Brazilian Cohort Fertility by Education and Color/Race: Trends and - PDF document

Brazilian Cohort Fertility by Education and Color/Race: Trends and Future Scenarios for Projections Eduardo L.G. Rios-Neto Adriana de Miranda-Ribeiro Paula de Miranda-Ribeiro I – Introduction In the first decade of this century, Brazil entered into the group of countries that had below replacement fertility. Below replacement fertility occurred about 40 years after the onset of fertility transition in the country. The transition process began in the late 1960s when the total fertility rate (TFR) was 5.8 children per woman. This transition accelerated during the 1980s, reaching an average of 2.4 children per woman by the end of the century. Data from the National Household Sample Survey (PNAD) of 2006 showed that the TFR in the country had reached two children per woman. The following PNAD confirmed this trend and, according to the 2010 Demographic Census, the TFR in Brazil was 1.9. The fertility decline in Brazil was accompanied by a fall in the mean age of childbearing (MAC), suggesting a rejuvenation of the fertility schedule. The PNAD of the second half of the 2000s and the 2010 Demographic Census both indicated that there was a reversal in this trend towards a rising MAC. The strong fertility decline first with a fall and later with a rise in the MAC makes it difficult to devise clear scenarios for future fertility in Brazil. We draw on our previous works calculating tempo and quantum fertility in Brazil to present a different approach calculating cohort fertility at ages 30, 35, and 40. We calculate cohort fertility by four education groups and color/race. Cohort fertility stratified by education presents greater variation among the different birth cohorts of women than in the case of women classified by color/race. This paper uses a dataset based on a methodology that we developed to build birth histories from Demographic Censuses. We add the 2010 census to the birth history already built to the 2000 Brazilian Demographic Census. We can move back as far as fifteen years with the calculated birth histories, a combination of this information with the total number of children ever born entail the calculation of cohort fertility at certain specific ages. After the calculation of cohort fertility, we perform some forecasting exercises with cohort fertility, TFR, adjusted TFR, tempo effect, and mean age at first birth. Some regressions of TFR with cohort fertility and tempo effect are performed to simulate predictions. A critical analysis of all results is performed to devise future fertility scenarios. II – Previous Findings In previous papers, we calculated the tempo effect and parity and tempo adjusted fertility for Brazil (Rios-Neto and Miranda-Ribeiro, 2015; and Miranda-Ribeiro, Rios-Neto, and Garcia, 2016). Following Figure 1, we can see that the Brazilian tempo effect is negative during the 1990’s, approaching zero at the end of this decade. The tempo effect becomes positive, between 5% and

10% during the first decade of this century. Over the whole period, the tempo effect goes from a 20% inflation of TFR in the early 1990s to a 10% deflation in late 2000’s. Figure 1 – Tempo effect (%) in Brazil, 1986 to 2010. 15 29 10 28,5 5 28 Tempo Effect (%) 0 27,5 -5 27 -10 26,5 -15 26 -20 -25 25,5 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 TEMPO (2000) TEMPO (2010) Mean age at childbearing Sources: IBGE – 1991, 2000 and 2010 Brazilian Censuses in Rios-Neto and Miranda-Ribeiro (2015). Figure 2 - Brazil, 1986 to 2010: observed TFR and adjusted PATFR. 3,50 3,00 Total Fertility Rate 2,50 2,00 1,50 1,00 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 PATFR_ADJUSTED TFR_OBSERVED Sources: IBGE – 2000 and 2010 Brazilian Censuses in Rios-Neto and Miranda-Ribeiro (2015).

We evaluate the distortions caused by tempo and parity composition effects on TFR by the comparing observed TFR with PATFR ADJUSTED . The PATFR ADJUSTED is the value of TFR without changes in the mean age at childbearing and in the parity composition. See the TFR and PATFR ADJUSTED series for Brazil in Figure 2. The difference between TFR and PATFR ADJUSTED diminishes over time, while both TFR and the pure index (quantum) reach replacement level. III – A Cohort Fertility Perspective Our technique to construct birth histories out of the period demographic censuses allows us to go back in time 15 years for period fertility. Thus, using the age-period-cohort identity, it is possible to calculate the number of children ever born at age 30 (CEBc30) for all women in the 30-44 age group. It is also possible to calculate CEBc35 for all women in the 35-49 age group and CEBc40 for all women in 40-49 age group. These cohort measures will refer to different single year cohorts at different periods or time references In Table 1, Figures 3 and 4, we observe a decline in cohort fertility. While the decline in cohort fertility at age thirty (CEBc30) may result from tempo and quantum fertility, cohort fertility at age forty (CEBc40) is clearly an indicator of quantum fertility. The youngest cohort fertility at age forty (for the 1970 birth cohort, age forty in 2010) is 2.34 (above replacement). The youngest cohort fertility at age thirty (for the 1980 birth cohort, age thirty in 2010) is 1.5. If we apply the ratio between cohort fertility at age forty and thirty for the 1970 birth cohort, we find 1.27. Applying this ratio to the 1980 birth cohort, we find a cohort fertility at age forty of 1.91 in 2020 (below replacement). Figure 3: COHORT FERTILITY AT AGES 30, 35, AND 40 BY BIRTH COHORTS - BRAZIL 4 3.5 3 Cohort Fertility 2.5 2 CEBc30 1.5 CEBc35 1 CEBc40 0.5 0 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 Birth Cohorts

Figure 4: COHORT FERTILITY AT AGES 30, 35, AND 40 BY BIRTH COHORT - BRAZIL 4 3.5 3 Cohort Fertility 2.5 2 CEBc30 CEBc35 1.5 CEBc40 1 0.5 0 BIRTH COHORTS

Table 1: Cohort Fertility at Ages 30, 35, and 40 by Birth Cohorts - Brazil Birth Cohort CEBc30 CEBc35 CEBc40 1951 3.219812183 3.626485498 1952 3.174274744 3.544310425 1953 3.089868857 3.423922755 1954 3.030046682 3.341859807 1955 2.988515947 3.277960611 1956 2.333149415 2.938857104 3.213732747 1957 2.303984496 2.85908914 3.122417035 1958 2.304564469 2.840243602 3.0930195 1959 2.216850973 2.711209692 2.945388884 1960 2.221954664 2.700628487 2.931480828 1961 2.158376353 2.633498119 2.748370792 1962 2.121837713 2.579278491 2.69652723 1963 2.098956799 2.539665747 2.641316747 1964 2.062763372 2.490369949 2.600298029 1965 2.002442669 2.329977059 2.509266603 1966 1.939051242 2.199765617 2.482466694 1967 1.930217616 2.185388296 2.469764345 1968 1.904483281 2.162484874 2.426488422 1969 1.868873178 2.117422178 2.375546979 1970 1.849622981 2.089574554 2.342615689 1971 1.718580985 2.0320373 1972 1.713219667 2.026090404 1973 1.711634458 2.019270805 1974 1.685976031 1.990509 1975 1.665241369 1.964023779 1976 1.651681508 1977 1.632020745 1978 1.595841927 1979 1.562460699 1980 1.502399085 Sources: IBGE – 2000 and 2010 Brazilian Censuses – Own calculation

IV – Cohort Fertility by Education and Color/Race Due to the large sample sizes in the demographic censuses, it is also possible to perform these calculations for specific attributes such as mothers´ education groups and race/color. With these calculations, not only we can highlight some socioeconomic gradients of low fertility, but we can also devise auxiliary tools for scenarios associated with future cohort fertility in Brazil. We can see in Figure 5 that cohort fertility at age 30 among all women in the 1956 birth cohort was 2.25 (a little above 2.1). Total cohort fertility in this cohort was almost the same level as the cohort fertility in the group of women with primary education (4-8 years of study). As we approach the cohort of women born in 1980, then total cohort fertility approaches the level of cohort fertility in the group of women with secondary education (9 to 11 years of schooling). As the cohort fertility for each education group declines less than cohort fertility for all women, we may infer that education composition played an important role in cohort´s fertility decline. Indeed, Figure 6 shows that roughly 30% of the 1956 birth cohort had secondary or tertiary education, while the number for the 1980 cohort corresponds to 66%. This major composition shift is likely to explain a great portion of the cohort fertility decline. Cohort fertility is more associated with “quantum” than with “tempo” effects, but it is also true that the cut line at age 30 may favor the operation of some tempo effect, particularly in the case of women with tertiary education. Figure 5: Cohort Fertility at Age 30 by Women´s Education Groups – Brazil Birth Cohort Fertility at Age 30 (CEB30Cohort) by Groups of Education - Brazil 3,5000 3,0000 2,5000 CEB30Cohort 2,0000 Total 0 to 3 Years of Study 1,5000 4 to 8 Years of Study 9 to 11 Years of Study 1,0000 12+ Years of Study 0,5000 0,0000 Birth Cohort Source: IBGE, 2000 and 2010 Demographic Censuses, Microdata, own calculations.