Using world fertility survey data to study household composition
De Vos, S.
Notas de Poblacion 25(45): 9-24
1987
ISSN/ISBN: 0303-1829 PMID: 12282297 Document Number: 363753
World Fertility Survey (WFS) data is used to study household composition in 6 countries in Latin America: Mexico, Costa Rica, the Dominican Republic, Panama, Colombia and Peru. WFS's household schedule is a census-like document that lists household members as present, absent and visitors. This issue is significant because between 10-26% of all households in Colombia, the Dominican Republic, Panama and Peru were affected; between 5-19% of the homes had an absentee member while 5-10% of the homes had a visitor. Visitors had fewer children and more young adults (15-24) than the present population. Absent members tended to be adult, males and either household heads or children of household heads. A recommendation is made to use a dejure definition that excludes visitors but includes absent members. Another critical issue in the WFS is the definition of head of household. The variation in defining this term in the 6 countries was between 66-76% because WFS classified households according to the number of couples they contained. However the Hammel-Laslett classification scheme, solitaire, no family, simple family, extended family and multiple family is recommended with WFS data. The issue of marital status is important and confusing because many of those in stable unions have never been legally or religiously sanctioned. These are counted as married. With regard to the quality of the data, Kabir's check on the internal consistency of data from 17 countries found errors in 2% (or less) of the households. A check for consistency between household and age revealed that consistency in information increased considerably when the age of groups were used in 5 year categories and not individually; 61% of those interviewed in Colombia had the same age, but when using 5 year age groups it rose to 88%.