Abstract
In countries of Mediterranean Europe, characterized by strong family ties, close residential proximity between kin facilitates frequent exchanges of free mutual aid. This article focuses on the challenges to proximity posed by the spread of cohabitation in Italy. We consider an extensive definition of kin networks, including siblings and, for women in a couple, the parents and siblings of the partner. We confirm that couples who cohabit or who have cohabited, and those who after having cohabited went on to marry, less frequently went to live near their kin. However, some caution should be taken before arguing that the particular density of kin networks in Italy is destined to disappear.
Introduction
Relations and mutual exchanges between kin are important not only in developing contexts but also in all advanced societies. 1 That said not all developed countries are the same: thirty years of research reveal that “there is something special about Mediterranean Europe.” 2 In this vast area—which includes Italy, Spain, Portugal, Greece, and Romania, Albania and the former nations of ex-Yugoslavia—strong blood ties are an “original trait” of social organization. 3 The decline in agricultural society along with massive internal migration, the general spread of the nuclear family, and secularization did not weaken the strength of these ties; they continue to shape and reshape the social systems of these countries, influencing their social and economic organizations. 4
One of the most important characteristics of these societies is the close living proximity between kin: the custom of residing near one’s own parents and siblings is much more common in the Mediterranean compared to Central or northern Europe. When looking at the average of the ten countries that participated in the Survey of Health and Retirement in Europe investigation in the year 2000 (Sweden, Denmark, Germany, Holland, France, Switzerland, Austria, Italy, Spain, and Greece), we observe that 45 percent of individuals with children aged 50+ had at least one offspring living within twenty-five kilometers of their home. But the differences between countries are enormous, ranging from 25 percent in Sweden to 95 percent in Italy. 5 In Mediterranean Europe, this close proximity between kin facilitates frequent exchanges of free mutual aid, often substituting economic transfers inter vivos and shaping the care system. 6 For example, living proximity between parents and adult children allows grandparents to care for grandchildren, permitting both parents to work without having to incur additional expenses for the care of their children but also slowing pressure “from below” for care service institutions for the very young. 7
In Italy, studies of medium and long durations have shown that proximity between relatives persists over time, “substituting”—in a manner of speaking—the extended family in the northeast and center regions, and changing very little in the regions of the south, the northwest, and in large cities where the nuclear family was the more diffused form of family even during preindustrial times, but where relatives—in the past as today—sought to live near one another. 8
Given the significance of residential proximity between kin in the organization of societies in southern Europe, it is important to understand what will happen in the near future. This article focuses on the challenges to proximity between kin posed by the spread of cohabitation, that is, one of the most important aspects of the so-called second demographic transition (SDT). 9 Even in countries of southern Europe, new modalities of family formation are becoming ever more common, whereas they were almost absent thirty years ago. With specific regard to Italy, the most populous among the European countries characterized by strong family ties, we aim to understand whether the spread of cohabitation is weakening residential proximity between kin.
Kin Ties and the SDT in Italy
Over the last thirty years, the SDT has also spread in Italy, see Table 1. A number of authors have studied the effects of these profound demographic changes on kin ties. 10 To begin with, prolonged periods of low fertility tend to thin kin networks, particularly horizontal kinship. 11 The average number of living siblings and cousins for a young Italian twenty years of age declines from 8.9 for those born in 1930 to 7.2 for those born in 1960 and then to 4.5 for those born in 1990. 12 Blood ties thus become less numerous and structured, even if those that do persist can become even more resilient: in a certain sense, quality can substitute for quantity. This aspect has not, however, been researched extensively by scholars of Mediterranean societies. It has, however, been demonstrated that mutual help between generations increases when there is strong asymmetry between the number of relatives in the respective generations. For example, if a non-self-sufficient elderly individual has many children, the latter can share care activities: in 1998 among Italians forty to fifty years of age, the probability of having a parent in a retirement home was double among only-children compared to those of the same age with at least one sibling. 13
Forty Years of Diffusion of SDT in Italy.
Source: Istat = The Italian National Institute of Statistics.
Note: SDT = second demographic transition. aFirst marriages in 1970–1979, 1980–1989, 1990–1999, and 2000–2009. Calculations on Istat survey data, 2009. bMarriages in 1969, 1979, 1989, and 1999. Estimates by Castiglioni and Dalla Zuanna, 2008.
There are numerous studies on the impact of conjugal separations on ties between kin, considering both short-term and long-term effects. 14 Focusing on residence proximity in Europe, Heylen and Mortelmans find that divorced individuals distance themselves from their parents if they form a new union, while they tend to move closer if they remain single, especially if they have children. 15 However, the countries considered by these authors (mainly included in the European MULTILINKS Project) do not include those of southern Europe. To date, the impact of divorce on Mediterranean (and Italian) family ties has only been roughly sketched, in part due to the fact that the increase in conjugal separations is relatively recent. 16 An exception is an article on Italy by Albertini and Saraceno who show that the proximity of adult children to separated parents is less intense compared to that among adult children and nonseparated parents, especially if the parent has formed a new family. 17
The present article does not focus on these important aspects of the SDT, rather marital disruption and number of siblings and parents (-in-law) are included as covariates, so as to better isolate the association between cohabitation and proximity between kin.
Recent attention has been paid to the possibility that cohabitations may weaken relations between cohabiting couples and their respective families of origin. A study on Italy using data gathered in 1998 shows that the then relatively few married couples who had cohabited before marriage tended to live further from their parents, and their partner’s parents. This characteristic remains even when controlling for several variables linked to the spread of cohabitation, such as the interviewee’s level of education and that of their partner, their parents and their partner’s parents’ level of education, and region of residence. 18
Nazio and Saraceno look at this issue from a broader perspective, comparing Italy and UK data from the early 2000s (UK 2001–2002, Italy 2003). 19 The authors find the following main results: (1) Italian couples live closer to their parents compared to English couples, (2) at equal distances, contact between parents and children is more frequent in Italy, (3) while distance from parents is the same for both married and cohabitating couples in the United Kingdom, in Italy there is a notably greater distance among cohabiting couples, confirming the results of the above cited study using data from 1998, (4) English cohabiting or married couples have similar frequency of contact with their parents, while in Italy, married couples see their parents more often, and (5) however, once several characteristics of the two partners are controlled for—including distance from the parents’ home for both the man and the woman—even in Italy, the intensity of contact with parents is the same for both married and cohabiting couples. In summary, less contact between parents and cohabiting couples compared to that of married couples in Italy is not a sign of weaker intergenerational ties but rather is linked to the fact that cohabitating couples live further from their parents compared to married couples.
Two ethnographic studies provide additional insight into the connections between the spread of cohabitation and intergenerational ties. Grilli and Zanotelli considered Valdelsa, an area of Tuscany where, in the recent past, the patrilocal complex family was very widespread. 20 In some cases, youth see the choice to cohabit as a way of accelerating detachment from their family of origin, looking for housing far from their parents. But—more generally—similar mechanisms of solidarity between young married couples and their parents are reproduced among cohabitating youth and their families of origin. The authors describe specific life stories in which young couples even begin cohabiting within the family of origin of one of the two partners (more often that of the man, reproducing the traditional patrilocal rule), not by constraint, but because it seems to everyone that this is the most “natural” choice. Even if the young cohabiting couple does establish a separate residence, the amount of time spent at the paternal or maternal home remains substantial: “… well, on Sunday there was the beef stew that was waiting for us … Every once in a while in the summer, we would sleep here, at their house, if it got a bit late.” 21 Striking, in the in-depth interviews gathered in the above-cited book, is the continual intertwining of parents, in-laws, uncles, cousins, with constant overlap in all areas of life (work, affections, and home). The lack of marital status seems—at the end of the day—of little relevance. Even the bonds that form between grandparents and grandchildren seem to abstract from the institutionalization of the couple.
Belletti et al. conducted in-depth interviews with 118 young couples (fifty-eight married and sixty cohabiting, 75 percent of whom were under the age of 35) living in six Italian regions, focusing on the differences between married and cohabiting couples in terms of their relationships with their families of origin. 22 Almost all of the cohabiting couples reported that they had never had a problem with their families of origin, even if to various degrees of enthusiasm: “sometimes the parents showed strong acceptance (they were happy), in others there was less obvious agreement, but rather the result of a mentality that keeps up with the times, while still others simply supported the choice without actually getting involved (as long as you are happy).” Sharing the choice was, in general, less easy when the parents were practicing Catholics, even if this aspect counted less than one might imagine. The case of a couple of Latium is exemplary, “My mother, being very Catholic, didn’t agree with cohabitation: I always said that I would one day cohabit, that I would never marry and she said “I won’t come visit you anymore!” But then she was really happy that I am with her [actual partner].” 23
A third ethnographic study paints a more conflicting picture. In order to analyze parental influence on cohabitation in Italy, Schröder conducted 56 semi-structured in-depth interviews in the cities of Bologna (in Emilia-Romagna, northeastern Italy) and Cagliari (in Sardinia) with cohabiting women who had married following cohabitation. 24 The analysis reveals different levels of parental acceptance toward cohabitation: more traditional-oriented parents tended to criticize the informal union of their adult daughters, even if the daughters themselves defined their cohabitation as constituting a premarital step. In contrast, parents with more modern values and attitudes accepted any kind of cohabitation, that is, cohabitation as an alternative to marriage or as a trial period. And they did so without imposing any conditions. According to that reported by daughters, both more “traditional” and more “innovative” parents—particularly mothers—heavily influenced their decision to cohabit and to marry after a certain period of cohabitation. It was difficult for daughters to make decisions relative to cohabitation and marriage in open contrast to the views of their mothers, especially when the latter had a good level of education.
This last result confirms that observed in a quantitative study by Di Giulio and Rosina. 25 These authors show that the choice to begin couple life with cohabitation rather than with marriage for couples who wed in Italy during the last two decades of the 1900s is statistically linked not so much to the level of education of the children but to that of their parents. Children of more educated couples cohabited with greater frequency than children of less educated parents. According to the above authors, this can be explained by two reasons. First, the strong links between parents and children in Italy make carrying out a decision in open contrast to the views of parents difficult. Second, more educated parents are more likely to accept change regarding wedding customs.
The strong influence of the external environment on youths’ choice to cohabit has also been observed by Barbagli et al. in a comparison of different studies of youth carried out in Italy between the 1980s and the beginning of this century. 26 Cohorts born in the 1960s and 1970s were equally favorable to the idea of cohabiting. However, while the younger cohort thought that cohabitation “is accepted by society,” the older cohort—who in practice had cohabited very little—perceived widespread hostility toward this type of living arrangement. These authors—after having analyzed other data on opinions expressed and on material and immaterial exchanges between their youth and their parents—suggest that the perception of hostility toward cohabitation referred not so much to society in general, as to their family of origin.
The findings reviewed above led to the research focus of this article. The distance between the residence of married/cohabiting adults and their kin network is clearly a key aspect of the strength of intergenerational ties. Distance is strongly correlated with other potential indicators concerning the density of material relations between kin (Figure 1), confirming the results of recent research on Italy and other social contexts. 27 Different distances shape opportunities and obligations, as well as the means and ways of facing daily life and emergencies. 28 If the distance is large, material exchanges can only partially be substituted for by communication by telephone or computer. Nazio and Saraceno have already shown how, in the first phase of the diffusion of cohabitations in Italy (before 2003), cohabiting couples lived farther from their respective parents than married couples. 29 With regard to the same period, we endeavor to extend these results, increasing the range of indicators employed. Employing a more extensive consideration of kin networks, we go beyond measures of distance between children and their parents, including siblings and, for women in a couple, the parents and siblings of the partner.
Frequency of meetings with mother by distance between the two homes. Italian women, aged 30–59 who do not live with their mothers, interviewed in 2003.
The population examined in this study includes women in a couple and single Italian women, born between 1944 and 1973, who were living in Italy in 2003. The value of our study is its descriptive nature. We neither support nor verify direct causal links between the decisions to leave home, cohabit, marry, and establish a new residence near or far from parents (-in-law) or siblings (in-law). In particular, we by no means argue that cohabitation per se weakens strong family ties among Italians. First, we observe how—generation after generation—the “availability” of kin and, generally, of kin residing nearby (within one kilometer) changes. Second, we observe whether a greater or lesser distance from kin (parents, in-laws, siblings, and siblings-in-law) is statistically associated with having cohabitated/cohabitating and/or marrying, controlling for other socioeconomic and demographic factors. Third, we discuss whether—in the years after 2003—the density of kin networks will actually be threatened by the pervasive spread of cohabitation.
Data and Methods
We focus our analysis on the initial phase of the spread of cohabitation in Italy (mainly the last twenty years of the twentieth century; see Table 1). We compare, with regard to ascendant proximity (to own parents or those of the partner) and collateral proximity (to siblings and siblings-in-law), couples and individuals who experienced cohabitation and those who followed more traditional paths, never having cohabited. Data come from the Multipurpose Survey on the Family, carried out by Istat (The Italian National Institute of Statistics) in 2003 on a sample of 49,451 individuals and 19,227 families, statistically representative of the population. We first extracted from the sample 11,018 women of Italian nationality between the ages of thirty to fifty-nine. 30 The survey—the first wave of the Gender and Generation Program in Italy—reconstructs histories of cohabitation, marriage, conjugal separation, and filiation. Each woman and (if present) her partner were also asked questions concerning presence, distance, and contact with noncoresident parents and siblings. It is thus possible to observe various aspects of the SDT, including cohabitation 31 in relation to the distance between the residence of the woman and that of ascendant kin (parents and partner’s parents, if present) and collateral kin (own siblings and those of the partner, if present), or of her kin, defined here as the sum of ascendant and collateral kin.
The decision to consider the distance between siblings and—for women in a couple—distance from the partner’s parents and siblings seemed particularly useful (even if somewhat more difficult to calculate), given that such data allow for the construction of the entire family network, compared to the majority of studies on this topic which consider only the distance between the interviewee and own parents or children. In particular, consideration of proximity between siblings and in-laws provides a much more realistic picture of kinship networks, given the extraordinary density of these types of ties, especially for cohorts born when fertility in Italy was higher. 32 The novelty of this reconstruction justifies the attention paid in this article to the presentation of indicators aimed at measuring the size and density of the complete family network for individuals of different ages and family conditions.
We describe in detail familial conditions and the number of living kin, differentiating by age and the presence of a cohabitation (actual or prior). In light of this analysis, it made sense to exclude from further study women who are neither head of the household (HH) nor partners of an HH, in that the great majority of this group consists of individuals who have never left their family of origin, and hence an analysis of proximity to relatives means something else altogether.
For the HH women without a partner and—separately—for women in an HH couple or who are the partner of an HH, we analyzed proximity with respect to relatives. Thanks to the availability of data on siblings and parents (own and partner’s parents), we describe proximity by reconstructing the family typology of each woman, using as reference the home in terms of the different distances provided by the survey:
As is obvious, as the reference distance lengthens, complex families grow (Figure 2). We use one kilometer as the criteria of distance: with this proximity, variability is quite interesting, in that within a distance of one kilometer from the home, 59 percent of women in a couple have at least one kin member, 46 percent at least one ascendant, and 39 percent at least one collateral kin. In addition, one kilometer is a much more precise measure of same municipality, which depending on the demographic and territorial dimensions of the municipality of residence varies from one woman to another, while for greater distances, nuclear families disappear. Finally, a distance of one kilometer guarantees the possibility of frequent contact, even between relatives who move on foot: 66 percent of Italian women aged 30–59, who do not live with their mothers but reside less than one kilometer away, see their mother every day, whereas the same proportion is only 39 percent among those living within the same municipality (but at a distance greater than one kilometer), and 25 percent among those living in different municipalities but less than sixteen kilometers away (see Figure 1).
Proximity of families classified by different distances. Women in a couple, aged 30–59, interviewed in 2003.
To complete the description of family networks and proximity, we also calculate the average number of relatives (ascendant and collateral kin), the average number of relatives who live within the range of one kilometer, and the proportion of kin who within live less than one kilometer away with respect to the overall total of relatives. These synthetic indicators are also differentiated by age, family situation, and experiences of cohabitation.
The last part of the analysis aims to better measure the statistical relationship between the variable of interest (cohabitation) and the response variable (distance between relatives). We use logistic regression models, analyzing separately women in couples and those HH but without a partner, considering four response variables:
To better specify the statistical association between response variables and cohabitation, other covariates are progressively included in the model: demographic variables (history of marital disruption and age), kin network (number of living ascendant kin, number of living siblings [-in-law], and the number of cohabiting children), and other socioeconomic variables (education, type of resident occupancy, typology and size of municipality of residence, and region of residence). 33
Results
Spread of Cohabitation in Italy and Family Situations of the Women Concerned
Of the women aged 30–59, 11 percent are neither HH nor partners of the HH, 15 percent are HH but do not live with a partner, 66 percent live with their husband and have never cohabited, 5 percent cohabited before getting married, and 3 percent are currently cohabiting. 34 We briefly describe several demographic characteristics of these five groups.
The proportion of women who are neither HH nor partners of the HH decreases notably with age. The youngest women are mostly daughters of the HH, but with age other kinship relations grow, such as daughters-in-law and—the very few—cohabiting with the son of the HH and sisters (or sisters of the partner) of the HH. This family situation is not simply the result of the lengthy continued residence of youth within the family of origin, or a residual of the old complex family, common in relatively recent times in large parts of Italy. Its diffusion has been revived by the SDT, given that 9 percent of these women have experienced at least one conjugal separation, and this percentage increases to 12 percent for women in the 40–59 age-class.
Within the group of women who are HH and do not live with a partner, there may be greater social malaise, such that it is important to analyze residential proximity to parents and siblings. Among these women, 48 percent have at least one cohabiting child and 21 percent have at least two cohabiting children. In addition, among the older women in this group, 38 percent are widows. One can, however, observe situations that—at least a priori—seem less problematic: 48 percent of these women live alone. Finally, 47 percent have experienced at least one conjugal separation, and this percentage increases to 58 percent for women 40–49 years of age, who—compared to those 10 years older—separated more often and had less time to establish new couple relationships. The large majority of the women who had experienced a conjugal separation live in this type of family situation.
The proportion of women who at the moment of the interview lived with their husband and had never cohabited surpasses 75 percent for older women (aged 50–59), while it falls below 60 percent for those aged 30–39, both because among this group, cohabitation is more common and because a higher proportion still live with their family of origin.
The number of married women who had cohabitated before getting married (not necessarily with their actual spouse or only with him) increases dramatically by age. The proportion is four times higher in the 30–39 age-class than in the 50–59 age-class.
The difference between women aged 30–40 and 40–59 is less marked with regard to cohabitation at the moment of the interview, influenced by the spread—among the less youthful—of cohabitation following a conjugal separation.
In concluding this preliminary analysis, two aspects should be highlighted. The large majority of the interviewed women have a life story that is—in a manner of speaking—traditional, unmarked by cohabitations. Thus, even if more uxorio cohabitations have spread rapidly, the end of the twentieth century saw relatively few Italian adults participating in such behavior. As we shall see in the concluding section, things began to change in the decade that followed. Nevertheless, even if not yet very widespread, there is a sufficient number of women interviewed in the Multipurpose Survey of 2003 who had cohabitated to allow for a statistically significant analysis of residential proximity to their own kin and those of their partner.
Cohabitation, Family Situation, and Overall Availability of Kin
The probability of living near kin also depends on the number of living relatives. We first consider the number of kin of women aged 30–59 resident in Italy in 2003 (Table 2). Women in couples had on average 6.9 kin; this values decreases with age (7.4 for the 30- to 39-year-olds, 7.2 for the 40- to 49-year-olds, and 6.0 for the 50- to 59-year-olds) due to the diminishment in the number of living ascendant kin. The number of siblings instead varies little, a combined and contrary effect of fertility decline among the cohorts of mothers, born in the first half of the twentieth century, and the mortality of siblings, born during the same time period as the women object of this study. The number of ascendant and collateral kin does not vary much with family situation: the differences observed in the last column of Table 2 could seem significant, but in reality derive from extremely differentiated compositions by age. For women not in a couple, the cohort trends are rather similar, but the number of relatives is roughly halved compared to women in a couple, in that without a partner, there are obviously no relatives on the male side.
Average Number of Relatives.
Note: Women aged 30–59 interviewed in 2003 by age-class, and family situation. HH = head of the household. aThe number of living siblings is not available for this group, but we do have the overall number of siblings born. The average number of living siblings was thus estimated by multiplying the overall average number of siblings born by three survival coefficients calculated by us for the respective cohorts using the Human Mortality Database (Berkeley-Rostock). The three survival coefficients are 0.81 (probability of surviving to 55 years of age for the 1944–1953 cohort), 0.90 (probability of surviving to 45 years of age for the 1954–1963 cohort), and 0.94 (probability of surviving to 35 years of age for the 1964–1973 cohort).
STD, Family Situation, and Availability of Kin Within the Range of One Kilometer
We now turn to the main focus of this study, crossing the family situation of women aged 30–59 who no longer live with their parents—gathered in different moments of their life course, differentiating by their exposure to cohabitation—with the presence of kin within a distance of one kilometer (Table 3).
Kin Network Within the Range of One Kilometer (Own and Partner’s Parents and Siblings).
Note: a.v. = absolute value. Women aged 30–59 interviewed in 2003, by age-class and family typology. aAverage number of ascendant/collateral kin (own and partner’s parents and siblings) living less than one kilometer from the woman’s home (only the three closest siblings of the woman and the partner are considered, in that data on proximity is not available for the others). bPercentage of ascendant/collateral kin (own and partner’s parents) living less than one kilometer away with respect to all living kin (only the three closest siblings of the woman and the three closest siblings of the partner are considered—additional data on siblings data are not available). cWe consider only the woman’s kin, in that there is no partner present. This row is comparable with that of the row above (Her kin).
We begin with women in a couple, considering both the kin of the woman and her partner. Of these women, 46 percent live less than one kilometer from at least one ascendant kin and 39 percent from at least one collateral kin. If the criterion for distance is fixed at one kilometer, only 41 percent of these women live in a nuclear family. On average, women in couples have 1.71 kin within a distance of one kilometer (25 percent of the total available kin). These data vary considerably by age, primarily due to the progressive and rapid thinning of ascendants (who for the 30- to 39-year-olds are about 60–75 years old; for the 50- to 59-year-olds are about 80–95 years old). With age, due to the death of ascendants, the number of nuclear and multiple families increases, while the number extended and extended + multiple families decrease. The proportion of kin who live within the range of one kilometer is highest among the youngest women.
For all of the age-classes, women that cohabit or have cohabited live further from their own and their partner’s parents. For example, among the 30- to 39-year-olds – —still relatively untouched by either conjugal separations or the death of ascendant kin—the average number of kin within the range of one kilometer declines from 2.17 among married women who have never cohabited (29 percent of the total of kin available) to 1.42 among women who cohabit or who—in the past—cohabited (19 percent of the total of kin available). Women in couples who have experienced a separation also live further from kin compared to women who have never been separated. This result is, however, not all that interesting, in that it concerns a number of women too small to specify by age. Finally, among the women in a couple, we observe a tendency to live near the partner’s family (especially to his siblings). The differences are, however, small, testimony to the decline in traditional patrilocality, already evident in 1998. 35
We now consider women who—although HH—do not live with a partner, such that only the woman’s kin are considered. The average number of own kin within one kilometer is virtually identical to that of women of the same age who live with a partner (0.82 compared to 0.86), and we observe a somewhat greater tendency to live close by (27 percent of their kin live within one kilometer compared to 23 percent of women who live with their partner). Almost half of these women have experienced a conjugal separation, but this experience does not seem to have much of an impact on proximity to kin. On the contrary, women without partners who have prior experience of cohabitation have a much less dense family network compared to those who have never cohabited. In the subsequent section, we test whether these results are confirmed when controlling for other confounding variables, foremost age and the number of available kin.
Searching for a Statistical Association between Cohabitation and the Availability of Kin Within One Kilometer
In light of the analyses above, cohabitation seems to be associated with greater distance from kin. We now consider whether these results are confirmed by a regression analysis where we control for a greater number of the characteristics of women and—for women in a couple—their partners.
We once again begin with women in a couple (Table 4). Cohabitation (present or past) is associated with less dense kin networks within one kilometer in all of the models considered here, both when only characteristics of the life course and kin dimensions are considered as covariates and when other socioeconomic variables are included in the model. 36 This last result is important, in that it means that greater distance from the kin network for those who cohabit or have cohabited is not due to a composition effect with respect to the other characteristics of women and their partners (as, e.g., education), characteristics that can change notably from one generation to the next.
Relative Risks from Logistic Regression on the Probability of Living in a Certain Form of Family Within a Range of 1 km from the Place of Residence.
Note: m-high = medium-high; D.o.F. = degree of freedom. Couples with women aged 30–59 interviewed in 2003.
*p < .05. **p < .01.
The other variables inserted in the model also give interesting results relative to the factors associated with proximity among kin, even if less directly linked to the specific research objective of this article (i.e., the connection between cohabitation and proximity).
First of all, we observe no statistical association between proximity to kin and having had a past experience of conjugal separation.
Younger women seem less inclined to live far from their kin, as suggested in Table 3. Differences by age, however, disappear when the presence of ascendants is included. This means that—with equal ascendant network—proximity does not vary within the large age-class of 30–59 (women born between 1944 and 1973).
The number of living ascendant kin is not associated with proximity to siblings (of partner). On the contrary, proximity to parents (of partner) increases with the decrease in the number of siblings and the increase in number of children. With this type of data, one cannot obviously test causal links: it may be that couples without children can more easily “afford” to live far from their kin; it may also be that those who live near parents or siblings can more easily afford to have children. Similarly, those with few siblings may feel more obligated to live closer to parents in order to provide assistance, as they grow older. Or, parents who have few children could have been more easily able to acquire, for each of their offspring, housing closer to their own.
If among more educated couples the number without siblings in the range of one kilometer is more numerous, not much difference is noted in the proximity of parents. As already highlighted by Barbagli et al., this last result is important in that it suggests that the strong increase, cohort after cohort, in the proportion of Italians with a high school or university degree should not bring about a weakening of the strength of ties between parents and children. 37
Proximity is at its highest for couples who live in a house under a usufruct agreement (usually with a relative) and is also very strong among home owners when compared to tenants. This result also holds for proximity to siblings (of partner) and to parents of the partner, as already observed by Barbagli et al. for proximity to parents. 38 Proximity is closely linked to the dimension of the municipality of residence: the lowest levels are found in urban areas and their outskirts, the highest in small villages.
Finally, the geography of proximity does not reproduce the past geography of Italian areas characterized by simple families alongside of areas where complex families were prevalent. 39 Generally speaking, proximity is stronger in the poor southern regions; however, two of the richest regions of Italy—Lombardy and Trentino-Alto Adige—are not places where proximity to kin is less widespread.
We now consider the 1,615 women in our sample who are HH and live without a partner (Table 5). For this group as well, cohabitation (in the past) is clearly associated with a lower presence of kin within one kilometer, confirming the results of bivariate analysis.
Relative Risks from Logistic Regression on the Probability of Being in a Certain Form of Family in the Range of One Kilometer from the Place of Residence.
Note: HH = head of the household. Women HH Without a Partner Aged 30–59 Interviewed in 2003.
*p < .05. **p < .01.
For the HH women not in a couple, we also consider the relation between proximity and other covariates. First, it is interesting to note that for a woman, marital disruption influences proximity, independent of the reason (conjugal separation or the death of the husband). Those who are separated or widowed live further from their own family members, mainly siblings. 40
Differences by age disappear when the presence of parents is included in the model; moreover, as with couples, proximity to siblings is not associated with the presence of parents while—differently than is true of women in a couple—there is no significant link between the number of siblings and proximity to parents, although the direction of the association is the same as for women in a couple (i.e., those who have more siblings are more likely to live far from their parents).
The effect of the other variables on proximity is not much different from that observed for women in a couple: proximity to parents is greater for HH women with children, who own or have a usufruct agreement for the home they live in, who live in small municipalities, and who live in the south of Italy. Proximity to siblings is weaker among more educated women and among those who live in the Center-North.
Discussion
Compared to other work on residential proximity, this study benefits from a more extensive consideration of kin networks, going beyond measures of distance between children and their parents to include siblings and, for women in a couple, the parents and siblings of the partner. This additional analytic effort allows for a measure of the size of familial networks, differentiating by distance from living kin. In 2003, 59 percent of Italian women in a couple aged 30–59 lived less than one kilometer from at least one relative, and Italian women aged 30–59 had, on average, 1.7 relatives within that same range. Among HH women not in a couple, the network is obviously weaker, in that it does not include that of a partner. That notwithstanding, the numbers remain notable: 40 percent live less than one kilometer from at least one parent or sibling and have, on average, 0.86 relatives within the range of one kilometer.
Observing the entire kin network and controlling for a number of covariates gives results that are empirically consistent with our initial hypothesis: during the first phase of the spread of cohabitation in Italy (before 2003), couples who cohabit or who have cohabited less frequently live near to their kin (own parents or partner’s parents and own siblings or partner’s siblings). The difference compared to couples who have never cohabited is statistically and substantially significant. Even those who after having cohabited went on to marry, less frequently went to live near their kin, compared to those who have never cohabited.
These results should not be interpreted as causal effects, as the data used do not allow us to verify whether the decision to live away from the family of origin was taken before or after the decision to cohabit. Women’s desire to cohabit may, on the one hand, push them away from their kin. On the other hand, once women live further away from home, cohabitation might become an easier option. Furthermore, in order to better understand the interrelations between cohabitation and kin network, the entire process of entry into adulthood should be considered—from school and work life (lived more or less in proximity to the family of origin) to affective and sexual relationships, controlling for possible common antecedent factors. Finally, cohabitation itself should be considered more as a process than a defined, specific event, as reflected by the qualitative studies cited in the first section of this article.
We conclude this article with a reflection on that which has occurred since 2003, and what the future may hold. Over the last decade, cohabitations in Italy have rapidly continued to spread: births outside of wedlock increased from 12 percent in 2000 to 28.7 percent in 2012; the new 2009 Istat survey shows that in the first ten years of the new century, 27 percent of first marriages were preceded by a cohabitation, compared to just 11 percent in the decade 1990–1999. If cohabiters continue to establish their homes further from their parents, the rapid spread of this new approach to family formation could push even Italy toward relationships of exchange and proximity between generations similar to those observable in central and northern Europe; even if—independent of distance—contact between adult children and their parents are similar for Italian cohabiting and married couples. 41
However, some caution should be taken before arguing that the particular density of kin networks in Italy is destined to disappear. The 2009 survey shows that among women who began living in a couple in 2000–2009, differences in kin proximity among those who married without cohabiting first, spouses who previously cohabited, and cohabiters are very small, especially if compared to women who began living in couple in 1994–2003 and were interviewed in 2003 (Table 6). This convergence depends on the reduction in proximity among married couples at interview. 42 Thus, beyond a general tendency to live further from parents, young married and unmarried couples behave more similarly compared to ten years ago. As suggested by some studies mentioned at the beginning of this article, with the “normalization” of cohabitation, this choice seems to lose the characteristic of a “rupture” in relationships between generations. 43
Proportion of Women in Couple Living Within the Range of One Kilometer from Their Mother.a
aWomen aged 15–69, by couple typology and year at the beginning of union.
Moreover, additional elaborations of the 2009 Istat survey data suggest that the behavior of cohabiting and married people is in fact converging. Before the year 2000, the steps of leaving the parental home were profoundly different between individuals who began their ménage with cohabitation and those upon marriage. Of the former, 44 percent left the parental home before a union, whereas the same proportion was only 10 percent among the latter. For young people who began living together in 2000–2009, the same indicator does not change for cohabiters (45 percent), whereas the percentage rather doubles for people who married directly (19 percent). These numbers show that for recent Italian cohabiters as well, living together is often preceded by a period lived outside of the parental home (mainly for study or work purposes). That said, however, the diffusion of this behavior among people who began a marriage without cohabiting shows that the choice of cohabiting is now less atypical. 44
Some demographic trends might bring kin closer to one another. Our data show that women in a couple who have few siblings or partner’s siblings tend to live closer to their own or partner’s parents, compared to those who come from larger families. If in the future these regularities persist, then couples and their ascendant kin are destined to move closer to one another as—due to the effect of thirty years of Italian very low fertility—couples with few collateral kin increase. In addition, our analyses show that more “modern” individuals are not those who live further from their parents. Distance from ascendants of more educated couples and women is the same as that observed among less educated couples and women. Moreover, if it is true that proximity networks are denser in the South and in small municipalities, it is also true that—ceteris paribus—proximity is at its highest in Lombardy, the region home to Milan, the richest and most economically dynamic area of Italy. Finally, with equal size and structure of the kin network, we do not observe any tendency—among the youngest women who have left home—to distance themselves from their families.
Setting up a home near to kin is thus not a legacy of the past, something traditional to shake off as soon as possible. It is too early to speak of the decline of the Mediterranean family.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 320116 for the research project FamiliesAndSocieties.