Seminar "Son preference and its impact on fertility in Central Asia and the Middle East: cohort comparison"
On May 23, PhD student of Stockholm University Konstantin Kazenin spoke at the seminar "Modern Demography" about the son preference and its impact on fertility in Central Asia and the Middle East in cohort comparison.
The seminar began with overview of the supporting factors of son preference in the countries of the Middle East and Central Asia. In addition to reducing and controlling the birth rate, many researchers write about the preservation of rigid gender asymmetries in society, i.e. the strong advantage of men in the labor market; formal and informal norms that make it necessary to have at least one male child; special responsibilities and reputational risks of parents of daughters, etc.
The speaker set a goal to compare the influence of son preference on decisions about the birth of the next child in different cohorts by year of birth. To do this, we used data from DHS (Demographic and Health Surveys) for countries where there is not a large prevalence of sex-based abortions and cohorts born in the 1940s – 1980s can be compared. The countries Egypt, Pakistan and Turkey were analyzed.
The author compiled models of proportional risks separately for the birth of second, third, and fourth children based on combined samples of all DHS for each country. The key independent parameter was the sex composition of living children. Due to the high child mortality, only living children were considered in each study period. The analysis included only women who had no deceased children at the beginning of the risk period. The beginning of the risk period was the month of the woman's life following the birth of the previous child. The analysis also included women who were married for the first time and the control parameters were the year of birth of the woman; education; age at birth of the first child.
The speaker showed two groups of models: for all cohorts, without the parameter of the sexual composition of children; separately for each ten-year cohort (1940-49, 1950-59, 1960-69, 1970-79, 1980-89), with the parameters of the sexual composition of existing children.
After analyzing the models, the author concluded that the role of the sex composition of existing children is growing from older to younger cohorts; only in the absence of a son in the family, the probability of the birth of 3 and subsequent children increases.