In this paper, we attempt to determine whether the life-cycle income (LCI) hypothesis can explain movements in the national savings rate using the panel data of countries and the semi-nonparametric approach. While relating movements in the population density function to movements in the national savings rate, we are able to estimate the age response function with a high level of precision, and the estimated age response function is hump-shaped, which is generally consistent with the prediction from the LCI hypothesis. Running time-series regressions separately for individual countries, we also demonstrate that the estimated age response functions are consistent with the LCI hypothesis in a large proportion of countries, despite limited observations in a variety of countries. Finally, our time-series and cross-sectional analysis results imply that the LCI hypothesis is more likely to hold in a country wherein the growth rate of per capita GDP and the growth rate of population are high.
Read full abstract