These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Argentina, Chile and Uruguay share several characteristics in terms of the historical composition of their population and the demographic and social trends that they have followed. The three countries also share social and cultural patterns that differentiate them from the rest of the region. These countries were not political or economic empires before the Spanish conquest, as were Mexico and Peru; instead, they were largely uninhabited territories that were progressively populated as the Spanish Crown expanded. The three countries have experienced a deep process of mestizaje since Colonial times, as did the rest of Latin America, but they were more ethnically homogeneous in terms of larger shares of Europeans (Frankema 2008) and smaller shares of Africans, who arrived as enslaved workers. The indigenous population did not have the salience that it had in other Latin American countries, especially in Argentina and Uruguay (Pellegrino 2010). In Chile, the native population had more importance historically, particularly regarding the reluctance of the mapuche people (the main native group) to surrender, first to the Spanish Crown and then to the Chilean government, but the group was confined to specific areas in the country’s south. As a result, the proportion of the indigenous population is currently small in all three countries as measured by self-identification (4–5 % in Chile and 2 % and 4 % in Argentina and Uruguay, respectively, according to census data from 2002, 2010 and 2011).

At the end of the nineteenth century, this region received important contingents of European migrants, mainly from Italy and Spain. The influx of European immigrants was not as large in Chile, but it existed and was encouraged by the government as a way to populate the country’s southern region. Immigration significantly influenced the cultural patterns and demographic characteristics of these countries. Argentina and Uruguay are well known as pioneers of the demographic transition in Latin America (Pantelides 2006), where the fertility decline followed the European path; in Chile, the fertility decline began in the mid-1960s, similar to the rest of Latin America (Chackiel and Schkolnik 1992). By the middle of the twenteith century, the total fertility rate in the three countries was three children per woman, which was half of the value of the sub-continent.

The early development of welfare states in the region also contributed to the introduction of modern behaviours. Argentina and Uruguay organized their welfare states at the beginning of the twenteith century, whereas Chile did so in the 1920s. In terms of education, the three countries experienced early expansions of their educational systems as the welfare state developed. The gross rates of enrolment in primary education were relatively high at the beginning of the twenteith century compared with other Latin American countries (except for Costa Rica, which also had relatively high rates) (Frankema 2008). Laws that established compulsory primary education were enacted in 1877 in Uruguay, 1884 in Argentina, and 1920 in Chile. Women had early access and similar rates of education as men since the beginning and during most of the twenteith century, which was similar to the situation in the US and the most advanced European economies. Gender equality was especially clear at the primary level, but there were comparatively low levels of gender inequality concerning access to secondary and tertiary education (Frankema 2008). Over the course of the twenteith century, the educational system expanded, similar to the rest of Latin America. Around 2010, of the population aged 25 years and older, approximately 40 % in Argentina, 52 % in Chile and 42 % in Uruguay had completed at least a secondary education (12 years of schooling or more).

The early creation of social security systems that covered the population in the formal sector of the economy, including retirement benefits, may be related to the low proportion of extended and composite households in the Southern Cone compared with the rest of Latin America (Arriagada 2002; García and Rojas 2002). In the three countries, nuclear households that include only one family are currently the rule, as 80 % or more of the population live in this type of household. The proportion of people who live in extended-family households has decreased sharply in the last two decades (Ullmann et al. 2014).

Despite these similarities concerning population composition and the development of the welfare state, there are differences in the countries that may shape the fertility and family formation patterns that they follow. Uruguay showed the earliest and highest level of secularization because divorce has been possible since 1907 (Caetano and Geymonat 1997). Although the State-Church division occurred at the end of the nineteenth century in the three countries, in practice, the influence of the Church continued to be important in public matters in Argentina and Chile (Torrado 2003). In these countries, divorce laws were approved late in the twenteith century in the case of Argentina (1987) and in the first decade of this century in Chile (2004).

Uruguay is on the cutting edge in terms of legal changes and recognition of the demands of civil society, which likely reflects its high rate of secularization and the diminishing power of the Church. Of the three countries, Uruguay is the only country where abortion is legal in all circumstances since the approval of a new law in 2012.

In recent years, the three countries have made some progress regarding the recognition of legal rights for consensual unions. Towards the end of the twenteith century (1985 in Argentina, 1998 in Chile and 2004 in Uruguay), changes in the ley de filiacion ended the privileges of children born within marriages, which blurred any differences in the rights of children who were born within and outside marriage in terms of inheritance and alimony. Additionally, the three countries introduced different legal measures to recognize informal unions in the first decade of this century, and same-sex marriages were legally recognized in Argentina and Uruguay (in 2010 and 2013, respectively).

These changes imply a recognition of diversity concerning individual and sexual identities, which contributes to greater tolerance and individual autonomy. In this vein, it is reasonable to consider an ideational change according to the postulates of the Second Demographic Transition (SDT).

2 Historical Trends in Cohabitation in the Southern Cone

The Southern Cone has historically had low levels of cohabitation compared with the rest of Latin America. The three countries appear at the bottom of the ranking by Quilodrán (2003) regarding the prevalence of informal unions in Latin America. This ranking is based on census data from 1960 to 2000, and the rates in the three countries are lower than 20 %.

Historical studies suggest that informal unions were not necessarily rare in the Southern Cone, but their overall prevalence was lower than the rest of the region. The social recognition and acceptance of these types of unions were also low. These studies typically indicate a prevalence of cohabitation that is higher in rural areas and among the poor (Pellegrino 1997; Barrán and Nahum 1979; Schkolnik and Pantelides 1974; Moreno 1997; Ciccerchia 1989, 1994)

Cohabiting unions have historically had great importance in Latin America, especially in Central America and the Caribbean, where they have coexisted with marriage as types of unions (Quilodrán 2003; De Vos 1998; Castro-Martin 2002). The existence of these two types of unions has created a “dual nuptiality system” in Latin America, where socioeconomic status, not individual preference, decides who marries and who cohabits. Although both types of unions were recognized as families and accepted as settings for childbearing and childrearing, they differed in social legitimacy and in the legal rights that they offered to women and children (Castro-Martin 2002).

In the Southern Cone, cohabiting unions were historically a minority practice. Some historical reports indicate that cohabiting unions may have been an important type of union at the beginning of the twenteith century. However, urbanization, modernization, and the actions of the incipient welfare state promoted the formalization of unions; therefore, marriage became the main type of union (Pellegrino 1997). Thus, marriage used to be the norm for union formation in the Southern Cone. The crude marriage rate in the three countries has followed a relatively erratic but overall increasing pattern during the first half of the twenteith century and peaked in Chile in 1930 (9‰). The crude marriage rates peaked in Argentina and Uruguay in the 1950s and reached approximately 7.5‰ and 8.5‰, respectively. These values were among the highest in the region (for instance, the crude marriage rate for Venezuela in 1970 was approximately 3.6‰). The decline in the marriage rate started slightly earlier in Argentina and Uruguay than in Chile, but the differences are small; the three countries converged towards similar rates at the beginning of the twenty-first century (Binstock and Cabella 2011). From 1970 forward, there was a clear decrease in the crude marriage rate in the Southern Cone, and it reached approximately 3.5‰ at the beginning of the twenty-first century in the three countries.

Simultaneously, the vital statistics for the three countries show an increase in the proportion of children who were born outside of marriage. This percentage fluctuated approximately 20–25 % during the 1970s, but it reached 68 % in Chile in 2010 (Salinas 2014), 50 % in Argentina in 2001 (the Office of Vital Statistics stopped gathering information regarding the marital status of mothers in that year) and 78 % in Uruguay in 2012.

Neither the decrease in the crude marriage rate nor the overall modest delay in union formation seems to reflect an open rejection of conjugal unions. These factors also do not seem to be related to significant changes in individual preferences concerning the timing of a co-residential union. On the contrary, these dynamics seem to reflect a change in the type of union that people choose to form rather than a change in the timing of union formation. Most couples choose cohabitation, not marriage, as the first type of union that they form. There is ample evidence that this choice is the case in Argentina and Uruguay (Binstock 2004, 2013; Cabella et al. 2005), and there is incipient evidence of this choice in Chile (Salinas forthcoming; Ramm 2013).

In recent decades, cohabiting unions have continuously increased. The first signs of the increase in cohabitation appeared in the mid-1970s in Uruguay and Argentina and in the 1990s in Chile. Compared with the rest of Latin America, the Southern Cone showed the greatest increases in cohabitation between 1970 and 2000. These increases were most noticeable among the most educated groups (Quilodrán 2011).

At the end of the 1980s, approximately 10 % of all unions were informal in Argentina and Uruguay. This proportion doubled in the next decade, and it doubled again in the decade after that. In Chile, the trends are similar, but the increase in cohabitation started in the 1990s. In approximately 2010, nearly half of Argentinian and Uruguayan women who were aged 20–44 and lived in a union were cohabiting instead of married, and the corresponding percentage in Chile was 40 % (Binstock and Cabella 2011).

Discussions regarding the reasons for the increase in cohabitation in the Southern Cone began in the mid-1990s, and scholars offered different arguments.

In Uruguay, two prominent sociologists, Ruben Kaztman and Carlos Filgueira, argued that the increase in cohabitation related to social disintegration and was a response to a male identity crisis. Changes in the labour market, including the worsening of employment opportunities for men and increases in female labour force participation, led men to question their ability to provide for their families. Men may have answered these challenges by avoiding stable or more committed relationships such as marriage (Kaztman 1992; Kaztman and Filgueira 2001; Filgueira 1996) (This interpretation was extended to the rest of Latin America by Kaztman in ¿Por qué los hombres son tan irresponsables? (Kaztman 1992). A minor proportion of the increase in cohabitation could be attributed to what scholars called “modern cohabitation”, that is, cohabitation among young, educated people, which are similar to European cohabitation traits. However, generally, the family changes that appeared during the 1990s (i.e., increases in divorce or union dissolution, increases in the proportion of children born outside of marriage, etc.) were interpreted in this perspective as a result of social malaise and manifestations of the inability of the family to fulfil its functions (Rodriguez 2004).

From another perspective, these family changes were interpreted as the emergence of new forms of unions that were a response to the deinstitutionalization of formal relationships. In this view, cultural or ideational changes were more important to explain the increase in cohabitation. This explanation is consistent with the postulates of the SDT. However, it has always been recognized that the SDT’s theoretical apparatus will not likely fit perfectly in societies that have still not solved the problem of material needs and must address these needs simultaneously as they begin to face higher order needs (Cabella et al. 2005; Salinas 2011; Ramm 2013)

At the end of the 2000s, the controversy between social disintegration and SDT as explanations for the increase in cohabitation became diluted. This dilution can probably be explained by the lack of appropriate data that link union formation patterns and ideational change. This dilution may also be because the trends that the labour market and the economy generally followed were not consistent with the theory of social disintegration. The increase in cohabitation was stable from year to year, which the data from household surveys show, and was independent of the economic and labour market conditions. Between 1990 and 2010, the Southern Cone countries experienced different economic cycles, including downturns, severe crises, recoveries, and sustained growth. These fluctuations are especially true for Argentina and Uruguay, whereas Chile experienced downturns and upturns of comparatively smaller magnitude and showed more economic stability. The decreasing trend in the crude marriage rate was unaffected by these changes, and cohabitation continued to increase in the years of economic crisis, in the years of economic growth and in periods of high unemployment and high employment (Esteve et al. 2012; Cabella 2009).

One of the main variables that marked the differences in the types of cohabiting unions in the Southern Cone is the timing of family formation (that is, the timing for starting co-residential unions and fertility). Although young people of all socioeconomic strata adopt cohabitation as their first type of co-residential union, they begin it at different times. These differences among socioeconomic groups have increased over time (Binstock 2010; Cabella 2009). The gap in the age of union formation or childbearing has increased because more vulnerable socioeconomic groups (with the lowest educational attainment) do not change the timing of union formation and childbearing between censuses, whereas more affluent groups (the most educated) postpone the age of union formation and their first births.

Observed in perspective, the explosive increase in cohabitation that registered between 1990 and 2000 again became a subject of discussion several years later. The spread of cohabitation as the mechanism for entering into conjugal unions in all social strata and as a universal practice among youths pulled the arguments towards cultural or ideational explanations (Cabella 2009; Peri 2004). From this cultural perspective, the increase in cohabitation is assumed to be related to the diffusion of new ideas concerning the relationships between men and women. However, cohabitation is also presumed to have different meanings for different social groups because different types of informal unions coexist, and the trajectories that different cohabiting unions are a part of may differ.

3 Census and Survey Analysis

3.1 Data and Analytical Strategy

For the empirical analysis, we use census data that were retrieved from IPUMSi for the census rounds of 1970, 1980, 1990, 2000 and 2010. Not all the variables from the 2010 census are available for Argentina; therefore, we complement the census data for that year with data from the Permanent Household Survey (2010) and the National Survey of Sexual and Reproductive Health (ESSR), which was conducted in 2013. The Permanent Household Survey is representative of the population who lives in large urban areas (70 % of the Argentinean population). The ESSR is representative of women aged between 14 and 49 years and men aged between 14 and 59 years who live in urban areas (of more than 2000 inhabitants).Footnote 1 The 2012 Chilean census suffered serious problems of implementation and coverage; thus, the government discarded it. Therefore, we use data from the 2011 Encuesta de Caracterización Económica Nacional (CASEN), which is the largest official household survey in Chile. The CASEN is representative at the national and regional levels for both rural and urban areas.

We restrict the sample to women aged 20–29 years, which are usually considered the principal years for union formation and childbearing. In the first section, we examine the general trends and how they differ by women´s educational attainment over the study period. We use educational attainment as a proxy of socioeconomic status to compare the most advantaged (postsecondary studies) with the most disadvantaged women. In the first censuses, the most disadvantaged group comprised women with primary education, but because access to education has expanded, women with incomplete secondary education or less more properly represent this group. Consequently, we conducted a preliminary analysis that distinguished two alternate groups as the most disadvantaged in educational achievement (women who have completed primary education or less and incomplete secondary education or less), and we obtained similar substantive conclusions. Thus, to simplify the presentation, the tables include the results that compare women who have an incomplete secondary education or less with women who have higher education (which includes tertiary and university).

In the second section, we restrict the analysis to married and cohabiting women to examine them across three aspects, namely, childbearing, labour market participation, and household arrangements. Childbearing distinguishes women who are mothers from women who are not. Labour market participation differentiates women in the labour force (including employed or unemployed) from women who are outside the labour force. Household arrangement is a dichotomous variable that has the value “nuclear” if the married or cohabiting woman is the head or partner of the head of household compared with “not nuclear”, which includes all other arrangements. Our motivation is to identify the extent to which young couples can form and manage an independent household or whether they co-reside with other relatives and/or non-relatives.

The analysis compares married and cohabiting women across these three dimensions to assess whether any differences, if they exist, are increasing or diminishing over time as cohabitation becomes more common. The analysis also controls for educational attainment to identify patterns according to socioeconomic status.

3.2 Results

3.2.1 Family Formation: When and How Do Women Start Conjugal Unions?

Figure 9.1 shows the proportion of women who are in a conjugal union (married or cohabiting) in each age group. The data suggest a slight delay of union formation in the three countries, particularly since the 1990s. The delay is similar in Argentina and Uruguay between 1970 and 2010 and reaches approximately 10 percentage points in the 20–24 and 25–29 age intervals. The delay is more marked in Chile, where the decline of women who are in a conjugal union reaches 20 percentage points in both age groups. As a result, the proportions of women who are in a conjugal union are currently lower in Chile than in Argentina and Uruguay, which have similar values.

Fig. 9.1
figure 1

Proportion of women aged 20–29 years in a conjugal union, 1970–2010 (Source: Authors’ tabulations based on census samples from IPUMS-International, except Chile 2011 which are based on Encuesta de Caracterización Económica Nacional (CASEN))

This general pattern of union postponement hides marked differences based on women´s educational attainment. As expected, Fig. 9.2 shows that in the youngest age interval (20–24), the proportion of women who are in a conjugal union is higher among the least educated than among the most educated in every census round. This result reflects the fact that many of the most educated women are more inclined to delay union formation. Among the highly educated women in Chile and Uruguay, the postponement of union formation between 1970 and 2010 is continuous and distinct. In the 20–24 age interval, the proportion in any type of union declines by approximately half between 1970 and 2011 and goes from 27 to 15 % in Uruguay and from 21 to 9 % in Chile. The trends in the 25–29 age interval are similar. Argentina, in contrast, shows a relatively stable pattern until 2010, when there is a noticeable postponement among the most educated women in both age groups. However, given that the information for that year is based on a complementary (and not fully comparable) data source, these results should be viewed cautiously.

Fig. 9.2
figure 2

Proportion of women aged 20–29 years in a conjugal union by education, 1970–2010 (Source: Authors’ tabulations based on census samples from IPUMS-International, except Chile 2011 which are based on Encuesta de Caracterización Económica Nacional (CASEN))

In contrast, the least educated group of women shows a relatively stable yet somewhat erratic timing of union formation. By the 2010s (the last available data period), there is a decline in the proportion of women who are in a conjugal union in both age groups in all three countries, but the decline is much smaller compared with the most educated group of women. That is, the least educated women in the Southern Cone changed the propensity and timing of their union formation very little. It is necessary to continue to monitor the timing of entry into conjugal unions to determine whether this trend continues.

3.2.2 The Evolution of Cohabitation

Figure 9.3 shows the well-known increase in cohabitation. Among women in a conjugal union, the proportion of cohabiting women was very low in the 1970 census round. There were virtually no differences according to age in the proportion of cohabiting women in Chile, whereas in Argentina and Uruguay, the youngest group had a relatively higher proportion of cohabiters. The increase in cohabitation is remarkable and nearly doubles between 1980 and 1990; although in the 1990 census round, the proportion of women in cohabiting unions was still a minority (a growing minority but still a minority). The largest increase in cohabitation is observed between 1990 and 2000, when it becomes the relationship where most women start their co-residential unions.

Fig. 9.3
figure 3

Share of cohabitation as a proportion of women who are in a conjugal union (Source: Authors’ tabulations based on census samples from IPUMS-International, except Chile 2011 which are based on Encuesta de Caracterización Económica Nacional (CASEN))

A pronounced increase in cohabitation occurred by the 1990s and continues to today. In the 2010s, the proportion of cohabiting women in the 20–29 age group generally doubled from the number that was observed in the previous census. Currently, cohabitation has become the norm for young people: between 77 % and 85 % of women who are 20–24 years old and are in conjugal unions are cohabiting. Cohabitation is still very high in the next age interval, with values that vary from 71 % in Uruguay to 57 % in Chile.

3.2.3 The Shift in Cohabitation by Educational Attainment

The novelty in this period is that the growth in cohabitation is more striking in the group of the most educated young women than in the group of the least educated young women.

As observed in Fig. 9.4, considering that the overall level of cohabitation was low, cohabitation in the 1970s was a type of union that a proportion of the least educated young women engaged in (in the 20–24 age interval), although this proportion was small. Conversely, among the most educated young women, cohabitation was practically non-existent (approximately 1–4 %).

Fig. 9.4
figure 4

Share of cohabitation by education, aged 20–29 years, 1970–2010 (Source: Authors’ tabulations based on census samples from IPUMS-International, except Chile 2011 which are based on Encuesta de Caracterización Económica Nacional (CASEN))

Clearly, the most significant change among the least educated women is the increase in the preference to cohabit as opposed to marry. Cohabiters represented between 10 % and 20 % of women between the ages of 20 and 24 years and between 8 % and 18 % of women aged 25–29 years in 1980. By 2010, these figures increased at extremely rapid rates and reached between 80 % and 86 % for women aged 20–24 years and 64 % and 73 % percent for women aged 25–29 years. These percentages closely mirror the percentages that were previously observed for all women, which indicates the influence of the least educated women in driving these trends.

The prevalence of cohabitation among highly educated women was extremely low until the 1990 census round. Between then and the next data collection, the increase was remarkable and approached the levels of their less educated peers. In fact, the most recent available data show similar patterns of cohabitation among women aged 20–24 years across educational groups. Additionally, the differences in conjugal preferences among women in other age groups have been declining.

The postponement of union formation is not a shared feature among all young women in the Southern Cone, but the election of cohabitation as the first type of conjugal union that they engage in is a shared feature of young women of all educational statuses. This result is not surprising. The cohabitation boom (Esteve et al. 2012) exists because nearly all members of certain cohorts choose this type of union.

3.2.4 Differences and Similarities Between Married and Cohabiting Women

The rationale, meaning and motivation to cohabit – as opposed to marry – has been a topic of intense and continuous debate in Latin America, particularly in the Southern Cone, where unmarried cohabitation was not previously a prevalent or common feature of the family system (Binstock and Cabella 2011; Quilodrán 2001; Rodríguez 2004; Filgueira and Peri 1993).

In this section, we move from the focus of examining the expansion in the incidence and preference to cohabit to the study of the similarities and differences between the dynamics of cohabitation and marriage in three specific dimensions: childbearing, labour market participation, and household arrangements. Again, we further control for educational attainment to assess whether cohabitation and marriage have different implications for women in different social strata. Given that cohabitation in the 1970s was extremely low (particularly among women with higher education), we begin the analysis in 1980.

3.2.5 Childbearing

The first panel of Table 9.1 shows that childbearing has been common among married and cohabiting women in each of the three countries, particularly until the 1990s. Afterwards, the pattern seems to have reversed (with the exception of Argentina), and childbearing becomes more common among married than cohabiting women.

Table 9.1 Women in conjugal unions aged 20–29 years

When we consider women´s educational attainment, we observe two contrasting trends. Women with low levels of education are mothers in a similar proportion whether they are married or cohabiting. The proportions of mothers in each conjugal group are high and remain stable across the observed period, particularly in Argentina and Chile. In Uruguay, the frequency of mothers among married women is slightly higher than among cohabiters, and this difference has somewhat increased over time in both age groups of 20–24 and 25–29 years. In Uruguay, compared with Argentina and Chile, childbearing seems to be more suitable in marriage among young, low-educated women.

The childbearing patterns among highly educated married and cohabiting women are very different. In general, and with only several specific exceptions, childbearing is more frequent among married women than cohabiters, which is consistent with the idea that marriage is still considered the more appropriate context to raise children. However, the trends are changing in a specific manner in each country.

In Argentina, the difference between cohabiting and married women’s childbearing behaviour has declined in both age groups of 20–24 and 25–29 years, which suggests a change in people’s conceptions of the two types of unions as an appropriate context for childbearing. In fact, this trend is consistent with the dramatic increase in births outside of marriage that mainly occur in cohabiting relationships. This result is also consistent with a lower and slower tendency for cohabiting couples to marry after the birth of a child.

In Chile, however, the childbearing differences between married and cohabiting women are also decreasing but only in the youngest age group, whereas among women aged 25–29, the pattern is more erratic. Among highly educated Chilean women, the youngest group differs from their married peers in terms of having and raising children within cohabitation, whereas in the older group, this tendency is less clear. Currently, the data from the next census is needed to evaluate the extent to which this pattern has continued or changed.

The situation among highly educated women in Uruguay shows a different yet interesting pattern. The decrease in the proportion of mothers was dramatic among both cohabiting and married women, as is the gap between the behaviours of these two conjugal groups. That is, the ratio of the proportions of highly educated cohabiting mothers and highly educated married mothers aged 20–24 declined from 1.2 to 0.5 between 1996 and 2011. The comparable proportion among these women aged 25–29 decreased from 0.97 to 0.57. The estimated ratios in 2011 are similar to the estimated ratios from 2001. Cohabiting and married educated women in 1985 exhibited a more similar reproductive profile than their reproductive profiles in the next two censuses. In the context of a general decline in the proportion of mothers among educated women, the reduction was significantly higher among cohabiters than among married women. A plausible explanation for this result is that younger highly educated cohabiting women may be transitioning to marriage as a response to motherhood more often than older highly educated cohabiting women.

3.2.6 Labour Force Participation

One of the most common explanations for the increase in cohabitation, particularly in European and highly developed countries, depends on people who behave based on values that are more oriented towards individualism and higher-order needs, as stated in the SDT schema. In this scenario, varying gender dynamics are expected based on the type of union in which people live. Consensual unions tend to be more egalitarian. Marriage is often a scenario for a more traditional division of gender roles in the family, where men are the main (or only) economic provider. In addition, if people choose cohabitation because it is a less restrictive type of union, it is likely that cohabiting women will be more inclined to work so that they can afford to live independently if the union dissolves. Therefore, cohabiting women should have higher rates of labour force participation than their married peers. An alternative scenario is that cohabitation is chosen because of the socioeconomic restrictions on marriage (Kaztman 1997). If this is the case, it is likely that cohabiting women will be less likely to work than their married peers.

The study period has witnessed increasing rates of female labour force participation that are independent of age, education and conjugal status (CEPAL 2014). Additionally, highly educated women consistently exhibit higher participation rates than their lower educated peers, which is not surprising given their better occupational opportunities and labour conditions.

The comparison of labour rates shows that by 1980, cohabiting women had somewhat lower rates of labour force participation than their married peers. This difference decreased as the years passed. The differences levelled off and even changed sign at the turn of the twenty-first century. By 2010, cohabiters generally showed higher rates of labour market participation. The differences are not very large, but the pattern is similar across countries and ages.

When we consider women´s educational attainment and we focus on the least and most educated groups, we find similar trends across both groups. Cohabiters have somewhat higher rates of labour force participation in Argentina and Chile, regardless of their age and educational level. In Uruguay, the pattern is more erratic, and cohabiters have slightly lower levels of participation than married women in the first two censuses. By 1996, the differences tended to either level off or revert, with more cohabiting than married women in the labour force, which continued to 2010 (see Table 9.2).

Table 9.2 Women in conjugal unions aged 20–29 years

3.2.7 Household Arrangements

One dimension that is frequently cited to account for the increase in cohabitation involves economic downturns or circumstances that lead young couples to postpone or avoid marriage. We lack the appropriate data to test this hypothesis for the Southern Cone, but it seems unlikely that this drastic and sustained increase across social groups in all three countries across such a long period is only or mainly a response to economic circumstances.

The eldest cohorts in the Southern Cone tend to be homeowners who do not depend on their children to live, which is not necessarily the case in the rest of Latin America. This result is also fuelled by the fact that pension systems in the Southern Cone achieved a high level of coverage very early compared with the rest of the region (Rofman and Oliveri 2011). Instead of promoting the incorporation of their children’s new families into the parental household, the eldest cohorts support the youngest cohorts in the establishment of their own (rented or owned) dwellings. This neo-local norm is highly accepted by the population (“el casado casa quiere”). Certain groups of the population, however, still depend on their relatives to solve their housing needs, which conforms to extended households that allow them to take advantage of economies of scale. This type of family arrangement is more common during economic downturns.

One consequence of good economic circumstances is the ability to fulfil a strong and long-established cultural preference for nuclear living arrangements. In addition, or alternatively, if cohabitation and marriage are considered essentially similar unions regarding commitment and expectations (i.e., reproduction, family organization, ownership, etc.), we would expect similar household organizational arrangements for both types of unions.

Cohabiting and married women live in nuclear arrangements in similar proportions in Argentina and Uruguay, which is a pattern that remained stable across the study period. In contrast, Chilean cohabiters lived in nuclear arrangements more often during the first three censuses. This difference levelled off by 2000 and reverted by 2010, when married women more often lived independently.

When we separately examine the household arrangements of women from different social sectors, we observe that low-educated women closely replicate the overall trend for all women. That is, the proportion of women who live in nuclear arrangements is similar among cohabiting and married women in Argentina and Uruguay. In Chile, the trend moves from nuclear arrangements that are somewhat more common among cohabiters to nuclear arrangements that are more common among married women (see Table 9.2).

The situation among more educated women is different. In Chile and Uruguay, it is more common for cohabiters to live in nuclear arrangements, whereas in Argentina, there are no differences, or these differences are restricted to the youngest group.

A tentative explanation for this finding considers that Chile has the highest incidence of extended arrangements in the Southern Cone, which correlates with a greater emphasis on more long-term, established Catholic family values. Accordingly, the first cohabiters, particularly the cohabiters with higher education, faced greater family resistance and opposition to co-residence as an unmarried couple. Alternatively, these cohabiters may have been more ready to confront the social norms that they did not share, such as extended household arrangements (accompanied by the economic ability to create an independent nuclear residence), and they may have placed greater value on couple intimacy (Table 9.3).

Table 9.3 Women in conjugal unions aged 20–29 years

4 Discussion

The objective of this chapter was to describe the changes in family formation in the Southern Cone by focusing on the spread of cohabitation and determining the differences and similarities between marriage and cohabitation. The objective was also to determine if the differences between these arrangements are increasing or decreasing and whether it is possible to identify groups of women in which either the old or new behaviours prevail. In general, the three countries clearly share patterns regarding forming unions and having children. Although there are nuances among them, it makes sense to distinguish this region as a whole.

There has been a change in the timing of union formation, and women show signs of delaying the age when they initiate their conjugal history. This change, however, has mainly occurred among highly educated women. Among the least educated group, conjugal union formation still occurs relatively early in life. In the future, the postponement of union formation may be expected to spread to groups with less socioeconomic resources as education expands.

There has also been a change in the modality of forming unions, and this change affects both the most and the least educated women. Cohabitation is becoming the typical way that women start their unions. Thus, entering directly into marriage is becoming more infrequent in the region.

Regarding childbearing, the proportion of women who have children has decreased but mainly among the highly educated. Most of the least educated women become mothers before they reach 25 years of age, whether they are married or cohabiting. Among the most educated women, there seems to be an increasing tendency to bear and rear children within cohabitation rather than within marriage in Argentina and Chile. In Uruguay, it seems that the most educated women are turning to marriage regarding childbearing and childrearing. These tendencies are recent and should be re-evaluated with more recent data, but with the results that were discussed above concerning the timing and modality of union formation, we can distinguish old and new behaviours among the least and most educated women. In the group with fewer socioeconomic resources, cohabitation starts early and is accompanied by early childbearing. For women with greater socioeconomic resources, cohabitation begins later, and fewer women have children.

With the increase in female labour force participation in the Southern Cone in recent years, the basic pattern is that the most educated women are more likely to be in the labour force than the least educated women. However, married women in the past were more likely to work than cohabiting women, whereas in recent years, this difference has levelled off or even reversed. Because this new pattern is recent and the difference in favour of cohabiters is small, we again need new data to determine whether this pattern is actually a trend. However, this pattern is another feature that may depict the emergence of more egalitarian behaviours in cohabitation, this time across groups with different socioeconomic statuses.

Our results concerning household arrangements are surprising. Among the least educated women, the tendency to live in a nuclear household is similar for both cohabiting and married women. Among the most educated and young women, in contrast, we observe a higher tendency to live in nuclear arrangements of cohabitation than marriage in Uruguay and Chile, whereas in Argentina, there are no major differences. Once the income that is required to afford independent living is met, we suggest that in the group of young women, cohabiters have a higher preference for independent living because it represents a setting where they face less questioning of their lifestyle (i.e., living with a partner and eventually having children without being married) by older relatives. This explanation makes more sense in Chile than in Uruguay because the conservative sector seems to wield more weight in Chilean society. Moreover, the household arrangement has not received much attention when examining marriage and cohabitation in the Southern Cone. What we know regarding families and household arrangements in Latin America is generally based on data in Central America and the Caribbean that were produced some years ago (De Vos 1987 and 1995). Our results are somewhat contradictory to the image that emerges from these studies, where extended arrangements appear to be characteristic of the region, especially among groups with few socioeconomic resources. More work should be conducted in this area to determine whether young and better-off cohabiters have a higher preference for independent living than their married peers and what such a preference implies.

Overall, we verify the expansion of cohabitation across socioeconomic statuses in the Southern Cone. However, when comparing cohabitation and marriage, our data suggest that married and cohabiting women in the lowest socioeconomic strata are more alike than better-off married and cohabiting women. Thus, cohabitation may be equivalent to marriage in the most deprived sectors of the population.

The tension between “modern” and “traditional” explanations of the increase in cohabitation has been present throughout the last two decades in Latin America. In the Southern Cone, and likely in the rest of the continent, it seems highly unlikely that we are witnessing a “traditionalization” of consensual unions. However, we probably cannot say that our societies are undergoing a “modernization” of consensual unions. Considering the strong social differences in the timetable of transitions in union formation and childbearing, we should focus on the interpretation of the social polarization of demographic behaviours.