In relation 1993, the proportion of mothers younger than 20 years old grew: in 2003, they were more than 20% of the total of mothers. However, in 2002, the percentage of young mothers was higher: 21.23%. Among men, in 2003, the rate of violent deaths was five times higher than among women. In relation to 1993, the contribution of violent deaths in relation to the total of deaths increased considerably, but, in relation to 2002, it decreased slightly. The number of marriages returned to the level seen in 1993 – almost 750 thousand – after it had decreased to 730 thousand in 2000. Also in relation to 1993, the number of legal separations and divorces grew 17.8% and 44%, respectively. However, under-registration, which is still considered high, fell from 23.4% in 1993 to 21.6% in 2003. The same happened to the late registrations, which fell from 25.6% to 22.5%, in the same period. IBGE releases the Civil Registry Statistics of 2003, which systematize information reported to the Institute by civil registry offices from all over the country. Such information is a rich data source covering weddings, divorces, legal separations, births and deaths registered in Brazil. But the records still poses serious problems, particularly in the North and Northeast Regions, and to a lesser extent, in the Central-West Region, except for the Federal District. Thus, data released through the Civil Registry Statistics should be examined with caution.

Statistics show that the majority of women have their children when they are 20 to 29 years old, and the highest rates are concentrated in the group of women aged 20 to 24. But the relative increase in the participation of the number of births of teenage and young mothers aged up to 19 has become a matter of concern. Studies reveal that the phenomenon is focused on teenagers of an economically lower class.

The trend must also be understood within the change in the fertility age patterns in Brazil; for, in the last two decades, its reduction was particularly seen among women aged 30 to 49, turning the Brazilian fertility pattern into a younger one. Therefore, the proportional increase in the number of teenage mothers and mothers younger than 20 also results from the change in the relative pattern.

In the Country as a whole, the contribution of births involving teenage mothers and youngsters below 20 years of age in the total of births moved from 17.3% in 1993, to 20.8% in 2003 - a 16.8% increase. The largest contributions of this group of age occurred in the North (25.8%), in the Northeast (23.3%) and in the Central-West (22.7%). Tocantins (28.3%), Pará (26.5%), Acre (26.2%), Rondônia and Mato Grosso do Sul (26%) registered the greatest percentages. On the other extreme, there were the Federal District (16.8%) and São Paulo (17.4%).

Death under-registration remains high

In Brazil, a significant number of Federation Units has poor-quality death registries. That implies difficulties in the construction of several population indicators, making it necessary to use alternative tools in order to come up with the estimates, for example, of infant mortality.

In the states of São Paulo, Rio de Janeiro, Minas Gerais, Espírito Santo, Paraná, Santa Catarina, Rio Grande do Sul, Mato Grosso do Sul and in the Federal District, the coverage of deaths, particularly of adults, is reliable. The biggest problem is still in the North and Northeast Regions, as shown in Graph 4.

Deaths of children aged under one decreased, but sub-notification remains

Concerning the deaths of children under one, the underreport in civil registry offices, for the country as a whole, was very high (48%). The biggest indexes were in the Northeast and North Regions, with average figures around 70% and 50%.

As a result, the differences in the values of infant mortality rate are very high when calculated by the direct method (relating infants' deaths with births recorded by the Civil Registry) and when compared with the rate estimated by indirect techniques.

Table 2 depicts the evolution of the contribution of infant deaths to the total of deaths for each Brazilian Region. There is a downward trend throughout the 1993/2003 period. For Brazil, the proportion of infant deaths in the total of deaths fell from 9.6% to 4.5%, a 53.1% reduction. The greatest drops were seen in the Northeast (58.5%), Southeast (52.4%) and South (49.3%).

But the reduction in the percentage of infant deaths seen in the Northeast might not describe the real situation of the Region's infant mortality levels, which go on being the highest ones in the country. The proportion found can be affected by the high underreporting levels of infant deaths, as mentioned before.

When the necessary measures are taken in relation to the data collected by the Civil Registry, it is possible to analyze the evolution of infant deaths, throughout the years, as well as the structure of infant mortality according to its components: early neonatal (deaths of children 0 to 6 days), late neonatal (deaths of children 7 to 27 days) and post neonatal (deaths of children 28 to 364 days).

Between 1993 and 2003, there was an expected concentration of mortality in the first days of life which constitute the early neonatal period. This concentration, though spread over all Regions, is even higher where infant mortality is relatively low, i. e., in the Center-South of the Country. In the Northeast, it is important to highlight that the post neonatal component is still high, a reflex of the worst social conditions and of the infrastructure of basic services. (Table 3)

Violent deaths

In 2003, in Brazil, the proportion of violent deaths was of 15.7% for men and of 4.1%, for women, i. e., violent deaths of men almost fourfold those of women (Table 5). In relation to 1993, there was a relative increase of 13.7% in the contribution of violent deaths among males in relation to the total of male deaths, whereas among women the percentages remain stable.

In relation to 2003, there was a slight rise in the percentages as shown in the following table. There were also drops in most of the regions, except for the Northeast, Central-West (where the violent deaths percentage had a slight increase in the period), and the Southeast (where the percentage of violent deaths among women rose 0.1 percentage points).

The Central-West Region presents the higher incidences of deaths related to violence; with percentages above 19% for men and 6% for women, with a downward trend as to the latter (it used to be 9% in 1996).   Due to the under-registration of deaths, particularly in the Northeast and North, it is necessary to relativize comparisons. Despite these restraints, collected data indicate the severity of the problem, especially in areas where the coverage is more complete.

Number of marriages is back to 1993's level

The Civil Registry Statistics do not count consensual unions. In 2003, there were 748 981 marriages, representing a return to 1993's figure (745 826). In this period, legal unions registered recovery partly because of collective weddings in several Federation Units, in a partnership between city halls and the Catholic Church, to legalize consensual unions. In 2003, there were more collective weddings than in the previous year, which contributed to a relative change of 4.7%.

A way of analyzing the evolution of marriages is to relate them to the total of the population in a marriageable age, estimating the overall rate of legal nuptiality¹. Comparing the 2003 rate (5.9 ‰) with that of 1993 (7.2 marriages for each thousand persons aged 15 or over), nuptiality is still low. But 2003 slightly offset 2001 and 2002 (5.7 ‰ in both).

In 2003, among women, the highest marriage rate occurred in the age group 20 to 24 (28.6‰). For men, the biggest rates were in the group 25 to 29 years of age (29.5‰). It is worth mentioning that women’s nuptiality rates are systematically lower than men's from the age group 25 to 29 on.

In 2003, on average, almost 10% of he spouses were 20 years old, whereas, among people older than 60, the marriage proportion was of 1.8%. In general, just 10.6% of the legal unions were among people aged 39. The average marriageable age, considering the set of legal unions, has been increasing systematically since the beginning of the 1990s. In 2003, the average age among women was 27.2 years old and, among men, of 30.6 years old. In 1993, the average ages were, respectively, 24 and 27.5 years of age (graph 9).

Number of legal separations and of divorces keep growing

The numbers of legal separations and of divorces keep gradually growing. From 1993 to 2003, the number of separations rose from 87 885 to 103 529 and of divorces, from 94 896 to 138 676 (or 17.8% and 44%, respectively).

The average age of legal separations and divorces has been increasing, following the pattern seen in the last decade. In graph 11, the average age of couples was higher in the cases of divorce than in the case of legal separations. That is probably due to the influence of the legislation in force, set by the 1988 Federal Constitution, which requires at least a year of legal separation or two years of actual separation so that the divorce process can begin.

In 2003, 77.9% of the legal separations and 68.7% of the divorces were consensual. In the legal separations and in the non-consensual divorces there were different patterns of petitioners. In non-consensual legal separations, the proportion of women petitioners (72%) was substantially superior to that of men (28%). However, in relation to divorce petitioners, the differences between men and women are significantly lower than those found in legal separations (46.6% to men and 53.4% to women). Probably, the growth of men wanting the divorce is related to the fact that they remarry in greater proportions than women.

The proportion of couples with minor children in legal separations (61.9%) was higher than in divorces (45.3%). Most of the times, mothers had the custody of minor children (91.4%), both in separations and in divorces (89.7%), followed by fathers (5.1% in separations and 6.1% in divorces). Only in 3.5% of the separations and in 4.2% of the divorces, did both parents share the custody of minor children.

___________________________

1 Nuptiality is obtained by dividing the number of marriages by the number of inhabitants and multiplying the final result by a thousand. This survey took into consideration marriages and population older than 15.