Fertility by kind of union in Brazil (1980-2010): educational and - - PDF document

1

Fertility by kind of union in Brazil (1980-2010): educational and religious differentials Joice Melo Vieira (Unicamp) Ana Paula Verona (UFMG) Paulo H. V. Martins (UFMG) Short abstract: This study explores the relationship between fertility and nuptiality in Brazil, using data from the demographic censuses of 1980 and 2010. We have found that women living in consensual unions have higher fertility levels than those who opt for marriage. However, we have detected a converging trend between the fertility levels of married women and those in consensual unions. We apply the decomposition of the fertility rate to establish the contribution of each type of union towards total fertility

• rate. With that, the share of consensual unions over time increases, as does the share of

this type of union in total fertility. The argument is that the meaning of having children has changed, and that even though consensual unions have become more common across all social groups, the type of union the woman is in when having children remains an important differential, according to her educational level and religion. Antecedents In the last 60 years, Brazil has undergone three sets of transformations that have been gradually changing the country's population composition: the demographic transition; the religious transition (Pierucci, 2004; Alves, Cavenaghi and Barros, 2014; Almeida, 2015) and the educational transition (Carnoy et al. 2013), all of which have had significant impacts on family life. These processes have intensified, particularly since the 1980s (see Table 1). The combination of two phenomena – fertility below the replacement level and the growth of consensual unions in lieu of marriages – is at the heart of this study. The pattern of Latin American unions has been described as dual, since consensual unions and marriages have always coexisted side by side in the region. The studies that claim that this dual pattern is prevailing emphasize that there has always been a strong association between union type and social status, with consensual unions being much more common among the less-privileged population – especially among lower-income and less educated groups – while marriage was the rule among those at the opposite end

• f the social hierarchy (Castro-Martín, 2002). These studies tend to express some concern

about consensual unions, based on the premise that informal marital relationships may

2

contribute to men being less involved with the family and possibly increase the risk of women and children ending up unassisted in the event of a separation1. The dual union type thesis has been revised, considering that, in recent decades, consensual unions have become much more common across all segments of Brazilian society, including among the most privileged social strata (Esteve et al., 2012). The explanations for such behavior tend to associate the phenomenon with ideational changes, a weakening of the role of institutions on everyday life, increased uncertainty in long- term decision-making, difficulty achieving financial stability and the elimination of legal barriers that, until re-democratization (1980s), guaranteed comparative advantages to married couples against those in consensual unions (Covre-Sussai and Matthijs, 2010; Covre-Sussai et al., 2014; Verona et al., 2015; Esteve et al., 2016; Vieira, 2016). Some authors tend to emphasize a single explanation. However, considering the complexity of the phenomenon, it is very likely that these diverse economic, cultural and political factors have combined to produce the significant increase in the share of consensual unions in lieu of marriages. A new generation of studies has begun to map the plurality of family formation process in Brazil. And, of course, for every trend there is a counter-trend. When it comes to consensual unions, the counter-trend seems to be represented by the Evangelical youth (Verona et al., 2015). Since the late 1990s, Evangelicals have emerged as a new political force in Brazil. While they are still a minority, they aspire to hold leadership positions in the national executive and legislative branches, based on a platform that calls for a defense

• f conservative values, especially with regards to family as an institution. Some of the

legislation proposals introduced by the Evangelical caucus in the Brazilian Congress is the “Family Act,” which seeks to largely reverse the progressive and pluralistic perspectives on family that have been in place since the country went back to a democratic rule in the 1980s.

1 In Brazil, however, it should be noted that, since the mid-1980s, legislation equated consensual

unions and marriages when it comes to mutual assistance rights and duties. Parenthood and conjugal status become independent entities in family law. Parents’ obligations towards their children do not depend on the context in which they were conceived or the type of relationship between the father and the mother. The popularity of DNA testing and its use as the decisive factor in defining who is responsible for child support has been widespread in Brazil, particularly in parental abandonment cases brought to trial, where it is often available free of charge.

3

Another social group in which marriage continues to be the norm are the highly

• educated. However, in this case, the social status markers associated with marriage seem

to prevail, something that Cherlin (2013) had already observed in the U.S. context. In international comparative studies, Latin America has been described as a region where it is socially acceptable to have and raise children in consensual unions (Castro- Martín et al, 2011; Laplante et al., 2015). But is this behavior widespread across all social segments? If unconventional cohabitation has spread to society as a whole, what family settings are children being born into? Is there a tendency to formalize the union before the children are born? Based on data from 1986, 1996 and 2006, Vieira (2016) affirms that the fertility rate of Brazilian women in consensual unions is slightly higher than that of married women, and the first group has children at a younger age. He notes that, despite a rapid increase in cohabitation also among women in more privileged social groups, the prevailing pattern for such women is to have children in wedlock. Given this situation, this study aims to contribute by comparing the levels and patterns of female fertility according to the type of union, religion and educational level. Then, the fertility rates are decomposed in order to explore the type of marital relationship maintained by the mothers near the birth of their children. Methodology The data used in this study were obtained from demographic census from 1980 and 2010, provided by the Brazilian Institute of Geography and Statistics (IBGE). The first step of the analysis refers to the construction of the specific fertility rates and the total fertility rate for the female population in formal unions (marriage) and in consensual unions. The purpose of this measurement as constructed herein is to explore the fertility level in each of these two subgroups when considered as two distinct populations: married Brazilian women and Brazilian women in consensual unions. Fertility estimates were calculated using the Brass P/F method, a variant developed by Trussell that is based on information about the female population of reproductive age, distributed by five-year age groups; number of children born alive in the last year by mother’s five-year age group; and total number of children born alive by mother’s five- year age group at the time of the interview. In general, when calculating fecundity by marital status, the average number of children per woman tends to be quite high, well above the total fertility rate (TFR) for the

4

entire population, especially due to the specific fertility rate in the 15-19 age group. Many women who are married or in a union in this age group have recently had children. However, when correcting for parturition/fertility of women in the 20-24 age group (P2/F2), the TFR values according to marital status are close enough to the TFR for the total population. Therefore, we adjusted the fertility rates for all five-year age groups based on the reality of 20-24 age group. That is done in order to obtain a more accurate estimate of present fertility, minimizing the impact of fertility among older birth cohorts. The specific fertility rates per five-year age group and the total fertility rate were obtained using the Population Analysis System (PAS), developed by the U.S. Census Bureau2. The second step of the analysis is to decompose the total national fertility rate according to the type of union the woman is in. This procedure allows us to determine which part of the total fertility rate can be attributed to women in formal unions, in consensual unions or who are not in a union. This second procedure allows us to determine the contribution of each type of union towards the total fertility rate in Brazil. What portion of total fertility depends on consensual unions? This procedure is described by Laplante and Fostik (2014). The third point of the analysis focuses on the contrast of fertility according to religion and educational level. Religion comprises the following categories: Catholic, Evangelical/Protestant, Other Religions and Non-Religious. As for educational level, the categories are: no education/some primary; primary completed/some secondary; secondary completed/some university; and university completed. In this step, the fertility rates are decomposed by union type, following the same procedures described in the second step, with a view to better understanding the marital context into which the children are being born. Results Between 1980 and 2010, important changes were made in the general profile of the Brazilian population. Illiteracy rate dropped from 25.5% to 9.6%; the percentage of individuals holding a college degree was still moderate in 2010 (11.3%), but it has expanded significantly compared to the 1980s, when only 3.3% of the population 25+ years old had completed higher education. Catholicism also lost strength during the period under study, but not only as a result of the expansion of Protestantism, especially

2 The Population Analysis System (PAS) is available for download at:

<http://www.census.gov/population/international/software/pas/>. Retrieved on: December 2, 2016.

5

• f entities from Pentecostal origins. The group of people who declared not being

associated with any religion also grew significant, as indicated in Table 1. Table 1 - Brazil, 1980 and 2010 (total population): selected indicators Indicator 1980 2010 Total fertility rate (children per woman) 4.4 1.9 Proportion living in consensual union (15+ years old) 11.8% 36.4% Catholics 89.2% 64.6% Evangelicals/Protestants 6.6% 22.2% Without religion 1.6% 8.0% Illiterate (15+ years old) 25.5% 9.6% University completed (25+ years old) 3.3% 11.3% Source: IBGE. Demographic censuses of 1980 and 2010. Table 2 - Brazil, 1980 and 2010: Total fertility rate- TFR (children by woman) Group 1980 2010 Religion Catholics 4,3 1,9 Evangelicals 4,6 2,0 Other religions 3,1 1,5 Without religion 3,3 1,9 Educational level University completed 1,4 1,1 Secondary completed/University uncompleted 1,8 1,3 Fundamental completed/Secondary uncompleted 2,6 2,5 Without instruction/fundamental uncompleted 5,1 3,0 Conjugal status Married 3,8 1,7 Cohabitant 4,5 2,2 Out of union 1,2 0,9 All Brazilian women 4,3 1,9 Source: IBGE, Demographic censuses of 1980 and 2010. In general, the boom of consensual unions is considered one of the main changes undergone by families in recent decades. Considering the total number of people 15+ years old living with a partner, the rate of individuals living together without formalizing the union before the Church or the State jumped from 11.8% in 1980 to 36.4%. In fact, the proportion of women of reproductive age (15-49 years old) who are not in a union remained virtually stable throughout the period under study: 43% in 1980 and 44.5% in 2010, while marriage was overtaken by consensual unions, dropping from 49.7% to 31.3% among women in this age group. On the other hand, consensual unions rose from 7.3% to 24.1%.

6

With regards fertility, there was a convergence towards replacement level fertility, both when breaking down the population according to religion and by type of union (consensual or marriage). Excluded from this scenario are women whose educational level is equivalent to secondary completed or university completed and those out of union whose fertility is below the replacement level. Upon examining specific fecundity rates according to religion in 1980 and 2010 (Figures 1 and 2), it can be observed that the difference between the two major religious groups – Catholics and Evangelicals – drops significantly between 1980 and 2010. Evangelical women present a higher fecundity rate than Catholic women, but this difference becomes significantly lower, remaining only in the 20-29-year-old range. Although the reduction of specific fertility rate is expressive in practically all age groups, it is interesting that such rate did not changed among teenagers without religion. The specific fertility rate is higher among these teenagers, remaining at the same level in the two years under study. Comparatively, differentials of fertility by educational level are much more pronounced than those ascertained according to the declared religion. There has been a noticeable aging of the fecundity standard of women with secondary completed or university completed between 1980 and 2010. The differences in fertility experiences between 15 and 34 years of age is significant, both in 1980 as in 2010. However, in 2010, the early years of pre-adulthood are even more critical to understand how women with different educational levels have contrasting average number of children at the end of the reproductive period: 3 children per woman for those without education and 1.1 children per woman for those with university completed. Figures 5 and 6 show the specific fertility rates by conjugal status in 1980 and 2010, respectively. Note that systematically, women in consensual unions present fertility rates higher than married women in all age groups. Undoubtedly, a composition effect takes place, given that in Brazil, consensual unions are historically more widespread among the socially underprivileged groups. Please note that the fertility rate of women who are not in a union is nearly one child per woman at the end of the reproductive period (1.2 in 1980 and 0.9 in 2010).

7

Figure 1 - Brazil, 1980: Total fertility rate (TFR) by religious group

Source: IBGE, Brazilian census 1980.

Figure 3 - Brazil, 1980: Total fertility rate (TFR) by educational level

Source: IBGE, Brazilian census 1980.

Figure 2 - Brazil, 2010: Total fertility rate (TFR) by religious group

Source: IBGE, Brazilian census 2010.

Figure 4- Brazil, 2010: Total fertility rate (TFR) by educational level

Source: IBGE, Brazilian census 2010.

0,00 0,05 0,10 0,15 0,20 0,25 0,30 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Catholic (TFR = 4,3) Evangelical/ Protestant (TFR = 4,6) Others (TFR = 3,1) Without religion (TFR = 3,3) All Brazilian women (TFR = 4,3) 0,00 0,05 0,10 0,15 0,20 0,25 0,30 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Without instruction/Fundamental incompleted (TFR = 5,1) Fundamental completed/Secondary incompleted (TFR = 2,6) Secondary completed/University incompleted (TFR = 1,8) University completed (TFR = 1,4) All Brazilian women (TFR = 4,3) 0,00 0,05 0,10 0,15 0,20 0,25 0,30 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Catholic (TFR = 1,9) Evangelical/ Protestant (TFR = 2,0) Others (TFR = 1,5) Without religion (TFR = 1,9) All Brazilian women (TFR = 1,9) 0,00 0,05 0,10 0,15 0,20 0,25 0,30 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Without instruction/Fundamental incompleted (TFR = 3,0) Fundamental completed/Secondary incompleted (TFR = 2,5) Secondary completed/University incompleted (TFR = 1,3) University completed (TFR = 1,1) All Brazilian women (TFR = 1,9)

8

Figure 5 - Brazil, 1980: Total fertility rate (TFR) by conjugal status

Source: IBGE, Brazilian census 1980.

Figure 6 - Brazil, 2010: Total fertility rate (TFR) by conjugal status

Source: IBGE, Brazilian census 2010.

0,00 0,05 0,10 0,15 0,20 0,25 0,30 15-19 20-24 25-29 30-34 35-39 40-44 45-49

Married (TFR = 3,8) Cohabitant (TFR = 4,5) Out of union (TRF = 1,2) All Brazilian women (TFR = 4,3)

0,00 0,05 0,10 0,15 0,20 0,25 0,30 15-19 20-24 25-29 30-34 35-39 40-44 45-49

Married (TFR = 1,7) Cohabitant (TFR = 2,3) Out of union (TFR = 0,9) All Brazilian women (TFR = 1,9)

9

Figure 7 - Brazil, 1980: Participation of each conjugal status in the fertility rate of women

Source: IBGE, Brazilian census 1980.

Figure 8 - Brazil, 2010: Participation of each conjugal status in the fertility rate of women

Source: IBGE, Brazilian census 2010.

Figures 7 and 8 show the contribution of each conjugal status to the total fecundity curve in Brazil in 1980 and 2000, respectively. Major changes can be highlighted. First, the contribution by married women to fecundity dropped significantly in this thirty-year period, going from 79% TFR in 1980 to 34.7% in 2010. At the same time, the contribution

10

to TFR by mothers who were not in a union or were in a consensual union at the time of the census has increased substantially. In 2010, 44.9% of fertility occurred among mothers who were in a consensual union, indicating that in Brazil, motherhood most often takes place in consensual unions. Another important result refers to the structure of fecundity for each conjugal

• status. Both in 1980 as in 2010, the largest contribution to fertility by married women is
• bserved between 25 and 29 years of age. On the other hand, the apex of the fecundity

curve is earlier (between 20 and 24 years of age) in women who were not in a union or were in a consensual union. This difference in relation to the age standard of the fertility curve is likely associated with the socioeconomic profile of women in each conjugal

• status. On average, women in consensual unions have lower educational level and income

compared to married women (Castro-Martin, 2002). Figure 9 shows the contribution of each conjugal status to the total fecundity curve in Brazil in 1980, according to religion. These results allow examining female fecundity levels and standards according to type of union and religion simultaneously. The group with the highest participation of married women in fertility are Catholic and Evangelical women, respectively. Of all Evangelical women who had children up to 12 months prior to the 1980 census, 87.4% were formally married. This year, the birth of children among women in a consensual union was more common among those who, at the time of the census, reported not being associated with any religion. Figure 10 shows the contribution of each conjugal status to the total fecundity curve in Brazil in 2010, according to religion. As for Catholic women in 2010, 31.6% of fertility occurred among married women and 47.3% among women in a consensual union. Please note the aged fertility curve among married Catholic women (the peak of the curve is found in the 30 to 34-year-old group) compared to early curve among Catholic women in a consensual union in 2010. Similar to 1980, Evangelical women represent the group with the greatest contribution of married women for fecundity in 2010. Earlier studies had already indicated this religious group’s preference of formalizing the union (Verona et al, 2015). Nonetheless, 52.7% of Evangelical women who had children up to twelve months before the 2010 census were, at the time of the census, in a consensual union (35.8%) or were not in a union (16.9%). Among women not associated with any religion who contributed to fertility in Brazil in 2010, only 15.9% were married.

11

Figure 9 - Brazil, 1980: Participation of each conjugal status in the fertility rate of women by religion

Source: IBGE, Brazilian census 1980.

12

Figure 10 - Brazil, 2010: Participation of each conjugal status in the fertility rate of women by religion

Source: IBGE, Brazilian census 2010.

13

Figure 11 - Brazil, 1980: Participation of each conjugal status in the fertility rate of women by educational level

Source: IBGE, Brazilian census 1980.

14

Figure 12 - Brazil, 2010: Participation of each conjugal status in the fertility rate of women by educational level

Source: IBGE, Brazilian census 2010.

15

Figures 11 and 12 show the contribution of each conjugal status to the total fecundity curve in Brazil in 1980 and 2010, respectively, according to educational level. The more educated a woman, the greater the proportion of fecundity within marriage in the two points in time studied. However, while in 1980 fecundity in Brazil depended mostly on married women regardless of their educational level, the situation is significantly different in 2010. That year, contribution by married women to the total fertility rate is more than 50% only in the group with university completed. Fecundity outside any type of union increased in all educational groups between 1980 and 2010; however, this event is more evident among women with lower educational levels. Especially in 2010, participation by married women to total fecundity is usually later, a characteristic that becomes even more pronounced the higher the women’s educational level. Finally, the breakdown of fertility by conjugal status according to educational level and religion indicates that differentials imposed by education appear to be more resistant over time. Contribution by married women to fecundity is also more pronounced in more conservative groups from the religious perspective, and in more privileged groups from the educational level perspective. Remarks This paper presents some of the main descriptive results of a larger study that aims to measure the contribution of women in different conjugal status in calculating total fertility in Brazil, and what factors could interfere for a child to be born within marriage, consensual unions, or outside any type of union. It also seeks to understand how the importance of conjugal status to the fecundity of different social groups can be different. While groups that follow more traditional religions and have more privileged social positions may prefer to have children within marriage, in other social sectors, there is an increased perception of consensual union as a common space for the constitution of

• ffspring. The interpretation of fertility outside the union is certainly more open to

ambiguity, as children may have been conceived within unions that had broken up by the time the child was born or even after. Moral and legal censure towards abortion could also contribute to the increased participation by women who are not in a union to total fertility rate in different social groups. However, at least for hypothesis for future studies, it should not be ruled out that independent voluntary reproduction could also account for part of this increase.

16

The fact that this study is based entirely on cross-sectional data to address issues that would be better understood from longitudinal data is a valid claim. However, given the lack of longitudinal data that allow studying the interaction between marital and reproductive trajectories of the Brazilian population, the results presented here could provide important clues. References Alves, J. E. D., Cavenaghi, S. M., & Barros, L. F. W. (2014). A transição religiosa brasileira e o processo de difusão das filiações evangélicas no Rio de Janeiro/The Brazilian religious transition and the process of diffusion of evangelical affiliations in Rio de Janeiro. Horizonte, 12(36), p.1055. Carnoy, M. et al. (2013). University expansion in a changing global economy: Triumph

• f the BRICs? Stanford University Press.

Castro-Martín, T. (2002). Consensual unions in Latin America: the persistence of a dual nuptiality system. Journal of Comparative Family Studies, vol. 33, n. 1, p. 35-55. Castro-Martín, T.; Cortina, C.; Martín García, T.; Pardo, I. (2011). Maternidad sin matrimonio en América Latina: un análisis comparativo a partir de datos censales. Notas de población, n. 93, Cepal, p. 37-76. Cherlin, A. J. (2013). Public and private families. New York: McGraw-Hill. 7ª ed. Covre-Sussai, M.; Matthijs, K. (2010). Socio-economic and cultural correlates of cohabitation in Brazil. 2010 Chaire Quételet Conference, Louvain-la-Neuve. Catholic University Leuven, Centre for Sociological Research, Leuven, Belgium. Covre-Sussai, M. et al. (2014). Disentangling the Different Types of Cohabitation in Latin America: Gender Symmetry and Contextual Influences (January 9, 2014). Available at SSRN: https://ssrn.com/abstract=2376739 or http://dx.doi.org/10.2139/ssrn.2376739 Esteve, A., Lesthaeghe, R., & López‐Gay, A. (2012). The Latin American cohabitation boom, 1970–2007. Population and Development Review, 38(1), 55-81. Esteve, A. et al. (2016). Cohabitation in Brazil: historical legacy and recent evolution. In: Esteve, A. e Lesthaeghe, R.J. (Eds.), Cohabitation and Marriage in the Americas: Geo- historical Legacies and New Trends, Springer Open. Laplante, B., Castro‐Martín, T., Cortina, C., & Martín‐García, T. (2015). Childbearing within marriage and consensual union in Latin America, 1980–2010. Population and Development Review, 41(1), 85-108. Laplante, B., & Fostik, A. L. (2014). The recent evolution of fertility within marriage and consensual union in two Canadian provinces: Disentangling the Québec fertility

• paradox. Paper presented at the Population Association of America Annual Meeting,

Boston, May 1–3, 2014. Pierucci, A. F. (2004). “Bye bye, Brasil": o declínio das religiões tradicionais no Censo

• 2000. Estudos Avançados, 18(52), 17-28.

17

Verona, A. P., Dias Jr, C. S., Fazito, D., & Miranda-Ribeiro, P. (2015). First conjugal union and religion: Signs contrary to the Second Demographic Transition in Brazil?.Demographic Research, 33, 985. Vieira, J. M. (2016). Diferenciales en la fecundidad brasileña según la naturaleza de la unión: algunas reflexiones sobre decisiones reproductivas y convivencia. Notas de Población, v. 102, p. 67-94.