Cohabitation in Spain – a rocket of a trend

Cohabitation in Spain: No Longer a Marginal Path
to Family Formation

by MARTA DOMINGUEZ-FOLGUERAS  – Universitat Pompeu Fabra
TERESA CASTRO-MARTIN  – Center for Human and Social Sciences

 Journal of Marriage and Family 75 (April 2013): 422 – 437
DOI:10.1111/jomf.12013

Background comment: For many generations Spain was the epitome of Catholic Europe with all the trappings such a status implies. It was the model of ‘family stability’ with divorce impossible and its marriages rates (along with Italy’s) leading all of Europe. Cohabitation was almost unknown and certainly not encouraged – as was illegitimacy. However, since the death of Franco, Spain has been gripped with a radicalism that must have been ‘pent-up’ during his reign. All the once stable indicators that marked out Spain from the rest of Europe have evaporated. Whenever there is a vacuum an alternative gas is always ready to rush in, and low fertility within the EU is now that vacuum all governments face. The following paper by Dominguez-Folgueras and Castro-Martin traces the developments within Spain since the 1980’s. During this epoch ‘radical feminism’ has increased at a time when active catholisim has been falling while underlying these events have been large scale surges in unemployment. Are these linked ? – RW

 

• Although many indicators reflect the marked retreat from marriage occurring in Spain since the 1980s, the diffusion of cohabitation has been slow. The confluence of very low and late fertility, latest-late marriage, and low cohabitation has been largely regarded as defying the predictions of the second demographic transition and has fueled a debate over the distinctiveness of the Mediterranean model of family formation. Comparative analyses based on the Family and Fertility Survey documented the marginal role of cohabitation in Spain and in the rest of southern European countries by the mid-1990s. In this research, the authors used more recent data from the 2006 Spanish Fertility, Family and Values Survey (N=5,750) to reveal that cohabitation has spread significantly among younger cohorts and hence can no longer be considered as playing a marginal role in the family formation process.

The process of family formation has undergone profound changes in all Western societies in recent decades (Billari, 2005). One of the key transformations has been the declining significance of marriage in family life and the spread of non-marital cohabitation, a process that is still unfolding (Bumpass & Lu, 2000; Smock, 2000).

Marriage, which was once part of the natural progression into adulthood, has lost much of its centrality in structuring young adult lives and has been gradually replaced by cohabitation, at
least as the initial stage of family formation (Seltzer, 2000; Smock & Manning, 2004). Yet the pace of incorporation of cohabitation into the family life course has not been uniform across
countries (Heuveline & Timberlake, 2004; Kiernan, 2001). Although the second demographic transition (SDT) theory initially expected Western countries to converge in their patterns of family formation (Lesthaeghe, 1995; van de Kaa, 1987), cohabitation has defied the predictions of rapid convergence. Several decades after its emergence as a social phenomenon, the role and meaning of cohabitation in the family system continue to diverge greatly across socioeconomic, institutional, and cultural settings.

Southern European countries have been frequently portrayed as a paradox (Dalla Zuanna
& Micheli, 2004) or as being at odds with the predictions of the SDT (Coleman, 2004) because
of their position as forerunners in lowest-low fertility in the early 1990s but laggards in cohabitation. Southern European countries were latecomers to the global process of family
change depicted by the SDT (Domınguez, Castro-Martın, & Mencarini, 2007), but once the
process was set in motion, the pace of change was generally faster, and its intensity greater,
than in other European societies (Lesthaeghe, 2010). Spain and Italy were the first countries to reach lowest-low fertility levels (total fertility rate [TFR] of 1.3 in 1991), and postponement of marriage and childbearing has been more pronounced than in western and northern Europe (Frejka & Sobotka, 2008). Nevertheless, the diffusion of cohabitation remained slow, at least until the mid-1990s.
The broad cross-national variation in the prevalence and nature of cohabiting unions was the basis for Heuveline and Timberlake’s (2004) classification of countries according to the role of cohabitation in the family formation process. In this classification, the role of cohabitation in Spain was labeled ‘‘marginal,’’ implying that institutional and cultural sanctions confined unmarried cohabitation to a relatively small minority of couples. Nevertheless, this description was based on the 1995 Fertility and Family Survey (FFS), which has become dated as a portrayal of current family patterns. As an illustration of this point, the percentage of non-marital births in Spain – as a share of total births – has more than tripled from 1995 to 2010, from 11% to 35.5%.
Our objective in this article is to reexamine the level of diffusion of cohabitation in Spain, taking advantage of a recent national survey, the 2006 Fertility, Family and Values Survey. The case of Spain presents a valuable opportunity to assess whether the diffusion of cohabitation is a universal feature of family change associated with the SDT or whether local contexts shape the pace and nature of family transformations.

The Spanish case could also provide important insights into the generality of patterns and correlates of cohabitation observed in low-fertility societies where this type of union emerged earlier. The structure of the  article is as follows: First, we review prior studies, in particular those referring to the Mediterranean pattern of family formation, and discuss secularization and attitudinal change in Spain. We then present a number of indicators that reflect recent changes over time in the characteristics of cohabiting unions, including changes in duration and dissolution patterns.
Next, we measure cohort changes in type of first union entry with life table estimates, and use event history models to examine its main correlates. Previous research based on the 1995 FFS showed that forerunners of cohabitation in Spain were a selective group of the population characterized by high educational attainment, high employment rates, and low religiosity (Dom´ınguez-Folgueras & Castro-Mart´ın, 2008; Meil Landwerlin, 2003).
Because there are multiple signs that the diffusion of cohabitation has accelerated in the past decade, we assess whether women opting for cohabitation comprise nowadays a less self-selective group (Manting, 1996). Finally, we discuss the findings and the foreseeable
evolution of cohabiting unions in Spain.

BACKGROUND

The decline and postponement of marriage and childbearing did not become apparent in Spain until the 1980s, a decade later than in northern Europe (Trost, 1978), but these processes became particularly intense in the 1990s (Munoz-Perez & Recano-Valverde, 2011). Hence, at the end of the 20th century Spain ranked lowest in fertility level and highest in age at marriage and age at first birth within the European context. Yet the decline in marriage was not  compensated by a parallel increase in cohabitation, as had been the norm in most European countries. Consequently, the percentage of Spanish women age 20 to 34 who had not yet entered their first union was among the highest in Europe in the 2000s census round: 62.2% (Castro-Mart´ın, Domınguez-Folgueras, &Martın-Garcıa, 2008).
Several family scholars have hypothesized that Western societies are going through a transition
in the way couples are formed (Kiernan, 2002; Prinz, 1995; Sobotka & Toulemon, 2008).

This partnership transition, encompassed in the SDT (second demographic transition), would have several stages. First, cohabitation emerges as a statistically rare behavior with a highly selective socio-demographic profile. Later on in the diffusion process, cohabitation is adopted by people from the various social strata, and it functions mainly as a prelude to marriage or a trial period to test the quality of the relationship, but childbearing remains confined to marriage. Gradually, cohabitation starts to supplant marriage: It lasts longer and becomes an acceptable context for parenthood. Finally, cohabitation and marriage become indistinguishable, which concludes the transition process.
All prior comparative studies on cohabitation concurred in positioning Spain in the lowest prevalence category – from a cross-sectional approach (Heuveline & Timberlake 2004)—or at the initial stage of the partnership transition—from a developmental perspective (Tob´ıo, 2001). Nevertheless, it must be noted that assigning one particular category or diffusion stage to a country is always problematic, because cohabitation may have more than one meaning at the same time in a given society.

In the United States, for example, it has been shown that cohabitation may play different
roles according to social class and racial and ethnic background (Choi & Seltzer, 2009;
Manning & Landale, 1996). Also, even in countries where the social meaning of cohabitation is
relatively homogeneous, this may change quite rapidly over time (Manting, 1996). In order to
avoid static interpretations of dynamic phenomena, it is essential to count on updated data
on union formation. A large amount of the literature on cohabitation in Europe was based
on analyses of the FFSs, which were carried out in the mid-1990s, but more recent data
sources, such as the Gender and Generations Surveys or the European Social Survey, have
shown that the meaning and patterns of cohabitation have changed considerably since then
(Hiekel, Liefbroer, & Poortman, 2012; Kasearu & Kutsar, 2011; Perelli-Harris et al., 2012).
Moreover, recent studies have revealed remarkable increases in cohabitation in societies long
regarded as having low prospects for diffusion of unmarried partnerships, such as Italy
(Castiglioni & Dalla Zuanna, 2009; Gabrielli & Hoem, 2010; Rosina & Fraboni, 2004), Poland
(Matysiak, 2009), or Japan (Raymo, Iwasawa, & Bumpass, 2009).
The ‘‘Mediterranean’’ Pattern of Family Formation and Attitudinal Change The Mediterranean, or southern European, pattern of family formation has been typically characterized by prolonged co-residence with parents; late transition to a conjugal union; predominance of marriage among first unions; and high synchronization among leaving the parental home, union formation, and first birth (Baizan,Aassve, & Billari, 2003; Billari, Castiglioni, Castro-Martın, Michielin & Ongaro, 2002).

The Mediterranean pattern is also highly responsible for the shift in the macro-level relationships between union dynamics and total fertility: Contrary to what was happening about 20 years ago, fertility is currently higher in countries with a larger share of cohabitation, non- marital births, and union disruption (Billari & Kohler, 2004).

Because early views of the SDT assumed that the decline in fertility would go hand in hand with the pluralization of family forms (van de Kaa, 1987), the co-existence of lowest-low fertility, latest-late-marriage, and relatively low incidence of cohabitation has been typically regarded as a challenge to the SDT premise of a common transition process encompassing both reproductive and partnership behavior.

Several explanations have been offered to account for the relative stability in partnership dynamics in southern Europe in the midst of rapid reproductive change. Economic factors, such as high youth unemployment (Ahn & Mira, 2001), increasing uncertainty linked to unstable job positions (Simo, Castro-Martın, & Soro, 2005), and tight housing markets (Holdsworth & Irazoqui Solda, 2002), have often been emphasized as barriers to union formation, regardless of union type. The lack of policies directed at youth, which reinforce their dependency on the family (Ferrera, 1996), and inconsistent gender relations in the public and private spheres (Esping-Andersen, 2009) have also been claimed to deter union formation. Another key element in the  Mediterranean family system is the strength of inter-generational ties (Dalla Zuanna, 2000; Reher, 1998), which lies behind the prolonged permanence of young adults in the parental home (Moreno Mınguez, 2003).
There is also the widespread idea that the lack of diffusion of cohabitation in southern Europe can be attributed to cultural values rooted in its Catholic inheritance. The Catholic Church certainly had a strong influence on Spanish society in the past, particularly during the Franco dictatorship (1936 – 1975), when it largely shaped politics, legislation, and the education system, endorsing a traditional conception of the family, asymmetric gender relations, and a strict sexual code for women (Nash, 1991). But as the process of democratization and socio-economic modernization unfolded, Spanish society experienced an intense secularization process (Requena, 2005).

A recent survey focusing on religiosity revealed that only 28% of the population over age 18 identified themselves as practicing Catholics, and this percentage dropped to 10% among those below age 40 (Centro de Investigaciones Sociologicas [CIS], 2008). Although the Catholic Church retains an important presence in the education system and in the provision of certain social services, its political, social, and cultural influence on society has clearly waned. The fact that   nowadays Spain is among the countries with more liberal legislation on abortion, assisted  reproduction, same-sex marriage, and adoption by same-sex couples confirms that the Catholic Church has lost its traditional power in shaping family-related legislation. The reduction in the percentage of newborns baptized (from 85% in 1990 to 64% in 2009) and the sharp decline in religious marriages (from 76% in 2000 to 42% in 2010) in favor of exclusively civil marriages are further manifestations of the Catholic Church’s declining role in legitimizing family transitions.

Despite broad social transformations and profound changes in values and lifestyles in Spanish society, the Catholic Church has maintained, unchanged, its traditional doctrine on marriage and the family, standing against premarital sex, contraceptive use, divorce, and abortion. As a consequence of its perceived detachment from the current reality of family life, it has lost its once-prominent role as provider of moral guidelines for sex and family matters. The process described by Laplante (2006) to explain the high cohabitation rates in Catholic Quebec is likely to hold also for Spain. According to Laplante, once a large number of Catholics reach the conclusion that Catholic morality has become obsolete and impracticable, they treat it as irrelevant.

Data from various opinion surveys attest to the declining significance of religious precepts in people’s lives: About 80% of individuals over age 18 declare that they ‘‘never’’ or ‘‘hardly ever’’ follow the Catholic Church’s recommendations regarding sexuality,marriage, and partnerships (CIS, 2004), and about 70% explicitly disapprove of the Church’s position on divorce and contraception (CIS, 2008). Hence, the Catholic view of cohabitation as ‘‘living in sin’’ probably has a negligible impact on young adults’ attitudes toward cohabitation and on their choice of living arrangements.

A loosening of social norms against family formation outside marriage is evident from responses to various surveys. For instance, in the 2003 International Social Survey Programme’s Survey on Family and Changing Gender Roles, 74% of Spanish respondents agreed with the statement ‘‘It is right for a couple to live together without intending to get married,’’ and 54% disagreed with the statement ‘‘People who want children ought to get married.’’ The corresponding figures for those younger than 35 were 89% and 75% respectively, and there were no significant gender differentials (authors’ calculations).

Acceptance of non-marital cohabitation—with and without children—is therefore nearly universal among Spanish male and female young adults, although this does not entail a devaluation of marriage, which continues to be a highly valued institution. From a comparative perspective, recent data from the 2008 European Values Survey confirm that the value profile of Spaniards concerning partnership and family issues is quite similar to that found in northern European countries.

For instance, the percentage of individuals agreeing with the statement ‘‘It is all right to live together without getting married’’ was 84% in Spain and 88% in Norway. The level of agreement with this attitudinal item was also high in Portugal (83%), but less so in Greece (70%) and Italy (54%). Yet the widespread social acceptance of cohabitation as a legitimate path to family formation had not translated into behavior in Spain; at least, it had not by themid-1990s.

This implies that ideational factors are a necessary but not sufficient condition to bring about a drastic transformation in partnership dynamics (Surkyn & Lesthaeghe, 2004). Nonetheless, the pervasive social tolerance toward nontraditional family forms signaled the readiness of Spanish society for an eventual rise in cohabitation (Liefbroer & Fokkema, 2008).

Some Indications That Cohabitation Might Be Taking Off

The significance and centrality of marriage for family formation have continued to erode during the first decade of the 21st century. The total first marriage rate – the sum of age-specific first marriage rates – dropped from 1 in 2000 to 0.63 in 2010, reflecting a marked trend toward marriage postponement and, possibly, a decline in the proportion of persons ever marrying during their lifetime. Concurrently, in the period 2000 to 2010, the mean age at first marriage increased from 28.1 to 31.1 among women and from 30.1 to 33.2 among men (Spanish Institute of Statistics; http://www.ine.es).

Recent studies have also shown that the probability of marital dissolution, traditionally low, has increased substantially among recent marriage cohorts (Bernardi & Mart´ınez-Pastor, 2011). Rising marital disruption is likely to have an impact on the prevalence of cohabitation because many divorced persons who repartner choose to cohabit rather than remarry (de Graaf & Kalmijn, 2003; Wu & Schimmele, 2005). Nevertheless, the weight of postmarital cohabitation on overall cohabitation has decreased over time in Spain.

According to census data, the proportion of cohabiting women who had been previously married declined from 48% in 1981 to 26% in 2001 (Garcıa Pereiro, Carella, & Pace, 2012).

Nuptiality studies based on the 1995 Spanish FFS underscored the low prevalence of cohabitation: Only 3.4% of women between ages 18 and 49 were living in a consensual union at the time of the survey, and 7.8% indicated that they had ever lived in a consensual union. Nonetheless, there were significant birth cohort differentials that suggested that cohabitation could be taking off among young cohorts (Domınguez-Folgueras & Castro-Martın, 2008).

Prior studies based on the 1995 FFS also found that the choice of cohabitation as first partnership was more common among urban, highly educated, and working young women (Meil Landwerlin, 2003). In general, individuals with greater material, social, and cultural resources at their disposal to make nontraditional choices regarding lifestyles acted as forerunners in the adoption of cohabitation.

The 2001 census, conducted 6 years after the FFS, revealed some changes in the spread of cohabitation. The percentage of unmarried women age 15 to 49 who were cohabiting with a partner at the time of the census was merely 4.3%. When the analysis was restricted to women currently in a union, however, important differences emerged across age groups: Nearly one third—32.8%—of women age 15 to 24 and 11.6% of those age 25 to 34 were in a cohabiting partnership compared to 4.6% of those age 35 to 49 (Castro-Martın & Domınguez-Folgueras, 2008). Although the transitional nature of most cohabitations and their limited duration makes it difficult to interpret these cross-sectional differences or to infer underlying trends, these data suggest that, despite low diffusion, cohabitation was losing its traditionally marginal position among the younger segments of the population. Furthermore, the 2001 census revealed an unexpected feature of cohabiting unions: Almost half – 46.3% – of all cohabiting couple households included minor co-resident children, and 39.3% of cohabiting couples had their own children.

Although this level is well below that for married couples, 80% of whom had children, it brings into question the prevailing assumption that cohabitation is predominantly a childless stage leading to marriage. The most persuasive indication that cohabitation can no longer be  considered a marginal path to family formation in Spain comes from the recent evolution of non-marital fertility. The percentage of births to unwed mothers rose significantly, from 3.9% in 1980 to 11.1% in 1995, but the increase afterward has been particularly sharp: from 11.1% in 1995 to 35.5% in 2010 (Spanish Institute of Statistics; http://www.ine.es).

This dramatic rise in non-marital fertility was largely driven by the growing number of births to cohabiting adults, a pattern also observed in many other countries (Raley, 2001). Spanish vital statistics bureaus started collecting data on cohabiting status for unmarried mothers in 2007,  and these register data confirm the strong connection between non-marital childbearing and cohabitation. In 2009, nearly two thirds of all non-marital births (62.8%) were born to a cohabiting couple, implying that childbearing in cohabiting families currently accounts for 21.7% of all births and 24.1% of all first births in Spain (Castro-Martın, 2010).

This  level is similar to that documented by Manlove, Ryan, Wildsmith, and Franzetta (2010) for the United States and suggests that cohabitation does not play a negligible role in current family dynamics.

METHOD

The Fertility, Family and Values Survey was carried out in 2006 by the CIS. The questionnaire replicated most of the earlier FFS questionnaire and included women’s partnership,  reproductive, and employment retrospective histories on a monthly time scale.

The sample was designed to be representative at the national and regional level, and it consisted of 9,737 women age 15 and over, who participated in a face-to-face interview. Because of  potential recall errors among older cohorts, we confine the analysis to women between ages 15 and 56 (i.e., birth cohorts from 1950 to 1991). After we had ‘‘cleaned’’ the data for inconsistencies in partnership biographies and applied the age selection, the analytical sample consisted of 5,750 women. Although the total number of women who experienced a transition to a first union  through cohabitation in the analytical sample (n = 1,140) is larger than in the FFS Cohabitation in Spain 427 (n = 319), it remains relatively small to perform separate analyses by birth cohort.

We first present some descriptive indicators of the characteristics of all cohabiting unions  – regardless of whether they are pre-marital or post-marital – in Spain. Then, to assess trends, we present life table estimates of the cumulative proportion of women in successive birth cohorts who opted for cohabitation or marriage as their first conjugal union.

The analysis is limited to the transition to first union because of the relatively small number of cases for second and higher order unions and because pre- and post-marital cohabitations often differ substantively from one another (Brown, 2000). Last, we apply discrete-time event history analysis to examine the socio-demographic correlates of distinct paths of first union formation. Cohabitation, marriage, and remaining single are modeled as competing outcomes, and  multi-nomial logit regression is used to estimate simultaneously the odds of these outcomes. This competing-risk approach allows us to identify factors whose effects work in similar or opposite directions in the transition to marriage and cohabitation.

Women are considered to enter the risk of forming a first union at the age of 14, and observations are included up to the month when they entered their first union or up to the month of the  interview, if no transition was made.

The multi-nomial logit models are based on person-months of exposure to the competing risks of marriage or cohabitation, a discrete time approach that facilitates the incorporation of  time-varying co-variates and that is analogous to continuous-time hazard regression (Allison, 1984). Robust standard errors were estimated to account for the non-independence of observations, and duration is controlled using dummy variables for each month.

The model takes the following functional form:

 

• Where Pijt is the conditional probability of experiencing either marriage or cohabitation (j = 1 for marriage, j = 2 for cohabitation, and j= 0 for no event occurring), for a single woman i at month t since her 14th birthday. The model includes m time-constant predictors and n time-varying covariates, described below.

Table 3 (below) presents the odds ratios derived from multinomial logit regression models predicting the conditional probability of entering a first union (regardless of union type) versus remaining single (column 1), entering marriage versus remaining single (column 2), entering cohabitation versus remaining single (column 3), and entering cohabitation as opposed to marriage (column 4). An odds ratio above 1 represents a positive effect, and a value below 1 indicates a negative effect on the transitions under study.

To assess trends in cohabitation and examine change over time in a multivariate framework, 10-year birth cohorts were included in the models.

Drawing on the marriage and cohabitation literature (Thornton, Axinn, & Xie, 2007), we also included a number of covariates, such as education, employment, and religiosity, documented to be particularly influential on the choice of union type. Educational attainment – a proxy for earnings potential, modern values, and higher demands for gender equality within  partnerships – is categorized into four levels: (a) primary, (b) lower secondary (which has been compulsory since 1990), (c) upper secondary, and (d) university. Although educational attainment can change over the life course, the survey analyzed does not contain educational biographies, so we use the highest educational level reported at the time of interview. Nevertheless, Spain has a pattern of very late union formation, and the vast majority of people have completed their studies by the time they enter a conjugal union (Coppola, 2004).

Religiosity is measured by religious adscription and practice as declared by the respondent and is grouped into four categories: (a) practicing Catholic, (b) nonpracticing Catholic, (c) other religious  ascription, and (d) not religious (if no religious ascription is acknowledged).

As an indirect measure of progressive attitudes, we include in the model information on  self-declared political ideology, measured using a 10-point scale, on which 1 – 4 is labeled as right, 5 – 6 as center, and 7 – 10 as left. Although this variable has not been commonly used in cohabitation research, prior research suggests that politically left-oriented individuals have less traditional attitudes toward family (Lye & Waldron, 1997). Both religiosity and political identification are measured only at the time of interview and, given that values and ideology are  responsive to life course experiences and family transitions, we cannot discard reverse causation.

Past research has shown that gender values and political beliefs tend to move toward  ‘‘traditionalization’’ along the life course (Davis, 2007); hence, our modelsmight underestimate the effect of secular and progressive attitudes on the choice of cohabitation.

Given the documented links between childhood living arrangements and subsequent union formation behavior (Teachman, 2003), having been raised in a nonconventional family is also controlled for in the analysis. We measured it with a dummy variable that denotes whether the respondent experienced parental separation before age 16. We also take into account whether the respondent had lived independently from the family of origin, either alone or with unrelated adults, during at least one year before coresiding with a partner. This is a relatively infrequent behavior in the Spanish context and has been shown to have a significant positive impact on the probability of choosing cohabitation over marriage as a first union (Dom´ınguez-Folgueras & Castro-Mart´ın, 2008).

Because of diverging union formation patterns by Spanish and immigrant women (Cortina, Bueno, & Castro-Mart´ın, 2010), nationality is also included in the models. In addition to this, we control for two specific time periods when Spain experienced very high rates of unemployment (over 20%), which may influence the timing or type of partnership formation: (a) the 1984 – 1986 crisis and (b) the 1993 – 1994 crisis, when unemployment reached 24%. Table A1, in the Appendix, presents the sample distribution for these variables.

We also include several time-varying co-variates in the analysis. Employment status, which has been shown to be positively associated with union formation in Spain (De la Rica & Iza, 2005), is  entered in the models as a dummy time-varying covariate coded 1 if the respondent was working in that particular month and 0 otherwise. The variable is not lagged in the models because we tried several specifications (1-, 3-, and 6- month lags) and results remained unaffected. Because of the strong interlinkages between union and childbearing behavior, data on whether a pregnancy or a birth occurred before union formation are also incorporated in the analysis as time-varying covariates.

The value of the pregnancy variable changes from childless (0) to pregnant (1) 8 months before the reported date of birth and then it changes back to 0. The birth variable takes the value of 1 one month after the first child is born. Because of common underreporting of terminated  pregnancies, we only identify conceptions that resulted in a live birth. The effect of these  variables should be interpreted with caution, because only 359 women (5.8%) were mothers before entering their first union and only 12.9% of first conceptions occurred before the formation of a co-residential partnership.

RESULTS

Several characteristics of cohabiting unions in Spain are described in Table 1. These indicators are calculated for 10-year birth cohorts, providing useful information on the changes that have occurred across generations. One can observe that cohabitation has gone from being rare to being a relatively common experience among recent cohorts. The percentage of women ever in a cohabiting union increased from 13.9% among those born in the 1950s to 37.1% among those born in the 1970s. The percentage was lower for the youngest cohort of women, who were 15 to 26 years old at the time of the survey, largely because of the late pattern of union formation in Spain. Cohort patterns are analogous when we consider current cohabitations, although the percentages are naturally lower from a cross-sectional perspective. Among current cohabiting unions, postmarital cohabitations outnumbered premarital cohabitations in the 1950 – 1959 birth cohort but represented a decreasing share of non-marital unions for successive cohorts. Although based on a small number of cases, the high proportion of postmarital cohabitations in  older cohorts points to the formerly married as pioneers in cohabitation in Spain.

The mean duration of cohabiting unions was around 5 years for women born in the 1950s and 1960s. This relatively long duration seems at odds with the low prevalence and marginal role of cohabitation in these cohorts, and it could be attributed to several factors.

On the one hand, women who cohabited in these cohorts were probably a highly selective group in terms of ideology and values, given the strong social norms favoring marriage that prevailed back then. On the other hand, more than half of cohabitations in those cohorts were preceded by a disrupted marriage, and postmarital cohabitation may be a long-term arrangement for women who had a negative experience with marriage, or even the only alternative if their marriages broke before 1981, when divorce became legal in Spain.

Table 1. Selected Characteristics of All Cohabiting Unions for Successive Birth Cohorts

 

The length of cohabitation declined for the following birth cohorts, although for the youngest one  this was partly due to limited exposure. Nevertheless, even for the 1970 – 1979 cohort, for which he observed prevalence of cohabitation was highest and the proportion of postmarital  cohabitation spells was low, the mean duration of cohabitating unions (nearly 4 years) was above what would be expected when cohabitation serves primarily as a prelude to marriage or a stage in the marriage process. In fact, the next column in Table 1 shows that the estimated proportion of cohabiting unions resulting in marriage within a period of 5 years for this cohort was only 38.4%.

The trends observed in the routes of exit from cohabitation suggest a decline in the likelihood of transition to marriage and a modest increase in the likelihood of separation across cohorts.

Life table estimates of the cumulative proportion of women who entered their first union through cohabitation or marriage for the birth cohorts under study are provided in Table 2. These estimates confirm the upward trend in cohabitation and the downward trend in direct marriage. By age 35, 38.8% of women born in the 1970s had entered their first conjugal union through  cohabitation, compared with 16.7% of women born in the 1960s and 6.4% of women born in the 1950s. Hence, the increase in cohabitation was particularly evident among women born in the 1970s, the majority of whom entered their first union after the turn of the 21st century. For this cohort, although marriage continued to be the main path to union formation, cohabitation was no longer a marginal living arrangement. The partnership trajectory of the youngest cohort we examined, that born in the 1980s, although incomplete, points toward a reinforcement of the upward trend in cohabitation. By age 25, nearly one third of women had already entered a cohabitating union. The bottom part of Table 2 also shows that there has been a general delay in partnership formation: By their 30th birthday, 86.9% of women born in the 1950s had entered their first union (whether marital or nonmarital), whereas only 77.5% of women born in the 1970s had doneNevertheless, by age 35, all cohorts show similar percentages of women having formed a conjugal union.

Table 3 presents the relative risks associated with selected covariates on the rate of transition to first union (regardless of union type), to first marriage, to first cohabitation and also the contrast of entering cohabitation versus entering marriage. Birth cohort effects suggest substantial changes in the patterns of union formation over time and confirm a clear upward trend in cohabitation. Whereas the likelihood of entering directly into marriage has declined markedly across cohorts, the likelihood of entering cohabitation has increased considerably. Compared with women born in the 1960s, those born in the 1970s were three times more likely to choose  cohabitation over marriage as their first union. The youngest women, those born in the 1980s, displayed very high odds of cohabiting instead of marrying (odds ratio = 17.2), but this strong effect may be partly due to selection, because the members of this cohort who had entered their first union by the survey date did so well before the average age of first partnership formation in Spain, so they may be a differentiated group.

Table 2. Life Table Estimates of the Cumulative Proportion of Women Entering Their First Union Through Cohabitation and Marriage for Successive Birth Cohorts

Despite the large increase in cohabitation, this was not large enough to compensate for the decline in marriage; hence, the overall rate of transition to first partnership, regardless of union type, was considerably lower among younger cohorts than older ones.

Women’s high educational attainment (a university degree) had a strong negative effect on the likelihood of entering a union and on the likelihood of entering marriage directly. This effect is consistent with Oppenheimer’s (1997) idea that marriage is losing centrality for women who can be economically independent. Educational attainment, however, had no significant effect on the likelihood of entering cohabitation instead of marriage. Although college-educated women were more prone to choose cohabitation over marriage in bivariate analyses (results not shown),  differentials among educational  groups were not statistically significant once other co-variates were introduced in the model.

This result differs from that obtained in previous studies based on the FFS, which documented that college education increased substantially the odds of entering cohabitation versus marriage (Domınguez-Folgueras & Castro-Martın, 2008). The change in the impact of education on entry into cohabitation suggests that, although highly educated women acted as forerunners in the adoption of cohabitation, this path to family formation has now spread to all educational strata.

Separate models of transition to first partnership for each of the birth cohorts under study (results not shown) confirmed that higher education increased the likelihood of cohabiting as opposed to marrying among women born before 1960, but not afterward; nevertheless, the interaction between cohort and educational attainment did not reach statistical significance (p = .09).

The effects of the rest of the co-variates on the choice of cohabitation versus marriage were  consistent with former evidence based on the FFS for Spain and other European countries. Those variables related to women’s ability to live independently were negatively associated with direct transition into marriage and positively associated with cohabitation.

Compared with women who were not employed, working women had a lower rate of direct marriage and increased likelihood of choosing cohabitation over marriage. The experience of living independently from the family of origin for at least one year was also strongly related to the choice of cohabitation over marriage as one’s first union.

Table 3. Odds Ratios From Multi-nomial Logistic Regression Models of Transition to First Marriage or First Cohabitation

Note: Ref. = reference category; NA= no answer. a Time-varying co-variate. ∗p<.05. ∗∗p<.01. ∗∗∗p<.001.

Given the prevailing late pattern of departure from the parental home, fueled by job instability, low salaries, and high housing costs (Jurado, 2001; MorenoMınguez, 2012), women who have lived independently from their family of origin can be considered a selected group with high aspirations of autonomy.

Consistent with the abundant literature showing that being brought up in a nontraditional family setting influences the type of first union chosen (Kiernan, 2001; Teachman, 2003), our results show that this is also the case in Spain.

Divorce became legal in Spain only in 1981, but ‘de facto’ separations were taking place before the divorce law. According to the results in Table 3, women who had experienced their parents’ separation were 2.6 more likely to choose cohabitation over marriage than women raised in a two-parent family.

With regard to the impact of pregnancy and motherhood on union formation, results are also consistent with former evidence from the FFS. Even though public opinion polls show very  tolerant attitudes toward families formed outside marriage in Spain, concerns about birth legitimation still influence decisions on union formation. Pregnancy increased considerably the risk of entering a co-residential union, in particular, marriage. The odds of pregnant women entering cohabitation instead of marriage were 64% lower than those of non-pregnant women.

Conversely, women who were already mothers had lower rates of marriage than childless women, a pattern also documented in other societies (Upchurch, Lillard, & Panis, 2001), and they were more likely to enter their first union through cohabitation rather than marriage.

The data do not allow us to establish paternal relationships, so we cannot ascertain whether mothers formed a coresidential union with the father of the child or with a different partner.

The results also pointed out important differentials in first union formation patterns by nationality. Immigrant women with no Spanish nationality were about twice more likely to enter cohabitation than Spanish women. This pattern is consistent with previous findings (Cortina, Esteve, & Domingo, 2008), but because the data do not contain migration histories, we do not know whether union formation took place in Spain or in the country of origin. Sample size limitations did not allow us to explore differentials among foreign women by region of origin, but the pattern observed could be possibly linked to the large share of immigrant women coming from Latin America and the high prevalence and widespread social recognition of consensual unions in that region (Castro-Martın, 2002; Cortina et al., 2010).

Although marriage to a Spanish citizen provides rapid access to Spanish nationality, there was no evidence in the data suggesting that women born abroad but with Spanish nationality at the time of the survey were more likely to have entered their first union through marriage.

Ideological factors were also found to be relevant for union formation. Women who declared themselves nonpracticing Catholics and those who did not identify themselves with any religious faith displayed higher probabilities of entering their first union through cohabitation, compared with practicing Catholics. The coefficient was not significant for women who had a religious faith other than Catholic. Regarding political orientation, the results suggest that women who situate themselves on the left of the political spectrum are more likely to choose cohabitation over marriage as their first union compared with women who position themselves in the center of the political scale. Nonetheless, differentials between politically right-oriented and center-oriented women regarding patterns of entry into cohabitation were not statistically significant.

As noted earlier, associations between the choice of union type and religious and political orientations must be interpreted with care, because the latter were measured at the time of the interview and may have changed over time.

Congruent with studies showing that economic recessions affect the dynamics of family formation (Sobotka, Skirbekk, & Philipov, 2011), our results confirm the importance of period conditions in affecting entry into union.

During the economic recession that took place in Spain in the early 1990s, which was  characterized by very high levels of unemployment (above 20%), the odds of entering both  marriage and cohabitation declined significantly. A previous recession in the mid-1980s showed a similar pattern but no statistically significant effects.

DISCUSSION

Family demographers have extensively documented that the significance and role of unmarried cohabitation within the family system is subject to rapid change over time (Seltzer, 2003). Provided that cohabitation represents a moving target in terms of social approval and societal diffusion, it is important to keep track of changing attitudes and behaviors through regularly updated indicators.

Spain, like other Mediterranean countries, has long been regarded as a country with a low  prevalence of unmarried cohabitation and low prospects of changing this feature. The confluence  of very low fertility, intense postponement of marriage and childbearing, and low diffusion of cohabitation also posed a challenge to the narrative of the SDT (second demographic transition), which underlined parallel and highly interconnected changes in the reproductive and partnership spheres. Being a country with a strong Catholic tradition, the low prevalence of cohabitation in spite of the increasing retreat from marriage was initially attributed to moral sanctions and social norms against cohabitation.

However, the rapid process of secularization and the marked shift in values concerning family  issues, documented through numerous attitudinal surveys, soon refuted the existence of religious or cultural barriers to the increase in cohabitation. In fact, the cultural and social preconditions for the diffusion of cohabitation were in place in the 1990s: low perceived legitimacy of religious institutions, recognition of individual autonomy in family decisions, sexual tolerance, high contraceptive use, and increasingly egalitarian attitudes about gender roles (Domınguez-Folgueras & Castro-Martın, 2008). Nonetheless, widespread social acceptance of nontraditional family forms had not translated into behavioral changes, and in 1995, when the FFS was fielded, cohabitation played only a marginal role in the family formation process.

Patterns of family formation outside the framework of marriage, which were in an emerging  stage in the mid-1990s, seem to have intensified since then. The dramatic rise in the non-marital fertility ratio (from 11.1% in 1995 to 35.5% in 2010), for instance, has been largely driven by the growing share of births to cohabiting couples. In this article, we have provided evidence that cohabitation has turned into an increasingly frequent partnership choice among young adults. Whereas the vast majority of women born in the 1950s followed the traditional pathway  prescribing a marital union followed  by motherhood, over one third of women born in the 1970s chose cohabitation as their first partnership. Multivariate results confirmed the marked increase in cohabitation among the younger cohorts, even after controlling for socio-demographic composition. Hence, although Spain has been a latecomer to the diffusion of cohabitation and unmarried partnerships emerged with a substantial time lag compared to other hallmarks of the SDT, cohabitation can no longer be considered a marginal path in the family formation process.

Furthermore, the fact that births to cohabiting couples currently account for one-fifth of all births suggests that cohabitation might be becoming an increasingly accepted context for childbearing and possibly child-rearing.

Concerning the socio-demographic factors that favor cohabitation over marriage, the results of our analysis suggest that education no longer plays a major differentiating role. Compared with prior studies based on the FFS, which portrayed college-educated women as forerunners of cohabitation, we found no statistically significant differentials by educational attainment. The waning effect of education may be interpreted  as an indicator of the diffusion of cohabitation across all social strata. Nevertheless, cohabitation is still selective of women with certain characteristics, namely, employed, secular, and politically left-oriented women. Additional selection traits are more related to personal experiences than to socioeconomic or ideological profiles: having experienced parental divorce, having lived independently for at least one year,  and having a pre-union child were found to be strongly associated with the likelihood of choosing cohabitation over marriage as first union.

Although the European ‘North – South’ divide in partnership dynamics has started to narrow, significant differences still remain, and it is too soon to anticipate whether cohabitation will eventually become the norm for first union formation and direct marriage the exception, as is already the case in many societies. There are several reasons to expect a further increase in the prevalence of cohabitation in Spain. The favorable attitudinal context, the powerful role of social networks and diffusion processes, and the natural generational replacement in the prime ages of union formation are likely to promote the spread of cohabitation.

On the other hand, the current economic crisis has put many plans for family formation on hold and has aggravated the difficulties faced by young adults to form a union. Unemployment  reached 24.4% in early 2012, twice the average for the European Union, and young people are among the groups hardest hit by the crisis—unemployment rates for the 16- to 24-year-old age group reached 52% in early 2012. In the present context of economic uncertainty, medium-term trends in cohabitation will be heavily conditioned by employment trends in the next years. One could argue that, because marriage is still associated in the social imaginarium with solid financial bases or prospects and cohabitation has less demanding prerequisites – in terms of home  ownership, savings and job stability—the latter might be better suited to youths’ circumstances,  at least as a temporary arrangement, and hence likely to increase in the future (Kalmijn, 2011;  Mills, Blossfeld, & Klijzing, 2005; Oppenheimer, 2003). Of course, it all depends on whether young adults respond to income uncertainty by remaining even longer in the parental home or venturing to emancipate and form a partnership even if this entails a decline in their standard of living with respect to the parental home.

Changes in the institutional and legal framework of cohabitation may also play a role in future developments (Perelli-Harris & Sanchez Gassen, 2012; Waaldijk, 2005). Recently, some steps have been made toward providing certain legal recognition to unmarried partnerships.

By the year 2000, most autonomous regions in Spain had established public registers for unmarried couples and had enacted legislation that extended some rights previously restricted to married couples to registered cohabitants, such as transfers of rental home tenancy or joint adoption. However, there is no national legislation regulating the rights and obligations of unmarried partnerships. Several laws have been enacted ad hoc, but they do not adopt a  consistent approach to the treatment of cohabiting couples. At present, cohabitants are treated  as married in some domains, such as health coverage or pension rights, although the  requirements to qualify for a pension are more demanding for cohabitants. In other domains, however, such as income tax or inheritance, cohabiting couples continue to be treated as unrelated persons rather than a family. In general, cohabitating couples remain more vulnerable, legally and financially, than their married counterparts in the case of a breakup or the death of one of the partners, but it is foreseeable that legislation will keep on adjusting to the demands of new family forms and that the gap in the legal protection of married and unmarried couples will continue to narrow. Increased institutional security for cohabitants may well contribute to further blur the boundaries between cohabitation and marriage.

NOTE

This research has been funded by the Spanish Ministry of Science and Innovation, Projects  CSO2009-11883 and CSO2010-17811.

 

Appendix  A1

REFERENCES

Ahn, N., & Mira, P. (2001). Job bust, baby bust: Evidence from Spain. Journal of Population Economics, 14, 505 – 521. doi:10.1007/s00148 0100093

Allison, P. D. (1984). Event history analysis: Regression for longitudinal event data. Sage University Paper Series on Quantitative Applications in the Social Sciences, No. 46. Thousand Oaks, CA: Sage.

Baizan, P., Aassve, A., & Billari, F. C. (2003). Cohabitation, marriage, and first birth: The
interrelationship of family formation events in Spain. European Journal of Population, 19, 147 – 169. doi:10.1023/A:1023343001627

Bernardi, F., & Martınez-Pastor, J. I. (2011). Divorce risk factors and their variation over time in Spain. Demographic Research, 24, 771 – 800. doi:10.4054/DemRes.2011.24.31

Billari, F. C. (2005). Choices, opportunities and constraints of partnership, childbearing and parenting: The patterns in the 1990s. In M. Macura, A. MacDonald, & W. Haug (Eds.), The new
demographic regime: Population challenges and policy responses (pp. 63 – 94). Geneva, Switzerland: United Nations.

Billari, F. C., Castiglioni, M., Castro-Martın, T., Michielin, F., & Ongaro, F. (2002). Household and union formation in a Mediterranean fashion: Italy and Spain. In E. Klijzing & M. Corijn (Eds.), Dynamics of fertility and partnership in Europe: Insights and lessons from comparative research (pp. 17 – 42). Geneva, Switzerland: United Nations.

Billari, F. C., & Kohler, H.-P. (2004). Patterns of low and lowest-low fertility in Europe.
Population Studies, 58, 161 – 176. doi:10.1080/ 0032472042000213695

Brown, S. L. (2000). Union transitions among cohabitors: The role of relationship assessments
and expectations. Journal of Marriage and the Family, 62, 833 – 846.

Bumpass, L., & Lu, H.-H. (2000). Trends in cohabitation and implications for children’s family contexts in the United States. Population Studies, 54, 29 – 41. doi:10.1080/713779060

Castiglioni, M., & Dalla Zuanna, G. (2009). Marital and reproductive behavior in Italy after
1995: Bridging the gap with Western Europe ? European Journal of Population, 25, 1 – 26.
doi:10.1007/s10680-008-9155-9

Castro-Martın, T. (2002). Consensual unions in Latin America: Persistence of a dual nuptiality
system. Journal of Comparative Family Studies, 33, 35 – 55.

Castro-Martın, T. (2010). Single motherhood and low birthweight in Spain: Narrowing social inequalities in health? Demographic Research, 22, 863 – 890. doi:10.4054/DemRes.2010.22.27

. . . / . . .

Seltzer, J. A. (2000). Families formed outside of marriage. Journal of Marriage and the Family,
62, 1247 – 1268. doi:10.1111/j.1741-3737.2000. 01247.x

Seltzer, J. A. (2003). Cohabitation and family change. In M. Coleman & L. H. Ganong (Eds.), Handbook of contemporary families: Considering the past, contemplating the future (pp. 57 – 78), Thousand Oaks, CA: Sage.

Simo, C., Castro-Martın, T., & Soro, A. (2005). The effects of the globalization process on the transition into adulthood in Spain. In H. P. Blossfeld, E. Klijzing, M. Mills, & K. Kurz (Eds.),
Globalization, uncertainty and youth in society (pp. 375 – 402). New York: Routledge.

Smock, P. J. (2000). Cohabitation in the United States: An appraisal of research themes, findings and implications. Annual Review of Sociology, 26, 1 – 20. doi:10.1146/annurev.soc.26.1.1

Smock, P. J., & Manning, W. D. (2004). Living together unmarried in the United States: Demographic perspectives and implications for family policy. Law & Policy, 26, 87 – 117.
doi:10.1111/j.0265-8240.2004.00164.x

Sobotka, T., Skirbekk, V., & Philipov, D. (2011). Economic recession and fertility in the developed world. Population and Development Review, 37, 267 – 306. doi:10.1111 / j.1728-4457.2011.00411.x

Sobotka, T., & Toulemon, L. (2008). Changing family and partnership behaviour: Common
trends and persistent diversity across Europe. Demographic Research, 19, 85 – 138. doi:10.4054/DemRes.2008.19.6

Surkyn, J., & Lesthaeghe, R. (2004). Value orientations and the second demographic transition
(SDT) in northern, western and southern Europe: An update. Demographic Research, 3, 43 – 86.
doi:10.4054/DemRes.2004.S3.3

Teachman, J. D. (2003). Childhood living arrangements and the formation of coresidential unions. Journal of Marriage and Family, 65, 507 – 524.

Thornton, A., Axinn, W. G., & Xie, Y. (2007). Marriage and cohabitation. Chicago: University
of Chicago Press.

Tobıo, C. (2001). Marriage, cohabitation and the residential independence of young people in Spain. International Journal of Law, Policy and the Family, 15, 68 – 87. doi:10.1093/lawfam/15.1.68

Trost, J. (1978). A renewed social institution: Non-marital cohabitation. Acta Sociologica, 21,
303 – 315.

Upchurch, D., Lillard, L., & Panis, C. (2001). The impact of non-marital childbearing on subsequent marital formation and dissolution. In L. L. Wu & B. Wolf (Eds.), Out of wedlock: Causes and consequences of nonmarital fertility (pp. 344 – 380). New York: Russell Sage Foundation.

Van de Kaa, D. J. (1987). Europe’s second demographic transition. Population Bulletin, 42, 1 – 59.

Waaldijk, K. (Ed.). (2005). More or less together: Levels of legal consequences of marriage, cohabitation and registered partnership for different-sex and same-sex partners. Documents de Travail de l’INED, 125.

Wu, Z., & Schimmele, C. M. (2005). Repartnering after first union disruption. Journal of Marriage and Family, 67, 27 – 36. doi:10.1111/j.0022- 2445.2005.00003.x

E N D