Life cycle measures of coresidence with relatives in Brazil, 1960-2010

Abstract

In the first stage of the life cycle, children depend on parents care and financial support. As children grow older, they start new families, get married and have their own children. During adulthood, they can divorce/separate, become widowed or remarry. Finally, during old age, individuals may decide to live alone or reside with their adult children and other relatives until death. However, these phases are not fixed, and neither is the time spent in each one of them. In the last decades, there have been changes in both the duration and the age distribution of the different types of coresidence in Brazil, and we still have a lot to learn about this process. Therefore, the primary objective of this thesis is to look at the duration of coresidence of an average person in Brazil with different types of relatives over their life cycle. We are particularly interested in measuring how coresidence has changed between 1960 and 2010, a period of intense transformations in the country (demographic and non-demographic). Also, we examine in more detail the coresidence with own mother. We measure not only the extent to which the actual agesex profiles of this type of coresidence have changed in the last fifty years, but disentangle the two main factors responsible for these changes: the availability of mothers and the propensity of children to reside with them. For this purpose, we draw data from 1960, 1970, 1980, 1991, 2000 and 2010 Brazilian population census, as well from 1993, 2003 and 2013 PNADs (a nationally representative Brazilian household survey). We calculate the mean duration of residence with own mother, father, spouse, children, other relatives and any relatives by using a method based on the works of Wolfbein (1949) and Sullivan (1971). The results showed that the duration of coresidence increased for all coresidence types, except coresidence with other relatives and between fathers and their children. Most of the increase in duration was due to changes in lifespan due to mortality gains over the life cycle. Except for children residing with mothers, the proportion of persons in all other forms of coresidence by age and sex has indeed reduced between 1960 and 2010, despite the increase in duration. In addition, our decompositions showed that the factor most responsible for the longer duration of the coresidence between children and their mothers was the higher propensity of coresidence during young adulthood, and not the higher availability of mothers because of survival gains. These findings confirm previous studies in showing that youngsters are postponing the transition to adulthood. Our study contributes to the literature by emphasizing the importance of disentangling demographic and non-demographic factors when studying changes in coresidence over time