Household consumption behaviour and fertility in Korea.

Families socioeconomic and demographic factors influence and shape their consumption and work-leisure behaviors and these behaviors in turn influence and reflect the correlates and proximate determinants of the families fertilities. Using data from the Korean 1980 census the authors studied 3 categories of variables: 1) situational 2) consumption and value preference and 3) fertility. There are 3 sample groups: 1) urban upper middle 2) urban poor and 3) rural. Results show that the rural group is poorer than the urban poor. Wives in the urban upper middle income group have a higher level of schooling than the other 2 groups. 98% of the urban middle hold white collar jobs or are self-employed. 11% of urban poor hold white collar jobs and 18% are self-employed. Most urban wives have no outside jobs. Rural households spend the greatest portion of the budget for food and housing and urban upper middle the least. 34% of the urban poor and 37% of rural households spend 40% or more of their budget on food and housing. The preference for a son over a daughter occurs in all groups when having 1 child. When having 2 children 65-78% of sample groups prefer having both a boy and a girl. 70.6% urban upper middle prefer a boy for psychological reasons; 43.9% of the urban poor prefer a boy to maintain family lineage and 39.7% for psychological reasons; 67.3% of rural households prefer a boy to maintain family lineage. The urban upper middle income group has the fewest number of children (the mode is 2); rural residents have the most (the mode is 4). Using regression analysis the authors show that those who spend a greater proportion of their household budget for food and housing tend to have more children than those who do not. Those who spend a greater proportion of the budget for items such as leisure clothing and medical care tend to have fewer children than those who do not. Those who expressed a preference for boy over a girl tend to have fewer children than those who do not. Rural-urban residence has the greatest impact on the number of children. Wives who do not have a job tend to have more children than those who do. The results of path analysis confirm and supplement the findings made from the regression analyses. It also shows that situational variables such as education income occupation and place of residence influence fertility by affecting consumption behavior and family beliefs expressed through sex preference of their offspring and the reasons for this preference.