Marriage in Europe: Ideal types of marriage in the first decade of the twenty-first century

Europeans today hold differing opinions on what they consider to be ideal marital relationships: they can last a lifetime or reach an expiration date, be necessary or optional, in vogue or outdated; the persons involved can refer to each other as spouses, partners or companions. What many Europeans define as an ideal marriage has been affected by the transformation of social values and shaped by individuals' external reality (Inglehart 1997; Herpin 2002). As a result, ‘Different rules and practices of marriage and sexuality have remained in the world’ (Therborn 2004: 315).

The current diversity of ideals and partnerships is not an exclusive characteristic of contemporary society (Boh et al.1989; Morgan 1991; Kuijsten 1996). As Burgess remarked in his well-known article in 1948: ‘Never before in human history has any society been composed of so many divergent types of families’ (Burgess 1948: 417). While emphasizing the diversity of the North American family, he pointed out those elements that conferred a certain unity on it: the tendency toward companionship appearing as an element providing identity to the American family (Burgess and Locke 1945; Burgess 1948).

In contemporary Europe, disagreements over what the marital relationship should be are proof of diversity, while similar ideas and beliefs reveal a certain unity. For example, opinion is divided over the legitimacy of divorce: on a scale of 1–10, with 1 being that divorce is never justified and 10 being that it is always justified, the European average is 5.96. In fact, very traditional postures such as ‘a bad marriage is better than no marriage’, ascribed to by 10% of Europeans surveyed in the International Social Survey Programme (ISSP 2002), coexist with less extreme views, such as ‘a stable relationship is necessary to be happy’, an opinion held by 60% of those interviewed, and more liberal views such as ‘divorce is always justified’, which 16% of those interviewed agree with.

The consensus emphasizes elements of identity: the importance of the marital relationship and the centrality of love. The percentage of respondents ‘in favour of marriage’ is over 80% in all the countries surveyed, reaching a maximum of approximately 95% in the Nordic countries (Eurobarometer 50.1 1998). Agreement over ‘the importance to live with someone as husband, wife or partner’ reaches similar levels, which shows that those surveyed identify marriage with an enduring relationship. The percentage that consider ‘having a relationship’ as very or quite important is greater than 80% in all the countries and reaches 91% in Sweden and almost unanimity in Greece, 98.6% (Eurobarometer 50.1 1998). The importance granted to interpersonal commitment and love in a stable relationship can be seen in attitudes toward faithfulness – 85% of Europeans seeing fidelity as important or very important for a successful marriage – (EVS 2010) and towards love – 78% of Europeans seeing ‘love one another deeply’ is very important ‘to live successfully with a partner’ – (Malpas and Lambert 1993). However, love not only defines the character of the relationship, but is also what fosters it: love and mutual attraction have reached, according to Buss et al. (2001), first place among the criterium for choosing a partner.

Differing ideas regarding what long-term coupling should be, reveal the complex diversity characteristic of contemporary society (Roussel 1989; Brines and Joiner 1999; Bengston 2001), which can be a source of conflict (Therborn 2004).

Over the twentieth century, two studies captured transcendental periods in the transformation of marriage: Burgess and Locke's study in the 1940s and Roussel's study at the beginning of the 1980s; both followed the same process, constructing ideal types portraying the marriage models of the time period. Burgess and Locke, in their empirical study, argued that ‘…the family in historical times has been, and at present is, in transition from an institution to a companionship’ (Burgess and Locke 1945: vii).

The typologies and analyses Burgess and Locke developed indicate the new direction the family was taking, but they also show that this new model was still not very prevalent throughout society. Thus, out of the five types of family integration they identify, the ‘dynamically unified family’, characterized by the voluntary participation of its members to achieve a common goal, did not enjoy great popularity, while older forms of integration based on patriarchal subordination (the highly solidified family) or habit (habit-bound family) remained widespread (Burgess and Locke 1945: 351).

Thirty-five years after Burgess and Locke's pioneering study of the American family, the demographer and sociologist Louis Roussel published the results of a study in which he analyzed the values and attitudes of French youth towards marriage in the 1970s (Roussel 1980). While other classifications were based on behavioural studies (Fowers and Olson 1989; Widmer et al. 2006), Roussel's consisted in constructing ideal types based on evaluative opinions. As well as certifying the end of institutional marriage, at least among the younger generations, he described three existing marriage ideals – the Alliance, Fusion and Association marriages – and identified the fusion marriage as the most prevalent in Europe (Roussel 1980).

The emergence of the alliance marriage (mariage d'alliance) implied a revolutionary transformation in conjugal relationships and rendered institutional marriage obsolete. The motivation behind the alliance marriage is the quest for happiness; individuals marry because they believe it is necessary to achieve happiness. However, obstacles for the breakup of the marital relationship remain as a vestige of the past. Fusion marriage (mariage-fusion), which refers to the ‘fusion of personalities’ characteristic of primary relationships, is based on affective solidarity. In such a relationship it is possible to revoke the bond if love disappears. Divorce sanctions a fait accompli, without there necessarily being any offense or a guilty party. In the association model (mariage-association ou compagnonnage) marriage is not considered to be essential but is seen as an association, as ‘a close-knit team in which each partner expects satisfying results’ (Roussel 1980: 1030). Its duration is necessarily connected to the rewards and benefits it provides: ‘Each one's happiness is what keeps the couple together’ (Roussel 1980: 1031).

A late twentieth-century trend towards a privatization and democratization of coupling was theorized by Anthony Giddens (1992) as an emergence of ‘pure relationships’ (Jamieson 1998, 1999; Gross and Simmons 2002), with confluent love as an alternative to romantic love, a change from a partnership in which a couple seek happiness together (fusion) to one in which individuals whose paths have converged at a point aspire to achieve happiness individually.

To define and construct the ideal types and calculate what proportion of the population of the European Union supports each posture we have used Roussel's typologies. This will permit us to construct a European map of ideal types of marriage (see Table 1).

To reconstruct the ideal types of marriages, we have used data collected in a comprehensive survey of values which has been carried out since 1981 by the European Values Study, a cross-national and longitudinal research project. The data analyzed comes from the fourth wave of the survey, conducted between 2008 and 2010, in which 47 European countries participated. Here we shall analyze the preferences and values of Europeans from 24 of the 27 countries that currently make up the European Union. Given that the typologies were constructed based on data from the 1999 edition, in which Cyprus did not participate and Romania and Bulgaria were not members of the EU, these three countries have been omitted. The total size of the sample is N=35,976 and, with few exceptions, the net sample size is 1500 respondents per country. Sampling procedure was based on a representative multi-stage or stratified random sample. The questionnaires were administered in face-to-face interviews in the appropriate national language in 22 of the 24 countries.

Among the questions on the extensive questionnaire, and after analyzing the validity of their content as measurements of the concepts that Roussel signals as characteristics of ideal marriages, we have selected five as indicators (Table 2). The nature and indissolubility of the bond, the social importance of the institution and the functionality of the relationship are concepts represented by the variables measured on a Likert-type scale, in which respondents are asked for their degree of agreement or disagreement with a statement, the importance of certain qualities or the extent to which certain behaviour is justifiable. Based on the answers we have been able to reconstruct the diverse definitions of ideal relationships as seen by Europeans. Although the term marriage is used in all of the questions to refer to a generic partner relationship, the rate of non-response, which would indicate a lack of identification with the type of relationship referred to in the questions, is low.

To discover the internal logic that connects the five variables we have used Categorical Principal Components Analysis (CATPCA). This data reduction technique consists in finding the dimension or dimensions underlying the grouping of categories of variables. After we selected two components (nature of the relationship and its functionality), we graphed the scores obtained for the categories of variables. The level of reliability of the analysis is acceptable; the total Cronbach's alpha is 0.796.

In Figure 1, showing scores by category, we can see the first dimension chosen (represented on the horizontal axis) defines the nature of conjugal relationships. The categories of the answers grouped on the left describe a marriage relationship that is more traditional and, above all, one that is indissoluble. In contrast, the categories on the right describe relationships that are voluntary and, above all, revocable. Based on Roussel's typology we have marked two cut-off points that mark the spatial limits of his three ideal types. The vertical axis is almost exclusively generated from categories which refer to the necessity (or contingency) of the marriage relationship. From the crossing of the two dimensions we obtain six subtypes. The most traditional one, the institutional model, appears in the upper-left quadrant, where divorce is ‘never justified’ and the respondents ‘strongly agree’ that ‘marriage is necessary’. The most liberal type, described as the ‘pure relationship’ by Giddens, is represented in the lower-right quadrant, with respondents saying divorce is ‘always justified’ and strongly disagreeing that ‘marriage is necessary’.

Through multivariable analysis we have identified three ideal types of marital relationships – alliance, fusion and association – and two variants of each one based on whether marriage is considered necessary or contingent. We have quantified the proportion of Europeans who currently identify with each of the three ideal types and six subtypes. Section 4.1 describes the three basic ideal types and quantifies their adherents. Section 4.2 specifies the characteristics of the more extreme subtypes: the institutional model, which 27.8% of the population still supports, and the ‘pure relationship’ type, supported by 8.4% of the population. In section 4.3, we provide an image of Europe based on the prevalence of the different ideal types in 24 countries of the European Union.

The first dimension that underlies the diversity of opinions that Europeans express on marriage concerns the nature of the relationship and particularly the reasons for its establishment and dissolution. The three columns in Table 3 show the three ideal types based on the response categories from the three segments of the horizontal axis (Figure 1). Slightly more than four out of 10 Europeans support the alliance marriage; 38.1% support the fusion marriage, and the association marriage is considered desirable by 19.5%.

A relative majority of Europeans, 42.4%, support the traditional ideal of an alliance marriage: one that is ‘for life’ (they do not agree it is outdated and divorce is rarely justifiable) with ‘the love of their lives’ (adultery is never justified). Although spouses are bonded by love in an alliance marriage, the institutional component is still of great importance: marriage is whatever society mandates it to be and there are no objective reasons for its dissolution. If we look at the responses of those who share an ideal type (Table 4), we find a high level of consensus among supporters of the alliance marriage. The almost unanimous agreement that marriage is not ‘outdated’ is an expression of the acceptance of the legitimacy of social norms regulating the relationship.

It is noteworthy that despite agreement on the importance of faithfulness ‘for a successful marriage’ among supporters of the alliance marriage (Table 4), this is not one of the defining elements of the model (Table 3). This, together with their absolute rejection of adultery, leads us to believe that infidelity (lack of faithfulness) and adultery are not synonyms, and that what is rejected is sexual infidelity. Thus, it is sexual fidelity, and not fidelity in the sense of a profession of faith in the other person, which defines the traditional marriage.

Fusion marriage is defined as a relationship where faithfulness – understood as an attitude of respect motivated by a feeling of pride – is very important for the marriage to succeed, or in other words, for it to last. The duration of the marriage depends on the state of the marital relationship, because it neither excludes the possibility of divorce, as supporters of the alliance marriage do, nor does it consider the possibility of the relationship ending for no reason, as supporters of the association marriage do. There are various reasons (5–9) for a fusion marriage to end, all of which can be summarized in one: ‘when it does not work’. Other than faithfulness (87% agreeing it is very important), there is no other social norm that regulates this marital relationship.

Supporters of the association model reject the social norms that tend to govern relations between spouses, because ‘marriage is an outdated institution’. By denying great importance to faithfulness, they create a new type of relationship. Saying that faithfulness is ‘quite important’ is the same as saying that it has ‘no importance’ whatsoever. If, as König (1994 [1974]) argues, divorce is a specular copy of marriage, and no justification is needed to terminate the relationship (divorce is always justified), it is because marriage is based on a weak and barely formalized agreement.

We should not expect those who support the association model, insofar as they are involved in determining the conditions of the association (Giddens 1991), to be in agreement. However, the one area in which they do agree is in seeing no obstacles to separation; 44% say that divorce is always justified (a score of 10, while the average is 8.4). That it is not even necessary for a lack of affect to exist reveals that no external party – society, for example –, can oppose or negatively sanction the breakup of the marriage, not even the spouse who has not taken the initiative to end the relationship can oppose it. Not only is it unnecessary to justify divorce, it can be a unilateral decision.

The data in Table 5 is the result of combining the three ideal types of marriage (alliance, fusion and association) with its functionality: is marriage necessary or contingent. From this analysis we are able to draw four noteworthy conclusions: First, that the attribution of additional functions to the marital relationship is independent of its nature; secondly, that necessity, the obligation that individuals feel to marry, protects against the breakup of the relationship and makes adultery more tolerable; third, that the combined reading of these two dimensions permits us to understand the specific utility individuals' attribute to marriage, and fourth, by describing the subtypes at the extremes of the spectrum we are better able to understand the ‘institutional’ model and the ‘pure relationship’.

Both the belief that marriage or a long-term relationship is necessary to be happy, as well as the belief that it is not, are independent of how individuals define their ideal of marriage. Within each ideal type we find a significant proportion of individuals with both views regarding the necessity of a long-term coupling. Only in the alliance model do we find that there is a certain correlation between the conservative definition of marriage and postulating its necessity (65% of respondents).

Hoping to satisfy their affective needs through marriage, some individuals are in a position of dependence in which they must be lenient with adulterous behaviour and accepting of circumstances they do not like in order to preserve the relationship. If we look at the averages for responses regarding the acceptance of adultery and divorce, we see that those who believe that a stable relationship is necessary are more tolerant of adultery and more likely to oppose divorce.

The specific function attributed to marriage depends on its definition. Among the supporters of alliance marriage, its function is materialistic, as marriage affects individuals' material well-being: it is related to obtaining resources, security and status. Among supporters of the fusion model, and resulting from the great importance they attribute to the subjective aspects of the relationship, what they seek is what the survey question explicitly asks about: happiness. The ‘necessity’ some supporters of the association model refer to is surely related to their search for identity (Allan 2001) or personal fulfilment. Today, people marry less for the social benefits they can obtain and more for the personal success it represents. Finding a partner and marrying is not easy; it represents an effort and doing so is considered an achievement (Cherlin 1981: 857).

The Institutional marriage model (Subtype 1) (Alliance-necessary) is supported by over one in four Europeans (27.8%) and is based, among other things, on the belief in the necessity and indissolubility of marriage. These ideas are vehemently defended by their believers. Their answers are maximalist because they allow no room for further moral reflections. Thus, 33% claim that divorce is never justified irrespective of the circumstances (1 on the scale of 1–10). The comparison of the averages for the responses on the questions regarding whether adultery or divorce can be justified for the two Alliance subtypes (S1-necessary and S2-contingent) provides us with interesting insight into the circumstances that might justify a divorce. The average place on the scale regarding whether divorce can be justified among subtype S2 respondents (average: 5.5) suggests that such an act cannot be approved or disapproved a priori, rather there must be specific reasons which can be evaluated and on which moral judgment can be passed. One possible justification is almost certainly sexual infidelity, of which there is an almost unanimous disapproval (average: 1.1). In contrast, respondents of the S1 subtype impose greater obstacles to divorce (average: 3.2) and are slightly more tolerant of adultery (average: 1.4). Both circumstances lead us to believe that one of the few scenarios in which divorce would be justified would be not mere infidelity, but infidelity that transcended the intimacy of the relationship and became public.

The ‘Pure relationship’ (S6) (Association-contingent) is the ideal for 8.4% of Europeans. Supporters of this type of relationship have something in common with the supporters of the S5 subtype; although they differ in the attributes they assign to the relationship, both believe that marriage is an outdated institution. A ‘pure relationship’ is a personal relationship between two individuals (faithfulness is very important: 88%), totally contingent (marriage is necessary: 4%) and revocable at any time (divorce is justified: average 9.0). However, those who see the marital relationship as contributing to the satisfaction of certain needs (S5) are more flexible toward faithfulness (only 17% judge it to be very important) and they impose more obstacles to the dissolution of the marriage (divorce is justified average: 7.8).

The percentages in Table 6 indicate the prevalence of the different marriage ideals in the countries analyzed. The data have been organized from the highest to lowest percentage of adherents to the Alliance model. Malta, the last EU country to pass legislation allowing divorce, is in first place on the list as 79% of Maltese are fervent supporters of traditional marriage. Sweden appears at the bottom of the list because there are the fewest supporters of the Alliance model (only 20%).

Three Catholic countries – Malta, Poland and Italy – top the list of proponents of ‘marriage for life’. They are followed by states from the centre and east of the European Union, among others, Hungary, Slovakia and Latvia. The middle of the list is occupied by Luxembourg, France, Belgium, Germany, Spain, Austria and Slovenia, countries in which the Fusion model has more supporters than the Alliance model. In Holland, Denmark, Finland and Sweden the Alliance type has little support – ranging from 31% in Holland to 20% in Sweden – and a significant segment of the population in each of these countries supports the Association type – from 21% in Denmark to 34% in Finland.

The overall percentage of those who believe that marriage is necessary – shown in the first column – varies from 77% among Greeks to 26% among the Swedish. As expected, the association between the definition of the marriage relationship and the attribution of functionality is low. If a correlation existed between them, the order of countries based on the percentage who believe in traditional marriage would be similar to the order of countries based on the percentage who believe marriage is necessary. However, this percentage is distributed in a random manner.

The populations of the Baltic States – Latvia, Lithuania and Estonia – and certain Central European states – Poland, Hungary and Slovakia – appear to hold similar traditional views as found in the southern states of Malta, Italy and Greece. Thornton and Philipov, in their study on the transformation of marriage in Central and Eastern Europe suggest the following explanation for this: For different reasons, the family values prevalent in these seven Eastern and Central European countries, are more similar to those of Western Europe than to those held by other states in their socio-political sphere (Thornton and Philipov 2009). They argue that these countries remained on the fringe of Soviet culture. In contrast to the marked opposition to Western culture that was prevalent in the USSR, the Baltic States were more moderate and more tolerant of Western cultural contagion. In addition, Poland, Hungary and Czechoslovakia were markedly more independent from Soviet influence than other countries behind the Iron Curtain. The singularity of Polish culture also stems from the important, although weakened, influence of the Catholic Church. If we examine the apparent similarity between these countries closely we find that, in general terms, central Europeans, in contrast to their southern counterparts, combine strong support of the necessity of marriage with a more lax definition of the contours of the institution itself. Drawing on Ronald Inglehart's (1997) post materialist theory, we can distinguish two sub-models. The first, represented by Malta, Poland, Italy and Greece, is strongly imbued with the traditional values of authority, religious in this case. The social values of the second sub-model, represented by Slovakia, Latvia and Estonia, have a materialistic character because they developed in a context of economic scarcity. While some defend ‘traditional family values’ because they are deeply rooted in the cultural and religious traditions of their country, others do so because of the benefits obtained in satisfying individuals' material well-being. This explains the almost unanimous support in these countries for the idea that ‘children need a home with both a mother and father to grow up happily’ and that ‘women need to have children to be fulfilled’.

In certain of Europe's most Western countries, such as Luxembourg, Belgium, Germany, Spain, Austria, France, as well as Slovenia, the most widespread ideal type is the model that represents the pure primary relationship in the strictest sense, the Fusion marriage.

The northern states – Holland, Denmark, Finland or Sweden – are at the bottom of the list because the dominance of the Fusion marriage is accompanied by significantly higher support of the Association model (e.g., Sweden: 33%) to the detriment of the Alliance model (Sweden: 20%). At the extreme, we find the Swedish case, with the highest proportion of defenders of the ‘pure-relationship’ model (22%, almost triple the percentage for Europe, 8.4%). In addition, in contrast to the other northern states, the proportion of Swedish who are supporters of the contingent version of the Association model (Association-contingent: 22%) doubles the proportion that support the necessity of the marital bond (Association-necessary: 11%). This is the most liberal model with its roots found in ‘[the] early démontage of patriarchal family legislation, which was a common Scandinavian initiative for the second and third decades of the twentieth century, … [and] firmly established an individualistic and egalitarian conception of the family’ (Therborn 2004: 196).

The particular combination of marital values in the UK (dominance of the alliance model and widespread belief in the contingency of marriage) and France (dominance of the fusion model and widespread belief in the necessity of marriage) contribute to the understanding of different relational logics. A slightly higher percentage of the British population desires a more traditional relationship than the overall population of the EU (UK: 45% vs. EU24: 42.4%), but this is a personal goal, not conditioned by perceived necessity (UK: 64% contingent vs. EU24: 44.2% contingent). The strong individualism in the UK is expressed in the concept of marriage as an agreement between two independent, free and equal individuals. In the absence of ‘because motives’, motives that arise from a state of necessity and that inevitably push us to do something, the British reveal themselves to be pulled by ‘in-order-to motives’ to construct their own ideals of conjugal relations.

France, which shares northern Europe's low level of adherence to the alliance model (33%) and a high level of preference for the association one (26%), differs, however, from those countries in that almost two-thirds of the population defends the necessity of marriage, much like in Italy, Poland and Hungary. In addition, among the proponents of the Association model, the proportion who believes marriage to be necessary is greater than the proportion who believes it to be contingent. A total of 62% believe that a stable relationship is necessary to achieve happiness, but only 22% (ISSP 2002) think that married people are happier than those who are not. This suggests that the French believe marriage satisfies certain needs but not necessarily that of achieving happiness. Marital relationships could be useful because they contribute to the construction of individual identity (Allan 2001), strengthen individuals' feelings of personal fulfilment (Cherlin 2004) and mitigate feelings of loneliness while the relationship lasts.

In this study we have identified different marriage ideals and calculated the level of adherence to each of them among Europe's population in the first decade of the twenty-first century. We have chosen two criteria for the construction of these ideal types: how the nature of the conjugal relationship is defined and, adding a second criterion to Roussel's original typology, whether functionality is attributed to it or not. The results are shown in Table 7, which synthesizes the conclusions of this study.

We cannot say that a specific European ideal of marriage exists because there is no single model of marriage that is clearly dominant. After analyzing the predominant values regarding marriage among the population of 24 EU member states we were surprised to find strong support for Institutional marriage, a model that many considered obsolete in the seventies (Roussel 1980); however, for 27.8% of Europeans, marriage is not only indissoluble, but should also fulfil certain functions related to material well-being. For those who support the ‘Contingent-Alliance’ model, the conjugal relationship is not established because it is necessary, but rather to attain happiness, which is identified with sexual fidelity. The only valid argument for divorce is the existence of an offense (infidelity, abandonment, etc.) and, therefore, a party responsible for the breakup.

In the first variant of the Fusion relationship, which is constituted for love (subtype S3), the relationship is also considered necessary for the spiritual well-being of the members of the family. As a consequence, along with the reasons for a breakup (essentially, that the relationship is bad), there are also reasons to avoid a rupture, such as ‘children need a home with both a father and mother to grow up happily’ or ‘married people are happier than unmarried people’. These beliefs act as a counterweight to the factors that might warrant termination of an unsatisfactory relationship. At the same time, 21.2% of Europeans idealize the model that defines marriage as the quintessential primary relationship: S4, ‘Contingent-Fusion’. If the marriage bond is grounded exclusively in love and there are no ulterior interests motivating the relationship, it would be perfectly acceptable for it to end if those involved cease to love each other or they are unhappy together. Legislation in most European countries on divorce explicitly states that it is enough for one of the partners to wish to terminate the relationship to obtain a divorce.

We have found two variants of Roussel's radical ideal type known as ‘Marriage Association’. Subtype S5, ‘Necessary-Association’ corresponds to his original. This ideal type refers to a weak and highly informal relationship, and is supported by individuals who desire to satisfy certain post-materialist needs through an intimate relationship, such as living a certain lifestyle, not being alone, or feeling a sense of accomplishment. But this model, which is relatively prevalent in French society, is markedly different from the model we have termed the ‘Pure relationship’ using Giddens’ concept. For 8.4% of Europeans, marriage is a relationship that can be reduced to its most essential element, a bond between two individuals, and has cast off the complementary functions it had acquired in the past, such as solidarity or mutual support (Bellah et al.1985; Pahl 2000), becoming totally contingent because it has no overt function in itself, neither for its members nor for society. A breakup can occur unexpectedly, no justification is required and neither is there a balancing of pros and cons, one of the members deciding they do not wish to perpetuate the relationship being sufficient. And given that it is unnecessary to think twice about it, some of Europe's most advanced legislation affords the possibility of integrating separation and divorce into a single legal act.

This study, designed to identify and describe how Europeans define the ideal marriage, has shown that different representations and ideals exist. This diversity reveals a complex panorama: we are as likely to hear contemporary versions of what we thought were out-dated traditional ideals, as we are to hear opinions idealizing novel forms of relationships. This complexity is most extreme because there is no agreement on what an ideal marriage is; not even on fidelity as important to the success of a marriage. This diversity and complexity may eventually lead to the emergence of contradictions and conflicts. However, this study has only verified that European society is fragmented when it comes to defining the ideal marital relationship: For some it is a relationship subject to norms and obligations; for others it is a primary relationship performing certain functions in society; yet for others, it represents a relationship through which individuals seek personal fulfilment, or it may even be a ‘pure relationship’ that implies no commitment.

This work was supported by Spanish Labour Ministry, Women's Institute [grant number 033/07], and by Spanish Ministry of Economy and Competitiveness [grant number CSO2012J35032].

Mercedes Camarero is Lecturer of Sociology at the Pablo de Olavide University. Her research interests are in the areas of family, social values, social capital, social quality and the developing of composite indicators. Currently, as a member of the research project entitled ‘Social Quality in Europe. Design and Development of Composite Indexes for the Measurement and Monitoring of the Quality of European Societies’, she is constructing an index of Family Social Capital.