Linking neighbors' fertility : third births in Norwegian neighborhoods

The aim of this paper is to gain more insight on the drivers behind geographical variations in family sizes by pointing out the role of neighborhoods and neighbors for two-child couples' transitions to third births. Couples' decisions about fertility behavior are influenced by their social context where immediate neighborhoods and neighbors may play a significant role. Furthermore, as neighborhoods are important contexts of childrearing, couples may sort geographically based on their fertility preferences. Using detailed geo-data from Norwegian administrative registers to locate couples in flexible ego-centered neighborhoods, this paper introduces a new dimension of spatial fertility variations. Results from regression models show that the family size of neighbors is positively related to each other. That is, the likelihood that two-child couples have a third child increases with the share of families with three or more children in the neighborhood. This relationship remains significant also after controlling for a range of couple characteristics, housing, neighboring women's educational level and time-constant characteristics of neighborhoods. It is also consistent for various neighborhood definitions which in this study range from the 12 to the 500 nearest neighbors. However, the strength of the association between neighbors’ fertility increases with the number of neighbors, providing evidence that residential sorting is a dominant driver.