Demographics as Determinants of Social Security

Abstract

Population demographics have been long debated as factors that influence the capacity of countries to manage their debt levels. Drops in fertility and replacement rate can endanger the viability of a country’s social security system. Reproductive declines pose a particular threat to pension plans. In some cases, declines in population can increase the burden borne by the state. Employing a series of econometric models and machine learning techniques, we find evidence that population, age dependency ratio, fertility rate, migration and unemployment determine the level of social security benefits. These relationships depend on the model and the proxy used to measure the level of social security or pension benefits. Machine learning models tend to assign more weight to governmental policies. All else being equal, machine learning models find, that countries offering generous health and social benefits also fund pensions generously, and vice versa.