Abstract
A method to compare fertility sequences, each of which represents a cyclical fluctuation around an underlying trend, is presented. Each fertility sequence is viewed as composed of a time trend in childspacing pattern and a corresponding trend in terminal fertility rate. The procedure involved in comparing each component of one fertility sequence with the corresponding component of another is described and illustrated. Using hypothetical data it is shown, among other things, that ifA andB are two cyclical fluctuations in fertility such that (i) the cycles in both are of the same length, (ii) the terminal fertility components of the two are identical, and (iii) the trend in the mean age at childbirth associated withA is higher than the corresponding trend associated withB, then, replacingB withA need not necessarily result in lowering the long-run rates of increase in the birth sequence.
