Fertility and financial development: evidence from U.S. counties in the 19th century

This paper uses data on fertility and financial development in 19th century U.S. to test the hypothesis that more developed local financial markets reduce the incentives for families to have a large offspring to provide for them at old age, the so-called old-age security hypothesis. We find that the presence of banks is associated to lower children-to-women ratios and crude birth rates even after controlling for a large set of socio-economic factors. To account for possible endogeneity of bank location we instrument for the presence of some banking activity in a given county in 1840 with the existence of at least a bank in that county in 1820. The results of using this identification strategy are in line with the OLS ones, namely that fertility in 1850 is negatively affected by financial development. Next we explore the relationship between banking activity and fertility in the state of Pennsylvania, where, by law, most banks were created before 1820. This allows us to treat banks in 1840 as exogenous and confirm the existence of a strong negative causal effect from financial development to fertility. Finally, we show that our results are robust to measuring banking activity with the number of cities with at least a bank in a given county.