Social support networks in Spain: The factors that determine models of choice

Abstract

Spanish

French

The main goal of this article is to understand social support models in Spain from the perspective of an individual facing a situation of need, whether domestic, economic or emotional. The data used come from the 2001 ‘Social Networks II’ survey of the International Social Survey Programme (ISSP). The study shows clear family-oriented models of preference, primarily between the closest members of the nuclear family of origin. The partner is especially important when domestic and emotional help are required, but parents come first when the need is economic. People’s behaviour also shows both traces of the more recent process of individuation in western society and differences based on the variables studied. The explanatory variables par excellence are gender, age, marital status and geographical distance.

Keywords

family network intergenerational exchanges social support Spain

Introduction

This article seeks to contribute to research in the field of social support networks. To date, the studies performed focus generally on aid received. The main goal of this investigation is to understand the models of social support from the point of view of an individual facing a situation of need that might occur at any point in life. More specifically, we wish to determine the factors that exert the most pressure on a person when he or she turns to a specific source of social support in cases of material and non-material need.

The data analysed come from the 2001 survey ‘Social Networks II’ of the International Social Survey Programme (ISSP). The module used includes a very important set of questions on the formation and structure of social networks and of informal support systems, key areas for the goal of our study. There are currently very few publications on social support in Spain.¹ The data used here come from the only publicly available survey to date that is representative at the national level for Spain.

The term social capital, which is deeply anchored in notions of social networks and social relations, has recently been gaining importance in sociology. Social capital is a multidimensional concept that has been defined differently depending on the perspective from which it has been considered. The different dimensions of social capital can in turn focus on two levels: the micro or individual level, developed, for example, in studies by Flap (1991), Lin (2001), Portes (1998) and Van Der Gaag and Snijders (2005) and the macro or collective level (Bourdieu, 1983; Coleman, 1988; Putnam, 2004). Since this article analyses the personal network of specific subjects and the individual’s preferences when in need of aid (the survey asks from whom the respondent would solicit aid in particular situations that any person might encounter), it makes sense to speak of social capital at the individual level. More specifically, we wish to indicate not use in itself but what Lin (2001) calls access to social capital, that is, an individual’s collection of potentially mobilizable social resources.

Following Lin (2001) on how to measure social capital and applying this method to social support, we find it is easier, more useful and more direct to focus on individuals’ ‘potential’ to access help. If we were to pause over ‘use’, we would have to take into account many additional elements (personal preferences, need for help, availability of institutional solutions, etc.) that might influence the measurement (Van Der Gaag and Snijders, 2005: 2). As to the kind of support analysed, we concentrate on three levels: aid in the home in case of sickness, economic aid and emotional aid. For example, the classification proposed by Van Busschbach (1996: 26–28), which has been gaining acceptance and is employed increasingly in social support research, refers especially to these dimensions. It indicates practical and instrumental support (services of a more mechanical and material nature), which in our case would be equivalent to ‘aid in the home in case of sickness’ and ‘economic aid’. We use the terms emotional support and companionship (support for psychological well-being and the subject’s affective need) here to refer to ‘emotional aid’.

Social support networks: Family and care

The social network is a theoretical concept applicable to any conceptual frame of reference. One of the definitions of ‘network’ includes all or some of the social units with which the individual or group is in contact. Mitchell (1969) labels this sense of the term a ‘personal’ or ‘ego-centred’ network. Our article develops around this meaning.

This study is concerned with the concept of network from the perspective of one function of social support or of the individual who needs help. We must thus distinguish between formal and informal sources of aid. For Bott (1971 [1957]: 142, 339), informal relationships have greater personal and emotional importance than the more specialized and formal networks. Studies since 1957 indicate that relatives have special significance in any kind of network.

Intergenerational transfers beyond the nuclear family still constitute a new research field.² According to Kohli and Künemund (2003: 123), it is only in the last decade that sociology has rediscovered the full reach of the family as a system of kinship that exceeds the nuclear household and includes numerous exchanges and flows between its members.

Intergenerational exchange in families is multidimensional. Parents and children share not only material but also emotional and physical support (Bengtson and Roberts, 1991: 861). Instrumental or physical help generally refers to tasks related to the ordinary functioning of the home. The help that the person in need of support receives comes mainly from one’s descendants, especially from those who live in the same home. Children are a very important source of assistance in most family-oriented environments (Hoff, 2007; Kalmijn and Sarraceno, 2008; Meil, 2004; Starrels et al., 1995).

Another frequent type of help is material aid, which involves the flow of money or goods within the family. Here, the person with the greatest capacity to lend support plays an important role – generally relatives in the older generation. Economic aid flows essentially from parents to children (Attias-Donfut, 2000; Freeman and Ruan, 1997; Kohli and Künemund, 2003; Meil, 2007; Silverstein and Marenco, 2001), from ascending to descending generations.

No less important is the emotional support provided by the person to whom one turns to confide personal matters or when feeling sad or depressed. Different studies show that relatives in the same generation are among the most common confidantes. Despite siblings’ role as a key link in the family network, most kinship studies focus more on intergenerational than on intragenerational relationships (Allan, 1977: 177). The role of friends is also relevant. Much research on friendship targets the elderly (Arth, 1980; Roberto and Scott, 1986; Zunzunegui et al., 2003).

For Kohli and Künemund (2003), it is crucial to understand the motives associated with such transfers within the family in greater depth, as they constitute a significant way of explaining the behaviour of exchange. However, this is a complex model in which different reasons for choices overlap and interact and diverse variables come into in play, among them the kind of relationship, style of socialization and sociodemographic factors.

Likewise, concepts of social network and social support pay special attention to the aspects of the social environment that influence people’s well-being. Researchers who examine social structures have documented, for example, how social networks define opportunities and limitations for the development of relationships and exchange of resources (Cox and Burchinal, 2002; Degenne and Forsé, 1994). According to Nadel (1957), who has an eminently normative conception of society, societies rest on rules and laws and thus on frequencies of behaviour of various types, which they call institutions or institutionalized modes of action, relations and groupings.

Applied to the area of care, although what we understand as ‘aid’ differs from one person to another based on the individual’s representation of social norms (Petite, 2005), in Mediterranean countries we find that providing solidarity and a culture of support are considered the norm in the overall population (Böhnke, 2007: 27; Freeman and Ruan, 1997; Lowenstein and Daatland, 2006).

It is clear, however, that a general social concern for informal care has arisen in recent years. We see this change in the establishment in both Europe and Spain of laws, plans and measures to regulate some of these questions, in particular the reconciliation of family and work life. According to Tobío (2005), the reality in Spain is that most women have entered the labour market – most of them full-time – yet their new role is still not recognized and accepted as it should be, in either family organization or social policies for the care of dependants.³ Recent data show that, even in homes where both members of the couple work, we see marked asymmetry between the rapid and passionate entrance of women into professional activity and men’s lagging behind in the area of family and the domestic. Because this change occurred later in Spain than in other European countries and is dominated by women of the younger generations, we are only beginning to see indices of attitudinal change in fathers concerning care tasks (especially childcare). There is still a lack of assimilation and unity between the new model of labour and the old model of family.

Research questions

We propose two research questions. First, we must ask whether there are differences between the family members to whom people turn when they need different kinds of help. That is, we must deepen our understanding of the social support network. We thus analyse three kinds of need, which we have defined as ‘domestic in case of sickness’ (help with the house, shopping, etc.), ‘economic’ and ‘emotional’. The second research question explores the explanatory factors for these decisions. We investigate whether choices vary based on the variables that prior work has shown to be most influential on behaviour from our literature review. These variables are gender, age, education, marital status and geographical distance.

Hypotheses and variables

H1: Gender

This article’s first study hypothesis is that women facing need at a specific point in their lives will turn more often than men to different members of the family network. Diverse studies that analyse the variable of gender maintain that a general pattern emerges in research that focuses on aid and exchanges between members of the family network at different stages of life. This pattern stresses the role of the woman as the main caretaker (Kalmijn, 2007; Tobío, 2005; Van Gaalen and Dykstra, 2006). Women are also portrayed as the subjects who receive support more often than men (Brandt et al., 2009; Meil, 2007; Roberto and Scott, 1986).

H2: Age

Patterns of aid within the individual’s network vary with age. We propose that, the younger the interviewees are, the more they will turn primarily to their parents when in need of support. In contrast, older interviewees will be more likely to choose other options (partner, friends, etc.). Different studies on the variable of age in intergenerational relationships (e.g. Hoff, 2007; Meil, 2007; Silverstein and Marenco, 2001) show this, particularly when one needs to borrow a specific amount of money.

H3: Education

Another variable that we predict may influence the model significantly is education. As to direction, the lower the education level, the greater the inclination to family. Numerous authors agree in confirming that low- and medium-level educational strata are more focused on family than are high (Allan, 1977; Böhnke, 2007; Kalmijn, 2006). In general terms, a high level of education is associated with more resources and thus less dependence on or expectation of aid from the family network (Kalmijn and Sarraceno, 2008; Lee et al., 1994). Further, a high level of education is usually associated with employment in a labour market restricted to specific areas, often requiring geographical mobility. Such constraints contribute to less contact with the family network and greater levels of individualism (Kalmijn, 2006; Lee et al., 1994).

H4: Marital status

We start from the premise that people without partners receive more aid from different members of the family network. This is seen most clearly, for example, in the case of the widowed population. Some research shows that widows receive support from their children in greater proportion than the population with partners (Kalmijn and Saraceno, 2008; Silverstein et al., 1995). Other authors, such as Lee et al. (1994), maintain that married parents receive less aid from their children.

H5: Geographical distance

We expect that geographical proximity among members of the family network implies family solidarity or greater amounts of aid among relatives. Studies focusing on intergenerational exchange of aid tend to consider geographical distance as an opportunity for exchange (e.g. Brandt et al., 2009; De Jong-Gierveld et al., 1991; Kalmijn, 2006; Meil, 2007; Silverstein and Bengtson, 1997).

Control variables

We have introduced the following variables as controls: occupation, habitat, family composition and family responsibility. These are variables in the bibliography consulted that the research predicts to be able to influence the choices in either direction.

Data and methodology

The database used is the 2001 survey ‘Social Networks II’ of the ISSP. The survey was administered in a total of 29 countries. In Spain, it was administered to a representative sample of the population over 18 years of age (1214 interviews) who live in private households.

As to the kind of methodology employed, we began with a descriptive analysis to provide a preliminary general view and to respond to the first research question. We then developed an explanatory analysis to answer the second research question: What affects the individual’s decision or the factors that determine the choice to seek a particular kind of support? Given that we have multiple choices in our dependent variable but only one decision between more than two non-ordered alternatives, we estimate a multinomial logit model for each kind of aid included in the questionnaire.

The three dimensions (domestic, economic and emotional) that act on each model respectively as dependent variables are formulated in the questionnaire as follows:

‘Domestic aid in case of sickness’: p. 67. We now wish to ask you how you would request aid in specific situations that any person might encounter. First, suppose that you have the flu and must stay in bed for a few days. You need help at home, with shopping and other housework. Whom would you ask for help first? (Please mark only one box)

‘Economic aid’: p. 69. Now imagine that you need to ask to borrow a large amount of money. To whom would you turn first to ask for help?

‘Emotional aid’: p. 71. Now imagine that you feel somewhat sad or depressed and would like to talk to someone about it. To whom would you turn first?

Given the broad mix of possible answers and the number of individuals who selected each choice, we recoded the responses into a smaller number of categories. For family members, our criterion consisted of grouping related people by the character of the link that defines their relationship: spouses, vertical relationships (based on kind of affiliation: members of the ascending or descending generation) and relationships within the same generation. We also considered close friends.

When we speak of conjugal relatives, descending and horizontal, we refer primarily to women (wives–daughters–sisters). A frequency analysis shows that the percentages assigned to women in these categories in the three modalities of aid studied are much higher than those for men (husbands–sons–brothers). Although gender roles are changing, there is still no evidence that men are taking the role of caretaker, especially in the Spanish society analysed here.

While the response alternatives were recoded to establish the same options for the three kinds of support, recoding was not possible for the modality ‘economic aid’. Unlike the other minority options, the survey’s option ‘bank’ could not be reformulated as ‘others’ because a high number of individuals (21.9% of the population) placed themselves in this category.

Due to the low percentage registered, the other responses (no one, other relatives, work colleague, psychologist/counsellor, social services, hired help, another person, etc.) were converted to missing.

We find one last nuance. In the case of domestic aid in Spain, the entrance of women into the labour market – combined with the phenomenon of immigration, particularly the benefit to immigrants after the regularization of immigration in 2005 – has encouraged the increase of employment of domestic service. According to data from the Spanish National Institute for Statistics, Spain has a greater number of domestic workers than any other EU country (close to 4% of the total population employed in the first trimester of 2009). This phenomenon has occurred, among other reasons, in response to the rapid growth in the need for attention to dependents. However, this article is not able to take domestic workers into account as one of the response choices, primarily because the data from the ISSP refer to the year 2001.⁴ At that time, the interviewees who said that they turn to this kind of aid constituted a very low percentage (0.4%).

Our choice of independent study variables was based on those discussed in the previous section. We used the same variables for the three models: sex (categories: ‘male’, ‘female’); age; education (measured in years of study);⁵ occupational status (recoded into the categories ‘employed’, ‘unemployed’, ‘inactive’); marital status (recoded into the categories ‘lives with a partner’, ‘does not live with a partner’); habitat (recoded into the categories ‘urban’, ‘rural’); number of members in the household; geographical distance (time it takes to reach the mother’s house⁶) (recoded into the categories ‘less than half an hour’, ‘from half an hour to three hours’ and ‘more than three hours’); and family duty, understood as the obligation children feel towards their parents.⁷

All of the qualitative variables or categories were transformed into dummy variables. The classifications of independent variables that were fixed to act as ‘reference categories’ for the other groupings established in the variables are as follows:⁸ for sex, ‘male’; for occupation, ‘unemployed’ and ‘inactive’; for marital status, ‘does not live with a partner’; for habitat, ‘urban’; and for distance from the mother’s house, ‘less than half an hour’ and ‘from half an hour to three hours’.

Results and discussion

Except in the case of economic aid, the descriptive analysis turned out not to agree with the results of our review of previous studies, which show that descending relatives are those primarily asked when in need of help in the home, whereas friends and siblings are preferred when one seeks emotional support. The analysis performed demonstrates that the partner is a key element of social support in Spanish society. This role is superseded only by ascending relatives when individuals require monetary support.

It is noteworthy that a formal source of support (the bank) appears under economic aid as the second choice. In a comparative, international study of interpersonal behaviour and role relationships, Freeman and Ruan (1997) refer especially to the rapid development of the market economy and, ultimately, the process of industrialization and urbanization experienced in western societies. Undergoing this process has prepared these societies to offer new support alternatives (for example, through well-established institutions or social services) and an increase in opportunities and resources beyond those of the family. As our data reveal, this is also true in the case of Spain.

We can now take a step further and examine the influence of different variables. After developing the logit models, we see how the variables selected influence the choice of one alternative or another, giving rise to interesting nuances (see Tables 1, 2 and 3).

Table 1.

Marginal effects of the probability of asking for ‘domestic aid’; multinomial logit model.

Variables	Spouse	Ascending generation	Descending generation	Relationships in same generation	Close friend
Woman	−.380***	.253***	.069***	.037**	.022*
	(.061)	(.057)	(.026)	(.018)	(.013)
Age	.003	−.006**	.003***	.001	−.000
	(.003)	(.003)	(.001)	(.001)	(.000)
Education	−.005	.009*	−.002	−.002*	.001
(years of study)	(.007)	(.007)	(.002)	(.002)	(.001)
Employed	.145**	−.137**	−.017	.006	.003
	(.067)	(.063)	(.017)	(.011)	(.008)
Lives with partner	.738***	−.302***	−.085*	−.156*	−.195**
	(.051)	(.105)	(.061)	(.084)	(.095)
Urban	.058	.010	−.019	−.038*	−.011
	(.071)	(.060)	(.021)	(.027)	(.015)
No. household members	−.000	−.015	.008*	.004	.005
	(.027)	(.024)	(.006)	(.005)	(.004)
Over 3 hours from mother’s home	.056	−.146***	.076*	.004	.010
	(.083)	(.059)	(.049)	(.022)	(.018)
Family duty (children’s obligation to care for parents)	−.058*	.041*	.009	.004	.005
	(.034)	(.030)	(.008)	(.007)	(.005)
No. observations	363
Log likelihood	−267.63381
Likelihood ratio test	226.13***
Pseudo-R²	.2970

Notes: *** indicates statistical significance at 1%, ** 5% and * 10%. The standard errors are shown in parentheses.

Source: ISSP, 2001. Developed by the author.

Table 2.

Marginal effects of the probability of asking for ‘economic aid’; multinomial logit model.

Variables	Spouse	Ascending generation	Descending generation	Relationships in same generation	Close friend	Bank
Woman	.012	.813	.004	−.023	.016	−.091*
	(.054)	(.064)	(.005)	(.034)	(.014)	(.056)
Age	.008***	−.015***	.001	.002	.001*	.004*
	(.003)	(.004)	(.000)	(.002)	(.001)	(.003)
Education	−.005	.003	−.001	.003	.000	−.000
(years of study)	(.006)	(.007)	(.001)	(.004)	(.001)	(.006)
Employed	.039	−.034	−.005	.011	.006	−.017
	(.056)	(.067)	(.006)	(.037)	(.012)	(.062)
Lives with partner	.266***	−.264***	.000	−.101*	−.067*	.166***
	(.041)	(.088)	(.004)	(.068)	(.048)	(.052)
Urban	.048	−.008	−.001	−.036	−.049*	.046
	(.054)	(.069)	(.004)	(.042)	(.029)	(.056)
No. household members	−.043*	.049*	−.002	.024*	.001	−.029
	(.025)	(.029)	(.002)	(.014)	(.005)	(.025)
Over 3 hours from mother’s home	.174**	−.067	.009	−.016	.019	−.119
	(.091)	(.091)	(.012)	(.045)	(.029)	(.066)
Family duty (children’s obligation to care for parents)	−.010	.009	.002	.032*	−.003	−.031
	(.029)	(.034)	(.002)	(.019)	(.006)	(.029)
No. observations	347
Log likelihood	−444.8586
Likelihood ratio test	127.16***
Pseudo-R²	.1250

Notes: *** indicates statistical significance at 1%, ** 5% and * 10%. The standard errors are shown in parentheses.

Source: ISSP, 2001. Developed by the author.

Table 3.

Marginal effects of the probability of asking for ‘emotional aid’; multinomial logit model.

Variables	Spouse	Ascending generation	Descending generation	Relationships in same generation	Close friend
Woman	−.256***	.079**	.007	.049*	.121***
	(.060)	(.035)	(.008)	(.034)	(.044)
Age	.003	−.003*	.001*	−.003*	.002
	(.003)	(.002)	(.001)	(.002)	(.002)
Education	.002	.002	−.001	−.006*	.003
(years of study)	(.007)	(.004)	(.001)	(.004)	(.005)
Employed	.098*	−.040	−.017	−.003	−.039
	(.066)	(.039)	(.013)	(.033)	(.046)
Lives with partner	.628***	−.049	−.079*	.018	−.517***
	(.064)	(.049)	(.055)	(.037)	(.091)
Urban	.012	.042	.000	.003	−.057
	(.066)	(.031)	(.008)	(.034)	(.052)
No. household members	−.028	−.014	.004	.011	.027
	(.026)	(.014)	(.003)	(.014)	(.017)
Over 3 hours from mother’s home	.067	−.082***	.028	.001	−.014
	(.079)	(.029)	(.025)	(.047)	(.055)
Family duty (children’s obligation to care for parents)	.011	.005	.006*	−.009	−.013
	(.031)	(.016)	(.005)	(.017)	(.021)
No. observations	346
Log likelihood	−315.20358
Likelihood ratio test	151.23***
Pseudo-R²	.1935

Notes: *** indicates statistical significance at 1%, ** 5% and * 10%. The standard errors are shown in parentheses.

Source: ISSP, 2001. Developed by the author.

By sex, the data show that H1 is supported; when women need help they would turn to a greater extent than men to different members in the family network (ascending and descending generation and siblings). This result attests to women’s greater investment in emotional and caring tasks in the family assumed throughout history, as a result of socialization and traditional division of gender roles. Because this dedication can become a strong nexus for union, especially with children, it influences the aid that women expect to receive later on. We must question, however, some of the data related to the social norms mentioned in the theoretical framework. For example, when the wife offers emotional support, she probably does not consider this as ‘aid’. Although this is difficult to see from the data, it is necessary to be aware of this kind of limitation.

On examining traditional gender roles in Spanish culture, we see the result of the model of the patriarchal family based on the division of labour by gender. Jobs performed by each sex receive different social characterization, giving rise to a cultural distinction that associates rationality with men and emotion with women (Comas, 2000). As Parsons (1949) states, the husband or father acts as the ‘instrumental leader’ and market provider and the wife or mother as the ‘expressive leader’, caregiver and nurturer.

We find another interesting contribution, namely a greater inclination to seek support from friends on the part of women respondents than men. This behaviour has previously been identified in some important studies which were principally focused on the elderly population. For example, Zunzunegui et al. (2003) show that women over 65 are more likely to maintain strong emotional links with friends or peers than are men. Men’s friendships tend to be based on shared activities, whereas women’s friendships are more intimate and intense, tending to focus on conservation and mutual support. Along the same lines, Roberto and Scott (1986) emphasize that friendships are more extensive and more significant for women than they are for men.

In the case of age, H2 is supported for the three kinds of aid analysed. The modality of aid in the individual’s social network varies with age. Indeed, the younger one is, the more often one’s parents are the main source of support. As the individual’s age increases, however, he or she turns to other options. In H2, we proposed that people would turn to different sources of support (partner, friends, etc.). It is important, however, that in the case of Spanish society primarily descending relatives emerge as responses. Spain is one of the countries where children live longest with their parents, primarily due to the precarious situation of the labour market (high unemployment rate and job instability). Living together seems to have its effect. Our study confirms the close relation between parents and children in this country.

Does a person’s educational level affect the choice of one member or another in the network? H3 is partially supported except in the case of economic aid. In general, interviewees with a lower level of education would turn to family, especially their siblings, in case of need. But even for higher education levels, there are interviewees who indicate the family network (e.g. parents in the case of domestic aid). In examining these results and applying them to the context of Spanish society, we conclude that more education is not necessarily associated with less preference for family. In fact, as stated above, cohabitation of children and parents in Spain is high, and geographical mobility among the population is low. Relatives usually live near each other. This would explain turning to the family network in any of these cases. We must also take into account the relatively low significance of this variable.

In the case of education, we can conclude primarily from studies that focus on aid received (e.g. Kohli, 2004; Starrels et al., 1995) that there are theoretical reasons to think that education may have a different effect on this variable according to gender. Given the possibility that there may be an interaction effect between these two variables in our research (that is, the effect of education on the need for aid may differ between the genders), we have estimated the different models including interaction between the two variables. The results remain almost unchanged, however. None of the interactions is significant except that of the likelihood of requesting domestic or emotional help from a close friend. Here the positive effect of education is lower for women. Likewise, the negative effect of education on the likelihood of requesting economic aid from one’s descendants is slightly less among women.

The marital status of the population is shown to be an especially influential factor. H4 is supported in all three cases: people with partners or spouses receive aid primarily from this person; those without partners obtain domestic, economic and emotional support from the different members of the family network (principally their parents). As Lee et al. (1994: 564) maintain, it is important to reflect on the expectations for filial responsibility and their relation to different dimensions of the parents’ resources. For example, parents who are married and have high education and income levels and better health have fewer expectations of receiving aid from their children than others in this group. Those who need less help expect less.

Further, the analysis shows the importance of close friends (in the case of emotional support they become the interviewees’ second choice over descendants), particularly in the group of the population without a partner. Although friends’ contributions are more public than those provided within the family, they are closer than the formal support mechanisms.

Yet many studies show that the intensity of the relationships or exchanges between family members is conditioned by multiple factors. Geographical distance between the components of the family network is especially important (Kalmijn, 2006). The industrial and professional system is constituted to make individuals mobile. As a result, relatives can be separated both socially and spatially. This fact fosters the creation of more heterogeneous social or family networks (Bott, 1971 [1957]: 159). However, as we have already indicated, the context is different in Spain. Although the likelihood that several generations live together in the same household is decreasing, it is still common for relatives to live nearby, a pattern that leads to a high density of family relationships. The results of the Survey of Health, Ageing and Retirement in Europe (SHARE)⁹ (2005) identify Spain as one of the countries in which the elderly population has more extensive family networks. Further, it is less likely that this age group lives alone, as the survey also registers greater presence of children living at home. In addition, the physical distance between parents and children is smaller than in other countries.

Based on what we expected and in agreement with H5, geographical distance is highly relevant in the model. People who live closer to their mother’s house turn more often to their parents when they experience any type of need than do those who live farther away. Respondents who live at a distance of more than three hours away turn more often to their partner, especially when they need money. They choose their children when seeking aid in the home. We were correct in predicting that H5 would be significantly confirmed in all of the support situations analysed in this study. In Spain, as in other countries, geographical proximity provides an opportunity for exchange. It is strange, however, that among those who live farther from the family network, children are a significant source of support, as in the case of domestic aid. This result reinforces once again the importance of prolonged cohabitation between parents and children in Spain.

Employment (respondent’s current occupational status) shows a significant marginal effect associated with spouse and the older generation when one needs help in the home. Individuals who work would be more likely to turn to their partner than would the unemployed and inactive. This situation also occurs in the case of emotional need. In contrast, those who are not working turn primarily to their parents, mainly because they live together and because of parental potential for support. The role of parents in the immediate family in Spain is crucial as the main safety net in difficult situations, especially in the case of unemployment.¹⁰

Although little research focuses on the influence of the variable habitat, the existing literature affirms that the family network is more extensive and finds a greater probability of turning to family and friends to ask for help in rural environments (Bott, 1971 [1957]; Requena, 2008; Scott and Roberto, 1987). Our research corroborates this situation in part. Habitat shows only a weak significance level in domestic and economic aid. The data here show that, for help at home, the probability of turning to siblings is 15.63% higher in rural environments than in urban. For economic support, the likelihood of turning to close friends is higher, probably due to the greater proximity and frequency of contact. In non-urban environments, the networks on which individuals rely are more extensive.

The same situation occurs when we analyse the number of members in the household. This is not a particularly influential variable. Bengtson and Roberts (1991) view number of children as a dimension of structural solidarity. More children provide more opportunities for interaction and exchange (Lee et al., 1994). In fact, our study shows that living together is in itself a form of aid. The greater the number of people in the household, the greater the support among the different members, enabling the possibility of turning to descending relatives in a situation of domestic help. In the case of economic aid, parents and siblings come to take a leading role.

Finally, although the influence of family duty is also weak in the model, we find the tendency we expected. People with less sense of obligation to their parents would be more likely to ask for help from a partner than would those with stronger responsibility for family. Individuals with a stronger sense of commitment would choose different members of the family network. For household aid, they are 4.14% more likely to choose the ascending generation than are those with less sense of duty. Siblings occupy first place when the interviewee needs money. In a study of the Spanish population in the Autonomous Community of Madrid, Meil (2004) states that siblings appear to be those cited most frequently as beneficiaries of money loans. When the aid solicited is emotional, children are the primary confidantes chosen.

The fit of the three models is good – highly significant; likelihood ratio test: 226.13***, 127.16*** and 151.23***. The Hausman test of each model indicates that they fulfil the IIA criterion (Interdependence of Irrelevant Alternatives). The alternate hypothesis is confirmed (see Tables 4, 5 and 6 of the Appendix).

Conclusions

The study shows models of preference concerning social support networks in Spain that are clearly family-oriented, especially between members who are closer to the original nuclear family. For example, the partner is especially important in the case of support in the home and for emotional aid. When we look at economic aid, primacy shifts to relatives in the ascending generation. Most of the older people support the younger members of their families in Spain. Further, the current situation in the Spanish labour market favours intergenerational solidarity in the face of unemployment among young people and the effects of unstable or precarious work situations. This is true mainly for those who have just obtained their first job. Young people continue to live with their parents, who provide them with various kinds of support. Children also hold important positions as potential sources of sustenance, and many people, especially women, turn to them when seeking aid.

In the case of economic need, the bank is the second option that people choose. Because western societies have undergone a process of industrialization and urbanization, they are increasingly better prepared to provide new alternatives for support. This finding fits with the more recent process of individuation in the nuclear family and the changes in behaviour occurring in western societies.

Our reflection on the first research question has deepened our understanding of the social support network in different situations of need in Spain, being the first study of this nature in the country. This investigation explores fields where no research has been carried out in the past. It has allowed us to establish facts that were not possible to find before, and offers the chance to compare our case with other countries. For example we could start to make comparisons with countries with similar or different welfare levels to see what similarities and differences exist between them.

We can now establish for the second research question, the factors that most influence these decisions, that gender (except for the case of economic need), age, marital status and geographical distance have been the explanatory variables par excellence in the three models analysed. This result confirms the predictions established in the hypotheses along the lines we expected. The exception is the variable education (H3), which shows a weak level of significance, and only in the cases of domestic and emotional aid. Thus, in Spanish society, surprisingly and unlike what happens in other countries, educational level is not a determinant in explaining the composition of a person’s social support network. The family continues to be the primary choice in the Spanish population, independently (regardless) of education, reinforcing the general idea of the important role that families have in this population.

This study has some limitations. Among these, the topic of social networks is a profoundly complex matter and requires further investigation. As to the data used, while the 2001 survey ‘Social Networks II’ of the ISSP contributes interesting information on the social networks of those surveyed, it does not offer enough data on the extent of exchanges between generations (Meil, 2007). The analysis offers only one perspective from one member in the network. It would be interesting to complement the data, for example, by asking each family in greater detail about the relationships of family members, friends and other components of the network. Further, there is no response choice ‘grandchild’ or ‘grandparent’, an addition that would provide interesting nuances from the perspective of intergenerational analysis. The study also lacks an in-depth analysis from the perspective of gender, specifically of the division of roles and the influence they have on care. Although the results of the analysis have shown these differences, the research goals were not directed along these lines.

Following Parsons (1949), we can say that there is bad integration between availability and resources in Spain. We see that sex is taken as the main criterion to determine roles. Although highly educated women are needed for the development of the economy in southern countries, the system used to assign roles still sees women as the caretakers. Today we can find highly educated women who are ‘pushed off’ the job market into the caretaker roles. While those who are fighting to stay in their jobs encounter difficulties in taking part in the socialization of their children. One possible solution may be the institutionalization of caretaking activities; in Spain, however, we do not find the actual structures that could do just that (Alexander, 1987). In countries with more developed welfare states women are not so depended upon by their families. In line with H1, we see that informal support is part of the Spanish culture.

In this study we present new information on explanatory factors that helps us understand the nature and value of the person who needs help at a specific moment in his or her life. As Hirsch (1981: 151) would argue, if we do not identify specific members in the network, we will be unable to obtain precise data concerning the relations between members of the network. The information provided can also be useful in designing policies and working in a cultural environment where informal aid is still preferred when seeking care. Finally, we have developed a new study in the field of the family as social group, an area still insufficiently studied. It would be interesting to advance in this line of work by comparing, for example, the case of Spain to other countries in different welfare state regimes.

Footnotes

Appendix

Table 6.

Hausman test of the assumption of Independence of Irrelevant Alternatives: Emotional aid.

Omitted	χ²	d.f.	p > χ²	Evidence
Spouse	−28.578	30	1.000	In favour of H₀
Ascending	−1.593	29	1.000	In favour of H₀
Descending	−2.410	30	1.000	In favour of H₀
Same generation	−5.016	30	1.000	In favour of H₀
Close friend	1.463	29	1.000	In favour of H₀

Source: ISSP, 2001. Developed by the author.

Funding

This article is part of the study PSI2008-01937, financed by the Ministry of Science and Innovation in the National Plan for R+D+I 2008-2011.

Acknowledgements

I gratefully acknowledge the many helpful comments and suggestions that I have received from the anonymous reviewers. I also thank Professors Jaime and Gamero for their comments and for their help during the preparation of this article, and Professors Gobernado, Requena and Ortega for their supervision of the final manuscript.

Notes

References

Alexander

(1987) Twenty Lectures: Sociological Theory Since War World Two. New York: Columbia University Press.

Allan

(1977) Sibling solidarity. Journal of Marriage and the Family 39(1): 177–184.

Arth

(1980) American culture and the phenomenon of friendship in the aged. In: Tibbitts

Donahue

(eds) Social and Psychological Aspects of Ageing. New York: Arno Press, pp. 529–546.

Attias-Donfut

(2000) Cultural and economic transfers between generations. The Gerontologist 40(3): 270–272.

Bazo

(2002) Intercambios familiares entre las generaciones y ambivalencia: una perspectiva internacional comparada. RES. Revista Española de Sociología 2: 117–127.

Bengston

Roberts

(1991) Intergenerational solidarity in ageing families: An example of formal theory construction. Journal of Marriage and the Family 53(4): 856–870.

Bengston

Giarruso

Mabry

. (2002) Solidarity, conflict, and ambivalence: Complementary or competing perspectives on intergenerational relationships. Journal of Marriage and the Family 64(3): 568–576.

Böhnke

(2007) Poverty and social integration in the enlarged Europe. Social Science Research Center/WZB Discussion paper, Berlin.

Bott

(1971 [1957]) Family and Social Network. London: Tavistock Publications.

10.

Bourdieu

(1983) Forms of capital. In: Richardson

(ed.) Handbook of Theory and Research for the Sociology of Education. New York: Greenwood Press, pp. 241–258.

11.

Brandt

Haberkern

Szydlik

(2009) Intergenerational help and care in Europe. European Sociological Review 25(5): 585–601.

12.

Coleman

(1988) Social capital in the creation of human capital. American Journal of Sociology 94: 95–120.

13.

Comas

(2000) Mujeres, familia y Estado del bienestar. In: del Valle

(ed.) Perspectivas feministas desde la antropología social. Barcelona: Ariel, pp. 187–204.

14.

Cox

Burchinal

(2002) Structural and supportive changes in couples family and friendship networks across the transition to parenthood. Journal of Marriage and Family 64(2): 517–531.

15.

Degenne

Forsé

(1994) Les Réseaux sociaux: une analyse structurale en sociologie. Paris: Armand Colin.

16.

De Jong-Gierveld

Liefbroer

Beekink

(1991) The effect of parental resources on patterns of leaving home among young adults in the Netherlands. European Sociological Review 7(1): 55–71.

17.

Flap

(1991) Social capital in the reproduction of inequality. Comparative Sociology of Family, Health and Education 20: 6179–6202.

18.

Freeman

Ruan

(1997) An international comparative study of interpersonal behaviour and role relationships. L’Année Sociologique 47(1): 89–115.

19.

Hirsch

(1981) Social networks and the coping process: Creating personal communities. In: Gottlieb

(eds) Social Networks and Social Support. London: Sage, pp. 149–170.

20.

Hoff

(2007) Patterns of intergenerational support in grandparent–grandchild and parent–child relationships in Germany. Ageing and Society 27(5): 643–665.

21.

Kalmijn

(2006) Educational inequality and family relationships: Influences on contact and proximity. European Sociological Review 22(1): 1–16.

22.

Kalmijn

(2007) Gender differences in the effects of divorce, widowhood and remarriage on intergenerational support: Does marriage protect fathers? Social Forces 85(3): 1079–1099.

23.

Kalmijn

Saraceno

(2008) A comparative perspective on intergenerational support. European Societies 10(3): 479–508.

24.

Kohli

(2004) Intergenerational transfers and inheritance: A comparative view. In: Silverstein

(ed.) Intergenerational Relations across Time and Place, Annual Review of Gerontology and Geriatrics, Vol. 24. New York: Springer, pp. 226–289.

25.

Kohli

Künemund

(2003) Intergenerational transfers in the family: What motivates giving? In: Bengtson

Lowenstein

(eds) Global Aging and Challenges to Families. New York: Aldine de Gruyter, pp. 123–142.

26.

Lee

Netzer

Coward

(1994) Filial responsibility expectations and patterns of intergenerational assistance. Journal of Marriage and the Family 56(3): 559–565.

27.

Lin

(2001) Building a network theory of social capital. In: Lin

Cook

Burt

(eds) Social Capital, Theory and Research. New York: Aldine de Gruyter, pp. 3–30.

28.

Lowenstein

Daatland

(2006) Filial norms and family support in a comparative cross-national context: Evidence from the OASIS study. Ageing and Society 26(2): 203–223.

29.

Meil

(2000) Encuesta sobre relaciones familiares. Universidad Autónoma de Madrid.

30.

Meil

(2004) Cambios en las relaciones familiares y en la solidaridad familiar. Arbor 702: 263–312.

31.

Meil

(2007) The consequences of the development of a beanpole kin structure on exchanges between generations. Journal of Family Issues 27(8): 1085–1099.

32.

Mitchell

(1969) The concept and use of social networks. In: Mitchell

(ed.) Social Networks in Urban Situations. Manchester: Manchester University Press, pp. 1–50.

33.

Nadel

(1957) The Theory of Social Structure. London: Cohen and West.

34.

National Institute for Statistics in Spain (various dates) Available at: www.ine.es.

35.

OASIS project (2000–2003) Available at: oasis.haifa.ac.il

36.

Parsons

(1949) The social structure of the family. In: Anshen

(ed) The Family, its Function and Destiny. New York: Harper, pp. 173–201.

37.

Petite

(2005) Les Regles de l’entraide. Sociologie d’une pratique sociale. Rennes: Presses Universitaires de Rennes.

38.

Portes

(1998) Social capital: Its origins and applications in modern sociology. Annual Review of Sociology 24: 1–24.

39.

Putnam

(ed.) (2004) Democracies in Flux: The Evolution of Social Capital in Contemporary Society. New York: Oxford University Press.

40.

Requena

(2008) Redes sociales y sociedad civil. Monografías collection, No. 256. Madrid: Centro de Investigaciones Sociológicas.

41.

Roberto

Scott

(1986) Friendships of older men and women: Exchange patterns and satisfaction. Psychology and Ageing 1(2): 103–109.

42.

Scott

Roberto

(1987) Informal supports of older adults: A rural–urban comparison. Family Relations 36(4): 444–449.

43.

Silverstein

Bengtson

(1997) Intergenerational solidarity and the structure of adult child–parent relationships in American families. American Journal of Sociology 103(2): 429–460.

44.

Silverstein

Marenco

(2001) How Americans enact the grandparent role over the life course. Journal of Family Issues 22(4): 493–522.

45.

Silverstein

Parrott

Bengtson

(1995) Factors that predispose middle-aged sons and daughters to provide social support to older parents. Journal of Marriage and the Family 57(2): 465–475.

46.

Starrels

Ingersoll-Dayton

Neal

. (1995) Intergenerational solidarity and the workplace: Employees’ caregiving for their parents. Journal of Marriage and the Family 57(3): 751–762.

47.

The Survey of Health, Ageing and Retirement in Europe (SHARE) (2005) Available at: www.share-project.org/

48.

Tobío

(2005) Madres que trabajan. Dilemas y estrategias. Madrid: Cátedra, Colección Feminismos.

49.

Van Busschbach

(1996) Uit het oog, uit het hart? Stabiliteit en verandering in persoonlijke relaties. Amsterdam: Thesis Publishers.

50.

Van Der Gaag

Snijders

TAB

(2005) The resource generator: Social capital quantification with concrete items. Social Networks 27: 1–29.

51.

Van Gaalen

Dykstra

(2006) Solidarity and conflict between adult children and parents: A latent class analysis. Journal of Marriage and the Family 68(4): 947–960.

52.

Zunzunegui

Alvarado

Del Ser

. (2003) Social networks, social integration, and social engagement determine cognitive decline in community-dwelling Spanish older adults. Journal of Gerontology 58B(2): S93–S100.