Distinguishing stopping behavior from spacing behavior with indirect methods.

A major controversy regarding fertility transitions in historical populations concerns whether the process consisted mainly of the diffusion of an innovation in stopping behavior termination of childbearing before the end of the reproductive lifespan or an adaptation to a new social and economic climate through more extensive spacing behavior deliberate prolongation of the intervals between births. The relative importance of stopping and spacing behaviors in fertility transitions also has implications for the general understanding of the way fertility-control decisions are made. This paper uses simulation methods to examine the ability of summary measures such as mean age at last birth and mean birth intervals to distinguish between the influence of increased stopping and spacing behaviors in fertility transitions. The author finds that while increases in some forms of spacing behavior do have a nonneglible negative impact upon mean age at last birth so that modest reductions in age last birth cannot on their own be used as conclusive evidence of increases in stopping behavior a technique such as McDonalds used together with changes in mean age at last birth can clearly differentiate between spacing behavior and stopping behavior. She also investigates the small-sample properties of mean age at last birth and finds that observed differences in mean age at last birth which are on the order of one or two years are unlikely to be due solely to the random variation which recurs in the small samples typically found in family reconstitution studies. The author found in subsequent analysis that examining changes in the pattern of increase of mean age at last birth with age at marriage is a more powerful way of distinguishing stopping from spacing than is examining changes in the overall mean age at last birth. Finally the author found that changes in the means of interbirth intervals prior to the last closed interval stratified by final parity are good diagnostic tools for distinguishing between stopping and spacing even when stopping behavior is less than 100% effective.