Skip to main content

How demographic changes shape intergenerational solidarity, well-being, and social integration: A multilinks framework

Final Report Summary - MULTILINKS (How demographic changes shape intergenerational solidarity, well-being, and social integration: A multilinks framework)

The objective of MULTILINKS was to investigate how changing social contexts, from macro-societal to micro-interpersonal, affect social integration, well-being and intergenerational solidarity across different European nations. Debates on ageing societies predominantly focus on the circumstances of the old. Our approach built from three key premises. First, ageing affects all age groups: the young, the middle-aged and the old. Second, there are critical interdependencies between family generations and between men and women. Third, one must recognise and distinguish analytical levels: the individual, dyad (parent-child, partners), family, region, historical generation, and country. Building from these premises, we examined: multiple linkages in families; multiple linkages across time; multiple linkages between, on the one hand, national and regional contexts, and on the other hand, individual behaviour, well-being and values. Throughout the project we tested, developed and used methodological strategies that enable sound policymaking.

To understand the family/state division of responsibility for the old and the young, three patterns in legal and policy frameworks were distinguished. These patterns pertain to the degree to which country-specific institutional frameworks support the desire to be responsible towards one's children and frail aged parents and/or support individual autonomy, thereby partially lightening intergenerational dependencies and the gendered division of family labour. The first pattern is 'familialism by default', where there are few or no publicly provided alternatives to family care and financial support. The second is 'supported familialism', where there are policies, usually in the form of financial transfers, which support families in keeping up their financial and caring responsibilities. The third is 'defamilialisation', where needs are partly addressed through public provision (services, basic income, pensions). The three patterns exemplify that public support may both be an incentive for and lighten private family responsibilities. A database with comparative indicators for all EU-27 countries (and Norway, Russia and Georgia) of legal and policy frameworks shaping financial and caring responsibilities in families (gendered intergenerational regimes) was developed.

The indicators represent the allocation of responsibilities to the state or to families for caring for children, financially supporting children, caring for frail older persons and financially supporting older persons. Care was taken to harmonise information across countries and to be explicit about decisions taken in quantifying the indicators. The usefulness of the MULTILINKS database was illustrated in several studies. A key message for policymakers is that national policies should seek to support intergenerational care regimes without reinforcing social class inequalities and gender inequalities. A consortium of nine partners organised in six teams carried out the project. All partners are involved in the Gender and Generations Programme, a system of nationally comparative surveys and contextual databases, which aims at improving the knowledge base for policymaking in UNECE countries. The consortium represented a wide range of substantive, methodological and policy expertise.

Project context and objectives

Debates on ageing societies predominantly focus on the circumstances of the old. In MULTILINKS, we felt that a change of focus was needed, which started from three key premises. First, population ageing is not only about older persons: it affects people of all ages. In debates on ageing societies, there seems to be an implicit assumption that demographic ageing primarily affects older persons, their economic situation, health, mobility, social integration, family support and care. Of course, increasing longevity and decreasing birth rates have resulted in larger numbers of older persons both in absolute and relative terms. Nevertheless, with dramatic shifts in the balance between old and young, the worlds of younger age groups are profoundly changed. The young are growing up in societies where they are a numerical minority and where they have several generations of family members 'above' them. These considerations suggest that attention should be given to people of all ages.

The new demographic circumstances in which members of multiple family generations share several decades together compel us to recognise that individuals are embedded in a complex web of vertical and horizontal ties. Thus, a second key premise is that there are critical interdependencies between family generations and between men and women in families, which are built and reinforced by social policies. These interdependencies should not be taken for granted. Rather, it is important to explicitly address the ways in which legal and policy arrangements constitute differential opportunities and constraints for men and women and across generations in families.

A third key premise is that to understand interdependencies in families, a spectrum of levels and units must be distinguished and recognised: country, historical generation, family, dyad (partners, parent-child) and the individual. Countries have disparate political, religious and economic histories and different welfare state arrangements. To understand the impact of demographic changes on people's lives, it is not sufficient to consider cross-national differences only. Regional diversity, including urban-rural differences and social change over time must also be considered. The key premises represent a MULTILINKS framework.

We examine:

(a) multiple linkages in families (e.g. transfers up and down family lineages, interdependencies between older and younger family members);
(b) multiple linkages across time (measures at different points in time, at different points in the individual and family life course);
(c) multiple linkages between, on the one hand, national and regional contexts (e.g. policy regimes, economic circumstances, normative climate, religiosity), and on the other hand individual behaviour, well-being and values.

The main objective was to investigate how changing social contexts, from macro-societal to micro-interpersonal, affect social integration, well-being and intergenerational solidarity across different European nations. To achieve this objective, MULTILINKS researchers have simultaneously worked on five partial objectives:

1. To gain insight into the consequences of demographic changes for family constellations in which Europeans are embedded

Changes in mortality, nuptiality and fertility patterns are reflected in living arrangements (households) and family structures (generational constellations). Questions addressed here were: How many generations surround individuals in different phases of life? How complex are today's families? How many have two generations with women only? Who is family-deprived, lacking intra- and intergenerational family ties, living alone? How common are only children? How do patterns differ across contexts? Knowledge about family constellations is important for forecasting and the creation of social policies and programs. When the implications of demographic developments for family change are considered, negative messages prevail. The following arguments tend to be put forward. Persistent low fertility means fewer children and siblings to call on for help. The concentration of demands for help in the oldest-old category means that children and siblings are too old to provide help by the time it is needed. The extension of life expectancy means that for several decades of adulthood individuals have family generations above and below them with competing needs. The rise in divorce means that fewer adults have partners to provide assistance, whereas more adults have histories of disruption.

Though these arguments are intuitively appealing, they are misleading, inaccurate and lack a sound empirical basis. In this project, we addressed the inaccuracies of popular views and improved the empirical evidence regarding the consequences of demographic change for European family constellations.

2. To gain insight into the linkages between family constellations and intergenerational solidarity, well-being, and social integration

Indicators of intergenerational solidarity in MULTILINKS were emotional and practical support, financial transfers, care and contact. Indicators of well-being were psychological and physical health, functioning and financial well-being. Indicators of social integration were embeddedness in networks, labour force participation and loneliness. Contrary to what is customary, we considered transfers 'up' and 'down' family lines together. The dominant perspective in ageing research is on the amounts of care for frail old people provided by family members, including spouses and adult children. We showed that the older generations often serve as significant sources of support and help for young families, through financial transfers, caring for young grandchildren and provision of practical help.

Regarding intergenerational solidarity, key questions included: what care strategies are chosen in different contexts? Do middle generations attempt to provide help equally up and down? To what extent do they prioritise the young over the old? Regarding well-being, key questions included: Do 'generational squeezes', i.e. facing care needs of young grandchildren or children at the same time that parents or grandparents are in need of help negatively affect health and well-being? Are singlehood and childlessness associated with lower levels of well-being? Regarding social integration, key questions included: what labour force participation strategies are chosen in different contexts? Do people refrain from working outside the home because of care responsibilities? Does help with caring tasks from spouses, adult children and parents enable people to hold jobs?

3. To identify policy regimes under which intergenerational constellations of care needs and resources in younger, middle-aged and older age groups constitute risks, and correspondingly, to identify policy regimes under which relationships and individual well-being seem positive

Both in policy and research discussions, care of old people and care for children are typically addressed in separate discourses. When 'family policy' or 'family friendly policy' is the topic, the focus is nearly exclusively on young children and their parents, adults who are struggling to meet family and work obligations. Eldercare, typically, is not linked to family policy, but discussed under 'health policy'. This separation is highly unfortunate, as it disregards complex interdependencies across generations. In MULTILINKS, we started with a consideration of how policies regarding public and private responsibility for childcare (parental leaves, day care and preschools) and available care provision for the frail old (home services, assisted living, institutional care and leaves for family members) combine into intergenerational care regimes in different societies and in different regions of a given country. Though several authors have attempted to combine policies regarding the young and those aimed at the old into typologies, no previous research had empirically explored family and individual consequences of living under a given type of regime. Subsequently, we focused on the links between intergenerational care regimes and strains at the dyadic (couples, child-parent) and individual level. By identifying intergenerational care regimes and their shortcomings, we contributed to the substantive understanding of the risks of becoming socially isolated and/or lacking necessary supports.

4. To gain insight into the linkages between values and intergenerational solidarity, well-being, and social integration

A focus on values draws attention to ideational and cultural determinants of behaviour. In analyses of demographic changes, structural and cultural narratives have lived side by side. Values played a three-part role in our project. First, their function as a driving force behind behaviour was relevant in analyses of family solidarity. Intensive exchanges of support were observed among those with strong feelings of responsibility towards family members. We confirmed the patterns reported in earlier work: marked east-west and North-South contrasts in views of responsibilities towards parents and children, as well as in views of the balance of public and private responsibility for care provision. Second, the function of values as an evaluation criterion proved to be relevant in analyses of well-being. Lower levels of well-being were more often observed among those who expect more from others than they are actually receiving and among those who give less support than they feel they ought to. Inconsistencies between values and behaviour seemed to create psychological stress, which in turn prompted changes in both behaviour and values. The consideration of subjective factors provided greater insight into well-being than the consideration of objective determinants only.

5. To gain insight into the role of different levels of context in shaping differences in intergenerational solidarity, well-being, and social integration

A range of contexts was considered: national policies, regional levels, family constellations, household and generational structures. A common problem with cross-national comparisons is a poor understanding of the source of observed between-country differences. It is often difficult to find out whether the differences are attributable to country characteristics, population composition and/or differential effects of individual characteristics by country. By employing multilevel models, we disentangled the effects of different levels of context, including variability on an intra-national level. We did not restrict our analyses to individual-level and country-level determinants. Where possible and feasible, regional levels were taken into consideration. An interesting theme concerned systematic differences between men and women in effects of different levels. Our findings suggest that women's lives and well-being are shaped more by micro-level contexts than is the case for men.

Project results

The demography of intergenerational family relationships

The conventional portrayal of family change under the influence of macro demographic trends is that the extension of life and the drop in birth rates result in 'beanpole' families with relatively many vertical ties and relatively few horizontal ties. Micro data from the generations and gender surveys (GGS) provide a more nuanced view. Contrary to popular belief, vertically extended families with four or five generations alive at the same time are not the norm. The majority of adults are members of three-generation families. However, we would find greater proportions of four- and five-generation families if we included the very old (greater than 80 years of age) and the very young (less than 10 years of age) in our samples. Note that mapping generational structures is a tricky task. For example, respondents might be part of a three-generational lineage on the paternal side of the family, and part of a five-generational lineage on the maternal side.

Increased longevity and postponed childbearing have opposing effects on the generational structure of families. The extended lifespan means, on the one hand, that older family members are living longer than they did in the past, which in turn suggests that three, four or even five generations of family members may be alive at the same time. Delayed childbearing means, on the other hand, that the age gap between generations is relatively large, which in turns reduces the likelihood that multiple generations are alive at the same time. Micro data make it possible to examine the opposing effects of increased longevity and postponed childbearing on the generational structure of families. Increased longevity and postponed childbearing have opposing effects on the generational structure of families. The extended lifespan means, on the one hand, that older family members are living longer than they did in the past, which in turn suggests that three, four or even five generations of family members may be alive at the same time. Delayed childbearing means, on the other hand, that the age gap between generations is relatively large, which in turns reduces the likelihood that multiple generations are alive at the same time.

Micro data make it possible to examine the opposing effects of increased longevity and postponed childbearing on the generational structure of families. For example, the proportions in one-, two-, three- and four-generation families are virtually identical in France and in Russia. The underlying demographic processes are quite different, however. In France, where people tend to live long lives, adults have relatively many ascending family generations. In Russia, where people tend to have children at a young age, adults have relatively many descending family generations. MULTILINKS' findings give little credence to the metaphor of the sandwich generation, the men and women caught between simultaneous responsibilities for their parents and children. Adults typically occupy middle-generation positions between the ages of 30 and 60. This is not a period in life when both young children and elderly parents are likely to need care. For those in the younger part of the age range (i.e. those with childcare responsibilities), parents are rarely at risk of frailty. For those in the older part of the age range (i.e. those caring for their parents), offspring will generally lead independent lives. Though the metaphor of a sandwich generation juggling care commitments towards parents and children is clearly a misconception of midlife, it continues to figure prominently in public and policy debates.

In gerontology, decreasing fertility rates over recent decades have received a considerable amount of attention, mostly motivated by a concern for parent care in years to come. It is important to keep in mind that standard measures of fertility are woman-based and therefore not perfectly indicative of the number of children in families with children. The decline in fertility means a decrease in average number of children per woman. It does not necessarily mean that the average number of children among mothers is dramatically lower. A critical factor to consider is patterns of childlessness.

Little attention has been paid to individuals who are vertically deprived in the sense that they have no children or grandchildren, or no surviving parents or grandparents. Whereas an examination of childbearing and mortality patterns informs us about the existence of biological kin, an examination of divorce and separation provides insight into a different form of vertical deprivation; having severed ties. Men are more likely to have broken family ties than women. One should not assume that all adults are part of multigenerational family structures. The focus on a presumed growth in multigenerational families has made researchers and policymakers overlook the substantial number of generational solos: individuals without any direct ascendant or descendant generational links.

Generational interdependencies

The new demographic circumstances in which members of multiple family generations share several decades together compel us to recognise that individuals are embedded in a complex web of interconnected ties. In MULTILINKS, we did not separate the old and young in families, but considered them jointly. In policy and research communities there tends to be a split between discussions on responsibilities for children and responsibilities for old people. 'Family policy' usually refers to young families. Issues related to the old come under different headings: 'ageing policy', 'long-term care policy', or 'caregiver burden'. The separation of care and financial policies in 'young' and 'old' sections is unfortunate because it disregards similarities between the young and old and overlooks interdependencies across generations. It also provides a 'chopped up' notion of what families are about. In all advanced societies, responsibility for the old and the young is shared in some manner between the family and the state. Interdependencies between generations (and between men and women) in families are built and reinforced by the legal and policy arrangements in a particular country. Laws define the relationships of dependence and interdependence between generations (and genders), whereas policies reward or discourage particular family patterns and practices. A consideration of legal norms and public policies draws attention to cultural specificity. Countries differ in their understanding of 'proper' intergenerational family relations. For that reason, it is difficult to disentangle culturalist and structuralist explanations of cross-national differences in patterns of intergenerational exchange in families.

To understand the family/state division of responsibility for the old and the young, three patterns in legal and policy frameworks have recently been distinguished in the context of MULTILINKS. These patterns pertain to the degree to which country-specific institutional frameworks support the desire to be responsible towards one's children and frail aged parents and/or support individual autonomy, thereby partially lightening intergenerational dependencies and the gendered division of family labour. The first pattern is familialism by default where there are few or no publicly provided alternatives to family care and financial support. The second is supported familialism, where there are policies, usually in the form of financial transfers, which support families in keeping up their financial and caring responsibilities. The third is defamilialisation, where needs are partly addressed through public provision (services, basic income, pensions). The three patterns in legal and policy frameworks go beyond the public / private responsibilities dichotomy. They make clear that public support may both be an incentive for and lighten private family responsibilities.

Policy indicators database

Comparative indicators for all EU-27 countries (and Norway, Russia and Georgia) of legal and policy frameworks shaping financial and caring responsibilities in families (gendered intergenerational regimes) have been developed over the course of the research programme. Note that the indicators represent policy frameworks, not service usage. The indicators represent the allocation of responsibilities to the state or to families for caring for children, financially supporting children, caring for frail older persons, and financially supporting older persons. The database focuses on 2004 (the year of GGP-data collections) and changes since then, using existing indicators as far as possible (OECD, Eurostat and Missoc). Information was partly collected through national informants. Care was taken to harmonise information across countries (e.g. financial support is related to the average net national income level). Care was also taken to be explicit about decisions taken in quantifying the indicators. The database has 71 indicators; 48 pertain to responsibilities for the young, and 23 pertain to responsibilities for the old. The database will become publicly available before the end of 2011. The usefulness of the database was illustrated in several studies.

Opportunity structure

Geographic proximity facilitates face-to-face contact, which in turn increases the likelihood of exchanges of help in kind. Frequent face-to-face contact not only reduces the costs of giving, but also helps to make support providers aware of recipients' needs. Exchanges of financial support are less affected by distance because they do not require interaction in person. Until recently, research rarely included East European countries, where coresidence of generations is widespread. Intergenerational co-residence (i.e. adults living with their parents) is among the strategies that can be adopted to organise support, economic and otherwise. Levels of coresidence are particularly high in Ireland, parts of Spain, Italy, Hungary and Poland. In those countries, over half of the 55-plus live with adult children. Generational economics, an interaction of generational interdependence and economic resources, are behind coresidence. Insufficient independent income precludes young adults and the elderly to maintain their own households. General economic malaise restricts this more widely, but specific housing markets may keep young people in their parental home and the absence of affordable public residential and home care may necessitate intergenerational coresidence for older adults in need of assistance.

When parents and adult children share a household, the direction of intergenerational transfers is not always clear. Who is supporting whom? Compared to previous data collection efforts, the GGS have the advantage that they include East European countries and have information on exchanges with family members both in and outside the household. We examined flows of assistance up and down generational lines in coresidential households. The direction of assistance tends to be downward. 'Assistance' pertains to providing help with household tasks, personal care and financial transfers. Most often, older parents are helping their adult children (over 60 % of cases). The older adult is the primary recipient of assistance in less than 5 % of coresident households. Contrary to popular belief, coresident living arrangements generally respond to the needs of adult children rather than those of the elderly parents. This finding is consistent with research on older adults living independently, which has repeatedly shown that the direction of intergenerational support flows tends to be primarily downward. Again, this is contrary to popular belief. Parents become net beneficiaries of help only at an advanced age (> 80 years). Parents want to be parents and maintain helping patterns long after the children grow up.

Normative obligations in families

Family obligations are generalised expectations about family members' responsibilities for each other. They are socially shared and have a normative component. Not only do they reflect the cultural climate in which people live, but also the individual circumstances in which they find themselves. Family obligations are of interest because they are predictive of support behaviour: they predispose people to behave in a certain way. Family obligations are also of interest because they serve as a source of information for policymakers. The answers to questions about people's wishes for care and about the types of care people are prepared to give provide insight into the extent to which policy measures are in keeping with public attitudes. They also offer tools for developing policy that enables or promotes the application of personal preferences. Support for norms of family obligation tends to be lower in generous welfare states. East Europeans are more likely than are West Europeans to endorse the statement that children should take responsibility for caring for their parents when their parents are in need. A stronger east-west contrast emerges for the item 'Children should adjust their working lives to the needs of their parents'. The latter item alludes to greater commitment and sacrifice on the part of children. Given the more limited public welfare system in East than in West European countries, it should not come as a surprise that Bulgarians, Russians, Romanians and Hungarians more strongly believe that it is important to provide help to family members in need than do the Dutch, Germans, French, and Norwegians.

Research carried out in the context of MULTILINKS revealed that a focus on attitudes provided a better understanding of intergenerational exchanges. There are three hypothetical types of families: in the first type, both mother and grandmother have traditional attitudes, in the second type, both the mother and the grandmother have modern attitudes, in the third type, the mother has modern attitudes but the grandmother has traditional attitudes. Traditional mothers are unlikely to work outside the home, whereas modern mothers are likely to have paid jobs. Traditional grandmothers are likely to help their offspring, by providing childcare, for example, but they are probably less in favour of mothers working outside the home. Modern grandmothers are unlikely to help and provide childcare, but they are probably supportive of female employment. It is assumed that traditional grandmothers would generate a negative estimate in estimates of the link between grandparental care and maternal employment. Their traditional attitudes would be linked with a greater willingness to care for grandchildren, but impose a restriction on daughters' working decision. Analyses showed a positive association between care provided by grandparents and maternal employment in each of the seven Generations and Gender Survey countries under investigation (France, Germany, the Netherlands, Hungary, Bulgaria, Georgia and Russia). However, as predicted, the strength of the association was biased by the attitudes of the mothers and grandmothers. After correcting for the attitudes, a 'true' effect was found only for France, Germany, Hungary and Bulgaria. A 'true' effect implies that care by grandmothers helped mothers to maintain jobs outside the home.

Supportive exchanges in families

Earlier work has shown that patterns of exchange in families tend to follow a north-south gradient. Intergenerational transfers of time and money among non-coresident family members tend to be less frequent in the Nordic than in the Southern European countries, with the continental European countries being somewhere in the middle. MULTILINKS research work has included East and Central European countries. In general, more support is found for the 'family resilience' than for the 'family decline' hypothesis. People living in families with multiple generations are more instead of less likely to provide emotional and practical support to their children. However, there is a tendency among the 'sandwich generation' to economise on emotional support given to their parents, compared with the support given to their children. Findings underscore the gendered nature of exchanges in families: daughters tend to be more heavily involved than sons in providing care, domestic assistance and emotional support to ageing parents. Gendered roles stressing daughters' kin keeping and daughters' presumed expertise in carrying out typically feminine tasks are among the underlying mechanisms.

Contrary to popular belief, the direction of intergenerational support flows tends to be primarily downward. Parents become net beneficiaries of help only at an advanced age (more than 75 years). Parents want to be parents and maintain helping patterns long after the children grow up. The focus on intergenerational dependencies has prompted new research questions. Studies on labour force exit, for example, have tended to look at the retiring generation in an isolated way ignoring the intergenerational ties that may play a role in the retirement decision. Researchers examined whether in countries with legal and policy arrangements allowing early retirement, grandchild care might encourage older workers to leave the labour force before the official retirement age. Their findings suggest that grandparenthood speeds up retirement, particularly for older women. Across all countries participating in the European Social Survey, just over 50 % of grandmothers who were previously active on the labour market had retired before reaching the age of 60, as compared to 37 % of the women without any grandchildren. Among men, the difference between grandfathers and men without any grandchildren was smaller: before reaching the age of 60, 27 % of the grandfathers had already retired and 23 % of the men without grandchildren. Another example of a new research question prompted by the consideration of generational interdependencies is a study on fertility decisions carried out by researchers. Adult children receiving help with childcare from their parents were more likely to have another child, but more strongly so in southern Europe than in Northern Europe. Moreover, siblings of the adult children receiving grandparental help were less likely to have another child, presumably because they were unable to turn to their already busy parents for help.

Mental well-being

One of the aims was to investigate how demographic changes in family structure affect people's mental well-being. Research focused on changes associated with parenthood and partnerships and found that certain family ties act as buffers while others have the effect of a stressor. Having ties with parents and with a partner positively affect mental well-being, whereas having (young) children negatively influences mental well-being. The presence of broken ties in the family and having been raised by a teenage mother negatively affect mental well-being. The impact of family ties appears to vary across the life course. Having ties with parents is important for mental well-being in young adulthood, whereas the impact of having a partner on well-being is weakest in that phase of life. Another focus was on how the impact of family ties on mental well-being is moderated by gender roles, family roles and work roles. Findings showed, among others, that the positive association between family ties and mental well-being is stronger for people who have strong feelings of obligation toward family members. Also, the positive impact of marriage and having a partner on mental well-being appears to be stronger for people with strong family norms.

Analyses using GSS-data revealed east-west contrasts with regard to levels of well-being. Older adults in eastern Europe tended to be lonelier and to feel more depressed than age peers in Western Europe. This east-west well-being differential is consistent with earlier work showing lower life satisfaction and poorer self-perceived health in former communist countries than among long-term members of the European Union (EU). An issue of debate is whether cross-national differences reflect real differences or cultural differences in the way people rate their experiences. We lean towards the former, given that the well-being measures used in the GGS have proven cross-national equivalence. The reliability, validity and structural characteristics of the measures are of high quality and allow intercultural comparison. The persistent east-west contrasts, which can be traced to differences in wealth, economic organisation, political traditions, cultural systems and policy arrangements, underscore the need to question whether determinants of well-being operate in a similar way across countries. A critical perspective on the portability, from 'West' to 'East', of different explanatory models is required.

Policies structure relations between generations

MULTILINKS provides illustrations of the ways in which country-specific policies structure family care. The data are from the 'Survey of health and retirement in Europe' (SHARE). The reliance on grandparents for childcare is particularly high in Spain, Italy and Greece. Care provided by grandparents can be viewed as an effort to improve the life circumstances of the middle generation (the adult sons and daughters). By providing childcare, the middle generation can be gainfully employed and can advance in their working career. The likelihood that grandparents provide childcare is strongly linked to the availability of public policy arrangements. The data on policies are from the MULTILINKS indicators database.

Another study on the structuring influence of policies was a cross-national analysis of positive and negative feelings towards the young and towards the old. Ageism was measured on a scale ranging from 0 (feeling extremely positive towards the young/old) to 10 (feeling extremely negative towards the young / old). Findings showed that Europeans generally feel less positive about the young than about the old. Of interest was whether policies shaping generational interdependencies help to prevent or reduce ageism. One view is that policies benefiting specific age groups represent 'compassionate ageism', a climate favourable towards the old or the young. An alternative view is that policies benefiting specific age groups are a source of intergenerational conflict, and thus contribute to ageism. Findings supported the first view. For example, in countries with high child allowances, people tended to feel more positively towards the young.

Methodological strategies

Throughout the project we have tested, developed and used methodological strategies that enable sound policymaking. A common problem with cross-national comparisons is a poor understanding of the source of observed between-country differences. It is often difficult to find out whether the differences are attributable to country characteristics, population composition and/or differential effects of individual characteristics by country. Another problem involves causality. The links between geographic proximity and intergenerational support are a case in point. Are intensive exchanges made possible because family members live nearby? Or, did family members move nearer in order to provide intensive assistance? Yet another problem pertains to unobserved influences of personal preferences. It is difficult to unravel whether support patterns in families are responses to needs, or they are motivated by self-interest. Explicit attention has been paid in a number of reports to statistical models where one needs to take into account hierarchical data structures, unobserved variance or issues of selection.

Potential impact

Messages for policymakers

A number of messages for policymakers emerge from the previous overview of findings from MULTILINKS. A first is the need to be aware that macro-level demographic patterns do not have simple extrapolates at the micro level of family structures surrounding individuals. For example, extended life expectancy does not necessarily mean a higher incidence of individuals in four- or five generation families. The age at childbearing also needs to be taken into account. Similarly, the drop in fertility levels does not necessarily mean fewer supports for ageing parents. Childlessness needs to be taken into account, as does the finding that children in small families tend to provide higher levels of support per individual child. A second message is that old adults do not become net beneficiaries of support in families until they are quite old. More support goes down generational lines than up generational lines. Coresident living arrangements generally respond to the needs of adult children rather than those of the elderly. This predominantly downward flow of help and assistance needs to be acknowledged in public and policy debates on intergenerational solidarity. Too often, older adults are portrayed as needy and dependent.

A third message is that interdependencies between generations and between men and women in families are built and reinforced by the legal and policy arrangements in a particular country. Laws define the relationships of dependence and interdependence between generations and gender, whereas policies reward or disincentive particular family patterns. Policymakers should critically examine the ways in which caring responsibilities for the young and the old have been allocated between the family and the collectivity. To what extent do country-specific institutional frameworks impose dependencies which limit the autonomy of individuals? To what extent do they support the choice to assume intergenerational obligations? Such a critical examination calls for a 'holistic' approach to policy making: a serious consideration of the ways in which public family provisions (or the lack thereof) create differential opportunities for individual autonomy for young and old, men and women. A fourth message is that national policies should seek to support intergenerational care regimes without reinforcing social class inequalities and gender inequalities. One of the issues is whether policies should involve payments for care, (paid) leaves or the provision of care services. The policy measures have different implications that need to be considered carefully. For example, when public support is offered in money rather than in kind, trade-offs between using it to buy services or to keeping it for the family budget while providing care directly are different for families in different socioeconomic circumstances. The strategy of staying at home to provide care is more likely to be adopted by members of the working-class (in practice women), reducing their ability to remain in the labour market and hence creating the conditions of old-age poverty for themselves. Another example pertains to a father quota in paid leave schemes. A 'use it or lose it' criterion promotes equal sharing of parenting responsibilities between men and women.

A fifth message is that women's integration in the labour force has taken place without fundamental changes to the formal and informal rules associated with the breadwinner model, which served as the basis for the organisation of the labour market and welfare states in many European countries. The structural discrepancy between the role of the breadwinner model in the organisation of paid work and unpaid care, on the one hand, and the increased labour force participation of women, on the other hand, is the source of tensions and stress, as witnessed in dropping fertility rates, marital instability, intra-family conflict and even emotional burn out. To help resolve this discrepancy, policies should consider how to get men more involved with caring. Men should not be discouraged from taking care leaves, and men should come to realise that intergenerational responsibilities throughout their working life are the norm, not the exception.

Dissemination activities

As MULTILINKS progressed, research findings were presented at academic conferences, and published in scientific refereed journals, working papers and other scientific reports. At the time of reporting, there were 21 peer-reviewed publications, 6 papers that had been accepted but not yet published, 14 manuscripts under review, and 11 reports (e.g. working papers). These publications exclude the 18 deliverables describing scientific results, and the six policy briefs that were released.

Final results will appear in a special issue of the open access journal Demographic Research edited by P. A. Dykstra. The title of the special issue is 'Intergenerational family ties in Europe: Multiple linkages between individuals, families and social contexts'. The list of dissemination activities covers over 90 activities. Clearly, the researchers involved in MULTILINKS have been productive. During the reporting period, MULTILINKS garnered considerable attention from national and European media (see the list of dissemination activities for details). Members of the consortium were regularly consulted by journalists on issues related to demographic change, intergenerational solidarity well-being and social integration. Project results made their way to consumers of the media through interviews on radio and television programmes, features in newspapers and popular magazines and the project website.

In accordance with the plans, two policy conferences were organised, that brought together governmental representatives, EU policymakers, end-users, members of the consortium and other researchers. Reports of the conferences are available through the website. In addition, researchers were invited to speak at high-level policy meetings:

- Chiara Saraceno was an invited speaker at the conference on 'Parental childcare and the employment policy-collision or complementarity' hosted by the Czech Presidency and supported by the European Commission (EC), on 5 - 6 February 2009 in Prague. The title of her talk was: 'Between care needs and equal opportunity goals, for women but also children'.
- Pearl Dykstra gave a keynote on 'Intergenerational relationships in ageing societies' at a meeting of the UNECE Working Group on Ageing attended by ministers and representatives from 32 UNECE Member States in Geneva, 23 - 24 November 2009.
- Pearl Dykstra was one of the speakers invited to the workshop 'Indicators of Active Ageing' organised by the Demography Network of the European Observatory on the Social Situation and Demography in Brussels, 23 February 2011. The title of her paper was: 'Survey based indicators: The contribution of the Generations and Gender Programme (GGP) and of MULTILINKS'.
- Pearl Dykstra was a keynote speaker at the conference 'Towards the European Year for Active Ageing and Solidarity between Generations (2012)' on 29 April 2011 in Brussels, held by the EC. The event was attended by more than 400 participants representing Member States, regional and local authorities, social partners and civil society organisations.

Findings from MULTILINKS were included in news alerts targeted at policymakers. The UN Economic Commission for Europe devoted their Eightth Policy Brief on Ageing 'Advancing intergenerational solidarity' in August 2010 to findings from the project (see http://live.unece.org/fileadmin/DAM/pau/_docs/age/2010/Policy-Briefs/8-Policy-brief_AIS.pdf online). The Socioeconomic Humanities Research for Policy News Alert Service published an alert about MULTILINKS in November 2010, titled 'Bridging the generational gap' (see http://www.scoopproject.org.uk/1bridging-the-generational-gap.aspx online). MULTILINKS researchers have regularly experienced that colleagues and policymakers refer to 'the MULTILINKS approach', implying an approach that does not separate the young and old in families and that considers multiple levels of influence on issues of social inequality, social integration and well-being. The active involvement of MULTILINKS researchers in scientific and policy networks has resulted in a wide dissemination of project findings.

For further information, see http://www.multilinks-project.eu

MULTILINKS partners

A consortium of nine partners organised in six teams has carried out the project. All partners are involved in the Gender and Generations Programme, a system of nationally comparative surveys and contextual databases, which aims at improving the knowledge base for policymaking in UNECE countries. The Erasmus University Rotterdam (EUR) became the new beneficiary in the project (No. 217523) as of 1 October 2009 (month 20). The EUR became the new beneficiary because the coordinator of MULTILINKS, Pearl A. Dykstra, left the Netherlands Interdisciplinary Demographic Institute (NIDI / KNAW) to take up the position of full professor of empirical sociology at the EUR. No competitive call was used because the change involved an employment transition of the coordinator. All beneficiaries in MULTILINKS expressed their support for the move to the EUR. With the arrival of the EUR, the NIDI/KNAW left the project (at the end of month 19).

List of partners

1. (Coordinator) Netherlands Interdisciplinary Demographic Institute KNAW/NIDI, the Netherlands. Aart Liefbroer. liefbroer@nidi.nl
2. Vrije Universiteit Brussel. VUB Belgium. Jan Van Bavel. jvbavel@vub.ac.be
3. University of Antwerp. UA Belgium. Dimitri Mortelmans. dimitri.mortelmans@ua.ac.be
4. Wissenschaftszentrum Berlin fur Sozialforschung. WZB Germany. Chiara Saraceno. saraceno@wzb.eu
5. Norwegian Social Research NOVA Norway. Britt Slagsvold.britt.slagsvold@nova.no
6. Utrecht University UU. the Netherlands. Aafke Komter a. e.komter@uu.nl
7. Estonian Interuniversity Population Research Centre. EKDK Estonia. LuuleSakkeus.luule@ekdk.estnet.ee
8. Dondena Centre for Research on Social Dynamics, Bocconi University. BU Italy. Arnstein Aassve. arnstein.aassve@unibocconi.it
9. (Coordinator) EUR. the Netherlands. Pearl Dykstra.dykstra@fsw.eur.nl