Analyzing birth intervals: Implications for demographic theory and data collection

In the last decade, the study of birth intervals has been greatly facilitated by a rapid expansion in data availability and by improvements in analytical techniques. Unfortunately, the results emerging from individual level, empirical studies of birth interval dynamics do not correspond with the predictions of standard demographic theory. This paper reviews a series of individual level studies that find substantial socioeconomic variation in childspacing after controlling statistically for the major intermediate or proximate variables. It then offers possible explanations for the lack of fit between theory and results at the micro level, concluding that the two most likely explanations are poor measurement of the four principal proximate determinants and the exclusion of additional proximate determinants.