ABSTRACT
Previous research points out that early life-course transitions affect subsequent parent–child relationships. This study examines the association between the leaving-home transition and later parent–child relations in Italy, by taking into account different timings and reasons to move out of the parental home. Using pooled data from two waves (2003 and 2009) of the Family and Social Subject Survey, Ordered Logistic regression models were adopted to analyze residential proximity and face-to-face contact between parents and their adult children. The findings show that the longer the time adult children spent in their parents’ home, the higher the propensity to reside near, and maintain frequent interaction with parents in later life is. Spending longer time in the parental home seems to provide much less benefit for daughters than for sons. The positive association between co-residence duration and later parent–daughter relations decreases until disappearing in concomitance with the normative age for leaving home. Considering the reasons for leaving home, adult children moving out to get married tend to have more intense intergenerational ties. Marriage continues to be the normative occasion to leave the parental family, particularly among adult daughters, who are subject to greater cultural expectations about family ties.
1. Introduction
Leaving the parental home is the first important step toward autonomy, self-reliance and adulthood. Over the life course, children evolve from a status of dependence on parents to a more independent position, adapting their relationship with their parents to new circumstances. Previous research has shown that young adults’ transitions to partnership and parenthood have a significant influence on the frequency of parent–child contact and parents’ well-being (Aquilino 1997; Bucx et al.2012; Kalmijn and De Graaf 2012).
Although several studies examine intergenerational relations during young adulthood, little is known about the role of nest-leaving processes in affecting later parent–child relationships. Examining this association may shed new light on how co-residence experiences and life course changes can shape family solidarity throughout the entire life course. Overviewing the whole European context, Leopold (2012) found that the time young adults spend in their parental home promotes intergenerational solidarity over the life course. Building on Leopold's work, the present study aims to contribute to the understanding of this phenomenon, by including different reasons for leaving the parental home. Marriage, the traditional route out of the parental home, has tended to be replaced by various forms of non-family living and non-marital cohabitation (Goldscheider et al.2014; White 1994). These pathways out of the parental home are often accompanied by a progressive importance of ethical values concerning individual autonomy and a less significant identification in traditional family life (Lesthaeghe 2010). Thus, the question addressed in this study is whether the heterogeneity of nest-leaving pathways affects later parent-adult child relationships in Italy.
Although peculiar for many reasons, Italy is an interesting case to understand whether a violation of norms concerning the socially accepted time and reason for leaving the parental home has a significant influence on later parent-child relations. In Southern Europe, young adults usually leave their home at a late age, when they get married (Billari 2004; Billari et al.2001); after moving out, they tend to reside close to their parents, and maintain frequent contact with them (Barbagli et al.2003; Kalmijn and Saraceno 2008; Reher 1998). In this context, characterized by strong family ties and traditional values, parents try to discourage their adult children from socially unaccepted behaviors with social and material sanctions (Di Giulio and Rosina 2007; Rosina and Fraboni 2004). Moreover, young adults, given their reliance on the family as provider of welfare, may experience intergenerational tensions when they do not conform to parental expectations. In this light, a violation of normative expectations about home-leaving may have greater consequences for parent–child relationships in Italy than in other societies where social policies and cultural norms foster individual autonomy.
Based on data from the Family and Social Subject Survey (2003 and 2009), this study investigates whether two peculiarities of Italian intergenerational bonds – high proximity and contact – are partly explained by the normative process of late home-leaving. Thus, I devote particular attention to theoretical approaches that mostly highlight the importance of normative expectations and individualistic/familialistic attitudes in Italy. I focus on the ‘middle years’ of the life course in which parent–adult child relationships are relatively free from need and responsibilities to provide care. During this life course phase, intergenerational family solidarity becomes particularly relevant for its affective and associative aspects. Residential proximity and parent–child contact also reflect the basis for future transfers of help and care during old age (Kalmijn and Dykstra 2006).
2. Background and hypotheses
2.1. Co-residence length
Co-residence length may be associated with a deeper involvement of children in their parents’ lives. The time spent in the parental home offers the opportunity for family members to share activities, interests and attitudes (Aquilino 1997; Aquilino and Supple 1991), thus fostering children's attachment to parents in later life. In addition, a long permanence in the parental home may facilitate children's integration in the collective life of the family, where norms and attitudes are promoted. Previous research has shown that early home leavers tend to develop attitudes about the importance of individual autonomy, whereas adult children who leave their parental home at a late age are generally characterized by ‘pro-family attitudes’ (Goldscheider and Lawton 1998; Ruggles 2007). These family norms and attitudes tend to vary according to different historical and social contexts in which families are embedded. To take these sources of heterogeneity into consideration, co-residence length is analyzed as a deviation from what the majority of people do in the same social group (by sex, regions and cohorts). Thus, it can be hypothesized that in each specific social group, the longer the time adult children spend in their parental home, the higher their propensity to reside near their parents and maintain frequent contact with them (Hypothesis 1).
The positive association between co-residence length and later parent–child relationships may be driven by two main confounding factors. First, high unemployment rates tend to translate into longer periods of intergenerational co-residence (Aassve et al.2013b), and late home leavers may feel obliged to maintain frequent intergenerational contacts in order to repay the support received early in life (Silverstain et al.2002). However, young adults who are forced to stay longer in the parental home may develop a resentment for co-residence, thus reducing their propensity of living near and having frequent visits with parents. It is not clear, therefore, whether young adults’ economic conditions may have a positive influence on co-residence experiences.
A second selection process is related to family climate during childhood/adolescence. The quality of family relations may affect whether a child will be an early or late home leaver (e.g. Bernhardt et al.2005), and these experiences can be carried over into later relationships. Adult children who left their home at an early age may have done so because of a less affective family climate that implies few obligations to maintain strong relations with parents. However, whereas early home leavers may be selected by a hostile family climate, an extended period of co-residence is not related to positive family relations (Ward and Spitze 2007). Thus, it is reasonable to assume that the possible influence exerted by the duration of co-residence on later parent–child relationships is not completely reduced to a mere selection effect.
2.2. Age norms
According to the family life course perspective (Elder 2003), deviations from the normative timing in the passage to adulthood may have adverse effects on later life. Social policies along with cultural norms contribute to the institutionalization of the life course, establishing a predictable timetable of roles and transitions for an individual. A violation of this timetable can weaken the relation between the individual and the socio-institutional context. Early nest-leaving may be perceived as a voluntary estrangement from the family of origin in contexts where people usually leave the parental home at a late age (Bordone 2009). Moreover, social norms set an ‘age deadline’ prescribing the upper end limit to leave one's parental home (Aassve et al.2013a; Billari and Liefbroer 2007; Liefbroer and Billari 2010; Settersten 1998). An extended co-residence may indicate difficulties in completing adulthood transition and an excessive dependency from parents’ resources. The literature suggests that an extended co-residence is often related to conflict and ambivalent feelings between parents and their children (Lüscher 2002, 2011; Ward and Spitze 2007). Family conflicts may emerge when their children infringe the normative unwritten expectation of reaching independence in a timely fashion (Pillemer and Suitor 2002). By assuming that family tensions can be translated into greater residential distances and less frequent parent–child contacts, early and late home leavers may exhibit a lower propensity to reside near and maintain frequent contact with parents, compared to those who move out of the parental home ‘on time’. Hence, I would expect to observe an inverse U-shaped association between co-residence length and later levels of residential proximity and parent–child contact (Hypothesis 2).
2.3. Reasons for leaving the parental home
Cultural norms may also prescribe the socially accepted reason for leaving the family nest. Historically, marriage was defined as the normative occasion to leave the parental home (Goldscheider and Goldscheider 1999; Goldscheider et al.2014) and the primary mechanism through which intergenerational family ties strengthened. During the twentieth century, the meaning of marriage changed in most Western countries. Previous research has shown that modern marriage is an expression of self-realization and functions as a ‘greedy institution’, i.e. an exclusive relationship that weakens parent–child relations (Bucx et al.2012; Sarkisian and Gerstel 2008). However, in Southern European countries, marriage continues to be an institution that fosters intergenerational ties and binds families together (Yahirun and Hamplová 2014). There is also evidence that, through the judgment of the family of origin, Italian young adults experience social pressures to marry (Vignoli and Salvini 2014). Thus, abandoning the nest in order to marry may fulfill normative expectations and may be related to strong intergenerational ties.
Other occurrences – such as cohabitation, pursuit of higher education or career, and desire for independence – are becoming widespread among highly educated young adults in Northern Italian regions (Barbagli et al.2003), and may favor a weaker involvement of children in their parents’ life. The literature has shown that cohabiting adult children are less prone to maintain frequent contact with parents than those who are married or single (Eggebeen 2005; Yahirun and Hamplová 2014). Attending advanced education courses and pursuing paid employment are motivations related to long-distance moves and weaker preferences for intergenerational contact (Bordone 2009; Kalmijn 2006). Moreover, leaving the parental home because of the desire for autonomy may be an expression of individualistic attitudes (Gierveld et al.1991), which in turn may negatively affect adult children's propensity to maintain close relations with parents in later life. Following from these findings, it can be hypothesized that adult children who leave the family of origin to get married are more prone to reside close and maintain frequent interaction with parents, than those who leave the nest for other reasons (Hypothesis 3).
The association between different reasons for leaving the parental home and the frequency of parent–child contact can be disputed, however. In Italy, non-traditional unions may occur mostly when supported by parental acceptance (Di Giulio and Rosina 2007). Nazio and Saraceno (2012) found that cohabitation is seen as a non-transgressive behavior that does not weaken intergenerational ties. Moreover, the effect of education on the frequency of parent–child contact may be to a large part indirect. The highly educated tend to move farther away to take advantage of geographically delimited opportunities (Bordone 2009; Kalmijn 2006). Thus, pursuing a higher education or a career may primarily affect the geographical distance between parents and their adult children. I take these ideas into account as part of my interpretation when describing the role of different reasons for leaving the parental home.
The motivations to leave the parental home appear to be somewhat related to later parent–child relations, and are also associated with different timings of leaving home. It is well known that in Italy, the prolonged stay of children in their parental home generally corresponds to marriage as the most common route out of the family nest (Billari et al.2001; Rosina and Fraboni 2004). Desires for independence and attending higher education courses, instead, tend to induce young adults to leave the parental home at an early age (Barbagli et al.2003). An early departure from the family of origin is often accompanied with experiences of non-family living, which encourage young adults to develop a ‘taste for independence’ (Michielin and Mulder 2007; Mulder and Clark 2002). Early and late home leavers are likely to follow different pathways in education, employment and union formation, which in turn may affect their chances and preferences for high levels of proximity and contact with parents. Thus, it can be hypothesized that the association between the co-residence length and later levels of proximity and contact with parents is partly explained by different reasons for leaving the family nest (Hypothesis 4).
2.4. Gender differences
The literature suggests that intergenerational co-residence may have greater consequences for daughters than for sons. Women tend to invest more in family relationships, assuming family responsibilities and providing informal services when necessary. Co-resident daughters are usually more involved in parents’ life, and they may be more prone to develop care-related obligations (Silverstein et al.2006). At the same time, daughters who stay in the parental home tend to be supervised more closely and have more family commitments than sons. This generally translates into greater costs and less benefits of co-residence, probably reducing their willingness for staying longer in the parental home (White 1994). Parents tend to attach greater importance to adult daughters, and a violation of social norms on their part may imply greater losses (Ward and Spitze 1992). Given these gender differences in the costs and benefits of co-residence, the following analysis examines parent–son and parent–daughter relationships separately.
3. Data and method
The empirical analysis is based on the last two waves of the Family and Social Subject Survey (the Italian component of Generations and Gender Survey) that took place in 2003 and 2009. These are five-year modules of the annual survey Indagine Multiscopo (Multipurpose survey), conducted by the Italian National Statistical Office (ISTAT). The survey offers a representative sample of the Italian population (the response rate is 85%). It involved more than 19,000 families (about 50,000 individuals) in 2003, and 18,000 families (about 44,000 individuals) in 2009. Both waves contain a retrospective section focusing on the transition out of the parental family, as well as union formation and trajectories of family life course. The survey also gathered detailed information about parents’ characteristics and intergenerational relations, including proximity and contact frequency.
Since the purpose of the present study is to examine parent–child relationships during the ‘middle years’ of the life course, the sample was restricted to children aged 30–45 who had at least one parent alive at the time of the survey. To reduce the selection bias due to problematic family relationships during childhood and adolescence, individuals who left their parental home before age 16 were excluded from the analysis. The final sample contains 7217 men and 8481 women.
3.1. Dependent variables
The two dependent variables refer to residential proximity and face-to-face contact between parents and their adult children. Proximity was measured on a six-point scale: ‘other city >50 km’; ‘other city 16–50 km’; ‘other city <16 km’; ‘same city >1 km’; ‘same city <1 km’; ‘other apartment, same building’. The unit of analysis is the respondent (or the child) if parents live together, whereas it is the parent–child dyad if parents do not live together. Thus, two observations are generated by one child having two divorced parents living in different households.
The frequency of face-to-face contact between parents and their adult children is measured on a five-point scale: ‘daily’; ‘several times a week’; ‘weekly’; ‘2 or 3 per month’; ‘once per month or less’. In line with previous research, Italians tend to maintain frequent contact with their parents (e.g. Kalmijn and Saraceno 2008): about 37% of adult children have daily visits with their parents (Table 1). Since information about contacts with mothers and fathers was collected separately, the analysis of family contacts is carried out on the parent–child dyads. The original sample includes two dyads for each individual, when both parents are alive, and one dyad in the case of widowed parents.
. | Sons . | Daughters . | . | Sons . | Daughters . |
---|---|---|---|---|---|
Dependent variables | Education degree | ||||
Residential proximity | Low | 83.8 | 82.5 | ||
Same building | 13.0 | 10.1 | Medium | 12.9 | 14.4 |
Same city ≤ 1 km | 24.9 | 23.6 | High | 3.3 | 3.1 |
Same city > 1km | 23.6 | 13.0 | Children's characteristics: | ||
Other city < 16 km | 13.5 | 15.2 | Birth cohort | ||
Other city 16–50 km | 7.4 | 10.0 | 1958–63 | 20.1 | 20.4 |
Other city > 50 km | 17.5 | 18.0 | 1964–68 | 36.0 | 33.8 |
Frequency of contact | 1969–73 | 32.8 | 34.0 | ||
Daily | 35.5 | 39.1 | 1974–79 | 11.1 | 11.8 |
Several times a week | 28.0 | 27.2 | Region | ||
Weekly | 11.9 | 10.0 | North | 44.3 | 43.2 |
2 or 3 a month | 9.7 | 9.0 | Centre | 17.3 | 17.2 |
Fewer than monthly | 14.9 | 14.7 | South | 38.4 | 39.6 |
Independent variables | Education | ||||
DGM (SD) | 12.6(5.3) | 13.6(4.7) | Low | 42.6 | 37.2 |
Reasons for leaving | Medium | 45.2 | 47.3 | ||
Marriage | 54.8 | 70.6 | High | 12.2 | 15.4 |
Cohabitation | 8.0 | 9.7 | Employment status | ||
Education/job | 20.6 | 11.6 | Employed | 94.1 | 60.5 |
Independence | 16.6 | 8.1 | Unemployed | 4.7 | 7.2 |
Parents’ characteristics: | Not in LM | 1.2 | 32.2 | ||
Sex (Father) | 44.7 | 44.8 | Homeowner | 62.2 | 65.1 |
Marital status | Marital Status | ||||
Living together | 81.7 | 81.1 | Never married | 15.5 | 10.7 |
Widowed | 13.8 | 13.6 | Married | 78.2 | 79.0 |
Divorced or separated | 4.5 | 5.3 | Divorced | 6.3 | 10.3 |
Poor health | 5.7 | 5.9 | Child <7 | 5.8 | 5.7 |
Father's occupation when the child was 14 | N. of siblings Years to find the first job | 2.0 | 2.0 | ||
Manager | 5.3 | 5.6 | 0 | 30.0 | 25.9 |
Clerk | 19.7 | 21.2 | 1 | 24.5 | 24.3 |
Manual worker | 44.8 | 44.6 | 2/3 | 25.4 | 21.0 |
Self-employed | 30.2 | 28.6 | 4 or more | 20.1 | 28.8 |
N. of individuals | 7217 | 8481 | N. of individuals | 7217 | 8481 |
N. of dyads | 12,306 | 14,474 | N. of dyads | 12,306 | 14,474 |
. | Sons . | Daughters . | . | Sons . | Daughters . |
---|---|---|---|---|---|
Dependent variables | Education degree | ||||
Residential proximity | Low | 83.8 | 82.5 | ||
Same building | 13.0 | 10.1 | Medium | 12.9 | 14.4 |
Same city ≤ 1 km | 24.9 | 23.6 | High | 3.3 | 3.1 |
Same city > 1km | 23.6 | 13.0 | Children's characteristics: | ||
Other city < 16 km | 13.5 | 15.2 | Birth cohort | ||
Other city 16–50 km | 7.4 | 10.0 | 1958–63 | 20.1 | 20.4 |
Other city > 50 km | 17.5 | 18.0 | 1964–68 | 36.0 | 33.8 |
Frequency of contact | 1969–73 | 32.8 | 34.0 | ||
Daily | 35.5 | 39.1 | 1974–79 | 11.1 | 11.8 |
Several times a week | 28.0 | 27.2 | Region | ||
Weekly | 11.9 | 10.0 | North | 44.3 | 43.2 |
2 or 3 a month | 9.7 | 9.0 | Centre | 17.3 | 17.2 |
Fewer than monthly | 14.9 | 14.7 | South | 38.4 | 39.6 |
Independent variables | Education | ||||
DGM (SD) | 12.6(5.3) | 13.6(4.7) | Low | 42.6 | 37.2 |
Reasons for leaving | Medium | 45.2 | 47.3 | ||
Marriage | 54.8 | 70.6 | High | 12.2 | 15.4 |
Cohabitation | 8.0 | 9.7 | Employment status | ||
Education/job | 20.6 | 11.6 | Employed | 94.1 | 60.5 |
Independence | 16.6 | 8.1 | Unemployed | 4.7 | 7.2 |
Parents’ characteristics: | Not in LM | 1.2 | 32.2 | ||
Sex (Father) | 44.7 | 44.8 | Homeowner | 62.2 | 65.1 |
Marital status | Marital Status | ||||
Living together | 81.7 | 81.1 | Never married | 15.5 | 10.7 |
Widowed | 13.8 | 13.6 | Married | 78.2 | 79.0 |
Divorced or separated | 4.5 | 5.3 | Divorced | 6.3 | 10.3 |
Poor health | 5.7 | 5.9 | Child <7 | 5.8 | 5.7 |
Father's occupation when the child was 14 | N. of siblings Years to find the first job | 2.0 | 2.0 | ||
Manager | 5.3 | 5.6 | 0 | 30.0 | 25.9 |
Clerk | 19.7 | 21.2 | 1 | 24.5 | 24.3 |
Manual worker | 44.8 | 44.6 | 2/3 | 25.4 | 21.0 |
Self-employed | 30.2 | 28.6 | 4 or more | 20.1 | 28.8 |
N. of individuals | 7217 | 8481 | N. of individuals | 7217 | 8481 |
N. of dyads | 12,306 | 14,474 | N. of dyads | 12,306 | 14,474 |
3.2. Independent variables
The main independent variable is the difference between the actual age and the normative age for leaving the parental home. The basic idea is that social norms about the appropriate timing of leaving home are shared and defined within specific social groups (Billari and Liefbroer 2007). Thus, the sample was divided into 24 groups according to cohorts of birth (1958–1963, 1964–1968, 1969–1973, 1974–1979), sex, and three macro-regions (North, Centrum, South). Age norm was defined as the median age at leaving home in each group. When computing the median of all reported ages for leaving the parental, survival functions of leaving home were estimated separately for each social group in order to account for censored observations. The group median age ranges from 22 years, among women born in the oldest cohort living in the South, to 30 years among youngest men residing in the North of the country.
The variable ‘Difference from the Group Median age’ (DGM) includes values from –14 to 23. Young adults who leave parental home ‘on time’ have value 0 in DGM. In order to understand the role of age norms, a quadratic measure of DGM was included in the multivariate analyses. If early and late home leavers violate social expectations about the ‘right’ time for leaving their parental home, the effect on their subsequent parent–child relationships should present a concavity point corresponding to young adults who left their parents’ home ‘on time’. To treat differently the experiences of early (negative values) and late home leavers (positive values), the range of the DGM was rescaled, by adding the minimum value of the total distribution.
Different motivations for moving out of the parental home were measured through the question ‘What was the main reason to start living on your own?’. Table 1 shows that among males, 55% left their parental home in coincidence with marriage, 8% with cohabitation, 20% in pursuit of career or education opportunities, and 17% left the nest driven by their desire for autonomy. Adult females, instead, are more likely to move out because of marriage (70.6%) and are less likely to leave their family to pursue higher education (11.6%) or for their desire for independence (8%). In Italian society, marriage is the most widespread reason for leaving the parental home, especially for adult females.
Selected parents’ and children's characteristics are used in the following analysis as control variables. Children's characteristics include age, residential region, education degree, employment status, housing tenure (homeowner), marital status, own parenthood and number of siblings. Parent's characteristics comprise sex, marital status, and limitations in everyday activities (‘poor health’). Notably, adult children from dissolved families leave their parental home at a significantly younger age (Goldscheider and Goldscheider 1999) and maintain less frequent family interactions in later life (Albertini and Garriga 2011). Thus, parental union dissolution can be considered an important confounding factor in the association between DGM and later parent–child relations. Moreover, to at least partly account for economic conditions during the period of co-residence, the following analyses include father's occupation when the child was 14 and the time taken to find the first job. As previously noted, unemployed children are likely to delay their transition to independence and may feel obliged to repay the support received early in life.
3.3. Analytical strategy
The analytical strategy consists in adopting ordered logistic regression models on the likelihood of living near and maintaining frequent visits with parents. The analyses are employed separately for sons and daughters in order to take into account different gender behaviors in the leaving-home transition as well as in intergenerational relations. A sensitivity analysis was conducted to make sure that the results were not influenced by the choice of statistical models. Multinomial logistic regression models yield similar results to those presented here. I chose to present ordered logistic regression models because they are more parsimonious and interpretable. To take into account the correlation between parent–child dyads within the same family, regression analyses include clustered standard errors. Moreover, to facilitate the interpretation of the results, predicted probabilities are illustrated in graphs by computing the average of the predictions.
4. Results
Predicted probabilities of living in the same city ≤1 km. (a). Table 2: Sons and (b). Table 2: Daughters.
. | Adult sons . | Adult daughters . | ||||||
---|---|---|---|---|---|---|---|---|
Model 1a . | Model 2a . | Model 1b . | Model 2b . | |||||
Coef. . | S.E. . | Coef. . | S.E. . | Coef. . | S.E. . | Coef. . | S.E. . | |
Children's characteristics | ||||||||
DGM | 0.130** | (0.017) | 0.014 | (0.020) | 0.136** | (0.021) | 0.053† | (0.022) |
DGM2 | –0.003** | (0.001) | –0.001 | (0.001) | –0.004** | (0.001) | –0.002† | (0.001) |
Cohort (Ref. 1958–63) | ||||||||
1964–68 | –0.011 | (0.063) | –0.033 | (0.063) | 0.004 | (0.057) | –0.019 | (0.057) |
1969–73 | –0.096 | (0.069) | –0.123 | (0.069) | –0.042 | (0.061) | –0.080 | (0.061) |
1974–79 | –0.338** | (0.100) | –0.374** | (0.101) | –0.114 | (0.090) | –0.160 | (0.091) |
Region (Ref. North) | ||||||||
Centre | 0.235** | (0.063) | 0.209** | (0.063) | 0.045 | (0.056) | 0.012 | (0.057) |
South | 0.924** | (0.051) | 0.900** | (0.052) | 0.827** | (0.050) | 0.783** | (0.051) |
Education (Ref. Low) | ||||||||
Medium | –0.188** | (0.048) | –0.124† | (0.049) | –0.129* | (0.047) | –0.048 | (0.048) |
High | –0.407** | (0.074) | –0.192† | (0.076) | –0.346** | (0.072) | –0.058 | (0.074) |
Marital Status (Ref. married) | ||||||||
Never married | 0.026 | (0.067) | –0.096 | (0.078) | 0.157† | (0.075) | –0.271* | (0.091) |
Divorced | –0.130 | (0.116) | –0.247† | (0.122) | 0.318** | (0.096) | –0.129 | (0.106) |
Years to find the first job (Ref. 0) | ||||||||
1 | –0.042 | (0.057) | –0.052 | (0.058) | –0.069 | (0.054) | –0.076 | (0.054) |
2/3 | –0.223** | (0.059) | –0.189* | (0.060) | –0.169* | (0.057) | –0.152* | (0.058) |
4 or more | –0.363** | (0.064) | –0.308** | (0.065) | –0.254** | (0.059) | –0.260** | (0.059) |
Parents’ characteristics | ||||||||
Marital status (Ref. living together) | ||||||||
Widowed | 0.044 | (0.054) | 0.039 | (0.055) | –0.002 | (0.049) | 0.001 | (0.049) |
Divorced or separated | –0.631** | (0.093) | –0.683** | (0.093) | –0.723** | (0.086) | –0.746** | (0.086) |
Father's occupation when the child was 14 (Ref. manager) | ||||||||
Clerk | 0.139 | (0.106) | 0.116 | (0.105) | 0.038 | (0.106) | –0.006 | (0.107) |
Manual worker | 0.216† | (0.102) | 0.158 | (0.102) | 0.074 | (0.104) | 0.023 | (0.104) |
Self-employed | 0.510** | (0.104) | 0.451** | (0.103) | 0.352** | (0.105) | 0.317* | (0.106) |
Reasons for leaving (Ref. marriage) | ||||||||
Cohabitation | –0.236* | (0.081) | –0.340** | (0.076) | ||||
Educ/job | –1.096** | (0.085) | –1.219** | (0.090) | ||||
Independence | 0.003 | (0.078) | –0.433** | (0.098) | ||||
N. of individuals | 7217 | 7217 | 8481 | 8481 | ||||
N. of dyads | 7487 | 7487 | 8837 | 8837 |
. | Adult sons . | Adult daughters . | ||||||
---|---|---|---|---|---|---|---|---|
Model 1a . | Model 2a . | Model 1b . | Model 2b . | |||||
Coef. . | S.E. . | Coef. . | S.E. . | Coef. . | S.E. . | Coef. . | S.E. . | |
Children's characteristics | ||||||||
DGM | 0.130** | (0.017) | 0.014 | (0.020) | 0.136** | (0.021) | 0.053† | (0.022) |
DGM2 | –0.003** | (0.001) | –0.001 | (0.001) | –0.004** | (0.001) | –0.002† | (0.001) |
Cohort (Ref. 1958–63) | ||||||||
1964–68 | –0.011 | (0.063) | –0.033 | (0.063) | 0.004 | (0.057) | –0.019 | (0.057) |
1969–73 | –0.096 | (0.069) | –0.123 | (0.069) | –0.042 | (0.061) | –0.080 | (0.061) |
1974–79 | –0.338** | (0.100) | –0.374** | (0.101) | –0.114 | (0.090) | –0.160 | (0.091) |
Region (Ref. North) | ||||||||
Centre | 0.235** | (0.063) | 0.209** | (0.063) | 0.045 | (0.056) | 0.012 | (0.057) |
South | 0.924** | (0.051) | 0.900** | (0.052) | 0.827** | (0.050) | 0.783** | (0.051) |
Education (Ref. Low) | ||||||||
Medium | –0.188** | (0.048) | –0.124† | (0.049) | –0.129* | (0.047) | –0.048 | (0.048) |
High | –0.407** | (0.074) | –0.192† | (0.076) | –0.346** | (0.072) | –0.058 | (0.074) |
Marital Status (Ref. married) | ||||||||
Never married | 0.026 | (0.067) | –0.096 | (0.078) | 0.157† | (0.075) | –0.271* | (0.091) |
Divorced | –0.130 | (0.116) | –0.247† | (0.122) | 0.318** | (0.096) | –0.129 | (0.106) |
Years to find the first job (Ref. 0) | ||||||||
1 | –0.042 | (0.057) | –0.052 | (0.058) | –0.069 | (0.054) | –0.076 | (0.054) |
2/3 | –0.223** | (0.059) | –0.189* | (0.060) | –0.169* | (0.057) | –0.152* | (0.058) |
4 or more | –0.363** | (0.064) | –0.308** | (0.065) | –0.254** | (0.059) | –0.260** | (0.059) |
Parents’ characteristics | ||||||||
Marital status (Ref. living together) | ||||||||
Widowed | 0.044 | (0.054) | 0.039 | (0.055) | –0.002 | (0.049) | 0.001 | (0.049) |
Divorced or separated | –0.631** | (0.093) | –0.683** | (0.093) | –0.723** | (0.086) | –0.746** | (0.086) |
Father's occupation when the child was 14 (Ref. manager) | ||||||||
Clerk | 0.139 | (0.106) | 0.116 | (0.105) | 0.038 | (0.106) | –0.006 | (0.107) |
Manual worker | 0.216† | (0.102) | 0.158 | (0.102) | 0.074 | (0.104) | 0.023 | (0.104) |
Self-employed | 0.510** | (0.104) | 0.451** | (0.103) | 0.352** | (0.105) | 0.317* | (0.106) |
Reasons for leaving (Ref. marriage) | ||||||||
Cohabitation | –0.236* | (0.081) | –0.340** | (0.076) | ||||
Educ/job | –1.096** | (0.085) | –1.219** | (0.090) | ||||
Independence | 0.003 | (0.078) | –0.433** | (0.098) | ||||
N. of individuals | 7217 | 7217 | 8481 | 8481 | ||||
N. of dyads | 7487 | 7487 | 8837 | 8837 |
Notes: DGM: Difference between individual age at leaving home and the Group Median age, ranging from 0 to 34. Other control variables: number of siblings, employment status, housing tenure, having a child aged 6 or less, parents’ education, parents’ health. Pooled models show a significant interaction term between children’ sex and DGM.
**p < .001.
*p < .010.
†p < .05
Figure 1a and b indicate that the propensity of adult sons and daughters to reside near their parents is positively connected to DGM. In line with previous research (Leopold et al.2012), late home leavers tend to maintain closer intergenerational proximity, compared to those who leave their family of origin at an early age. This is particularly evident when comparing early home leavers to adult children who move out of the parental home ‘on time’ or later. The positive correlation between DGM and the likelihood of living close to parents becomes smaller as co-residence length reaches the normative age for leaving home.1
Different reasons to leave the family nest are included in models 2a and 2b. Pathways other than marriage are chosen by highly educated children; thus, the coefficient of high education decreases when moving from model 1a-b to model 2a-b. In line with hypothesis 4, the association between DGM and residential proximity is partly explained by different reasons for leaving the parental home (Figure 1a and b). Early and late home leavers are generally characterized by different pathways to independence. Late home leavers are likely to move out of the parental home because of marriage and tend to reside close to their parental home. Early home leavers, instead, are more prone to leave their family of origin to go after better job and educational opportunities, which are related to long-distance moves. The average probability of living in the same city is 21 (Coef. −1.09) percentage points lower for adult sons who follow educational and job opportunities, than for those who leave the parental home to marry. Desires for independence may be also related to an early departure from the parental home, but it is not significantly associated with parent–son residential proximity (Model 2a). Among adult daughters (Model 2b), all motivations to leave the parental home other than marriage are negatively associated with the likelihood of living near parents. This finding supports hypothesis 3, that is, that marriage as the normative pathway to independence is correlated to closer residential proximity between generations, than other routes out of the parental home.
Predicted probabilities of having more than weekly contact with parents. (a.) Table 3: Sons and (b). Table 3: Daughters.
. | Adult sons . | Adult daughters . | ||||||
---|---|---|---|---|---|---|---|---|
Model 3a . | Model 4a . | Model 3b . | Model 4b . | |||||
Coef. . | SE . | Coef. . | SE . | Coef. . | SE . | Coef. . | SE . | |
Children's characteristics | ||||||||
DGM | 0.044† | (0.017) | 0.019 | (0.019) | 0.118** | (0.021) | 0.080** | (0.022) |
DGM2 | –0.001 | (0.001) | –0.000 | (0.001) | –0.003** | (0.001) | –0.002** | (0.001) |
Cohort (Ref. 1958–63) | ||||||||
1964–68 | –0.191** | (0.065) | –0.198** | (0.065) | 0.046 | (0.062) | 0.049 | (0.062) |
1969–73 | –0.134 | (0.069) | –0.143† | (0.069) | 0.100 | (0.064) | 0.099 | (0.064) |
1974–79 | –0.298* | (0.092) | –0.313** | (0.093) | 0.067 | (0.084) | 0.073 | (0.084) |
Region (Ref. north) | ||||||||
Centre | 0.145† | (0.067) | 0.138† | (0.067) | 0.044 | (0.063) | 0.028 | (0.063) |
South | 0.411** | (0.057) | 0.411** | (0.058) | 0.358** | (0.055) | 0.330** | (0.055) |
Education (Ref. low) | ||||||||
Medium | –0.143* | (0.054) | –0.123† | (0.054) | –0.205** | (0.055) | –0.171* | (0.056) |
High | –0.350** | (0.081) | –0.288** | (0.083) | –0.320** | (0.079) | –0.193† | (0.083) |
Marital status (Ref. married) | ||||||||
Never married | –0.145* | (0.070) | –0.191† | (0.079) | 0.141 | (0.076) | –0.182† | (0.089) |
Divorced | –0.191 | (0.116) | –0.240† | (0.119) | 0.078 | (0.100) | –0.246† | (0.109) |
Years to find the first job (Ref. 0) | ||||||||
1 | 0.006 | (0.064) | 0.005 | (0.064) | 0.005 | (0.062) | –0.014 | (0.063) |
2/3 | –0.132† | (0.062) | –0.124† | (0.062) | –0.083 | (0.065) | –0.085 | (0.065) |
4 or more | –0.168† | (0.070) | –0.154† | (0.070) | –0.272** | (0.064) | –0.288** | (0.064) |
Parents’ characteristics | ||||||||
Father | –0.070** | (0.018) | –0.072** | (0.018) | –0.186** | (0.017) | –0.188** | (0.017) |
Marital status (Ref. living together) | ||||||||
Widowed | –0.145† | (0.059) | –0.146† | (0.059) | –0.143* | (0.055) | –0.140† | (0.055) |
Divorced or separated | –0.798** | (0.105) | –0.812** | (0.105) | –0.836** | (0.098) | –0.828** | (0.099) |
Father's occupation when the child was 14 (Ref. manager) | ||||||||
Clerk | 0.235† | (0.106) | 0.234† | (0.106) | 0.136 | (0.107) | 0.113 | (0.107) |
Manual worker | 0.266† | (0.104) | 0.255† | (0.104) | 0.105 | (0.106) | 0.080 | (0.106) |
Self-employed | 0.540** | (0.106) | 0.528** | (0.106) | 0.232† | (0.106) | 0.214† | (0.107) |
Proximity (Ref. Other city >50 km) | ||||||||
Other city ≤ 50 km | 2.692** | (0.085) | 2.629** | (0.087) | 2.622** | (0.071) | 2.579** | (0.072) |
Same city > 1 km | 3.243** | (0.087) | 3.179** | (0.088) | 3.431** | (0.076) | 3.381** | (0.077) |
Same city ≤ 1 km | 4.674** | (0.093) | 4.613** | (0.095) | 4.814** | (0.084) | 4.768** | (0.085) |
Reasons for leaving (Ref. marriage) | ||||||||
Cohabitation | –0.060 | (0.091) | –0.279** | (0.081) | ||||
Educ/job | –0.313** | (0.080) | –0.544** | (0.084) | ||||
Independence | –0.026 | (0.080) | –0.491** | (0.100) | ||||
N. of individuals | 7217 | 7217 | 8481 | 8481 | ||||
N. of dyads | 12,306 | 12,306 | 14,474 | 14,474 |
. | Adult sons . | Adult daughters . | ||||||
---|---|---|---|---|---|---|---|---|
Model 3a . | Model 4a . | Model 3b . | Model 4b . | |||||
Coef. . | SE . | Coef. . | SE . | Coef. . | SE . | Coef. . | SE . | |
Children's characteristics | ||||||||
DGM | 0.044† | (0.017) | 0.019 | (0.019) | 0.118** | (0.021) | 0.080** | (0.022) |
DGM2 | –0.001 | (0.001) | –0.000 | (0.001) | –0.003** | (0.001) | –0.002** | (0.001) |
Cohort (Ref. 1958–63) | ||||||||
1964–68 | –0.191** | (0.065) | –0.198** | (0.065) | 0.046 | (0.062) | 0.049 | (0.062) |
1969–73 | –0.134 | (0.069) | –0.143† | (0.069) | 0.100 | (0.064) | 0.099 | (0.064) |
1974–79 | –0.298* | (0.092) | –0.313** | (0.093) | 0.067 | (0.084) | 0.073 | (0.084) |
Region (Ref. north) | ||||||||
Centre | 0.145† | (0.067) | 0.138† | (0.067) | 0.044 | (0.063) | 0.028 | (0.063) |
South | 0.411** | (0.057) | 0.411** | (0.058) | 0.358** | (0.055) | 0.330** | (0.055) |
Education (Ref. low) | ||||||||
Medium | –0.143* | (0.054) | –0.123† | (0.054) | –0.205** | (0.055) | –0.171* | (0.056) |
High | –0.350** | (0.081) | –0.288** | (0.083) | –0.320** | (0.079) | –0.193† | (0.083) |
Marital status (Ref. married) | ||||||||
Never married | –0.145* | (0.070) | –0.191† | (0.079) | 0.141 | (0.076) | –0.182† | (0.089) |
Divorced | –0.191 | (0.116) | –0.240† | (0.119) | 0.078 | (0.100) | –0.246† | (0.109) |
Years to find the first job (Ref. 0) | ||||||||
1 | 0.006 | (0.064) | 0.005 | (0.064) | 0.005 | (0.062) | –0.014 | (0.063) |
2/3 | –0.132† | (0.062) | –0.124† | (0.062) | –0.083 | (0.065) | –0.085 | (0.065) |
4 or more | –0.168† | (0.070) | –0.154† | (0.070) | –0.272** | (0.064) | –0.288** | (0.064) |
Parents’ characteristics | ||||||||
Father | –0.070** | (0.018) | –0.072** | (0.018) | –0.186** | (0.017) | –0.188** | (0.017) |
Marital status (Ref. living together) | ||||||||
Widowed | –0.145† | (0.059) | –0.146† | (0.059) | –0.143* | (0.055) | –0.140† | (0.055) |
Divorced or separated | –0.798** | (0.105) | –0.812** | (0.105) | –0.836** | (0.098) | –0.828** | (0.099) |
Father's occupation when the child was 14 (Ref. manager) | ||||||||
Clerk | 0.235† | (0.106) | 0.234† | (0.106) | 0.136 | (0.107) | 0.113 | (0.107) |
Manual worker | 0.266† | (0.104) | 0.255† | (0.104) | 0.105 | (0.106) | 0.080 | (0.106) |
Self-employed | 0.540** | (0.106) | 0.528** | (0.106) | 0.232† | (0.106) | 0.214† | (0.107) |
Proximity (Ref. Other city >50 km) | ||||||||
Other city ≤ 50 km | 2.692** | (0.085) | 2.629** | (0.087) | 2.622** | (0.071) | 2.579** | (0.072) |
Same city > 1 km | 3.243** | (0.087) | 3.179** | (0.088) | 3.431** | (0.076) | 3.381** | (0.077) |
Same city ≤ 1 km | 4.674** | (0.093) | 4.613** | (0.095) | 4.814** | (0.084) | 4.768** | (0.085) |
Reasons for leaving (Ref. marriage) | ||||||||
Cohabitation | –0.060 | (0.091) | –0.279** | (0.081) | ||||
Educ/job | –0.313** | (0.080) | –0.544** | (0.084) | ||||
Independence | –0.026 | (0.080) | –0.491** | (0.100) | ||||
N. of individuals | 7217 | 7217 | 8481 | 8481 | ||||
N. of dyads | 12,306 | 12,306 | 14,474 | 14,474 |
Note: DGM: Difference between individual age at leaving home and the Group Median age, ranging from 0 to 34. Other control variables: number of siblings, employment status, housing tenure, having a child aged 6 or less, parents’ education, parents’ health. Pooled models show a significant interaction term between children’ sex and DGM.
**p < .001.
*p < .010.
†p < .05.
Differently from the pattern observed for sons, the positive association between DGM and the frequency of later parent–adult daughter visits decreases and ultimately disappears as the co-residence duration reaches the group median age of moving out (Figure 2b). This finding partly supports both hypotheses 1 and 2. At an early age, the time spent in the parental home promotes frequent intergenerational contact, whereas late departures from the family nest are not positively associated with the frequency of later parent–child visits.
In model 4a, the coefficient of DGM disappears when including adult sons’ reasons for leaving home. Late home leavers tend to move out of their parental home to get married and are more likely to maintain frequent interactions with parents, compared to those who leave the family of origin to seek education and job opportunities. Leaving home to cohabit or because of desires for independence is not significantly related to the likelihood of having frequent parent–son visits. For women (Model 4b), instead, leaving the parental home for reasons other than marriage is linked to less frequent face-to-face contacts with parents. Consistent with hypothesis 3, desires for independence, going after better education or job opportunities and cohabitation, instead of marriage are related to less frequent parent–daughter contacts. Moreover, the association between DGM and the frequency of parent–daughter visits remains significant even after considering different reasons for leaving the parental home in the analysis. Marriage appears a necessary but not sufficient condition to maintain strong family ties.
5. Discussion
The literature suggests that intergenerational relationships are shaped by early life-course transitions. Leaving the parental home and establishing a new household is often viewed as a crucial step in reaching the adult status, and, on such occasion, the parent–child relationship is redefined and shaped according to the different pathways chosen to acquire independence. Previous research has shown that late nest-leaving is positively related to parent–child involvement (Aquilino 1997), ‘pro-family attitudes’ (Goldscheider and Lawton 1998) and intergenerational solidarity in later life (Leopold 2012). The results presented here indicate that, in a given social context (by region and cohort), an extended period of co-residence with parents tends to promote close proximity and frequent contacts in later parent–child relationships. This finding is consistent with the idea that co-residence length offers the opportunity to share interests and activities, thus fostering pro-family attitudes and close family relations. It is also possible that familialistic attitudes encourage young adults to postpone leaving home in order to marry as well as to maintain high levels of solidarity with parents in later life. I suggest that the timing and pathways out of the parental home contribute to the development of individualistic/familialistic attitudes, rather than being mere consequences of pre-established family orientations. People who have experienced non-family living early in adulthood may be more prone to acquire a ‘taste for independence’ (Mulder and Clark 2002), which, in turn, favor less intensive family relations. Instead, leaving the parental home at a late age in order to marry can help to develop attitudes and preferences for high levels of proximity and contact with parents.
The redefinition of family relations during the transition to adulthood seems to depend on gender-specific expectations. Parent–adult daughter relationships are more likely to be affected by social expectations concerning both the normative age and the appropriate circumstance for moving out. As noted by Liefbroer and Billari (2010), Dutch women are more likely than men to perceive an age deadline for leaving the parental home. The results for Italian daughters reveal a ceiling effect of late nest-leaving on the frequency of later parent–child visits. Spending longer time in the parental home appears to provide much less benefit for Italian daughters than for sons. Age norms tend to define the time limit after which the general positive association between co-residence length and later parent–daughter relations ceases. However, a violation of the socially accepted time for leaving home is not significantly correlated with greater distances and fewer parent–child contacts in later life. Consistent with Settersten (1998)'s argument, violating the cultural age deadline for leaving home is not accompanied by interpersonal tensions and sanctions. Internalized age norms may play an important role in shaping life course decisions even if they are not backed up by external sanctions (Billari and Liefbroer 2007). In addition, as suggested by the literature on intergenerational ambivalence, parents can feel obliged to maintain frequent interactions even when the quality of parent–child relationships is low (Lüscher 2011). It is plausible that the detrimental consequences of extended co-residence on the quality of parent–daughter relations are compensated by parents’ feelings of intergenerational solidarity.
In line with previous research, this study indicates that marriage appears as the normative occasion for young adult daughters to leave the family nest (Goldscheider and Goldscheider 1999). Postponing independence to get married seems to comply with a culture of strong family ties, whereas other routes out of the parental home negatively affect geographical proximity and the frequency of parent–daughter contacts. Married daughters conform to norm-guided behavior and tend to receive approval from the family of origin in countries where traditional values are prevalent (Vignoli and Salvini 2014). Whereas in Nordic and Continental European countries, marriage tends to weaken parent–child relations, in strong family cultures, married individuals maintain more frequent intergenerational contacts than those who are single (Sarkisian and Gerstel 2008; Yahirun and Hamplová 2014). The results presented here underline the importance of considering the timing and synchronization between marriage and home-leaving in the study of later parent–child relations. It is also important to note that in some occasions, leaving the parental home to get married follows a previous long-term relationship with the partner. This may be related to a deeper involvement of parents in the couple relationship and may also facilitate feelings of closeness between generations.
My results also contribute to the debate on residential proximity, suggesting that early home leavers tend to move away from their parental home mainly because the structure of the labor market and the education system provides geographically circumscribed opportunities. Consistent with previous research (Bordone 2009; Kalmijn 2006), this finding supports the idea that, beside normative expectations, structural constraints have long-term consequences for later residential spatial moves and, indirectly, for parent–child contacts.
Three limitations of this study might guide future research agenda. First, age norms are investigated only in an indirect way, by using the group median age of leaving home. Future research could examine the role of social expectations directly expressed by parents and their children. Second, one could ask to what extent youth economic vulnerability may shape the meaning of intergenerational co-residence and later parent–child relationships. Due to data availability, I cannot observe a comprehensive set of indicators of economic pressures during the period of co-residence. Nonetheless, I am confident that my control variables about father's occupation and the first job in life can account for important aspects of economic conditions. Thus, the results suggest that the association between home-leaving age and later parent–child relations is not completely economically driven. Third, another question remains unanswered: can different experiences of intergenerational co-residence be partly explained by family climate during childhood and adolescence? Overall, more research on young adults’ life-course transitions is required to provide further insights into how intergenerational relations play out in their later life.
Acknowledgments
I wish to thank Marco Albertini, Michael Gähler, Carlo Buzzi, Milena Belloni, and the anonymous reviewers for their helpful comments on previous versions of this article. A draft of this manuscript was presented at the Espanet conference (University of Oslo, 2014) and the SUDA Demographic Colloquium (University of Stockholm, 2014).
Disclosure statement
No potential conflict of interest was reported by the author.
Notes on contributor
Marco Tosi is a sociologist of the family. He received a Ph.D. in Sociology and Social Research from the University of Trento in April 2016. His research interests include the transition to adulthood, the intergenerational exchange of social and economic support, and family relationships in ageing societies. His recent work on these issues has been published in Acta Sociologica.
Footnote
Late nest-leaving may be also associated with close proximity to parents, because it implies less time for additional spatial moves (Michielin and Mulder, 2007). However, similar results were found by including the variable ‘time since leaving the parental home’ in the analysis.