Time period versus generation: how should trends in fertility be measured

Abstract

This paper starts by presenting the contrast between two approaches to depicting time-trends in fertility - the period and cohort perspectives - and discusses why it is that the existence of two approaches presents a problem for fertility analysis. It then considers the arguments for and against the period and cohort approaches to representing fertility trends. The conclusion reached is that there are clear grounds for preferring the period approach to the cohort perspective, but that the period measures currently used to depict fertility trends are defective and need to be replaced. I go on to examine what the criteria are for a sensible system of measurement for period fertility and present a (partly) new approach to constructing fertility indicators that meets these requirements. Recent trends in fertility in Ireland are considered briefly in the context of the potential uses of data collected by the current Irish birth registration system. Finally, some objections to the view advocated in the paper are discussed. The paper?s aims are, thus, both destructive and constructive, arguing against the currently conventional approach and suggesting an alternative system of measurement.