Does fertility affect growth? Evidence and simulation results from alternative quantile regression estimators

Brazil’s declining total fertility rate may enhance economic growth by mitigating the dilution of the capital stock, raising investment in human capital quality, and temporarily increasing the labor force relative to the whole population. As a lower total fertility rate implies a smaller family size in the coming years, we expect this change to improve an individual’s quality of education with favorable effects on income. This paper employs instrumental variables methods in conditional quantiles across 5564 Brazilian municipalities to identify the heterogeneous response of economic growth to declining fertility. The distribution of the economic benefits of lower birth rates is heterogeneous across quantiles in Brazilian cities. Conditional on the other places’ characteristics, the gains in income growth in reducing births are higher in locations where the speed of economic development is low. Our Monte Carlo simulation results indicate the sample sizes and error distributions for which three alternative instrumental variables quantile regression estimators are shown to be consistent. The convolution-smoothed approach with the triangular kernel performs better in finite samples in most scenarios.