This paper analyzes the role of increased schooling, especially for women, in the onset and subsequent pace of Brazil's rapid fertility decline. The authors begin by documenting several important changes in the distribution of education in Brazil in recent decades. First, there has been a substantial increase in the mean level of education for both men and women, even though the mean for both sexes remains low by international standards. Second, the increase in mean schooling has been associated with a reduction in inequality in schooling for both men and women. Finally, there has been a convergence in the educational attainment of men and women, with women's education rising at a faster rate than that of men for all of the last four decades. Women's educational attainment surpassed that of men beginning with cohorts born in the 1950's, a trend which the study's results indicate is continuing.
Using the retrospective fertility histories in Brazil's 1984 PNAD, the authors reconstruct the fertility behavior of cohorts of Brazilian women back to the 1930 cohort. Comparison of fertility profiles across cohorts indicates that fertility began to decline with the 1940 birth cohort. The authors identify a substantial decline in period fertility beginning around 1965. This fertility decline appears to be strongly dominated by cohort effects. Cohorts born before 1940 do not exhibit substantial changes in their fertility compared to earlier cohorts, even after significant declines in age-specific fertility are observed at younger ages in the 1960's. The cohorts that appear to initiate the dramatic fertility decline are cohorts that also exhibited the most dramatic increases in schooling.
In order to quantify the role of education in the timing and pace of Brazil's fertility decline, the authors estimate a series of regressions with fertility as a function of the schooling of the husband and wife. They use these regressions to predict fertility across cohorts, making it possible to estimate the relative importance of increased schooling in Brazil's fertility decline. Their results indicate that we can explain from 40 percent to 80 percent of the fertility decline across recent cohorts by changes in the schooling of women and their husbands. Furthermore, they are able to quite accurately predict the timing of the onset of the fertility decline and the major increases in the pace of the decline based on the increases in education for both women and men across cohorts.
In contrast to previous studies that have emphasized "economic pressure" and relative economic deprivation of poorer classes as a driving force in Brazil's sustained fertility decline, these results suggest a more optimistic interpretation of recent Brazilian demographic history. Increases in schooling, especially for women, appear to be strong predictors of both the timing and pace of the decline. The authors interpret these increases in schooling as indicating substantial improvements in the social and economic status of women, improvements that have induced large declines in fertility. Although a number of mechanisms may be responsible for the strong link they observe between increases in schooling and reductions in fertiltiy, they find substantial evidence consistent with an economic interpretation, with increased education leading to increased wages and labor market opportunities for women, and thus to decreased fertility.