‘Stay home’: age and gender frames and limits of residential family solidarities in the times of lockdown (a French example)

In France, the lockdown started on 17 March 2020 and finished on 11 May 2020. The residents were invited to stay at their homes, as the President said in his speech of 16 March: ‘While staying at home, take care of your close ones, staying in your apartment or your house. Give news, ask for news’.¹ On the one hand, this speech represents family as a group of people living under the same roof and a group of people sharing a household are alluded to as ‘close ones’. However, family and household are not necessarily the same thing and taking one or the other as measurement unit influences public policy (Bauman 1999; Hill 1995). Besides, the social changes of the past decades have led ‘many people, children especially, to have two (or more) homes’ which is also the case of ‘highly mobile professionals’ (Coast et al. 2016: 168).

On the other hand, we find here a strong representation of family, prevailing within French family policies since the 1980s as well as in sociological studies, depicting family as a shelter in case of difficulties (Martin 2002). The first research in social sciences did insist on the integrating role of the family institution (Durkheim 1951 (1897)) and it remains a source of ‘solidarity’ in contemporary world (Bengtson 2001; Martial and Segalen 2019; Silverstein and Bengtson 1997). However, families invest into their members in a very unequal way (Albertini and Radl 2012; Albertini and Kohli 2013). Solidarity with one's relatives depends on economic and cultural levels (Albertini and Radl 2012; Silverstein and Bengtson 1997; Laé and Murard 2011), as well as family structure and personal relationships (Petite 2005; Silverstein and Bengtson 1997). European, and especially French sociological researches focus on services, and financial and care exchanges between households (Börsch-Supan et al. 2013; Heady 2012; Fontaine, Gramain and Wittwer 2007; Renaut 2003). But they seldom question the circulation of individuals between residences within the kinship.

The corona crisis challenges family sociologists. It raises the crucial question of the role of family social integration in uncertain times. It also challenges social and spatial dimension of a household: with whom were people confined? Who stayed alone? Who hosted whom? We will thus address the ways families faced the lockdown in France through these two questions – housing and inequalities in being together. The shutdown questions precisely with whom individuals lived during this exceptional period showing what kind of kin relationships can be activated (Firth 1956) in case of crisis. It also reveals social factors that impact solidarities, especially age and gender. Indeed, childhood, youth and old age are seen by sociologists as the stages of life when people receive more help from relatives (Bengtson 2001; Silverstein and Bengtson 1997; Van de Velde 2007). Gender is also one of the principle analyzers of family relationships as researches underline a matrifocal orientation of European families (Déchaux 2009; Silverstein and Bengtson 1997). Thus, we will explore how these two factors are at play in residential lockdown configurations. The paper will present the first results of an on-line survey activated from 15 April to 10 May 2020 in France.

We carried out our research through an on-line survey that questioned the composition of the household, kinship and care relationships before and during the lockdown. It was constructed using Lime-survey platform, and consisted of 49 closed and 5 open-ended questions organized into 4 themes: socio-demographic data, residence before the confinement, residence during the shutdown, and care received and given during the lockdown. The questionnaire was diffused mainly through university lists and social media. The 4300 exploitable answers were processed by SPSS software. The sample was a volunteer one, so not representative. However, this is partly compensated by the numerical importance of the sample, besides, it gives a high diversity of situations as for ages (up to 88), education level and households before and during the lockdown. However, the type of dissemination and the subject of the questionnaire resulted in several constrains. First of all, women represent ¾ of the sample while they constitute a bit more than a half of the French population (INSEE 2019). This can be explained by the family topic of the research in which women are usually more interested than men (Slauson-Blevins, and Johnson 2016) and an overrepresentation of women among French students (INSEE 2019). Then, even if the survey embraces people aged up to 88, the sample remains rather young: the median age is 32 and the average age is 36, compared to 41 and 42 years for the French population (INSEE 2020a). This is due to the diffusion through university lists and social media. 40% of the sample are students (compared to 11% of the French population over 15 (INSEE 2020b), almost half of the sample is active (47% compared to 74% of the 15–64 in France) (INSEE 2020b), though the retirees are sufficiently represented for a young sample – 8% (7% among the 15–64) (INSEE 2020b). Finally, the diffusion mode also created a rather educated sample, which is related to the overrepresentation of young people and women among them (INSEE 2019). 38% of our sample has secondary education level or less, which is the case of only a quarter of those who finished their studies. This is a huge underrepresentation compared to the French population over 15 (70% of high school diploma or less) (INSEE 2020b) but corresponds to the education level of those under 40 (INSEE 2019). 46% of the sample live in couple and 44% have children. 63% of our sample live alone (compared to 16% of the French population, INSEE 2019) which is due to the fact that younger people, overrepresented in our sample, live alone more frequently (INSEE, 2015). This bias of our sample is an advantage for this paper because we will focus here on the shutdown experience of those individuals.

In this paper, four hypotheses will be tested. Firstly, as policies and sociological literature suppose that families can serve as a shelter in case of difficulties (Martin 2002), we will check if households welcomed relatives for the lockdown. As in continental welfare, co-residence between generations is not uncommon as form of family support (Albertini and Kohli 2013; Renaut 2003), we will see if those who live alone tended to spend the lockdown with relatives. Then, we will test the effects of two factors, age and gender, on joining a family household. The second and third hypotheses state that women, as well as the young and the elderly of both sexes, who usually benefit more from financial help and services within kinship (Bengtson 2001; Silverstein and Bengtson 1997; Déchaux 2009), were more likely to activate family residential solidarities. The forth hypothesis suggests that the level of education has an impact on the residential patterns. Researches show that intergenerational co-residence is more common for socially vulnerable young people (Albertini and Kohli 2013) while financial help and services are frequent within middle and high classes (Renaut 2003). We can thus suppose education level to be positively correlated with residential solidarities during the shutdown.

The lockdown represented a time respondents preferred to spend with their families. Firstly, 98% of those who lived as a couple or family prior to the lockdown remained in this configuration. Secondly, only 16% of the respondents spent the confinement alone (while one-person households constituted 31% of our sample before the shutdown) (Figure 1). Almost all those who had housemates during the lockdown stayed with one or several family members (97%). Thirdly, the part of flatshare also diminished during the shutdown (3% instead of 6% before). Thus, the lockdown was the occasion of migration from one-person households or flatshares to a family household.

Moreover, the respondents who explain joining their family³ emphasize social ties: they wished to avoid being alone (30%), to be with their family (15%), parents (16%) or partner (5%). Almost half of them say that they moved to a more comfortable place (46%): a better housing (36%) and a better social and natural environment (23%). Leaving one's home for another one was also a question of circumstances (15%), such as work or impossibility to move because of the shutdown.

Thus, family appeared as a better place to live through this exceptional period for at least a part of our sample. Although profoundly transformed, family ties remain fundamental assets in France (Martial, Segalen 2019). Our data confirm the hypothesis of family being a shelter, but regrouped households are not necessarily stable over time. Indeed, the households that were not modified for the shutdown were most stable during this period. Only 2% of those who stayed at their place moved whereas 12% of those who went elsewhere changed place during the lockdown.

The possibility of joining the residence of someone is not equally distributed through the respondents of our sample. Several factors should be underlined here.

Among those who live alone, only the respondents under the age of 39 mention the fear of loneliness as a reason for being locked-down elsewhere. In a similar way, the youngest in our sample cite more often the will of joining a relative for going elsewhere if they lived alone before. 59% of those under 21 and only 19% of those over 38 having changed location for this period give this rationale.

The trend of the youngest to be hosted by their parents is partly compensated by the fact that older adults entertaining a Living-Apart-Together (LAT) relationship were more likely to join their partner than younger and elderly respondents.

These results show that residential solidarities are related to age. Young adults are more likely to join their parents for the shutdown than parents to go to their grown-up children's place.

Secondly, gender is an important factor in family reunification. Only 46% of women living alone stayed alone through the lockdown against 59% of men (Table 1). In general, women were locked-down with 2.03 persons against 1.85 for men. Young men living alone joined their family's residence less often than young women: 70% of women aged from 21 to 25 joined their family or a partner, compared to 50% of men. Older women living alone welcomed their children more frequently than men, especially those between the ages of 39 and 51 (Table 2). It should be noted that women are 11% more to have children in this age group of our sample (which corresponds to national statistics (INSEE 2019)) and having children has an impact on not being locked-down alone. Among those over 51 living alone, the rate of shutdown with the family passed from 4% to 10% if they have children, and from 5% to 31%, for those aged from 39 to 51. Nonetheless, having non-resident children does not predict being locked-down with them and gender plays a crucial role here. Among the 39–51, 91% of women, who usually live alone but were locked-down with housemates, spent the lockdown with their children, against only 40% of men. By contrast, middle-aged men, living alone but locked-down with a kin, were more likely to join a partner than women: 67% of men over 40 against 9% of women living alone but shutdown with a housemate. These findings confirm the central position of women in vertical kinship solidarities (Déchaux 2009; Silverstein and Bengtson 1997). They also confirm that middle-aged men are more likely to rely on their partner than women (Costemalle, 2015).

Thirdly, the education level has also an impact on lockdown households. Among the persons who have already finished their training, those who stopped at high school lived this period with 1.35 persons on average while those who have spent at least 2 years in higher education shared the dwelling with almost 2 persons. As for those who lived alone before the lockdown, the risk of staying alone through this period decreased with the education level. 93% of those who have not accomplished the secondary cycle remained in a one-person household against 75–76% of those who have a Master degree or more. The part of LAT students who lived through the analyzed period with their partner was 25%. Among those who have already finished their studies, the more educated LAT (Master degree and more) joined more often their partner (43%) than those who have only a high school diploma (17%). Young people under 25 living alone tended to stay alone if they work and to join their family or partner if they study. This result tends to confirm that the most advantaged profit more from family solidarities (Renaut 2003) (residential in our case).

The findings of our survey suggest discussing four points.

Firstly, our data propose to reconsider the standard European and American definition of family as ‘persons related by blood, marriage, or adoption, and living in the same residence’ (Tillman and Nam 2008) from two points of view. On the one hand, having two homes is not uncommon, especially for students and non-cohabiting adult couples. However, these two homes are not equal in comfort and relationships they imply. On the other hand, the circulation between residences for the lockdown suggests a kind of porosity of the frontiers of a household, much more than classical sociology asserts (De Singly 2017). We observe here not only an exchange of goods (Albertini and Kohli 2013) and services (Attias-Donfut, Segalen and Lapierre 2002) between households but also a circulation of relatives. Family in France, especially the vertical and couple ties, works as a ‘latent matrix of kin connections’ (Riley and Riley 1993). These bonds can thus be reactivated in a situation such as the Covid-19 crisis. Residential solidarities then consist in not leaving one of the relatives but in sheltering him or her in a more comfortable home and in the company of his or her family members.

Secondly, our results show that kinship and couple do not work as a ‘safety net’ (Riley and Riley 1993) for everybody. This security appears to be assured for some men through partnership whereas children and parents – the lineage – appear more as the basis of solidarity for women. This confirms the matrifocal orientation of French kinship networks regarding contacts and services (Déchaux 2009). The effects of parent's gender on housing solidarities are coherent with existing research that revealed more service exchanges (Chan 2008; Silverstein and Bengtson 1997) as well as closer relationships of adult children with their mothers (Silverstein and Bengtson 1997). Our data show, furthermore, a stronger effect of adult child's gender, young daughters being more likely to go to their parents’ place.

Thirdly, our results corroborate the life cycle variations in intergenerational solidarity hypothesis (Silverstein and Bengtson 1997). If parents tend to provide more financial and practical help and advice to their non-resident children than they receive (Renaut 2003; Chan 2008; Tillman and Nam 2008), the same trend can be stated as to housing through a difficult period, such as the lockdown. Young people benefit from an in-between status that facilitate their come-back to parental residences (Van de Velde 2007), while adults and seniors follow a residential independence standard (Attias-Donfut, Lapierre and Segalen 2002). The hypothesis of elderly persons joining their family was not confirmed, which should be taken cautiously given not only the underrepresentation of the elderly in general within our sample but especially of those having health issues.

Finally, students and people with high education level tended more to join their family or partner than working young people or low educated adults. Our results show that among young people, students seem to be more legitimate to join their parents and partner than working and unemployed young. It challenges the thesis that within the continental type of welfare, co-residence with young adult children is a kind of support addressed to the most vulnerable ones (Albertini and Kohli 2013). However, the effects of social background should be looked into in a more detailed way, which is the prospect of our further analysis.

Our findings confirm that family households are likely to change over time (Hill 1995), including even very short periods, like such ‘crisis’ as a lockdown. This point also remains to be explored in further publications.

No potential conflict of interest was reported by the author(s).

Virginie Vinel is full professor of Sociology and Anthropology at the University Bourgogne Franche-Comté (LaSA UBFC). Her research interests are age of life and age transitions, gender, health and family. She aims at understanding the course of life by articulating the dimensions of age, gender, family and body. Her main works focus on childhood, adolescence, mid-life and aging. Among her recent publications: Vinel V. and Zaltron F. (Ed.), 2020, « Enfants. Contraintes et pouvoir d’agir », Revue des Sciences sociales, n°63, 2019, on line: https://journals.openedition.org/revss/4727; Diasio N and Vinel V. (Ed.), 2017, Corps et preadolescence. Intime, privé, public, Rennes, PUR, 2017; Vinel, V., & Kessler-Bilthauer, D., 2017, Se soigner en zone rurale lorraine: Domestication du système de soin et encadrement familial des trajectoires individuelles. Enfances Familles Générations. Revue interdisciplinaire sur la famille contemporaine, (28) online: http://journals.openedition.org/efg/1609. http://lasa.univ-fcomte.fr/pages/fr/menu4048/le-laboratoire/vinel-virginie-19802-17681.html

Veronika Kushtanina is assistant professor in sociology at the University Bourgogne Franche-Comté (LASA-UBFC). Her PhD thesis explored grandparenting in France and Russia. Her current research is focused on different ages of life and biographical narratives, analyzed from the perspective of family roles and relationships, chronic illness and disorders, as well as caregiving. Among her recent publications: COLLECTIF.B. (É. DION, V. KUSHTANINA, E. LAGIER, E. PAPE, C. PERRIN-JOLY, J. PLé, P. ROBIN, B. SAVINEL, R. SCHLAGDENHAUFFEN) (EDS.), 2020, PARLER DE SOI: MéTHODES BIOGRAPHIQUES EN SCIENCES SOCIALES, ÉDITIONS DE L’EHESS; Vinel, V., & Kushtanina, V. (2019). Agency or disempowerment: Children with learning disorders in France. Salute E Società, vol. 3, pp. 107-18; V. Kushtanina, 2019, Transition de l’époque soviétique à la période post-soviétique au prisme de biographies singulières. Clio. Femmes, Genre, Histoire, n°49, pp. 239–60. http://lasa.univ-fcomte.fr/pages/fr/kushtanina-veronika-18751.html