Working Paper: CEPR ID: DP1832
Authors: Zvi Eckstein; Pedro Mira; Kenneth I. Wolpin
Abstract: This paper analyses the relationship between age-specific fertility, mortality and real wages in Sweden during the demographic transition. We take an overlapping generation?s model of life cycle fertility and fit it to actual Swedish time-series data over the past two and a half centuries. The model fits the data well, accurately portraying the total fertility decline from more than four children per female before the mid-19th century to about two children today. About 80% of this decline was in fertility that occurred at female ages over 30. The fitted model implies that reductions in child mortality over this period is the most important factor explaining the fertility decline, while real wage increases can explain only less than one-third of the decline in fertility. Their combined effect was, however, considerably larger than a simple summing-up would predict. The fertility decline was also magnified by the combination of increasing real wages and rising adult survival rates. In addition, we find that a model that is estimated using only pre-transition data would actually overstate the subsequent fertility decline.
Keywords: demographic transition; fertility; survival; overlapping generations
JEL Codes: 121; J11; N33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Reduction in child mortality (J13) | Fertility decline (J19) |
| Real wage increases (J39) | Fertility decline (J19) |
| Reduction in child mortality + Real wage increases (J39) | Fertility decline (J19) |
| Fitted model using pre-transition data (C51) | Overstatement of fertility decline (J11) |