Affection and Conflict in Intergenerational Relationships of Women in Sixteen Areas in Asia, Africa, Europe, and America


Affection and Confl ict in Intergenerational Relationships of 
Women in Sixteen Areas in Asia, Africa, Europe, and America

Bernhard Nauck

Abstract: Studies of intergenerational solidarity in affl uent societies suggest that 
relationships between generations consist simultaneously of both emotional close-
ness and confl icts. This analysis extends the standard model of intergenerational re-
lationships, which until now has been applied only to countries with bilineal kinship 
systems, to culturally and economically diverse areas with varying kinship systems. 
Latent class analysis was applied to measure affection and confl ict in the ongoing 
relationships of young and middle-aged women with their mothers (7,522 relation-
ship pairs) and fathers (5,338 relationship pairs). The empirical analysis was based 
on standardised oral interviews with mothers from areas in China, Indonesia, North 
and South India, South Africa, Palestine, Israel, Turkey, Russia, Estonia, Poland, East 
and West Germany, France, Jamaica, and the United States (n = 8,756). The best 
fi tting model of relationship differences consisted of four latent classes: „amicable“ 
(45 percent), “detached“ (28 percent), “ambivalent” (22 percent), and “disharmoni-
ous“ (5 percent). Based on a cross-culturally largely invariant measurement model, 
results revealed signifi cantly different distributions for fathers and mothers and 
across areas. Multinomial three-level regression analysis was used to analyse the 
complex cross-level interdependence of area effects, individual characteristics, and 
the respective relationship on class membership. In patrilineal societies, relation-
ships of women with their biological parents are more likely to be ambivalent, less 
likely to be detached, and very likely to become disharmonious in case of spatial 
proximity. In affl uent societies, the relationships are less likely to be disharmonious 
and most likely to be harmonious. Whereas the frequency of contact decreases the 
likelihood of detached or disharmonious relationships in affl uent societies, func-
tional exchange with the parents increases the likelihood of disharmonious relation-
ships.

Keywords: Intergenerational relationships · Emotional closeness · Confl ict · 
Solidarity · Cross-cultural

Comparative Population Studies
Vol. 39, 4 (2014): 647-678 (Date of release: 30.10.2014)

© Federal Institute for Population Research 2014  URL: www.comparativepopulationstudies.de
DOI: 10.12765/CPoS-2014-16en    URN: urn:nbn:de:bib-cpos-2014-16en3



•    Bernhard Nauck648

1 Introduction

The intergenerational solidarity paradigm – a comprehensive scheme for describ-
ing sentiments, behaviours, attitudes, values, and structural arrangements in adult 
intergenerational relationships – has become the “gold standard” model for assess-
ing intergenerational relationships (Silverstein et al. 2010: 1007). The initial model 
postulated that intergenerational solidarity consists of six distinct components: 
emotional (closeness), associational (social contact), structural (geographic proxim-
ity), functional (supportive behaviour), normative (fi lial obligations), and consensual 
(attitudinal agreement) (Bengtson/Schrader 1982; Bengtson/Roberts 1991). In ac-
knowledgment of the fact that intergenerational relationships can be simultaneously 
warm and antagonistic, the model was extended recently to include “ambivalence” 
(Luescher/Pillemer 1998; Luescher 2002). Originally developed for the description 
and better understanding of intergenerational relationships in the United States, 
the model subsequently demonstrated considerable explanatory power in other 
countries and, more recently, in cross-national comparisons that have included the 
United States, the United Kingdom, Norway, Germany, Spain, and Israel (Silverstein 
et al. 1998; Daatland/Herlofson 2003; Lowenstein et al. 2005; Lowenstein/Daatland 
2006; Lowenstein 2006; Gans et al. 2009; Silverstein et al. 2010). Thus, the scope of 
cross-national comparisons so far has been limited to affl uent societies with a simi-
lar cultural heritage. Moreover, most of the empirical knowledge available is based 
on a very limited number of data sets, and only one of these was designed using 
the intergenerational solidarity paradigm (OASIS (Old Age and Autonomy: The Role 
of Service Systems and Intergenational Family Solidarity; Lowenstein et al. 2002). 
Given this situation, there is legitimate reason to question the external validity of the 
intergenerational solidarity paradigm when applied to societies of varying cultural 
heritage.

All societies so far included in cross-national comparisons based on the inter-
generational solidarity paradigm have similar institutions of bilineal kinship. In these 
societies, the intergenerational relationships of married women are relatively free 
of institutional regulations that restrict choices regarding residence, social contact, 
mutual support, control rights, or inheritance. As a result, women have much lati-
tude to live out their individual preferences and choices. Unilineal kinship systems, 
be they patrilineal or matrilineal, are quite different as they impose restrictions 
that can impinge strongly on intergenerational relationships. In patrilineal kinship 
systems, for example, parents-in-law may have the same responsibilities to their 
daughters-in-law that parents have to their daughters in bilineal kinship systems. 
These responsibilities are expected to be reciprocated by fi lial obligations of func-
tional solidarity. This situation challenges the relationship with the family of origin 
and can provoke feelings of ambivalence.

Extending the intergenerational solidarity paradigm to culturally diverse soci-
eties has far-reaching conceptual and methodological consequences. “Research 
within single societies often fails to reveal the infl uence of culture as a social force 
[...]. Norms and values distinguish family and state responsibilities in meeting the 
needs of the aged. Future research must recognise the relationship between cul-



Affection and Confl ict in Intergenerational Relationships of Women    • 649

tural belief systems and individual understandings of roles and obligations” (Bengt-
son/Putney 2000: 281). This challenge can be resolved adequately only through a 
multi-level approach (Kohli 2005: 270) that takes into account both the variability of 
intergenerational relationships within individual societies and the institutional vari-
ability of family and kinship systems across societies that differ in terms of control 
rights, descent, and belonging.

The following empirical analysis contributes to research based on the inter-
generational solidarity paradigm in two respects. First, it applies the model to 16 
cultural areas in China, North and South India, Indonesia, Palestine, Israel, Turkey, 
South Africa, Russia, Estonia, Poland, East and West Germany, France, Jamaica, and 
the United States. The empirical analysis thus comprises widely varying economic 
conditions and diverse kinship systems and their institutions. Second, it comple-
ments cross-national comparative research on intergenerational solidarity and am-
bivalence (Silverstein et al. 2010) with the perspective of young and middle aged 
women towards their biological fathers and mothers. The model of Silverstein et al. 
(2010) focusses on the perspective of parents towards their children and is based 
on samples from six countries (England, Germany, Israel, Norway, Spain, and the 
United States). It suggests a typology, derived from latent class analysis, of four dis-
tinct relationship types: “amicable” (high level of emotional closeness and low level 
of confl ict), “detached” (low levels of both), “disharmonious” (low level of emotional 
closeness, high level of confl ict), and “ambivalent” (high levels of both).

For extending the intergenerational solidarity and ambivalence paradigm to 
culturally diverse contexts, a strategy of replication and controlled extension was 
chosen. Replication follows the design of categorical measurement models that are 
able to identify meaningful types of intergenerational relationships using measures 
of affection and confl ict derived from the solidarity-confl ict paradigm (Silverstein et 
al. 2010: 1007). In the analysis below, the paradigm is extended through a) the in-
vestigation specifi cally of the relationships of young and middle-aged women with 
their mothers and fathers, b) the question of whether the distribution of relationship 
types varies across areas with diverse institutional settings and economic condi-
tions, and c) the question of how much variance can be accounted for by individual 
resources and opportunity structures (Nauck/Arránz Becker 2013).

2 Intergenerational relationships in cross-cultural comparison

2.1 Theoretical background

By itself, the typology of intergenerational relationships based on the solidarity-
confl ict model provides no argument regarding which conditions favour which 
combination of affection and confl ict. This theoretical gap can be addressed using 
models of reactions to given relationship states. One model, with striking similari-
ties to that of Silverstein et al. (2010), was developed by Rusbult et al. (1982, 1986, 
1991; Rusbult/Van Lange 2003). These authors take up the distinction between exit, 
voice, and loyalty originally developed by Hirschman (1970) to distinguish possible 



•    Bernhard Nauck650

reactions of members to decline in formal organisations. To this they add a fourth 
type, neglect, and apply it to the analysis of partnerships, marriages, and other inti-
mate relationships. They postulate that their four types of reactions to a (dissatisfac-
tory) relationship state are distributed in a two-dimensional space defi ned by the 
axes of active versus passive and constructive versus destructive. Relating this ty-
pology to interdependence theory, they conclude that the reaction is determined by 
(a) the level of previous satisfaction with the relationship, (b) the extent of previous 
investments in the relationship, and (c) the quality of alternatives to the relationship 
(Rusbult et al. 1986). Previous satisfaction and investments are positively related to 
loyalty and also to accommodation in case of confl ict; they are lower in less interde-
pendent relationships (Rusbult et al. 1991). The likelihood of exit increases with the 
quality of alternatives. Dependency, especially mutual dependency, increases exit 
costs. It thus decreases the likelihood of exit and increases the likelihood of voice or 
neglect (Drigotas/Rusbult 1992).

Whereas the model of Rusbult et al. (1982) describes individual reactions to 
states of social relationship, the solidarity-confl ict model of Silverstein et al. (2010) 
describes the social relationship itself (as a result of the previous actions of its par-
ticipants). From their perspective, amicable relationships are the product of high 
investments accompanied by high relationship satisfaction, resulting in a high gain 
– low cost situation of relationship maintenance. Detached relationships are the 
result of a previous exit, resulting in a low gain – low cost situation of maintaining 
a “non-relationship.” Disharmonious relationships are those with low gains – high 
costs, which may arise when there is no exit option. Ambivalence characterises a 
relationship that is close and rewarding, based on high previous investments, but at 
the same time confl ictual because of normative disagreement or distribution con-
fl icts; this results in a high gain – high cost situation. Ambivalent relationships are 
not negative; they are in fact the richest and most demanding kind of relationship 
because they imply frequent and intense contact and intergenerational exchange 
in combination with ongoing confl ict management as long as the exit option is not 
chosen. However, ambivalence also “calls for a resolution either in the direction of 
a mutually supportive tie or in the direction of drifting apart” (Schenk/Dykstra 2012: 
124), i.e. a shift toward an amicable or detached relationship. These assumptions 
are supported by the fi nding of van Gaalen et al. (2010) that ambivalence increases 
in intergenerational relationships if adult children have few exit options, for example 
when they are socially isolated or have few siblings.

Additionally, differences between types of social groups and formal organisa-
tions should be considered. In the case of intimate relationships, active exit or pas-
sive neglect usually equates to destroying the relationship or waiting for its termi-
nation. In the case of formal organisations, they usually survive the exit of (some) 
members. Kinship relationships, in which those of adult married women with their 
parents are embedded, are located somewhere in between intimate relationships 
and formal organisations. The kinship system as such “survives” the exit of single 
members, especially if it has the size of large clans, but is more sensitive to individ-
ual reactions than formal organisations, as the relationships are based on complex 
exchange patterns.



Affection and Confl ict in Intergenerational Relationships of Women    • 651

The challenging task of cross-cultural research in this realm is to develop bridg-
ing hypotheses that relate properties of the macro-level, i.e. variations between 
societies or cultural areas, to properties on the micro-level, i.e. systematic varia-
tions of affection and confl ict in intergenerational relationships. Previous studies 
of intergenerational relations relevant for the countries selected for this paper have 
analysed intra-societal variations in a single country or have compared a small num-
ber of countries. These included India (Datta et al. 2003; Mishra et al. 2005; Singh 
2005), China (Chen/Silverstein 2000; Chu et al. 2011; Chu/Yu 2010; Cong/Silverstein 
2008, 2011, 2012a/b; Guo et al. 2009, 2012; Nauck 2009; Silverstein et al. 2006), 
Indonesia (Nauck/Suckow 2006; Schwarz et al. 2010; Trommsdorff/Schwarz 2007), 
Israel and Palestine (Kagitcibasi et al. 2010; Katz/Lavee 2005; Lowenstein et al. 2005; 
Silverstein et al. 2013), and Turkey (Kagitcibasi/Ataca 2005; Kagitcibasi et al. 2010; 
Nauck/Klaus 2008).

Most approaches in this domain, especially if only a small number of countries 
are compared, identify pecifi c socio-structural or cultural characteristics of the so-
cieties included and relate them to outcome variables on the individual level. The 
empirical analysis of Silverstein et al. (2010: 1009) is guided by “knowledge of the 
political economies and family cultures of nation-states to speculate about the na-
ture of cross-national variations in the emotional ties between older parents and 
their adult children.” This approach becomes unfeasible – and to some extent un-
necessary – when analysing multiple societies, because a larger number of country 
cases makes it possible to test cross-level assumptions about the effects of societal 
characteristics on intergenerational relationships. Analyses of helping, care, and liv-
ing arrangements between generations using data sets that include larger numbers 
of European countries (Hank 2007) often group countries together using typolo-
gies such as Esping-Andersen’s (1990) typology of welfare states, or they introduce 
quantitative indicators for the description of societal-level variables in multi-level 
approaches (Brandt et al. 2009). The following empirical analysis does not seek 
primarily a better understanding of the specifi c conditions of solidarity and confl ict 
in intergenerational relationships “within” individual societies using an “emic” ap-
proach. Instead, it follows an “etic” approach in looking for differences “between” 
societies, posing the research question of whether general characteristics of socie-
ties explain some of the variation in intergenerational relationships

2.2 Hypotheses

This analysis focusses on two macro-level dimensions: the institutional structure of 
the kinship system and the opportunity structure of the society, measured in terms 
of its welfare level. It is assumed that intergenerational dependence decreases as 
the welfare level in the respective area increases. It is also assumed that more alter-
natives to a specifi c relationship – both within and outside the kinship system – are 
available in a bilineal than in a unilineal kinship system. Accordingly, expected so-
cietal differences in kinship relationships are formulated as a lineage hypothesis (a) 
and a welfare hypothesis (b).



•    Bernhard Nauck652

(a) An important characteristic of social institutions is related to whether kinship 
relationships are organised unilinearly or bilinearly. Unilinear kinships systems 
provide a clear structure of belonging, as all individuals are members of just 
one lineage. These systems are also based on relatively strong expectations 
of normative solidarity. In patrilineal kinship systems, intergenerational wealth 
fl ows only between male members of the kinship system as part of functional 
solidarity; associational and affective solidarity is also primarily lineage based. 
Unilinear kinship systems also provide clear (typically highly ritualised) rules of 
inclusion. For females in patrilineal societies in case of marriage, this implies that 
they will have to leave their lineage of origin and will be wholly included in the 
lineage of their husbands and are then part of it. Thus, for women relationships 
with the lineage of origin may become detached after marriage. Previous re-
search found that this detachment was only related to obligations and duties and 
thus to functional solidarity, whereas emotional bonds were not affected (Nauck/
Arránz Becker 2013; Nauck 2012). However, whether strong normative solidarity 
toward the husband’s lineage results in confl ictual or ambivalent sentiments to-
ward biological parents remains an open empirical question. A different pattern 
of intergenerational solidarity should be observable in matrilineal kinship sys-
tems, where normative obligations are allocated most clearly within the mother-
daughter dyad and intergenerational wealth is transmitted between females. As 
this relationship is lifelong, it should be associated with strong emotional bonds, 
but because of the lack of exit options it should also be prone to confl ict and thus 
also to emotional ambivalence. Bilinear kinship systems provide a balanced loy-
alty to both lines of descent; thus, individuals are not tied to a specifi c lineage, 
but rather have their own, “individual” kindred. This reduces normative solidar-
ity toward parents specifi cally and the signifi cance of intergenerational relation-
ships generally because it makes the marriage unit and the nuclear family the 
primary unit of solidarity. This may then result in a higher proportion of detached 
intergenerational relationships or – especially in the case of expected functional 
help – in high proportions of ambivalent or disharmonious relationships.
Lineage hypothesis: Unilineal kinship systems provide clear borderlines and thus 
fewer alternatives outside the lineage than bilineal kinship systems. Emotionally 
unrewarding relationships are maintained as confl ictual and not transformed 
into detached relationships because exiting the relationship is rarely possible. 
In conjunction with high investments, emotionally rewarding relationships are 
more likely to be transformed into ambivalent ones than in bilineal kinship sys-
tems.

(b) Societies also differ considerably in terms of alternatives to the functional soli-
darity of the kinship system. In societies in which all welfare production is kin-
ship based and intergenerational relationships are the only source of insurance 
against the risks of life, investments in and maintenance of functional solidarity 
is a very salient part of kinship relationships and may also have a strong im-
pact on emotional bonds and intergenerational confl icts. Silverstein et al. (2010: 
1009) argue that fi lial obligations tend to vary inversely with the degree of wel-



Affection and Confl ict in Intergenerational Relationships of Women    • 653

fare development and that family values are generally stronger in societies “in 
which the state assists citizens only when they have exhausted their personal 
resources” than in societies, “in which all citizens are incorporated under a single 
universal insurance system.” “For that reason,” they continue, “older parents in 
nations with more evolved welfare systems may also have less confl ict with their 
children than those in nations with more residualist policies. In contrast, older 
parents in nations with weaker social policies may exhibit both more affection 
and more confl ict because of their greater involvement with and dependence on 
adult children in such nations.” These arguments were developed to differenti-
ate between welfare regimes within relatively affl uent societies. It is uncertain 
whether they apply also to poor societies with weak or non-existent welfare in-
stitutions, where the state has no means to compensate exhausted personal 
resources and individuals have “no choice” but to rely on their relatives. This 
situation should give rise to a more frequent combination of strong emotional 
bonds and confl icts, and hence ambivalence.
Welfare hypothesis: Affl uence and well-developed welfare states provide alter-
natives to functional solidarity within intergenerational relationships. Relation-
ships in affl uent societies “survive” as amicable, if based on long-term rewarding 
emotional closeness, or as ambivalent, if long-term investments are not met by 
emotional rewards. Disharmonious intergenerational relationships are “opted 
out” and transformed into detached relationships.

Additionally, interaction effects between the societal and the relational charac-
teristics may be hypothesised. High levels of functional exchange between women 
and their family of origin might result in a much more confl ictual situation in patri-
lineal kinship systems as compared to bilineal ones, as such exchanges might be 
perceived as a signal of disloyalty to the patrilineage to which they belong after 
marriage. However, this challenge may be reduced if both kinships live very close to 
each other and have frequent contact. Mutual functional intergenerational exchange 
may also have different effects on affection and confl ict depending on whether it 
takes place in a poor or affl uent society. Whereas functional intergenerational ex-
change is normatively favoured in less affl uent societies, it is much less favoured 
in affl uent societies, where the generations primarily rely on their own resources. 
Accordingly, disharmony and confl ict should increase with functional exchange in 
affl uent societies.

Although the mechanisms covered by the two broad “lineage” and “welfare” hy-
potheses are theoretically independent, it may be diffi cult to separate them empiri-
cally. Lineage-based societies tend to have a strong tendency to rely on kinship with 
regard to the production of welfare and especially with regard to security against 
the risks of life. Bilinear societies, for obvious reasons of decreased normative soli-
darity and ambiguous modes of belongingness, provide strong incentives to look 
for alternatives to the kinship system for insurance and welfare. These theoreti-
cal considerations underscore that, depending on societal conditions, intergenera-
tional relationships are very different intermediate goods in the social production 
function for the maximisation of individual welfare (Lindenberg 1989, 1996; Nauck 



•    Bernhard Nauck654

2001, 2007a/b; Nauck/Klaus 2007; Ormel et al. 1999). Intergenerational relationships 
in affl uent societies are means for the pursuit of welfare benefi ts primarily in the 
realm of social esteem, as they provide communication, exchange of emotions, and 
behavioural confi rmation. In contrast, intergenerational relationships in less affl u-
ent societies are means to obtain substantial welfare benefi ts also in the realm of 
physical well-being, as they are an essential resource for mutual services and the 
production and distribution of material goods. Accordingly, one may speculate that 
intergenerational ambivalence has different causes in both societal types. In wel-
fare societies, its likelihood increases when normative obligations to have a “good” 
emotional relationship are undermined by personal confl ict; in poor societies its 
likelihood increases when normative obligations to have a “good” functional rela-
tionship are undermined by shortages in individual resources.

3 Method

3.1 Samples

Data for the following empirical analysis were gathered in a larger research project 
on the value of children in cross-cultural comparison (Trommsdorff/Nauck 2005, 
2010). The overall goal of the project was to better understand cross-cultural varia-
tions in the perceived benefi t of having children and its impact on fertility and inter-
generational relationships. Data for the study were collected through standardised 
face-to-face interviews with mothers of preschoolers and mothers of young ado-
lescents. Data collection was undertaken in cooperation with academic teams from 
local universities and was based on a common sampling design. Area sampling, a 
well-established strategy in cross-cultural psychology (Whiting 1968), was applied 
to enhance cross-national comparability. As nationally representative samples were 
neither feasible nor intended, variance within otherwise culturally sometimes quite 
heterogeneous societies was minimised by collecting data from three similar geo-
graphic-demographic areas in each country: an urban middle-class area, an urban 
lower-class area, and a rural area. In two countries, multiple regions were included 
because of important historical legacies. Included separately were East and West 
Germany because of their long political division (Szydlik 1996) and North and South 
India because of differences in the institutionalisation of kinship regimes (Klaus/
Tipandjan in press). Sampling was based on register data, where possible, or on 
multi-stage cluster sampling with random walk. Data were collected in 2002/04 in 
the People’s Republic of China (n = 614), Indonesia (n = 600), North India (n = 600), 
Palestine (n = 499), Israel (n = 408), Turkey (n = 622), South Africa (n = 688), East 
Germany (n = 204), West Germany (n = 409), and France (n = 200), in 2005/08 in 
Russia (n = 548), Poland (n = 678), and the United States (n = 337), in 2009/10 in Es-
tonia (n = 300) and South India (n = 599), and in 2011 in Jamaica (n = 615). A total of 
8,756 interviews were conducted. Because “culture,” as captured by the respective 
area, is an important unit of analysis, the sub-samples for each area are weighted to 
adjust for size differences in the multivariate parts of the analysis.



Affection and Confl ict in Intergenerational Relationships of Women    • 655

An important strength of our data is that each respondent evaluates relation-
ships with both father and mother, if they are alive, summing to a total of n = 12,860 
relationships of 7,882 respondents with at least one parent alive. As the informa-
tion is based on female respondents, the data are especially suitable for identifying 
institutional differences. In patrilineal societies, women are affected by patrilocal 
residence and the resulting increased interaction with the lineage of their husbands. 
In matrilineal societies, women remain in their relationship with their mothers. In 
bilineal societies, neolocal residence prevails.

3.2 Measures of affection and confl ict in intergenerational relationships

The women in the sample were asked a series of questions about the nature of their 
relationship with their biological mother and father. Questions were originally de-
veloped in English and translated for use in non-English-speaking countries. The ac-
curacy of translations was ensured through back-translation, extensive pre-testing 
of the instruments, and controls for measurement equivalence (Nauck/Klaus 2007). 
The key measures in this analysis captured affectual and confl ictual dimensions of 
the solidarity-confl ict model. For this purpose, sub-scales of the Network of Re-
lationships Inventory (NRI), developed by Furman and Buhrmester (1985) and as-
sessed by Furman and Buhrmester (1992) and Schwarz (2000), were used. Furman 
and Buhrmester (1992) reported a mean alpha of .81 for the NRI subscales. In a Ger-
man study on divorced mothers, the scales confl ict and affection revealed internal 
consistencies between .76 and .86 (Schwarz 2000).

Respondents were asked to rate the respective relationship from (1) “never” to 
(5) “always”. “Confl ict” was rated with three items (How often do you and your fa-
ther/mother disagree and quarrel? How often do you and your father/mother argue 
with each other? How often do you and your father/mother get upset with or mad 
at each other?). “Affection” was also rated with three items (How often does your 
father/mother let you know that you’re good at many things? How often does your 
father/mother like or approve of the things you do? How often do you feel that your 
father/mother admires you?).

3.3 Statistical procedures

3.3.1 Correction for culture-specifi c response styles

One well-known issue in cross-cultural research is cultural variation in response 
styles to attitude items (van de Vijver/Leung 1997; Fischer 2004; Smith/Fischer 
2008). Whereas especially East Asian respondents show a “modesty bias” in pref-
erably using middle categories, respondents from the Near East and Central and 
South America, for example, tend to show “strong opinions” and preferably use 
the extreme categories. In order to correct for response styles, the proportion of 
extreme responses for all 126 attitude items with various contents was calculated 
and used to correct for individual response styles. In fact, while an unbiased use of 
the extreme categories would have been 40 percent, the found proportion of an-



•    Bernhard Nauck656

swers on extreme categories ranged from 67 percent in South India, 57 percent in 
South Africa, 52 percent in Jamaica, and 49 percent in Israel to 32 percent in China, 
27 percent in Estonia, and 25 percent in Russia. Extensive checks revealed that the 
correction for response styles changed level-differences between areas, but did not 
have any substantial effect on the multivariate results.

3.3.2 Factor analysis

A fi rst attempt to replicate the intergenerational solidarity and confl ict model with 
data from culturally diverse areas was made with regard to the underlying factor 
structure. In fact, the two-dimensional structure previously found for relationships 
between parents and their adult children (Silverstein et al. 2010) was largely con-
fi rmed for father and mother relationships in the sixteen samples studied: Disagree-
ment ( .88), arguing ( .86) and being upset ( .87) loaded on the fi rst factor; appraisal 
( .87), approval ( .87), and admiration ( .87) loaded on the second independent fac-
tor (r = .15), with 76 percent explained variance by these two factors. This result 
suggests that the underlying theoretical dimensions of the solidarity-confl ict model 
were captured by the chosen indicators and that both dimensions were independ-
ent of each other.

Additionally, the cross-cultural validity of the constructs was checked. The ulti-
mate aim was to compare in pairs the pooled solution with the respective country 
solutions. The computation of target rotation (van de Vijver/Leung 1997: 88-99) was 
performed, which provides factor-specifi c agreement coeffi cients. The proportion-
ality coeffi cient was used to assess structural equivalence. The results revealed that 
the structural equivalence of the measurement of both “affection” and “confl ict” 
across areas was extremely high. The square root of mean squared difference per 
factor only rarely exceeded .10, namely for both constructs in France (.11) and the 
United States (.11) and for “confl ict” in South India ( .13). The proportionality coef-
fi cient, Tucker’s phi, for “affection” was lowest for France and the United States with 
.98, the proportionality coeffi cient for “confl ict” was lowest for South India, France, 
and the United States ( .98). Thus, based on the factor structure of the multiple 
measurement of the constructs, cross-cultural equivalence was decisively estab-
lished.

3.3.3 Latent class analysis

The next step in testing the solidarity-confl ict model in culturally diverse areas was 
to replicate the types of relationships by latent class analysis. The basic idea of this 
approach is that “relations can be characterized as a circumscribed set of ‘ideal’ 
types that are empirically manifested by combinations of observed variables” (Sil-
verstein et al. 1997: 437). For the purpose of replication, the six indicators were 
dichotomised along the median value to achieve highest comparability. A series of 
latent class analyses was run to test whether the same set of unobserved classes 
accounted for the association of the same categorical variables in culturally diverse 
intergenerational relationships. Considerably more variation in the relationships 



Affection and Confl ict in Intergenerational Relationships of Women    • 657

was expected. Higher variation would imply that a multitude of associations ex-
ist between the manifest variables and thus would present a strong challenge to 
the measurement model. In the test, latent classes were added successively until 
an acceptable fi t to the data was reached. The selection of an acceptable model 
was based on the inspection of the likelihood ratio chi-square test statistic (L²), the 
Bayesian Information Criterion (BIC), the dissimilarity index (DI), and the proportion 
of classifi cation errors (CE).

Results revealed that the L² values dropped from 23,267 for a one-class model 
to 30 for a six-class model, whereas the BIC-values went from 22,727 to -206 in the 
four-class model and then increased again, when allowing for more classes, thereby 
confi rming the superiority of the four-class model. From this perspective and fol-
lowing the “elbow criterion” (Collins/Lanza 2009), the four-class model was selected 
as the most acceptable, with a DI of .019 and a CE of .055. Thus, LCA based on 
dichotomous variables reproduces the reported factor structure based on ordinal 
scales, which indicates some robustness of the measurement.

The next step in the analysis was to test which class solution also results in “a 
well-defi ned, descriptively meaningful, and generalizable typology” (Silverstein et 
al. 2010: 1012) of kinship relationships. The upper panel of Table 1 shows the result-
ing measurement model in terms of the conditional item probabilities constitut-
ing the four-class solution, computed across mother and father relationships and 
countries, along with their respective prevalence. The response probabilities on the 
six indicators closely matched the theoretical class defi nitions, thus supporting the 
underlying latent class measurement model.

The fi rst class, prevalent in one-fi fth of the relationships, was characterised by 
a high probability of both affection and confl ict, i.e. “ambivalence.” It should be 

Tab. 1: Latent class probabilities and average latent class distribution across all 
areas and both relationships (corrected data; respondents n = 8,756; 
Relationships n = 12,860)

* Latent class probabilities greater than .75 are considered relatively high and are shown 
in bold. 

Source: VOC-study

Latent class  Probabilities* ambivalent amicable disharmonious detached

appraisal .96 .89 .10 .11

approval .97 .94 .21 .26

admiration .96 .89 .16 .14

disagreement .93 .10 .88 .06

arguing .93 .11 .82 .04

being upset .82 .05 .76 .04

Total 22 % 45 % 5 % 28 %



•    Bernhard Nauck658

noted that this measurement of ambivalence differs from approaches that direct-
ly assess perceptions of ambivalence by asking respondents the degree to which 
their feelings toward their parents are mixed (Pillemer/Suitor 2002). It also differs 
from approaches that capture ambivalence by separately measuring positive and 
negative feelings toward individuals (Fingerman et al. 2004; Ferring et al. 2009). It 
resembles most the approach of van Gaalen and Dykstra (2006), Silverstein et al. 
(2010), and Schenk and Dykstra (2012) because it focuses on contrasting dimen-
sions of intergenerational relationships, wherein ambivalence refl ects “high levels 
of both solidarity and confl ict (the intense ties)” (van Gaalen/Dykstra 2006: 949). 
The second class had high probabilities on affection items and low probabilities 
on confl ict items, thus suggesting an amicable type of relationship. The third class 
was characterised by low affection and high confl ict probabilities and was labelled 
disharmonious. Finally, the fourth type had low probabilities on both affection and 
confl ict items, implying an emotionally detached type of relationship.

Although based on different sets of items, the latent class analysis revealed 
the same types of intergenerational relationships as the analysis provided by Gia-
rrusso et al. (2005) for a US sample and by Silverstein et al. (2010: 1014) based on 
data from “six developed nations” from the parental perspective. Signifi cant differ-
ences, however, were apparent in the relative importance of the respective types. 
In the study of Giarrusso et al. (2005) with parents under 65, ambivalent relation-
ships ranked fi rst, followed by disharmonious and amicable ones. In Silverstein et 
al.’s study of parents (2010), amicable relationships prevailed with 61 percent and 
ambivalent ones ranked last (8 percent). In the data summarised in Table 1, which 
refl ects the daughters’ perspective, amicable relationships also ranked fi rst, but 
only for 45 percent of the sample. Detached relationships ranked second, but am-
bivalent relationships ranked ahead of disharmonious ones. This fi nding confi rmed 
the “intergenerational-stake hypothesis” (Bengtson/Kuypers 1971; Giarusso et al. 
1995, 2004; Kopp/Steinbach 2009; Trommsdorff/Schwarz 2007), suggesting a more 
amicable evaluation of intergenerational relationships by the older generation and 
a more distant evaluation by the younger generation. The fi nding also refl ects the 
different levels of affection and confl ict in the respective stages of the life course: 
as the average age of both daughters and parents is younger as compared to those 
in the study of Silverstein et al. (2010), a higher level of confl ict is to be expected 
(Aquilino 1999).

3.4 Predictors of relationship types

In subsequent multivariate regression models predicting types of relationships, a 
number of theoretically important covariates were considered. Table 1 and Table 2 
of the Appendix display the descriptive characteristics of the sixteen samples. Sig-
nifi cant cross-area differences were found in the distribution of all variables.



Affection and Confl ict in Intergenerational Relationships of Women    • 659

3.4.1 Individual level

On the individual respondent level, several social characteristic variables captured 
individual variation in resources and opportunities. To differentiate between the 
opportunity structures of rural and urban settings, a measure of rural background 
was used, indicating whether the respondent lived in a village during her school-
age years, at time of her wedding, or at the time of the interview (0 = never lived in 
rural areas; 3 = lived in rural areas at all three time points). Following the research 
tradition of the Demographic Health and Fertility Surveys, the welfare level of the 
household was measured using an eleven-point index based on the possession of 
durable consumer goods and property (0 = lowest, 10 = highest). This provides a 
larger degree of cross-cultural comparability than measures of monetary income. 
The educational level was coded with fi ve categories: no schooling at all (1), some 
schooling without a primary school degree (2), a primary school degree but no sec-
ondary degree (3), a secondary degree with or without some additional vocational 
training (4), and a tertiary degree or at least some college or university training (5). 
Low levels of education characterised the Indian samples, whereas the highest pro-
portions of women with tertiary education were in the Israeli, the Eastern and West-
ern European and the US samples. A seven-point index measured formal inclusion 
in the labour force, based on items such as the respondent’s gainful employment (at 
the time of her wedding and at the time of the interview), voluntary work, workplace 
outside the home, employment by a non-family member, monetary payment, blue-
collar versus white-collar worker, and full-time versus part-time work. The mothers 
in the Indian samples had the lowest employment status, the mothers in Eastern 
and Western Europe and in the United States had the highest.

Relevant individual attitudes of the respondents were captured with three vari-
ables. Value of children (VOC) was measured in two dimensions, namely the instru-
mentality of children for increasing the parent’s comfort and the instrumentality of 
children for providing stimulation and affect. The measurement itself was based 
on a well-validated instrument (Nauck 2007a; Nauck/Klaus 2007). The fi rst factor 
contained four items, such as “to have one more person to help the family economi-
cally” or “children can help you when you are old,” and was interpreted as comfort 
benefi ts through children. The internal consistency of the resulting additive scale 
across all countries was a = .84. The second factor contained four items, such as 
“it is fun to have young children around the house” or “the pleasure you get from 
watching your children grow,” and was interpreted as affection benefi ts through 
children. The reliability of the scale was a = .82. A 10-item scale of normative family 
obligations was constructed based on selected items from the family value scale of 
Georgas (1989, 1991; Georgas et al. 1996), and a short version of the self-construal 
scale (Singelis 1994) was used after modifying the wording (Schwarz et al. 2005) to 
make the items refer to the respondent’s family. Participants indicated on a 5-point 
scale how strongly they disagreed (1) or agreed (5) with the statements. The reli-
ability of this scale was á = .80, and thus superior to the interdependence subscale 
(á = .71) and the family value scale (ranging from á = .56 to .77).



•    Bernhard Nauck660

3.4.2 Relational level

On the relational level, the age of the parents is considered to be an indicator of vul-
nerability and need for personal assistance. Proximity indicates the spatial distance 
between the respondent and the respective parent, ranging from 1 (overseas) to 5 
(in the same household). Especially the matrilineal society of Jamaica showed sig-
nifi cant differences between mothers and fathers, which also resulted in signifi cant 
differences in the frequency of contacts and mutual help. Contacts ranged from 1 
(less than once a month) to 5 (daily). Mutual help is a combined index of receiving 
help with daily chores and providing help with daily chores, measured on a three-
point scale each, ranging from 1 (almost never) to 3 (regularly). High levels of mutual 
help were found in the South African sample, whereas relatively low levels (with the 
biological parents of married women) were found in patrilineal societies like India 
(Nauck 2010; Nauck/Arránz Becker 2013), and in Germany and the United States.

3.4.3 Area level

Three societal-level indicators were considered. Following the basic theoretical as-
sumptions in the lineage and welfare hypotheses, the type of kinship system in the 
respective area and the societal welfare level were expected to be central determi-
nants.

The gross national income per capita (GNI) of 2005 in the purchasing power of 
2000 was taken as a proxy indicator for each area’s welfare level. The United States 
(37,206 USD), France (23,688 USD), Germany (23,747 USD) were by far the most af-
fl uent societies, whereas India (409 USD) and Indonesia (910 USD) ranked last.

Patrilineal, matrilineal, and bilineal kinship systems have different consequences 
for women’s relationships with their family of origin. Whereas matrilineal and bi-
lineal kinship systems do not limit access, patrilineal kinship systems may draw a 
clear demarcation between married women and their family of origin. An attempt to 
measure the prevalence of lineages within the respective area was made by com-
paring the relationships of the respondents with their biological parents and their 
in-laws. The comparison was based on proximity, the frequency of contact, and 
mutual help. The sum of indicators for the in-laws was subtracted from those for the 
biological parents within each area. If the relationships are balanced, the resulting 
indicator equals zero, a value above zero indicates a higher prevalence of matrifocal 
relationships, whereas a value below zero indicates a higher prevalence of patrilin-
eal relationships (within the lineage of the husband). The areas with the strongest 
prevalence of patrilineages were found in India, Palestine, and China. The areas with 
a strong orientation toward the family of origin were in South Africa, Indonesia, and 
Jamaica. Accordingly, these areas are labelled as having a high prevalence of matri-
focal kinship systems.

Individual attitudes are related to the respective cultural background. For shap-
ing intergenerational relationships, culturally institutionalised normative expecta-
tions toward children are central, with instrumental utilities of children showing 
the highest variation between countries (Nauck 2006). Moreover, multi-level analy-



Affection and Confl ict in Intergenerational Relationships of Women    • 661

ses have demonstrated that whereas fertility intentions are negatively related to a 
comfort utility value of children on the cultural level, they are positively related to 
comfort utility expectations on the individual level (Mayer/Trommsdorff 2010). In 
order to capture this dynamic, the area mean of comfort utility VOC of individual 
responses was introduced as culture of comfort utility of children on the area level.

4 Results

4.1 Latent class distribution

Table 2 shows gender-specifi c and area-specifi c distributions of the four latent 
classes. Relationships with mothers and fathers were characteristically different. 
Whereas the relationship with the mother was much more likely to be ambivalent or 
amicable than the relationship with the father, the reverse was true for detached re-
lationships. No gender-specifi c differences were found for disharmonious relation-
ships. This fi nding refl ects the closer and much less “avoidable” mother-daughter 
relationship, which entails greater investments in contact and mutual help and thus 
more opportunities for confl ict, whereas the relationship with fathers is more likely 
to be distant and on a low level of interaction. Areas in Table 2 are ranked according 
to their GNI (gross national income per capita) in 2005.

The most amicable relationships between daughters and their mothers were in 
the United States, Palestine, Germany, and Russia. The lowest levels were in South 
India, China, and Jamaica. Detached relationships with mothers were most common 
in China, Estonia, and Indonesia and least common in South Africa, India, Palestine, 
and Israel. Disharmonious relationships were most common in Estonia, Russia, and 
France and least common in South Africa, Palestine, China, and South India. Am-
bivalent relationships were most prevalent in India, South Africa, Jamaica, Israel, 
Turkey, and Palestine but rare in China, Estonia, Russia, and Germany. These results 
showed some signifi cant differences to analyses by Silverstein et al. (2010: 1014) 
based on the OASIS (Old Age Service Systems and Intergenerational Solidarity) and 
LSOG (Longitudinal Study of Generations) data, as the most extreme cases in the 
cross-area comparisons were not to be found in the western European and North 
American context but rather in East Asia, Africa, and the Near East, whereas the 
countries, which were included in both studies, turned out to be rather “moderate”.

Smaller differences in the relationships with mothers and fathers in individual 
areas were largely overshadowed by general trends. For example, relationships 
with fathers were more likely to be detached in all areas. Other signifi cant differ-
ences were to be found in ambivalent relationships. Without exception, the mother 
relationship was more likely to be ambivalent than the father relationship, but there 
was large variance in degree among the areas. Whereas the percentage differences 
were low in Indonesia, China, France, Estonia, and Poland, they were high in Jamai-
ca (16 percent), Israel (13 percent), Turkey (10 percent), India (8 percent), and South 
Africa (8 percent), indicating strong gender differences in intergenerational relation-
ships in these countries.



•    Bernhard Nauck662

Tab. 2: Estimated latent class distribution for relationships and areas (percent)

ambivalent amicable disharmonious detached

Mother (M) 24.9 47.7 4.6 22.9
Father (F) 19.2 41.7 5.1 34.0

North India M 50.9 40.0 2.6 6.6
F 42.0 38.7 1.8 17.5

South India M 69.7 26.8 1.8 1.6
F 62.4 32.4 2.4 2.7

Palestine M 26.7 64.8 1.2 7.3
F 22.1 64.1 2.5 11.3

Indonesia M 12.9 47.7 2.8 36.4
F 13.7 42.1 2.7 41.4

China M 3.7 30.7 1.8 63.8
F 3.2 24.4 1.3 71.1

Jamaica M 44.6 36.0 5.8 13.5
F 28.5 32.6 10.6 28.2

Russia M 11.9 58.7 9.9 19.5
F 8.6 45.0 10.0 36.4

Turkey M 29.7 49.0 5.6 15.7
F 19.7 47.6 5.7 26.9

South Africa M 43.7 51.6 0.9 3.8
F 35.5 52.7 0.9 10.9

Poland M 14.0 50.9 7.3 27.8
F 11.1 45.1 10.6 33.1

Estonia M 9.0 42.9 10.2 38.0
F 4.5 24.0 9.7 61.7

Israel M 34.8 51.7 4.8 8.7
F 21.5 56.3 2.6 19.6

France M 15.4 47.8 9.9 26.9
F 12.1 37.1 10.9 40.5

East Germany M 11.7 64.2 3.9 20.1
F 6.3 52.1 6.3 35.2

West Germany M 13.4 60.9 6.6 19.1
F 7.0 47.6 7.4 38.0

United States M 15.9 67.2 4.4 12.5
F 9.6 60.1 3.8 26.4

Source: VOC-Study (respondents n = 8,756; relationships n = 12,860)



Affection and Confl ict in Intergenerational Relationships of Women    • 663

With regard to the initial research questions, even the descriptive results indicat-
ed that relationships of adult daughters with their biological parents were strongly 
infl uenced by institutionalised kinship systems and their implications for gender 
relations. The inspection of area differences revealed that detachment is a result 
of “opting out,” especially in bilineal kinship systems, where disharmonious rela-
tionships were relatively uncommon and amicable relationships were relatively fre-
quent.

The clear differences among the sample’s strongest patrilineal societies (India 
and China) are puzzling. Whereas ambivalent relationships were extremely fre-
quent in North and South India, they were relatively rare in China, where amicable 
or detached relationships prevailed. These differences could not be explained well 
by systematic socio-structural differences among the respondents (Appendix, Ta-
ble A1) or by differences in the intergenerational interaction patterns (Appendix, Ta-
ble A2). Although the highest proportion of women living in an extended household 
(with their in-laws) was found in North India (49 percent), the share in South India 
was lower (37 percent) and not much different from China (30 percent). However, 
the share of women with less than primary education was much higher in North 
India (29 percent) and South India (29 percent) than in China (7 percent), which – 
together with their much lower labour force participation – may indicate a much 
weaker position in the joint household with the in-laws. An analysis of the effect of 
relational characteristics was similarly inconclusive. Of all areas studied, the North 
Indian women had the lowest proximity, the lowest contact level, and the lowest 
level of mutual help with their parents; however, South Indian women did not differ 
signifi cantly from their Chinese counterparts on these variables. These inconsist-
encies in the fi ndings could be related to specifi c aspects of the institutionalised 
patrilineal kinship system favouring both high levels of ambivalence (in India) and 
high levels of detachment (in China). Uncovering why the effects vary unexpect-
edly between countries requires more detailed analyses from an “emic” perspective 
(which would be beyond the scope of this paper), i.e. taking an immanent culture 
specifi c perspective. It needs to be investigated whether low negotiation power, 
which is most likely to be associated with low educational level, results in low inter-
action with biological parents in combination with high levels of both affection and 
confl ict.

4.2 Multinomial logistic regression

A series of three-level multinomial logistic regression analyses was undertaken 
to test the relationship between class membership and relational, individual, and 
societal characteristics. Amicable relationships as the most common class were 
chosen as the reference. As the number of areas included (level 3) was limited, 
separate analyses were run with either only the “matrifocal kinship system” (and 
related interaction terms on level 1) or the “gross national income per capita” (and 
related interaction terms) included. However, results diverged only marginally from 
the analysis that included both variables, with the coeffi cients never exceeding a 
difference of >.05 and with no changes in the signifi cance tests.



•    Bernhard Nauck664

Tab. 3: Three-level multinomial logistic regression predicting latent class 
membership of intergenerational relationships (n = 12,860)

relationship class (reference: amicable)

exp (b) detached disharmonious ambivalent

Fixed effects

Level 3 (areas)

Matrifocal kinship system 1.09 1.07 .94

Gross national income per capita .88 .70* .83

Culture of comfort utility of children .72 .38** 1.06

Level 2 (respondents)

Rural background 1.06* 1.01 .90***

Education .94 .99 .94

Workforce inclusion .97*** .98 .99

Family welfare level .93*** .90*** .99

Comfort VOC .85** .81* 1.36***

Affection VOC .79** .86 .61***

Normative family obligations .94 .28*** .13***

Level 1 (relationship)

Parent’s age .99** .97*** .99***

Parent’s sex (reference: father) .55*** .83* 1.36***

Parent’s proximity 1.22*** 1.55*** 1.05

Contact to parent .77*** .77*** .95

Functional exchange with parent .91*** .85*** 1.09**

Matrifocal * proximity .98 .94*** .99

Matrifocal * contact .99 1.04* 1.01

Matrifocal * exchange 1.00 1.00 .99

GNI * proximity .98 .95 .94*

GNI * contact .91*** .93* .99

GNI * exchange 1.00 1.10** 1.01

Random effects

Variance components M0
a M1 M0 M1 M0 M1

Relationships (n = 12,860) π²/3 π²/3 π²/3 π²/3 π²/3 π²/3

Respondents (n = 8,756) 1.29 1.45 1.42 1.35 1.46 1.35

Areas (n = 16) .79 .88 .46 .42 .74 .38

Interclass Correlation ICC (null/full model)

Level 2 .24 .26 .27 .27 .27 .27

Level 3 .15 .16 .09 .08 .13 .08

* p < .05; ** p < .01; *** p < .001
a refers to a three-level random intercept model without covariates, M1 refers to the three-

level model with level 3 covariates

Source: VOC-study



Affection and Confl ict in Intergenerational Relationships of Women    • 665

The fi rst block of variables in the fi nal model (Table 3) contains the societal-level 
characteristics: predominant kinship system, welfare level measured as GNI, and 
the culture of the instrumentality of children as indicated by aggregated variations 
in the comfort utility of children (level 3). The second block contains personal char-
acteristics of the daughter such as her position in the social structure, her personal 
expectations toward children, and her self-perceived normative family obligations 
(level 2). Additional tests revealed that neither the household structure (i.e., whether 
the daughters lived in an extended household or not) nor the number of their chil-
dren had an additional effect on class membership; both variables were omitted. 
The third block contains the characteristics of the relationship between the daugh-
ter and her parents as well as interaction terms between societal characteristics and 
relationship characteristics (level 1). Additional tests revealed no interaction effects 
between the parents’ age and sex and societal characteristics. Table 3 displays the 
odds ratios for detached, disharmonious, and ambivalent relationships with parents 
in relation to an amicable relationship as the reference category.

This analysis revealed the following general empirical regularities within the broad 
cross-cultural context, providing the kind of empirical evidence previously available 
only for a limited number of countries located mostly in Europe and America.

Relationship Characteristics and Characteristics of the parents
The older the parents were in this study, the more likely it was that the relationship 
was amicable and not ambivalent, detached, or disharmonious. As already shown 
in the bivariate results, the relationship with the mother was more likely to be am-
bivalent or amicable, whereas the relationship with the father was more likely to be 
disharmonious or detached. Living in close proximity to the biological parents and 
high contact frequency were counteracting forces with regard to affection and con-
fl ict in the relationship. Whereas proximity as such increased the risk of a disharmo-
nious or even a detached relationship, this was counterbalanced by the frequency 
of contact, which generally varies positively with opportunity. Thus, if the contact 
with parents was frequent, the relationship was less likely to be disharmonious or 
detached.

A higher level of functional exchange in the intergenerational relationship de-
creased the likelihood that the relationship was disharmonious or detached, but 
it increased the likelihood that the relationship was ambivalent. This was a further 
indication that intergenerational dependence is a strong predictor of loyalty and 
amicable relationships but also for confl ict (“voice”) and ambivalence if returns on 
high and complex investments in the relationship do not meet expectations. It also 
confi rmed fi ndings of Silverstein et al. (2010: 1015) that parents who received help 
with household chores from a child were more likely to have ambivalent relation-
ships with that child.

Individual Characteristics of the Daughters
Compared to institutionalised relationship characteristics, individual expectations 
and perceived normative obligations, the socio-structural positioning of the women 
had relatively small effects on the quality of intergenerational relationships.Women 



•    Bernhard Nauck666

with a rural background were more likely to have detached relationships with their 
parents and were less likely to have ambivalent relationships because they were 
more likely to have moved to an urban environment and to have had less opportu-
nity for frequent contact.

Respondents’ educational level had no effect on class membership. Their formal 
involvement in the labour force reduced the likelihood of detached relationships, as 
did the welfare level of the household in which they were then living. In other words, 
if a woman had secure economic resources, the relationship with her parents is 
more harmonious and rewarding. On the other hand, relative poverty in the societal 
context made confl icting or detached relationships more likely. In contrast, expec-
tations about intergenerational relationships had a strong impact on the quality of 
intergenerational relationships.

The expectation that children are instrumental for increasing physical well-being 
signifi cantly increased the likelihood of ambivalent intergenerational relationships 
and decreased the likelihood of detached or disharmonious relationships. This, 
again, underscores the dependency effects on ambivalence. Having the expecta-
tion that children are instrumental for receiving affection was strongly related to 
amicable intergenerational relationships and reduced the likelihood of ambivalent 
and detached relationships. The same effect was associated to an even greater 
degree with perceived strong normative family obligations, reducing the likelihood 
of ambivalent and disharmonious intergenerational relationships. This fi nding con-
tradicts previous research on intergenerational ambivalence, which has suggested 
that ambivalence is strongly related to normative obligations. This contradiction in 
fi ndings may, however, stem from differences in theoretical constructs. Whereas 
ambivalence research focuses on moral dilemmas and contradictory normative 
structures (Lüscher/Pillemer 1998; Pillemer et al. 2007), the self-construal scale 
(Singelis 1994) used here seemed to refl ect more the affective dimension of be-
longingness (Nauck/Arránz Becker 2013).

Societal Characteristics
Central for the guiding hypotheses is the relationship between societal character-
istics and relationship types. Although low case numbers reduced the chances of 
achieving statistical signifi cance, some of the observed differences were strong 
enough to test societal effects aftercontrolling for relational and individual charac-
teristics.

• Bilineal and matrilineal kinship systems, which allow for the maintenance 
of relationships with the woman’s family of origin, were associated with a 
higher prevalence of detached and disharmonious intergenerational rela-
tionships and with a lower prevalence of ambivalent relationships (relative 
to amicable relationships). In other words, ambivalent relationships with bio-
logical parents are more likely for women in patrilineal kinship systems. This 
fi nding supports existing arguments that patrilineal kinship systems affect 
primarily functional solidarity, wealth fl ows, and inheritance while leaving 
emotional relationships with the family of origin untouched (Nauck 2010; 
Nauck/Arránz Becker 2013).



Affection and Confl ict in Intergenerational Relationships of Women    • 667

• Intergenerational relationships of adult women also varied with affl uence 
on the societal level, as a higher welfare level decreased the likelihood of 
detached, ambivalent, and especially disharmonious relationships even if 
all the relationship and individual characteristics of the women were con-
trolled for, i.e., intergenerational relationships in affl uent societies were very 
likely to be more harmonious than in poor societies. This fi nding supports 
the general assumptions about “opting out” of disharmonious relationships 
under affl uent conditions, which are in turn related to issues of decreased 
dependency and increased choice regarding the resulting relationship char-
acteristics.

• A counterbalancing force to the effects of the societal welfare level was a 
culture of instrumentality in intergenerational relationships as indicated by 
collectively shared high comfort utility expectations toward children. In ar-
eas where comfort utility VOC was generally high, the likelihood of dishar-
monious intergenerational relationships was very low. Because high com-
fort utility expectations on the individual level also signifi cantly decreased 
disharmonious intergenerational relationships, societal and individual level 
effects reinforced each other.

Signifi cant interaction effects between societal and relationship levels provided 
additional insight into the preconditions of affection and confl ict in the intergenera-
tional relationships of women. Whereas functional exchange generally decreased 
the likelihood of disharmonious relationships ( .85***), they became more likely in 
affl uent societies ( 1.10**). Intergenerational support is generally accepted under 
conditions of scarce resources, but it indicates a precarious situation in affl uent 
societies, which increases the likelihood of confl icts (in the absence of affection). 
Frequent intergenerational contact made detached and disharmonious relation-
ships especially unlikely in affl uent societies and thus again boosted effects on the 
individual level. Similar reinforcing effects were found with regard to the institu-
tionalised kinship system. As mentioned before, proximity and contact generally 
counterbalanced each other. The counterbalancing effect was stronger in patrilin-
eal societies and weaker in matrifocal kinship systems. Thus, in patrilineal kinship 
systems, there is an especially high potential that women’s relationships with their 
family of origin will become confl ictual.

5 Discussion

Silverstein et al. (2010: 1019) concluded in their analysis that “the evidence in this 
investigation points to the importance of national context in structuring emotional 
ties between older parents and their adult children. Identifying the ecological condi-
tions responsible – welfare state structure, economic development, and/or cultural 
values – will require a larger sample of nations on which to map these multiple path-
ways.” The present investigation was designed to help fi ll this gap. The latent class 
structure of two core dimensions of intergenerational relationships – affection and 
confl ict – were examined in an analysis of the quality of the emotional ties of adult 
women to their fathers and mothers. Respondents lived in eighteen different areas 
in Asia, Africa, Europe, and North America and were thus embedded in diverse eco-



•    Bernhard Nauck668

nomic development contexts and in a variety of institutionalised welfare regimes 
and kinship systems. The analysis demonstrated that the four-class model of am-
bivalent, amicable, disharmonious, and detached relationships was also suitable for 
describing the perceptions of the younger generation regarding their relationships 
with their mothers and fathers and that the model can be extended to the analysis 
of intergenerational relationships in a much larger variety of ecological conditions. 
This provided further empirical evidence that the structure of emotional ties be-
tween generations is robust not only within affl uent societies, but also across levels 
of economic development and institutionalised kinship systems.

The results of this study complement previous fi ndings about exchange pat-
terns within the kinship systems of mothers from work related to the functional 
and structural dimensions of the solidarity model. Among economically developed 
societies, previous studies found the lowest level of kinship usage in countries like 
the United States and Germany, resulting in a high proportion of detached relation-
ships. This study confi rmed these fi ndings. However, an even more relevant obser-
vation from previous work was that cross-societal differences were most clearly 
notable in the proportions of “obligatory” kinship relationships: “These were widely 
absent in Israel, the United States, and Germany, whereas they were signifi cantly 
more prevalent in countries like Russia, Turkey, and Palestine” (Nauck/Arránz Becker 
2013: 589). This suggests a self-selection effect in kinship relationships: under af-
fl uent conditions relationships “survive” only, if based on strong emotional bonds, 
whereas in the absence of emotional bonds, opting out is much more likely. In less 
economically comfortable situations, relationships will be maintained regardless 
of their emotional quality in order to secure survival through the maintenance of 
functional exchange.

In kinship relationships generally, individuals have the freedom to opt out and 
therefore to have the relationship classifi ed as “detached”. Yet, in the special case 
of intergenerational relationships this choice is restricted, especially when strong 
norms of intergenerational obligations are imposed and internalised. Only in the 
case of patrilineal systems, women can take the (unlikely) choice of shifting inter-
generational obligations from their family of origin onto the in-laws. Thus, inter-
generational relationships are much less likely to be detached (“exited”) but much 
more likely to be characterised simultaneously by strong emotional bonds, con-
fl icts, and disputes (“voice”) and, hence, ambivalence. However, women’s mother 
relationships are much less “avoidable” for them than their father relationships, a 
fact refl ected in different gender-specifi c distributions of ambivalent and detached 
relationships.

If the intergenerational stake hypothesis holds, then the level of ambivalence 
should be higher as compared to the older generation. In fact, the main result of the 
descriptive analysis was that, although two thirds of the intergenerational relation-
ships of the women were affectionate, about one half of these are “ambivalent”. 
This simultaneous inclusion of affection and confl ict thus adds additional insights 
into their cross-cultural comparison (Trommsdorff/Mayer 2012: 326-327), within 
which the general prevalence of closeness was stressed, especially with regard to 
societies with collectivistic family cultures.



Affection and Confl ict in Intergenerational Relationships of Women    • 669

The latent class analysis revealed that the highest prevalence of ambivalent re-
lationships with mothers (from the perspective of the daughter) were to be found 
in India, Jamaica, South Africa, Israel, Turkey, and Palestine (in that order) and were 
thus less common to the affl uent and individualised societies of the West. Multi-
nomial regression revealed that the likelihood of ambivalence in intergenerational 
relationships increases the more intense the functional exchange and the stronger 
comfort expectations toward children are, whereas the likelihood decreases with 
stronger affective expectations toward children and feelings of strong family obli-
gations. Finally, interaction effects between societal and relational levels revealed 
that confl icting relationships with biological parents increase with proximity in pat-
rilineal societies and when functional exchange is performed in affl uent societies.

The fact that only mothers were included in the sample may limit the fi ndings. It 
remains unclear how intergenerational relationships are structured before the tran-
sition to parenthood and how exactly parenthood or marriage impacts intergen-
erational solidarity in different contexts. It is also unlikely that the fi ndings would 
be similar for fathers, especially because intergenerational relationships interact 
strongly with gender in kinship systems, be they matrilineal, patrilineal, or bilineal. 
Possibly, men’s relationships with their fathers are less detached than women’s fa-
ther relationships, especially in patrilineal kinship systems. Without a doubt, includ-
ing male respondents would make for a more comprehensive test of the lineage 
hypothesis, but it would also increase the complexity of analyses considerably. A 
further potential limitation stems from the use of area sampling. Whereas this ap-
proach is appropriate for cross-regional comparison, which was the primary aim 
of the study, it cannot yield accurately representative pictures of the countries to 
which the regions belong. Moreover, the analysis could not rely on independent 
measurements on the area level. Instead, country level measures were used to 
measure the welfare level, and aggregated individual data were used to measure 
the institutionalised kinship system.

Despite these limitations, the replication of the solidarity-ambivalence model 
across very different cultures proved to be analytically fruitful. Its integration in a 
multilevel framework is promising because it not only takes immediate individual 
opportunities into account but also long-standing, historically rooted institutional 
regulations and cultural beliefs concerning the relevance of intergenerational soli-
darity.

Acknowledgements
This study is part of the research project “Value of Children, Fertility and Intergener-
ational Relationships in Cross-cultural Comparison” (P.I.: Bernhard Nauck, Chemnitz 
and Gisela Trommsdorff, Konstanz, Germany), supported by the German Research 
Foundation. The regional collaborators in the VOC-study include: Gang Zheng (Peo-
ple’s Republic of China), Ramesh Mishra and Arun Tipandjan (India), Lieke Wisnubra-
ta (Indonesia), Karl Peltzer (South Africa), Asher Ben Arieh (Israel), Muhammad M. 
Haj-Yahia (Palestine), Cigdem Kagitcibasi (Turkey), Zaretkhan M. Saralieva (Russia), 
Kairi Kasearu (Estonia), Katarzyna Lubiewska (Poland), Colette Sabatier (France), 



•    Bernhard Nauck670

Annekatrin Bock (Jamaica), Wolfgang Friedlmeier (United States). The author thanks 
Fons van de Vijver (Tilburg), Boris Mayer (Bern) and two anonymous reviewers for 
valuable comments and suggestions on previous drafts of the paper.

References 

Aquilino, William S. 1999: Two Views of One Relationship: Comparing Parents’ and 
Young Adult Children’s Reports of the Quality of Intergenerational Relations. In: Jour-
nal of Marriage and the Family 61: 858-870.

Bengtson, Vern L.; Kuypers, Joseph A. 1971: Generational Difference and the Develop-
mental Stake. In: Aging and Human Development 2: 249-260.

Bengtson, Vern L.; Putney, Norella M. 2000: Who Will Care for Tomorrow’s Elderly? 
Consequences of Population Aging East and West. In: Bengtson, Vern L.; Kyong-Dong, 
Kim; Myers, George C.; Eun, Ki Soo (Eds.): Aging in East and West. Families, States, 
and the Elderly. New York: Springer: 263-286.

Bengtson, Vern L.; Roberts, Robert E. L. 1991: Intergenerational Solidarity in Aging 
Families: An Example of Formal Theory Construction. In: Journal of Marriage and the 
Family 53: 856-870.

Bengtson, Vern L.; Schrader, Sandi S. 1982: Parent-child Relations. In: Mangen, David 
J.; Peterson, Warren A. (Eds.): Handbook of research instruments in social gerontol-
ogy. Vol. 2: Social roles and social participation. Minneapolis: University of Minnesota 
Press: 115-185.

Brandt, Martina; Haberkern, Klaus; Szydlik, Marc 2009: Intergenerational Help and Care 
in Europe. In: European Sociological Review 25: 585-601 [doi: 10.1093/esr/jcn076].

Chen, Xuan; Silverstein, Merril 2000: Intergenerational Support and the Psycho-
logical Well-Being of Older Parents in China. In: Research on Aging 22: 43-65 [doi: 
10.1177/0164027500221003].

Chu, Cyrus; Xie, Yu; Yu, Ruoh, Rong 2011: Coresidence With Elderly Parents: A Compar-
ative Study of Southeast China and Taiwan. In: Journal of Marriage and Family 73,1: 
120-135 [doi: 10.1111/j.1741-3737.2010.00793.x].

Chu, Cyrus; Yu, Ruoh Rong 2010: Understanding Chinese Families. A Comparative 
Study of Taiwan and Southeast China. Oxford: Oxford University Press [doi: 10.1257/
jel.49.3.719.r16].

Collins, Linda M.; Lanza, Stephanie T. 2010: Latent class and latent transition analysis. 
With applications in the social, behavioral, and health sciences. Hoboken, NJ: John 
Wiley & Sons [doi: 10.1002/9780470567333.ch1].

Cong, Zhen; Silverstein, Merril 2008: Intergenerational Support and Depression Among 
Elders in Rural China: Do Daughters-In-Law Matter? In: Journal of Marriage and Family 
70: 599-612 [doi: 10.1111/j.1741-3737.2008.00508.x].

Cong, Zhen; Silverstein, Merril 2011: Intergenerational Exchange Between Parents and 
Migrant and Nonmigrant Sons in Rural China. In: Journal of Marriage and Family 73: 
93-104 [doi: 10.1111/j.1741-3737.2010.00791.x].

Cong, Zhen; Silverstein, Merril 2012a: Caring for grandchildren and intergenerational 
support in rural China: a gendered extended family perspective. In: Ageing & Society 
32: 425-450 [doi: 10.1017/S0144686X11000420].



Affection and Confl ict in Intergenerational Relationships of Women    • 671

Cong, Zhen; Silverstein, Merril 2012b: A Vignette Study on Gendered Filial Expecta-
tions of Elders in Rural China. In: Journal of Marriage and Family 74: 510-525 [doi: 
10.1111/j.1741-3737.2012.00970.x].

Daatland, Svein O.; Herlofson, Katharina 2003: ‘Lost solidarity’ or ‘changed solidarity’: a 
comparative European view of normative family solidarity. In: Aging and Society 23: 
537-560 [doi: 10.1017/S0144686X03001272].

Datta, Prishnee; Poortinga Ype H.; Marcoen, Alfons 2003: Parent Care by Indian and 
Belgian Caregivers in their Roles of Daughter/Daughter-in law. In: Journal of Cross-
Cultural Psychology 34: 736-749 [doi: 10.1177/0022022103258589].

Drigotas, Stephen M.; Rusbult, Caryl E. 1992: Should I Stay or Should I Go – a Depend-
ence Model of Breakups. In: Journal of Personality and Social Psychology 62: 62-87.

Esping-Andersen, Gösta 1990: The Three Worlds of Welfare Capitalism. Cambridge: Pol-
ity Press.

Ferring, Dieter; Michels, Tom; Boll, Thomas; Filipp, Sigrun-Heide 2009: Emotional rela-
tionship quality of adult children with ageing parents: on solidarity, confl ict and ambiv-
alence. In: European Journal of Ageing 6: 253-265 [doi: 10.1007/s10433-009-0133-9].

Fingerman, Karen L.; Hay, Elizabeth L.; Birditt, Kira S. 2004: The Best of Ties, the Worst 
of Ties: Close, Problematic, and Ambivalent Social Relationships. In: Journal of Mar-
riage and Family 66: 792-808 [doi: 10.1111/j.0022-2445.2004.00053.x].

Fischer, Ronald 2004: Standardization to Account for Cross-Cultural Response Bias. A 
Classifi cation of Score Adjustment Procedures and Review of Research in JCCP. In: 
Journal of Cross-Cultural Psychology 35: 263-282 [doi: 10.1177/0022022104264122].

Furman, Wyndol; Buhrmester, Duane 1985: Children’s perceptions of the personal rela-
tionships in their social networks. In: Developmental Psychology 21: 1016-1024.

Furman, Wyndol; Buhrmester, Duane 1992: Age and sex differences in perceptions of 
networks of personal relationships. In: Child Development 63: 103-115.

Gans, Daphna; Silverstein, Merril; Lowenstein, Ariela 2009: Do Religious Children Care 
More and Provide More Care for Older Parents? A Study of Filial Norms and Behaviors 
across Five Nations. In: Journal of Comparative Family Studies 40: 187-201.

Georgas, James 1989: Changing Family Values in Greece - from Collectivist to Individu-
alist. In: Journal of Cross-Cultural Psychology 20: 80-91.

Georgas, James 1991: Intrafamily Acculturation of Values in Greece. In: Journal of 
Cross-Cultural Psychology 22,4: 445-457.

Georgas, James et al. 1996: Acculturation of Greek Family Values. In: Journal of Cross-
Cultural Psychology 27: 329-338.

Giarrusso, Roseann; Silverstein, Merril; Gans, Daphna; Bengtson, Vern 2005: Ageing 
Parents and Adult Children: New Perspectives on Intergenerational Relationships. In: 
Johnson, Malcolm L. (Ed.): The Cambridge Handbook of Age and Ageing. Cambridge/
New York: Cambridge University Press: 413-421 [doi: 10.2277/0521826322].

Giarusso, Roseann; Feng, Du; Bengtson, Vern L. 2004: The intergenerational-stake phe-
nomenon over 20 years. In: Annual review of gerontology and geriatrics 24: 55-76.

Giarusso, Roseann; Stallings, Michael; Bengtson, Vern L. 1995: The “intergenerational 
stake” hypothesis revisited: Parent-child differences in perception of relationships 20 
years later. In: Bengston, Vern L.; Schaie, Warner K.; Burton, Linda M. (Eds.): Adult 
intergenerational relations: Effects of societal change. New York: Springer: 227-263.



•    Bernhard Nauck672

Guo, Man; Aranda, Maria P.; Silverstein, Merril 2009: The impact of out-migration on 
the intergenerational support and psychological wellbeing of older adults in rural Chi-
na. In: Aging and Society 29: 1085-1104 [doi: 10.1017/S0144686X0900871X].

Guo, Man; Chi, Iris; Silverstein, Merril 2012: The Structure of Intergenerational Relations 
in Rural China: A Latent Class Analysis. In: Journal of Marriage and Family 74: 1114-
1128 [doi: 10.1111/j.1741-3737.2012.01014.x].

Hank, Karsten 2007: Proximity and Contacts Between Older Parents and Their Chil-
dren: A European Comparison. In: Journal of Marriage and Family 69: 157-173 [doi: 
10.1111/j.1741-3737.2006.00351.x].

Hirschman, Albert O. 1970: Exit, Voice and Loyalty. Responses to Decline in Firms, Or-
ganizations and States. Cambridge, MA: Harvard University Press.

Kagitcibasi, Cigdem; Ataca, Bilge 2005: Values of Children and Family Change: A Three-
Decade Portrait from Turkey. In: Applied Psychology 54,3: 17-337 [doi: 10.1111/j.1464-
0597.2005.00213.x].

Kagitcibasi, Cigdem; Ataca, Bilge; Diri, Aysesim 2010: Intergenerational Relation-
ships in the Family: Ethnic, Socioeconomic, and Country Variations in Germany, Is-
rael, Palestine, and Turkey. In: Journal of Cross-Cultural Psychology 41: 652-670 [doi: 
10.1177/0022022110372193].

Katz, Ruth; Lavee, Yoav 2005: Families in Israel. In: Adams, Bert N.; Trost, Jan (Eds.): 
Handbook of World Families. Thousand Oaks/London: Sage: 486-506.

Klaus, Daniela; Tipandjan, Arun (in press): Son Preference in India: Shedding Light on 
the North-South Gradient. Comparative Population Studies 40,1.

Kohli, Martin 2005: Intergenerational Transfers and Inheritance: A Comparative View. In: 
Silverstein, Merril; Schaie, Warner K. (Eds.): Intergenerational Relations Across Time 
and Place. Vol. 24 Annual Review of Gerontology and Geriatrics: 266-289.

Kopp, Johannes; Steinbach, Anja 2009: Generationenbeziehungen. Ein Test der interge-
nerational-stake-Hypothese. In: Kölner Zeitschrift für Soziologie und Sozialpsycholo-
gie 61: 283-294 [doi: 10.1007/s11577-009-0054-9].

Lindenberg, Siegwart 1989: Choice and Culture: The Behavioral Basis of Cultural Impact 
on Transactions. In: Haferkamp, Heinz (Ed.): Social Structure and Culture. Berlin: de 
Gruyter: 175-200.

Lindenberg, Siegwart 1996: Continuities in the theory of social production functions. In: 
Ganzeboom, Harry B.; Lindenberg, Siegwart (Eds.): Verklarende Sociologie: Opstellen 
voor Reinhard Wippler. Amsterdam: Thela Thesis: 169-184.

Lowenstein, Ariella 2006: Solidarity-confl ict and ambivalence: Testing two conceptual 
frameworks and their impact on quality of life for older family members. In: Journal of 
Gerontology Social Sciences 62B: S100-S107.

Lowenstein, Ariella; Daatland, Svein O. 2006: Filial norms and family support in a com-
parative cross-national context: evidence from the OASIS study. In: Aging and Society 
26: 203-223 [doi: 10.1017/S0144686X05004502].

Lowenstein, Ariella; Katz, Ruth; Daatland, Svein O. 2005: Filial norms and intergenera-
tional support in Europe and Israel: A comparative perspective. In: Silverstein, Merril 
(Ed.): Annual review of gerontology and geriatrics. Intergenerational relations across 
time and place. Vol. 24: 200-223.

Lowenstein, Ariella; Katz, Ruth; Mehlhausen-Hassoen, David; Prilutzky, Dana 2002: 
The research instruments in the OASIS project. Haifa, Israel: Center for Research and 
Study of Aging at the University of Haifa.



Affection and Confl ict in Intergenerational Relationships of Women    • 673

Lüscher, Kurt 2002: Intergenerational Ambivalence: Further Steps in Theory and 
Research. In: Journal of Marriage and Family 64. 585-593 [doi: 10.1111/j.1741-
3737.2002.00585.x].

Lüscher, Kurt; Pillemer, Karl 1998: Intergenerational Ambivalence: A New Approach to 
the Study of Parent-Child Relations in Later Life. In: Journal of Marriage and the Family 
60: 413-425.

Mayer, Boris; Trommsdorff, Gisela 2010: Adolescents’ Value of Children and Their Inten-
tions to Have Children: A Cross-Cultural and Multilevel Analysis. In: Journal of Cross-
Cultural Psychology 41: 671-689 [doi: 10.1177/0022022110372195].

Mishra, Ramesh C. et al. 2005: The Value of Children in Urban and Rural India: Cultural 
Background and Empirical Results. In: Trommsdorff, Gisela; Nauck, Bernhard (Eds.): 
The Value of Children in Cross-cultural Perspective. Case Studies from Eight Societies. 
Lengerich: Pabst: 143-170.

Nauck, Bernhard 2001: Der Wert von Kindern für ihre Eltern. “Value of Children” als 
spezielle Handlungstheorie des generativen Verhaltens und von Generationenbezie-
hungen im interkulturellen Vergleich. In: Kölner Zeitschrift für Soziologie und Sozi-
alpsychologie 53: 407-435 [doi: 10.1007/s11577-001-0073-7].

Nauck, Bernhard 2006: Value of Children and Fertility Strategies in Cross-cultural Com-
parison. Ideal Family Size and Targeted Fertility in Eleven Societies. In: Gomes, Cris-
tina (Ed.): Social Development and Family Changes. Newcastle: Cambridge Scholars 
Press: 300-344.

Nauck, Bernhard 2007a: Familiensystem und Kultur. In: Trommsdorff, Gisela; Kornadt, 
Hans J. (Eds.): Theorien und Methoden der kulturvergleichenden Psychologie. Vol. S1: 
Enzyklopädie der Psychologie. Göttingen: Hogrefe: 407-486.

Nauck, Bernhard 2007b: Value of Children and the Framing of Fertility: Results from a 
Cross-cultural Comparative Survey in 10 Societies. In. European Sociological Review 
23: 615-629 [doi: 10.1093/esr/jcm028].

Nauck, Bernhard 2009: Patterns of Exchange in Kinship Systems in Germany, Russia, 
and the People’s Republic of China. In: Journal of Comparative Family Studies 40: 
255-278.

Nauck, Bernhard 2010: Intergenerational Relationships and Female Inheritance Expecta-
tions: Comparative Results From Eight Societies in Asia, Europe, and North America. 
In: Journal of Cross-Cultural Psychology 41: 690-705 [doi: 10.1177/0022022110375161].

Nauck, Bernhard 2012: Value of Children and Intergenerational Solidarity. In: Bertram, 
Hans; Ehlert, Nancy (Eds.): Family, Ties, and Care. Family Transformation in a Plural 
Modernity. Opladen: Barbara Budrich: 299-314.

Nauck, Bernhard; Arránz Becker, Oliver 2013: Institutional Regulations and the Kinship 
Solidarity of Women – Results from 13 Areas in Asia, Africa, Europe, and North Amer-
ica. In: European Sociological Review 29: 580-592 [doi: 10.1093/esr/jcr110].

Nauck, Bernhard; Klaus, Daniela 2007: The Varying Value of Children: Empirical Results 
from Eleven Societies in Asia, Africa and Europe. In: Current Sociology 55: 487-503 
[doi: 10.1177/0011392107077634].

Nauck, Bernhard; Klaus, Daniela 2008: Family Change in Turkey: Peasant Society, Islam, 
and the Revolution “From Above”. In: Jayokody, Rukmalie; Thornton, Arland; Axinn, 
William (Eds.): International Family Change. Ideational Perspectives. New York: Law-
rence Erlbaum Associates: 281-312.



•    Bernhard Nauck674

Nauck, Bernhard; Suckow, Jana 2006: Intergenerational Relationships in Cross-Cultural 
Comparison: How Social Networks Frame Intergenerational Relations Between Moth-
ers and Grandmothers in Japan, Korea, China, Indonesia, Israel, Germany, and Turkey. 
In: Journal of Family Issues 27: 1159-1185 [doi: 10.1177/0192513X06288125].

Ormel, Johan; Lindenberg, Siegwart; Steverink, Nardi; Verbrugge, Lois M. 1999: Sub-
jective Well-Being and Social Production Functions. In: Social Indicators Research 46: 
61-90 [doi: 10.1023/A:1006907811502].

Pillemer, Karl; Suitor, Jill J. 2002: Explaining Mothers’ Ambivalence Toward Their 
Adult Children. In: Journal of Marriage and Family 64: 602-613 [doi: 10.1111/j.1741-
3737.2002.00602.x].

Pillemer, Karl et al. 2007: Capturing the complexity of intergenerational relations: Explor-
ing ambivalence within later-life families. In: Journal of Social Issues 63,4: 775-791.

Rusbult, Caryl E.; Johnson, Dennis J.; Morrow, Gregory D. 1986: Determinants and Con-
sequences of Exit, Voice, Loyalty, and Neglect – Responses to Dissatisfaction in Adult 
Romantic Involvements. In: Human Relations 39: 45-63.

Rusbult, Caryl E.; Van Lange, Paul A. M. 2003: Interdependence, interaction, and re-
lationships. In: Annual Review of Psychology 54: 351-375 [doi: 10.1146/annurev.
psych.54.101601.145059].

Rusbult, Caryl E. et al. 1991: Accomodation Processes in Close Relationships – Theory 
and Preliminary Empirical Evidence. In: Journal of Personality and Social Psychology 
60: 53-78.

Rusbult, Caryl E.; Zembrodt, Isabella M.; Gunn, Lawanna K. 1982: Exit, Voice, Loyalty, 
and Neglect – Responses to Dissatisfaction in Romantic Involvements. In: Journal of 
Personality and Social Psychology 43: 1230-1242.

Schenk, Niels; Dykstra, Pearl A. 2012: Continuity and change in intergenerational family 
relationships: An examination of shifts in relationship type over a three-year period. 
In: Advances in Life Course Research 17: 121-132 [doi: 10.1016/j.alcr.2012.01.004].

Schwarz, Beate 2000: Frauen aus verschiedenen Familienformen und ihre alten Eltern 
– Beziehungsqualität und wechselseitige Unterstützung. In: Zeitschrift für Soziologie 
der Erziehung und Sozialisation 20: 410-424.

Schwarz, Beate et al. 2010: Intergenerational Support and Life Satisfaction: A Compari-
son of Chinese, Indonesian, and German Elderly Mothers. In: Journal of Cross-Cultural 
Psychology 41: 706-722 [doi: 10.1177/0022022110372197].

Schwarz, Beate; Trommsdorff, Gisela; Albert, Isabelle; Mayer, Boris 2005: Adult Parent-
Child Relationships: Relationship Quality, Support, and Reciprocity. In: Applied Psy-
chology 54: 396-417 [doi: 10.1111/j.1464-0597.2005.00217.x].

Silverstein, Merril; Bengtson, Vern L.; Lawton, Leora 1997: Intergenerational Solidar-
ity and the Structure of Adult Child – Parent Relationships in American Families. In: 
American Journal of Sociology 103: 429-460 [doi: 10.1086/231213].

Silverstein, Merril; Burholt, Vanessa; Wenger, G. Clare; Bengtson, Vern L. 1998: Par-
ent Child Relations among very old Parents in Wales and the United States: A Test of 
Modernization Theory. In: Journal of Aging Studies 12: 387-409 [doi: 10.1016/S0890-
4065(98)90026-3].

Silverstein, Merril; Cong, Zhen; Li, Shuzhuo 2006: Intergenerational Transfers and Liv-
ing Arrangements of Older People in Rural China: Consequences for Psychological 
Well-Being. In: Journal of Gerontology: Social Sciences 61B: S256-S266.



Affection and Confl ict in Intergenerational Relationships of Women    • 675

Silverstein, Merril et al. 2010: Older Parent - Child Relationships in Six Developed Na-
tions: Comparisons at the Intersection of Affection and Confl ict. In: Journal of Mar-
riage and Family 72: 1006-1021 [doi: 10.1111/j.1741-3737.2010.00745.x].

Silverstein, Merril et al. 2013: Intergenerational Support and the Emotional Well-being 
of Older Jews and Arabs in Israel. In: Journal of Marriage and Family 75: 950-963 [doi: 
10.1111/jomf.12041].

Singelis, Theodore M. 1994: The measurement of independent and interdepend-
ent self-construals. In: Personality and Social Psychology Bulletin 20: 580-591 [doi: 
10.1177/0146167294205014].

Singh, J. P. 2005: The contemporary Indian Family. In: Adams, Bert N.; Trost, Jan (Eds.): 
Handbook of World Families. Thousand Oaks/London: Sage: 129-166.

Smith, Peter B.; Fischer, Ronald 2008: Aquiescence, Extreme Response Bias and Cul-
ture. A Multilevel Analysis. In: van de Vijver, Fons J. R.; van Hemert, Dianne A.; Poort-
inga, Ype (Eds.): Multilevel Analysis of Individuals and Cultures. New York/London: 
Lawrence Erlbaum Associates: 285-314.

Szydlik, Marc 1996: Parent-child relations in East and West Germany shortly after the fall 
of the wall. In: International Journal of Sociology and Social Policy 16: 63-88.

Trommsdorff, Gisela; Mayer, Boris 2012: A Cross-Cultural Study of Intergenerational Re-
lations: The Role of Socioeconomic Factors, Values, and Relationship Quality in Inter-
generational Support. In: Bertram, Hans; Ehlert, Nancy (Eds.): Family, Ties, and Care. 
Family Transformation in a Plural Modernity. Opladen: Barbara Budrich: 315-342.

Trommsdorff, Gisela; Nauck, Bernhard (Eds.) 2005: The Value of Children in Cross-Cul-
tural Perspective. Case Studies from Eight Societies. Lengerich, Germany: Pabst.

Trommsdorff, Gisela; Nauck, Bernhard 2010: Introduction to Special Section for Journal 
of Cross-Cultural Psychology: Value of Children: A Concept for Better Understanding 
Cross-Cultural Variations in Fertility Behavior and Intergenerational Relationships. In: 
Journal of Cross-Cultural Psychology 41: 637-651 [doi: 10.1177/0022022110373335].

Trommsdorff, Gisela; Schwarz, Beate 2007: The “Intergenerational Stake Hypothesis” in 
Indonesia and Germany. Adult Daughters’ and their Mothers’ Perception of their Rela-
tionship. In: Current Sociology 55: 599-620 [doi: 10.1177/0011392107077641].

van de Vijver, Fons J. R.; Leung, Kwok 1997: Methods and data analysis for cross-cultural 
research. Thousand Oaks: Sage.

van Gaalen, Ruben I.; Dykstra, Pearl A. 2006: Solidarity and confl ict between adult chil-
dren and parents: A latent class analysis. In: Journal of Marriage and Family 68: 947-
960 [doi: 10.1111/j.1741-3737.2006.00306.x].

van Gaalen, Ruben I.; Dykstra, Pearl A.; Komter, Aafke E. 2010: Where is the exit? Inter-
generational ambivalence and relationship quality in high contact ties. In: Journal of 
Aging Studies 24: 105-114 [doi: 10.1016/j.jaging.2008.10.006].

Whiting, John W. M. 1968: Methods and Problems in Cross-Cultural Research. In: Gard-
ner, Lindzey; Aronson, Elliot (Eds.): The Handbook of Social Psychology. 2 ed., Vol. 2. 
Reading, MA: Addison-Wesley: 693-728.



•    Bernhard Nauck676

Date of submission: 28.02.2013  Date of acceptance: 06.01.2014

Prof. Dr. Bernhard Nauck (). Department of Sociology, Technische Universitaet 
Chemnitz. Chemnitz, Germany. E-Mail: bernhard.nauck@soziologie.tu-chemnitz.de 
URL: http://www.tu-chemnitz.de/hsw/soziologie/Institut/Profi le/bna/index.php



Affection and Confl ict in Intergenerational Relationships of Women    • 677

A
p

p
e

n
d

ix

T
a

b
. 

A
1:

 
D

e
sc

ri
p

ti
v
e
 s

ta
ti

st
ic

s 
fo

r 
th

e
 s

tu
d

ie
d

 s
am

p
le

s 
fr

o
m

 1
6
 a

re
a
s:

 in
d

iv
id

u
a
l c

h
a
ra

c
te

ri
st

ic
s 

(n
 =

 9
,3

0
3
 f

e
m

a
le

 
re

sp
o

n
d

e
n

ts
)

  
In

d
ia

 N
o

rt
h

In
d

ia
 S

o
u

th
P
al

e
st

in
e

In
d

o
n

e
si

a
C

h
in

a
J
am

ai
ca

R
u

ss
ia

Tu
rk

e
y

A
g

e
 o

f 
re

sp
o

n
d

e
n

t
3
3
.0

3
1
.8

3
3
.1

3
3
.0

3
5
.4

3
3
.1

3
4
.1

3
5
.4

N
u

m
b

e
r 

o
f 

ch
ild

re
n

2
.6

2
.1

3
.7

2
.3

1
.4

2
.3

1
.6

2
.2

R
u

ra
l b

ac
kg

ro
u

n
d

 (
0
-3

)
1
.7

1
.6

1
.7

1
.6

1
.5

1
.2

2
.1

1
.2

E
x

te
n

d
e
d

 h
o

u
se

h
o

ld
 (

%
)

4
8
.7

3
7
.1

3
.5

4
5
.5

3
0
.2

1
8
.8

2
6
.9

1
1
.2

E
d

u
ca

ti
o

n
 <

 p
ri

m
ar

y
 (%

)
5
1
.3

2
0
.3

4
.0

4
.2

6
.6

1
.2

0
.2

8
.1

E
d

u
ca

ti
o

n
 t

e
rt

ia
ry

 (%
)

2
8
.9

8
.8

2
3
.4

7
.6

1
2
.6

1
0
.6

4
5
.9

2
4
.6

La
b

o
r 

st
at

u
s 

(0
-6

)
0
.7

0
.8

1
.2

1
.6

3
.3

2
.8

4
.1

1
.6

H
o

u
se

h
o

ld
 S

E
S

 (
0
-1

0
)

4
.2

2
.8

4
.3

3
.7

3
.4

3
.2

5
.0

4
.6

N
o

rm
at

iv
e
 O

b
lig

at
io

n
s 

(1
-5

)
4
.4

4
.7

4
.6

4
.7

4
.2

4
.3

4
.2

4
.3

C
o

m
fo

rt
 V

O
C

 (
1
-5

)
3
.8

3
.6

3
.7

4
.0

2
.4

3
.0

2
.8

2
.5

A
ff

e
ct

io
n

 V
O

C
 (

1
-5

)
4
.2

4
.8

4
.5

4
.4

4
.2

4
.2

4
.4

4
.6

S
o

u
th

 A
fr

ic
a

P
o

la
n

d
E

st
o

n
ia

Is
ra

e
l

Fr
an

ce
G

e
rm

an
y
 E

as
t

G
e
rm

an
y
 W

e
st

U
n

it
e
d

 S
ta

te
s

A
g

e
 o

f 
re

sp
o

n
d

e
n

t
3
2
.7

4
1
.7

4
1
.0

3
5
.8

4
4
.3

3
6
.1

3
9
.1

4
4
.7

N
u

m
b

e
r 

o
f 

ch
ild

re
n

2
.6

2
.3

2
.4

2
.6

2
.7

1
.8

2
.0

3
.2

R
u

ra
l b

ac
kg

ro
u

n
d

 (
0
-3

)
2
.8

1
.5

1
.4

0
.3

0
.3

0
.5

1
.6

1
.5

E
x

te
n

d
e
d

 h
o

u
se

h
o

ld
 (

%
)

2
1
.2

1
8
.0

1
0
.0

1
.4

2
.7

1
.1

1
.4

2
.1

E
d

u
ca

ti
o

n
 <

 p
ri

m
ar

y
 (%

)
1
0
.9

0
.3

0
.4

0
.0

0
.5

1
.1

0
.3

0
.0

E
d

u
ca

ti
o

n
 t

e
rt

ia
ry

 (%
)

2
6
.0

4
8
.5

3
3
.3

6
6
.5

6
4
.6

3
7
.2

4
4
.6

5
4
.6

La
b

o
r 

st
at

u
s 

(0
-6

)
2
.3

3
.9

4
.9

3
.6

3
.7

4
.0

3
.2

4
.8

H
o

u
se

h
o

ld
 S

E
S

 (
0
-1

0
)

2
.8

5
.1

4
.9

4
.9

7
.0

4
.1

4
.9

6
.9

N
o

rm
at

iv
e
 O

b
lig

at
io

n
s 

(1
-5

)
4
.7

4
.3

4
.0

4
.3

4
.0

4
.0

3
.8

4
.2

C
o

m
fo

rt
 V

O
C

 (
1
-5

)
3
.7

2
.5

2
.6

2
.9

1
.7

1
.7

1
.6

1
.8

A
ff

e
ct

io
n

 V
O

C
 (

1
-5

)
4
.0

4
.3

3
.8

4
.4

4
.3

4
.0

4
.0

4
.3

S
o

u
rc

e
: 

V
al

u
e
 o

f 
C

h
ild

re
n

 S
u

rv
e
y



•    Bernhard Nauck678

T
a

b
. 

A
2
: 

D
e

sc
ri

p
ti

v
e
 s

ta
ti

st
ic

s 
fo

r 
th

e
 s

tu
d

ie
d

 s
am

p
le

s 
fr

o
m

 1
6
 a

re
a
s:

 r
e

la
ti

o
n

a
l c

h
a
ra

c
te

ri
st

ic
s

  
In

d
ia

 N
o

rt
h

In
d

ia
 S

o
u

th
P
al

e
st

in
e

In
d

o
n

e
si

a
C

h
in

a
J
am

ai
ca

R
u

ss
ia

Tu
rk

e
y

A
g

e
 M

o
th

e
r

6
0
.2

6
0
.1

6
2
.4

5
9
.4

6
5
.1

5
8
.9

5
8
.6

6
4
.0

A
g

e
 F

at
h

e
r

6
1
.5

6
1
.6

6
3
.6

6
1
.6

6
5
.7

6
0
.3

6
1
.2

6
4
.8

P
ro

x
im

it
y
 t

o
 F

at
h

e
r 

(1
-5

)
2
.6

3
.0

3
.1

4
.1

3
.3

2
.9

3
.4

3
.1

P
ro

x
im

it
y
 t

o
 M

o
th

e
r 

(1
-5

)
2
.6

3
.3

3
.0

4
.2

3
.4

3
.8

3
.8

3
.1

C
o

n
ta

ct
 t

o
 F

at
h

e
r 

(1
-5

)
2
.0

3
.1

3
.1

3
.7

3
.1

3
.2

3
.5

3
.0

C
o

n
ta

ct
 t

o
 M

o
th

e
r 

(1
-5

)
2
.0

3
.4

3
.3

3
.9

3
.2

4
.3

4
.1

3
.4

M
u

tu
al

 h
el

p
 F

at
h

er
 (1

-6
)

2
.9

3
.3

3
.5

3
.4

3
.9

3
.2

3
.6

3
.3

M
u

tu
al

 h
e
lp

 M
o

th
e
r 

(1
-6

)
3
.1

3
.6

3
.6

3
.8

4
.1

4
.2

4
.3

3
.5

S
o

u
th

 A
fr

ic
a

P
o

la
n

d
E

st
o

n
ia

Is
ra

e
l

Fr
an

ce
G

e
rm

an
y
 E

as
t

G
e
rm

an
y
 W

e
st

U
n

it
e
d

 S
ta

te
s

A
g

e
 M

o
th

e
r

6
0
.9

6
8
.3

6
7
.5

6
3
.2

7
2
.3

6
1
.9

6
7
.3

8
0
.0

A
g

e
 F

at
h

e
r

6
3
.7

6
9
.9

6
8
.6

6
4
.8

7
3
.0

6
2
.7

6
8
.1

8
0
.9

P
ro

x
im

it
y
 t

o
 F

at
h

e
r 

(1
-5

)
4
.1

3
.3

3
.0

2
.8

3
.1

2
.9

2
.9

3
.0

P
ro

x
im

it
y
 t

o
 M

o
th

e
r 

(1
-5

)
4
.2

3
.5

3
.2

3
.1

3
.3

3
.0

3
.0

3
.0

C
o

n
ta

ct
 t

o
 F

at
h

e
r 

(1
-5

)
3
.3

3
.4

2
.8

3
.9

3
.2

3
.0

3
.2

3
.1

C
o

n
ta

ct
 t

o
 M

o
th

e
r 

(1
-5

)
3
.5

3
.7

3
.6

4
.3

3
.6

3
.7

3
.6

3
.6

M
u

tu
al

 h
el

p
 F

at
h

er
 (1

-6
)

4
.1

3
.6

3
.3

3
.0

3
.2

2
.8

2
.7

2
.6

M
u

tu
al

 h
e
lp

 M
o

th
e
r 

(1
-6

)
4
.4

3
.8

3
.7

3
.4

3
.6

2
.9

2
.9

2
.8

S
o

u
rc

e
: 

V
al

u
e
 o

f 
C

h
ild

re
n

 S
u

rv
e
y



Published by / Herausgegeben von
Prof. Dr. Norbert F. Schneider

Federal Institute for Population Research 
D-65180 Wiesbaden / Germany

Managing Editor /
Verantwortlicher Redakteur
Frank Swiaczny

Assistant Managing Editor /
Stellvertretende Redakteurin
Katrin Schiefer

Language & Copy Editor (English) /
Lektorat & Übersetzungen (englisch)
Amelie Franke

Copy Editor (German) /
Lektorat (deutsch)
Dr. Evelyn Grünheid

Layout / Satz
Beatriz Feiler-Fuchs

E-mail: cpos@bib.bund.de

Comparative Population Studies

www.comparativepopulationstudies.de

ISSN: 1869-8980 (Print) – 1869-8999 (Internet)

Scientifi c Advisory Board / 
Wissenschaftlicher Beirat
Paul Gans (Mannheim) 
Johannes Huinink (Bremen) 
Michaela Kreyenfeld (Rostock)
Marc Luy (Wien) 
Clara H. Mulder (Groningen) 
Notburga Ott (Bochum) 
Peter Preisendörfer (Mainz)
Zsolt Spéder (Budapest)

Board of Reviewers / Gutachterbeirat
Martin Abraham (Erlangen)
Laura Bernardi (Lausanne)
Hansjörg Bucher (Bonn)
Claudia Diehl (Konstanz)
Andreas Diekmann (Zürich)
Gabriele Doblhammer-Reiter (Rostock)
E.-Jürgen Flöthmann (Bielefeld)
Alexia Fürnkranz-Prskawetz (Wien)
Beat Fux (Salzburg)
Joshua Goldstein (Berkeley)
Karsten Hank (Köln)
Sonja Haug (Regensburg)
Aart C. Liefbroer (Den Haag)
Kurt Lüscher (Konstanz)
Dimiter Philipov (Wien)
Tomáš Sobotka (Wien)
Heike Trappe (Rostock)

© Federal Institute for Population Research 2014 – All rights reserved