Childlessness: a further look at cohort estimates based on survey time-series data

Maire Ni Bhrolchain, University of Southampton
Eva Beaujouan, University of Southampton
Michael Murphy, London School of Economics and Political Science (LSE)

Survey data on fertility are an important source for monitoring trends in many fertility indicators, particularly where vital registration systems are incomplete with respect to past history. In contemporary low fertility societies, survey data have particular significance where vital registration systems do not record information on the true order of birth, this resulting from a number of lacunae in the collection of information on order of birth. Of particular importance, both for policy and for theoretical and explanatory purposes, is the proportion of women who have at least one birth, and its complement, the proportion childless. In the UK, estimates of the proportions childless have been based largely on survey data, the annual General Household Survey in particular, since the registration of births collects information only on the order of birth within marriage, thus omitting non-marital births from the classification of the order of a registered birth. This is a particular shortcoming when 44% of all births currently (according to the latest, 2007, national figures) occur outside marriage. While survey data on fertility in developed countries have generally been considered reliable, recent evidence of intra-cohort increases with rising age in the reported proportions childless (Murphy, Population Studies 2009) has highlighted major difficulties in this respect. The present paper examines this anomalous finding in more depth, and compares estimates based on direct questioning and on the own-child method. The paper looks at alternative explanations, including the association between reported childlessness and characteristics of individuals and their households, together with aspects of interview schedules and procedures.