What Can We Learn About Fertility Transitions From the New York State Census of 1865?

Abstract

Aggregate variations in average children ever born and parity progression ratios across counties in New York, as reflected in the state census of 1865, show that while some of the inter-areal variation is due to differences in age structure and marital status, social characteristics of the counties are also quite important. In particular, housing values or living standards strongly depress the probability of additional births. This finding ties in well with other studies of the correlates of low fertility in nineteenth-century France and Massachusetts. The present study also finds that the decision to have a first child (not to be childless) has some different areal correlates than the decision to proceed from a moderate to large family; in particular, areal emphasis on schooling and Baptist religious affiliation are associated with high childlessness. Some previously emphasized social causes of the American decline, such as the shortage of available farmland, are found to be unimportant in New York State when other variables are included in the analysis.