Measuring persistence of the world population: a fractional integration approach

Abstract

This paper uses fractional integration to measure the degree of persistence in historical annual data on the world population over the period 1800-2016. e analysis is carried out for the original series, and for its log transformation and its growth rate. e results indicate that the series are highly persistent; the estimated values of the diencing parameter are above 1, which implies that shocks have permanent e ects. Endogenous break tests detect one main break shortly a er WWII. e evidence based on the corresponding sub-sample estimation indicates a sharp fall in the degree of dependence between the observations in the second sub-sample. Although the original data and their log transformation still exhibit explosive behaviour in that sub-sample, the growth rates are meanreverting, and thus shocks will only have transitory effects; moreover, there is a negative time trend. is has implications for the design of policies aimed at containing population growth.