financial bequests among family members. If this uncertainty is concentrated among lower socio-economic groups, poorer individuals may not only receive smaller transfers from their parents and grandparents than the rich; they may also be less able to anticipate the timing of those transfers. Given the importance of these transfers for the recipients’ socio-economic conditions, higher lifespan inequality may strengthen the transmission of disadvantage across generations, reducing intergenerational social mobility.

The significance of including lifespan inequality in policy and research

Increasing the average age at death alone is not enough to guarantee a more equitable distribution of the length of life. An important message of recent research is that policies and interventions can affect this relationship by reducing mortality at ages that matter the most (Aburto et al., 2020). In terms of policy interventions, one key question to ask is how different types of healthcare investments, such as preventive as opposed to end-of-life care, may relate to inequality in lifespans. It is also important to understand how the strength of the relationship between life expectancy and lifespan inequality varies across societies and periods characterized by different levels of social inequality. There is reason to believe that more equal or socially cohesive societies may be more effective at reducing lifespan inequality, for example by reducing violent crime or providing better access to preventive healthcare across all socio-economic groups. These are promising avenues for future research linking demographic phenomena to the broader societal context in which they take place, with important implications for social policy.

*The authors thank Jim Oeppen for thoughtful comments and suggestions.

References

Aburto, J. M., Villavicencio, F., Basellini, U., Kjærgaard, S., & Vaupel, J. W. (2020). Dynamics of life expectancy and life span equality. Proceedings of the National Academy of Sciences.

Bloom, D. E., Canning, D., Moore, M., & Song, Y. (2006). The effect of subjective survival        probabilities on retirement and wealth in the United States. NBER Working Paper, 12688.

García, J., & Aburto, J. M. (2019). The impact of violence on Venezuelan life expectancy and lifespan inequality. International journal of epidemiology.

Smits, J., & Monden, C. (2009). Length of life inequality around the globe. Social Science & Medicine, 68(6), 1114-1123.

van Raalte, A. A., Sasson, I., & Martikainen, P. (2018). The case for monitoring life-span inequality. Science, 362(6418), 1002-1004.

The post On the (changing) relationship between life expectancy and lifespan inequality appeared first on N-IUSSP.