Dual career couples and their inclinations regarding work and family: the Brazilian scenario

This study aims to explore the database of the National Household Sample Survey (PNAD) produced by the IBGE (Brazilian Institute of Geography and Statistics) in order to investigate how couples from a dual-career perspective approach the demands of home and work. The study uses a quantitative-descriptive analysis of secondary data to update the model proposed by Berlato (2015), in which typologies were delineated from a sample of dual-career couples who were former students of a state university in São Paulo. From a sample of 890 individuals derived from the PNAD, it was possible to observe the predominance of traditional characteristics among Brazilian couples. Despite the suggestion of social change that comes with the dual-career phenomenon, there are still some features that accentuate gender disparities in the country. percentages indicate that the single biggest group (46.4%) is acrobats, while the proportions of familists (25.8%) and careerists (27.8%) are


INTRODUCTION
Within the broad discussion of the themes of work and family, it is clearly important and relevant to understand the main demands made on couples in these areas. The variables involved are changeable and need to be revised to confirm or refute impressions obtained in previous studies. Nowadays, it is necessary to take into account the continuous metamorphosis of family reconfigurations, and this implies a revision of both the "modus operandi" and the databases used by researchers. This study, despite its limitations, offers a novel characterization of dual-career couples and their approaches to work and family within a context that, because of the sampling methods used, may be taken as representative of the entire Brazilian population.
Because of the scarcity of research into dual careers in specific national contexts, it is interesting to approach this theme from the Brazilian perspective, since each country has particular characteristics that influence its conditions. It was with this aim that Berlato (2015) tried to draw profiles of Brazilian couples according to their specific circumstances, identifying five types characterized by their priorities in relation to the demands of work and family. However, Berlato's findings were based on a selective sample of individuals, which prevents further generalization and justifies the expansion of the scope of research. Further, a more comprehensive investigation of the types of couples that are representative in the country would make use of a more general database that includes variables related to the family and the dual career, such as the National Household Sample Survey (PNAD in Portuguese). The aim of the present study is therefore to use data from the National Household Sample Survey to expand the investigation of how couples from a dual-career perspective in Brazil approach the demands of home and work.
The use of a secondary database as a source of evidence related to the theme of work is not common, but neither is it new in the context of international research (ALLEN and FINKELSTEIN, 2014;MOTTE-BAUMVOL, BONIN and BELTON-CHEVALLIER, 2015;MYSÍKOVÁ, 2014;MINNOTTE, MINNOTTE and THOMPSON, 2016). However, in the national literature, there is no research using data of this type to investigate the theme of dual careers. In Brazil, research has been limited to surveys related to the composition of the family, and these have so far prioritized description over critical analysis (BARROS, ALVES and CAVENAGHI, 2008;CAMARANO, MELLO and KANSO, 2009;LEONE, MAIA and BALTAR, 2010).
The use of PNAD was crucial for this study because of a number of factors: its national validity (it contains information about all the Brazilian states); its composition (it covers significant variables regarding characteristics of work and family); its periodicity (in contrast to other national surveys, it is carried out annually); its sampling system; and the opportunities it provides for data manipulation according to the needs of the researcher. The PNAD is widely used in some fields, such as economics and the social sciences, but in the field of administration, and especially in studies concerning dual-career couples, it is less common. Nevertheless, some authors have made attempts to use it in investigations of the relationship of career to the family sphere (e.g., BRUSCHINI, 2007;BARROS, CARVALHO, FRANCO et al., 2009;POCHMANN and AMORIM, 2009;BONETTI and FONTOURA, 2009;MEDEIROS and SOUZA, 2016;REINDEL and PEREIRA, 2014). Thus, there is relevance in the proposed objective and the potential contributions of this survey to the academic community, particularly in terms of studies related to careers.

THEORETICAL FRAMEWORK
The aim of this section is to situate the reader in the dual-career phenomenon, clarifying its roots and identifying relevant typologies created by researchers investigating the choices and priorities of couples.

The dual career and some typologies
Until the advent of feminist theory, and more specifically the effective entry of women into the labor market, the social scenario was marked by the so-called patriarchy. The family was essentially characterized by the presence of a man as breadwinner and a woman as housewife, with or without children (BRUSCHINI, 1990;ABRAMO, 2007).
The implications of this traditional model are many; however, for this study, it is interesting to highlight the effects of this type of family configuration on gender stereotypes, mainly in the roles played by women and men as part of a couple. In the sense that individuals are surrounded by a pre-established pattern of behavior, those who do not follow this pattern may suffer discrimination (KIMMEL, 2011). In other words, a couple who do not conform to the standard behavior usually suffer social pressure, at least initially. This is what happened at the beginning of the participation of women in the labor market, an event that inaugurated the idea of the dual-worker family in which both spouses have an occupation outside the home (RAPOPORT and RAPOPORT, 1976). It is from the field of the dual-worker family (RAPOPORT and RAPOPORT, 1976) or dualemployed couples (STOLTZ-LOIKE, 1992) or two-career couples (HALL and HALL, 1978) that the concept of the dual-career couple emerged.
For Rapoport and Rapoport (1976, p. 9), among the pioneers in this area, "[…] the term dual-career family was coined to designate a type of family structure in which both heads of household, husband and wife, pursue active careers and family lives". This structure varies according to the presence or absence of children (RAPOPORT and RAPOPORT, 1969). It is important to emphasize that, for these authors, the concept of career is closely related to a high degree of commitment and development at work.
Over the years, the term gained a wider scope. For example, in the definition proposed by Hall and Hall (1978), it could also be assigned to homosexual couples, regardless of any legal or religious regulations, as long as the partners live together in the same house. Even with this updating of the term, some elements are retained from the proposal of Rapoport and Rapoport (1976): the high degree of commitment, continuous development, and belief in their occupations are essential characteristics of the dual-career household.
On the basis of these definitions, some authors sought to identify variations within the dual-career group, emphasizing that this group is not homogeneous and that couples vary in their perceptions of the relationship between the household and work spheres. It is also important to understand how gender roles are shared between men and women. Even when both spouses have occupations outside the home, women continue to be associated with the domestic domain and men with the professional domain. This conception has lasted for a long time and is still rooted in society (RAPOPORT and RAPOPORT, 1971;JAITLY, RANA and SHARMA, 2014;MATIAS and FONTAINE, 2015).
Several authors have followed this line of enquiry. Hall and Macdermid (2009) considered the division of household tasks between men and women in the US and found three profiles of couples: (1) the counterbalanced couple, in which the woman is more dedicated to household chores and works less outside the home; (2) the second-shift couple, in which the woman has a full-time career and performs most of the housework; and (3) the parallel couple, in which the man and the woman better allocate the hours they spend between home and work. Helms, Walls, Crouter et al. (2010) found four profiles of couples in the US based on attitudes toward provision of family income: (1) the main-secondary profile, where one spouse is the main provider and the other the secondary provider; (2) the coprovider profile, characterized by a more equitable division between the man and the woman of the provision of family income; (3) the ambivalent coprovider profile, in which there is no consensus between the spouses on the issue of provision; and (4) the mismatched profile, in which each spouse has a different view on the issue of provision. Allen and Finkelstein (2014) determined six family stages based on the age of the couple and the presence or absence of children: (1) the establishment stage, with individuals under 35 years of age without children; (2) the very young children stage, with individuals who have children under three years of age; (3) the preschool children stage, with individuals whose youngest child is between the ages of three and five; (4) the elementary schoolchildren stage, with individuals with children between six and 12 years of age; (5) the teenage children stage, with individuals with children between 13 and 18 years; and (6) the empty nest stage, with individuals over 54 years of age with no children at home.
In the Brazilian context, unlike in the international literature, only a few studies have focused on the theme of the dual career. Given the specific characteristics of the country, it is necessary to adapt any theory that has been developed in a different context, at least until country-specific studies gain force and develop a local response. To this end, authors such as Berlato (2015) have tried to outline whether and how Brazilian couples fit into the existing profiles. From a sample of 340 participants, all of whom were former students of a public university in the country, the author attempted to develop a profile of the Brazilian dual-career couple, identifying five types based on the way the couples managed their time between the domains of work and family. Berlato (2015) initially identified three major groups: the "familists," the "careerists," and the "acrobats." In the first of these groups, both spouses are more dedicated to the family than to their careers. In the second group, both spouses are more Dual career couples and their inclinations regarding work and family: the Brazilian scenario 498-508 Cad. EBAPE.BR, v. 17, nº 3, Rio de Janeiro, July/Sept. 2019.

Heliani Berlato Thais Fernandes
Daielly Melina Nassif Mantovani dedicated to their careers than to the family. In the last group, there is a balance between the two domains of family and work. In the second stage of Berlato's research, the familists and the careerists were split in two groups based on the variable "sex." The women in the familist group were classified as "coordinated familists" and the men as "conventional familists." The women in the careerist group were classified as "conventional careerists" and the men as "coordinated careerists." Five categories were therefore obtained: coordinated familist, conventional familist, coordinated careerist, conventional careerist, and acrobat (BERLATO, 2015).
The relevance of Berlato's (2015) study for understanding the dynamics of the Brazilian couple is clear, given the scarcity of country-specific literature on this theme. However, it is not possible to generalize from these results, because the sample is centered in a very specific group of couples. It is for this reason that the present study aims to expand on these findings by using a richer database (i.e., the PNAD data) in order to support a possible update of Berlato's model.
The PNAD data are national. Their use in the present research may be considered as a novel contribution, since data of this type (microdata) are not frequently used in dual-career research. Scholars have generally preferred to gather primary data (e.g., LIU, NGO and CHEUNG, 2015;UNGER, SONNENTAG, NIESSE et al., 2015;MATIAS and FONTAINE, 2015;SCHMIDT, AZEVEDO BOLZE, VIEIRA et al., 2015;SCHOOREEL and VERBRUGGEN, 2016;MADALOZZO and BLOFIELD, 2017). We used PNAD data, since they can provide the researcher with an overview of the national context, which is useful for the identification of trends and emerging issues that may shape future research (MYSÍKOVÁ, 2015;MINNOTTE, MINNOTTE and THOMPSON, 2016;AISENBREY and FASANG, 2017).

Selection of the data source
The data source used was the PNAD, a database supplied by the Brazilian Institute of Geography and Statistics (IBGE in Portuguese). The PNAD consists of secondary data, collected to provide researchers in the country with a basis for their studies on themes such as family, income, work, education, health, and fertility (Methodological Notes of the PNAD, 2014).
In the international context, the use of a secondary database as a source of evidence for the dual-career phenomenon is not common, but it is not new. Authors including Wheatley (2013) In the national context, however, there is no evidence of the use of this kind of data in the dual-career literature.
The PNAD was chosen for the present study for its comprehensiveness (it provides data from all Brazilian states), its frequency (unlike other databases, it is updated annually), its variables (most of which are related to the theme of this study), its sampling system (which makes it possible to generalize from the results based on the data), and the possibility of manipulating the data according to the needs of the researchers.
A quantitative-descriptive study was therefore conducted using the PNAD secondary data. As the goal was to identify the characteristics and approaches of couples in Brazil, the model proposed by Berlato (2015) was used as a starting point. From this, the data were used to specify an appropriate sample and to select only the variables related to the dual-career theme.

Treatment of the data
In the 2014 edition (the most recent edition available at the time of this study), the PNAD database consisted of information from 1,100 municipalities, divided into 9,166 sectors, making a total of 151,291 households or 362,627 people (PNAD, 2014). Using Stata/SE 12.0-64 software, the microdata were extracted and filters were applied with the purpose of selecting the sample required for this study and all relevant variables related to the dual-career theme (Box 1). The PNAD data were divided into two different sets, one consisting of information relating to people and the other of information relating to households. The two sets were linked by means of "control number" and "serial number," variables designed to identify households and their occupants. By means of this link, it was possible to apply the first filter, retaining the data for households with people living in them and deleting the data for empty houses.
As the goal was identification of couples, a second filter was applied, using the "total residents" variable to remove households occupied by a single individual. Using the "type of family" variable, a third filter was applied to retain only households consisting of couples, with or without children. At this point, 261,929 cases remained in the sample.
Concerning the "type of family" variable, it is necessary to point out that no account was taken of children from other marriages nor of those who did not reside in the same household as the parents. This means that the sample only included resident children of the household who were born from the current union. This scenario was chosen to provide a more homogeneous sample, as the presence of children from a previous marriage changes the dynamic of the couple, and because a couple's way of life in households with no children is similar in cases where the children have left and in cases where there were no children.
As the dual-career phenomenon is particularly common in urban environments, a fourth filter used the "census sector code" variable to remove households located in rural areas (thus excluding 42,567 cases). After that, a fifth filter based on the "family situation" variable was used to select only those people classified as "reference person" or "spouse," thereby limiting the sample to couples. To reinforce the presence of couples, a sixth filter was applied using the variable "Do you live with a Dual career couples and their inclinations regarding work and family: the Brazilian scenario 500-508 Cad. EBAPE.BR, v. 17, nº 3, Rio de Janeiro, July/Sept. 2019.

Heliani Berlato Thais Fernandes
Daielly Melina Nassif Mantovani spouse or partner?" Only those who answered "yes" to this question were retained, but there were no further exclusions. At this point, the sample contained 131,108 cases.
Moving forward, as a proxy for career development (regarded as an elementary characteristic of the dual-career couple), it was necessary to use the variable "highest level of education reached." This criterion was chosen because it is clear from the dual-career literature (JAITLY, RANA and SHARMA, 2014;BERLATO, 2015;UNGER, SONNENTAG, NIESSE et al., 2015;MCFALL and MURRAY-CLOSE, 2016) that the search for better qualifications characterizes the engagement of individuals in their professional trajectories, a context that frames them in the dual-career perspective. Thus, an eighth filter was necessary to remove all the people who did not have a college education, leaving 3,763 cases in the sample.
As dual-career couples are composed of two people who have careers and are developing those careers, it was important to keep in the sample only households where both spouses were college graduates. Cases that did not fulfill this condition were therefore excluded. Households containing more than one couple were also removed to maintain the homogeneity of the sample. After the filtering process, the final sample consisted of 890 cases.
The great reduction of the sample size, from 362,627 cases to only 890, is notable. However, the low number of cases, which includes only those couples in which both spouses have a college degree, is in line with data concerning the education of the Brazilian population generally. The PNAD (2014) indicates that only 8.57% of individuals in the population have reached higher-level education. The number of such people who are also married can only be lower.
The first step of the data analysis involved a descriptive discussion of the cases. Then, to confirm the relationships established in the first step, a correspondence analysis was carried out, creating perceptual maps based on the correspondences between all variables. The technique is therefore based on the relationships observed in contingency tables, creating a map in which proximity indicates the level of association between the categories. It should be observed that correspondence analysis is applicable only when the relationship between the variables in the contingency table is verified using the chi-square test (HAIR, BLACK, BABIN et al., 2009).

Characteristics of the sample
From the PNAD data for 2014, taking into account only couples with a college-level education, the sample consists of 890 observations. Box 2 gives the frequencies of the profile variables in this sample. The family composition is among the elements that draw attention to the profile of this sample. First, the predominance of couples without children indicates a tendency to have fewer or no children (BARROSO, ALVES and CAVENAGHI, 2008;LEONE, MAIA and BALTAR, 2010). Second, the large number of couples with children under 14 (40.4%), given that the average age in the sample was 32 years, may indicate a tendency to delay maternity/paternity.

Box 2 Profile of the sample
Another notable point emphasizes the association of the caretaker role with women; women are mostly responsible for taking care of household chores. This fact reflects what Elias (1994, p. 185) considers as a constraint in which the "[…] civilizing process does not follow a straight line". Even if it is possible to identify progressive trends, there are still barriers to social progress, and the progress identified here is the shortest way to greater equality between the sexes. Thus, even if the dual-career phenomenon is an indicator of social change (RAPOPORT and RAPOPORT, 1976), the data still reveal traces of the traditional model, in which women are linked to the domestic sphere and men to the public sphere.

Attitudes of Brazilian couples
As a proxy for classifying couples according to Berlato's model (2015), the variable "[…] hours usually worked per week in all jobs in the reference week" was used. This criterion is commonly used to measure the individual's involvement with their work. Wheatley (2013), Unger, Sonnentag, Niessen, et al. (2015), and Ferguson, Carlson, Kacmar, and Halbesleben (2016), are recent examples of this usage. In the present study, those who did not work or worked up to 20 hours per week were classified as familists. Those who worked between 20 and 40 hours a week were classified as acrobats. Those who worked more than 40 hours a week were classified as careerists. By crossing these categories with the variable "sex," the following results were obtained (Figure 1).  Women classified as familists (BERLATO, 2015) make up the coordinated familist type (20.7%). In this group, there was a predominance of white women (60%), of legally married women (67.6%), and of women not exercising any paid activity (67.6%). The fact that the vast majority of them did not work outside the home partially accounts for the small proportion (23.8%) who were classified as reference persons for their households. The vast majority (94.1%) claimed to take care of domestic tasks, with 47.6% spending more than 20 hours a week on such activities. Among women who declared themselves as performing paid activities (32.4%), the majority (76.7%) had monthly incomes of up to R$1,500, and 53.3% had been in the same job for a year or less. When looking at household income in per capita terms, 74.7% of the households in this group had incomes above R$1,500; that is, income from these women did not appear to be the only source of income for the household.
Regarding children, a large number of these women were in the category of couples without children (41.6%) or with children under 14 years old (42.7%). It is interesting to observe that, among the women who declared themselves as not working, 41.6% claimed not to have children, a finding that contradicts the idea that women often stop working because of maternity (ANDRADE, 2015).
Male respondents classified as familists (BERLATO, 2015) were of the conventional familist type (5.1%). Among them, blacks and browns predominated (75.6%), as did those who were legally married (51.1%) and those who were not undertaking paid work (64.4%). However, unlike the women in the same group, 86.7% of these men were classified as the reference person of their household. Although 75.6% of familist men said that they took care of household chores, 53% spent a maximum of 10 hours on these activities, while familist women generally spent more than 20 hours a week.
Of those men who reported working in the reference week, 50% had been in that job for a year or less, 62.5% had a maximum monthly income of R$1,500, and 64.4% had a per capita household income higher than R$1,500. This finding indicates that these men's individual incomes carry great weight in their households; compared to the women, the proportion of families with incomes over R$1,500 declined along with the proportion of individual income of these men.
As in the case of the women, those without children were the single biggest group (41.6%). This raises the question of why, although most of these men were not engaged in paid work, they nevertheless performed fewer domestic tasks than the women.

Heliani Berlato Thais Fernandes Daielly Melina Nassif Mantovani
The analysis of multiple correspondence for the group of familists indicated that, apart from gender, what differentiated them was the time dedicated to domestic tasks and the position occupied in the family unit. Most of the coordinates dedicated more than 20 hours per week, whereas the conventional ones devoted a maximum of 10 hours per week. The conventional coordinates were more often the reference person for their household (p < 0.01 in the chi-square test) (Figure 2). It is possible to observe that, in the group of familists, although most individuals of both genders affirmed that they were not working in the reference week of the survey, the man was more likely to be the household reference person. This is in addition to the significance of their individual income in the composition of the family income. As indicated above, even if such men generally performed household chores, they performed them less than women. These findings reflect the traditionalist character of this group; even on a smaller scale, men are still seen as the breadwinners, despite being unemployed. The income of the women in this group is still a secondary consideration, as some models with traditional traces have noted (HILLER and DYHOUSE, 1987).
Turning attention to the careerist group, it is clear that the careerist conventional type includes some women (1.7%) (BERLATO, 2015). Most of these women were white (56.7%) and legally married (52.9%). Despite their high workload (more than 40 hours a week), they tended to be the spouse at home (72.1%) and dedicated to domestic tasks (97.1%). This group is more career-inclined than family-inclined; however, as in the case of the familists, the presence of couples without children may represent a general tendency rather than something related to how couples approach the balance between home and work (BARROS, ALVES and CAVENAGHI, 2008;CAMARANO, MELLO and KANSO, 2009;LEONE, MAIA and BALTAR, 2010).
The monthly income range of up to R$1,500 was most common for women in this group (64.4%), taking into account the cases in which the family income per capita was over R$3,000 (60.6%). Like the familists, the individual income of these women was mostly in the range up to R$1,500, although, in the second case, the family income was much higher. This observation indicates that women's wages, even when they have a high workload, may be of secondary importance to their household's income.
In contrast, the coordinated careerist type includes male respondents classified as careerists (16.1%) (BERLATO, 2015). In this group, there was a prevalence of individuals of black and brown race (51.8%) and of those who were legally married (63%). Unlike female careerists, male careerists were predominantly reference persons (75.5%). This is in line with the findings relating to the familists; although the careerist women had high workloads, in most cases men were the reference persons of their households. Couples with children under 14 were most prevalent in this group (46.2%), followed by couples without children (42.7%). The fact that careerist men had a greater presence of children than careerist women may be connected to the fact that the largest proportion of people who were not involved in housework (25.9%) were in this category. That is, in households with children, it appears that most of the responsibilities attached to them were transferred to the partners of careerist men.
The predominant range of monthly income was relatively high compared to that of women in the same group, ranging from R$1,500 to R$3,000. As the household income per capita was in most cases higher than R$3,000 (61.5%), it appears that male careerists have higher incomes than female careerists, even if both have high workloads. This finding is consistent with the literature on the gender gap in earnings (ROOS and GATTA, 1999) and the income gap between genders. Careerist men usually receive more money than women, even when their circumstances are similar.
The analysis of multiple correspondences for careerist individuals indicated a relationship between the careerist profile and the situation at home, dedication to household chores, and individual income (p < 0.01). The map (Figure 3) indicates that the women (conventional careerists) were the spouses at home, had lower incomes, and were more dedicated to household tasks than coordinated careerists, who were usually the person of reference and were less dedicated to household tasks. Finally, the acrobat group, which represents the majority of the sample (46.4%), showed balanced proportions in terms of gender (men 53.8% and women 46.2%), as well as condition in the family unit (54% were reference persons and 46% were spouses).
In this group, whites (53%), the legally married (65.4%), and households without children (45.5%) or with children under 14 years old (40%) predominated. Most of the acrobats took care of domestic tasks (84.5%), with a predominance of workloads of up to 10 hours per week (40.2%). In relation to income, the most common monthly income range was up to R$1,500 (45.3%), followed by R$1,500 to R$3,000 (30.8%). Household income, in per capita terms, was in most cases above R$3,000 (58.1%).
It was necessary to apply multiple correspondence analysis in order to verify differences of behavior between the genders within the group of acrobats. The chi-square test indicated a relationship between genders only for position in the family unit, individual income, and dedication to domestic chores (Figure 4). The male acrobat had the characteristic of being the reference person of the domicile and of dedicating a maximum of 10 hours a week to the activities of the house. The women in the group were mostly the spouses and presented income of up to R$1,500 per month.  Evaluation of the groups jointly by means of analysis of multiple correspondences ( Figure 5) shows that the coordinated familists stood out as having the greatest dedication to domestic tasks. The coordinated careerists were notable for having higher income and less dedication to domestic tasks and for being the reference person of the household. The conventional careerists were characterized by lower incomes and intermediate levels of dedication to domestic tasks. Acrobats were the group showing the most balance in relation to the study variables. Dual career couples and their inclinations regarding work and family: the Brazilian scenario 506-508 Cad. EBAPE.BR, v. 17, nº 3, Rio de Janeiro, July/Sept. 2019.

Heliani Berlato Thais Fernandes Daielly Melina Nassif Mantovani
As in Berlato (2015), the careerist and familist groups were mostly in second position after the acrobat group, since the proportions between them were very similar to those of the original model. In the case of the subgroups, in Berlato (2015) the coordinated careerists predominated, but here the coordinated familists were more prevalent. Likewise, the smallest group in Berlato (2015) was the conventional careerists, whereas here the conventional familists occupied that position.

FINAL CONSIDERATIONS
Despite the variations within the dual-career phenomenon, it is possible to identify recurring characteristics across the various groups. For example, men made up the majority of household reference persons, and women in all groups performed most of the household tasks. The findings indicate that the profile of the Brazilian couple (from the database), even from the perspective of the dual-career household, presents several traits consistent with traditional family models. This applies even in cases in which women seemed to be pursuing their careers; when the role of their income within the family income was examined, its secondary character became clear.
Among the groups that were identified, the acrobats showed the best balance between genders, whereas the familists and the careerists showed considerable disparities between men and women. As the acrobats were the largest group, it is possible to infer that, even given the presence of some characteristics that reflect the old models, there is clear progress toward achieving greater balance.
Finally, despite the limitations of the database (such as the absence of variables that addressed homosexual relationships and the inability to identify who was married to whom), it was possible to identify certain trends, especially with regard to the relationship between genders within the dual-career perspective. With the improvement of this technique, PNAD still has much to contribute to the literature, especially regarding the factors that permeate the conceptions of each gender in relation to the spheres of family and work.
On the basis of these findings, a suggestion for future research is an exploration of the database to investigate how couples, starting from the point of view of each sex, perceive the balance between the various demands on them, including family and work. It should be emphasized that there is no need to restrict research efforts to the spheres of work and family. Characteristics related to migration, fecundity, income, schooling, health, and welfare, among others, can usefully be the focus of analysis and will bring relevant information to enrich understanding of the social context of the couple in Brazil.